- '4835 / 195 DEFINING AND MEASURING SUSTAINABILITY THE BIOGEOPHYSICAL FOUNDATIONS Edited by Mohan Munasinghe and Walter Shearer The United Nations University (UNU) and The World Bank DEFINING AND MEASURING SUSTAINABILITY The Biogeophysical Foundations Edited by Mohan Munasinghe and Walter Shearer Distributed for the United Nations University by The World Bank Washington, D.C. Copyright © 1995 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W. Washington, D.C. 20433 U.S.A. This publication may be reproduced in whole or in part and in any form for educational or nonprofit uses, without special permission from the copyright holder, provided acknowledgment of the source is made. Copies may be sent to the Environment Department, The World Bank, 1818 H Street N.W., Washington, D.C. 20433, U.S.A. and to the UNU, 53-70 Jingumae, 5-chome, Shibuka-ku, Tokyo 150, Japan. No use of this publication may be made for resale or other commercial purpose without the prior written permission of the copyright holders. The designations of geographical entities in this book, and the presentation of materials, do not imply the expression of any opinion whatsoever on the part of the World Bank or the UNU concerning the legal status of any country, territory, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. The interpretations and conclusions in this report are those of the authors and do not necessarily represent the views of the World Bank or the UNU. Library of Congress Cataloging-in-Publication Data Defining and measuring sustainability: the biogeophysical foundations / edited by Mohan Munasinghe, Walter Shearer. p. cm. Includes bibliographical references. ISBN 0-8213-3134-5 1. Sustainable development. 2. Natural resources-Management. 3. Sustainable development-Case studies. I. Munasinghe, Mohan, 1945- . 11. Shearer, Walter. HC79.E5D442 1995 333.7-dc2O 95-2865 CIP Contents Foreword ix Acknowledgments xi Contributors xiii An Introduction to the Definition and Measurement of Biogeophysical Sustainability xvii Mohan Munasinghe and Walter Shearer Part A. Background Papers 1 1. The Meaning of Sustainability: Biogeophysical Aspects 3 John P. Holdren, Gretchen C. Daily, and Paul R. Ehrlich Biogeophysical sustainability in context 4 Definitions of environmental sustainability 6 Biogeophysical sustainability in theory and practice 10 Contrasting views about the sustainability of human activities 11 The causes and character of environmental damage 12 Ignorance, knowledge, and uncertainty 14 2. Key Concepts and Terminology of Sustainable Development 19 Mohan Munasinghe and Jeffrey McNeely A historical perspective 20 Poverty and environmental degradation 21 General ideas about sustainable development 23 Concepts and definitions of sustainable development 24 Economic approach 24 Biophysical approach 27 Sociocultural approach 30 Reconciling different approaches to operationalize sustainable development 33 Economic approaches at the local or project level 33 Safe minimum standards 34 Valuation of environmental assets and impacts 35 Multicriteria analysis 38 Economic approaches at the macro level 40 Some practical steps at global, regional, and local levels 41 Sustainability indicators 42 Conclusions 44 Appendix 2-1. Current strategies for ensuring biodiversity 47 Appendix 2-2. Cost-benefit analysis 49 Defining and Measuring Sustainability: The Biogeophysical Foundations 3. Limits to Sustainable Use of Resources: From Local Effects to Global Change 57 Peter M. Vitousek and lane Lubchenco Human activity and global change 59 Human response to global change 61 4. Sustainability: The Cross-Scale Dimension 65 C. S. Holling A cross-scale journey 66 Ecosystem function 67 Dynamics of hierarchies 68 Two puzzles of sustainability 69 The ecosystem organization puzzle 69 The ecosystem management puzzle 71 Conclusions 73 5. Cumulative Effects and Sustainable Development 77 Gordon Beanlands Definitions, concepts, and approaches 78 Generic focus 79 Regulatory focus 80 Focus on vulnerable resources 80 Cumulative effects: A specific type of change 81 The decisionmaking dimension 84 Comparing cumulative effects and sustainable development 85 6. Managing Landscapes for Sustainable Biodiversity 89 H. Ronald Pulliam Landscapes of suitable patches 89 Habitat-specific demography: Sources and sinks 90 Spatially explicit models 92 Managing for diversity 93 Fragmentation and species diversity 95 Economic considerations 96 Sustainable landscape management 98 Comments, Eduardo R. Fuentes 100 7. Scale and Sustainability: A Population and Community Perspective 103 Simon A. Levin Mosaics and fragmented environments 105 Food webs 105 Global change in climate and ecological models 106 Pattern and scale 706 Conclusions 109 Comments, Charles H. Peterson 115 iv Contents 8. Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate 117 Meinrat 0. Andreae and Robert E. Dickinson Clianging physical climate and carbon dioxide 119 Changing chemical climate 126 Stratospheric ozone and ultraviolet radiation 127 Photochemical smog and acid rain 128 Global oxidation capacity 130 Detecting change 131 Monitoring chanige 132 Collclusionis 132 Comments, Eneas Salati and Reynaldo Luiz Victoria 134 9. Sustainability at Landscape and Regional Scales 137 R. V. O'Neill, C. T. Hunsaker, D. Jones, J. M. Klopatek, V. H. Dale, M. G. Turner, R. H. Gardner, and R. Graham Measuring pattern at landscape scales 138 Research necds in landscape and regional studies 140 Conclusions 147 Part B. Case Studies 145 10. Indicators of Biophysical Sustainability: Case Study of the Chaco Savannas of South America 147 Enrfti1je H. Bucher The Gran Chaco 147 Indicators of sustainability 150 11. The Sustainability of Natural Renewable Resources as Viewed by an Ecologist and Exemplified by the Fishery of the Mollusc Conchiolepas concholepas in Chile 153 I uat Carlos Custilla T he mechanistic price system and the main tenets of natural renewable resources 154 Two siml-ple contrasting models 155 T he example: The fishery of the mollusc Concholepas concholepas in Chile 156 The tragedy of the locos 157 12. Sustainable Development and the Chesapeake Bay: A Case Study 161 Christopher F. D'Elia Brief description and history of the Chesapeake Bay 163 The Chesapeake as a managemenit case study 164 Sustainability and the Chesapeake 165 Sustainability and growth 165 Monitoring and understanding sustainability 167 Crucial questions 171 Conclusions 174 V Defining and Measuring Sustainability: The Biogeophysical Foundations 13. Restoration of Arid Lands 177 G. Pickup and S. R. Morton General characteristics of Australia's arid zone 177 Sustainable pastoral management 178 Functioning of pastoral ecosystems 179 Measuring sustainability in the biological system 182 Restoring and maintaining sustainability in grazing lands 184 Management for conservation 185 Functioning of the natural ecosystem 186 Measuring sustainability 186 Implementing a sustainable conservation management scheme 188 Conclusions 189 14. Currencies for Measuring Sustainability: Case Studies from Asian Highlands 193 P. 5. Ramakrishnan Shifting agriculture, upland forests, and sustainable development: Case study of northeastern India 193 Development of the Philippine highlands 202 Case studies of other shifting agricultural systems 203 Conclusions 204 15. Large Marine Ecosystems and Fisheries 207 Kenneth Sherman Large marine ecosystems as global management units 210 Historical perspective 210 Perturbations and driving forces in large marine ecosystems 213 Management considerations 216 Ecosystem assessment and monitoring 218 Changing ecosystem states and health indexes 223 Present and future efforts 225 Part C. Managed Ecosystems 235 16. Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement 237 Nigel 1. H. Smith and Donald L. Plucknelt The special challenge of marginal lands 238 The pest and disease treadmill 239 Weeds as friend and foe 240 Tree crops as pillars of sustainability 241 The land use mosaic 242 Intensification: A pantropical imperative 243 A blend of traditional knowledge and modern science 243 An odyssey without end 243 Comments, R. Maria Saleth 248 vi Contents 17. Biophysical Measurement of the Sustainability of Temperate Agriculture 251 C. Lee Campbell, Walter W. Heck, Deborah A. Neher, Michael 1. Munster, and Dana L. Hoag Sustainability, agroecosystems, and monitoring for status and trends 251 Assessment and research monitoring programs 252 Measurements for monitoring sustainability 257 Research needed to develop and implement biophysical measurements of sustainability in agroecosystems 269 Conclusions 270 Comments, D. W. Anderson 274 18. Measuring Sustainability in Tropical Rangelands: A Case Study from Northern Kenya 277 Walter 1. Lusigi Tropical rangeland ecosystems 278 The IPAL study 289 Conclusions 302 Comments, Lee M. Talbot 304 19. Indicators of Grassland Sustainability: A First Approximation 309 Paul G. Risser Grassland ecosystems 309 Indicators and end points 310 Grassland indicators of ecosystem sustainability 311 Aboveground primary production 312 Species diversity 314 Soil organic carbon 315 Nitrogen contenit of the vegetation 315 Discussion 316 20. Sustainability in Tropical Inland Fisheries: The Manager's Dilemma and a Proposed Solution 321 Peter B. Bayley Approaches to predicting yield 321 The multispecies fishery and the manager's dilemma 322 A proposed solution: Progressive pulse fishing 323 Discussion 324 Conclusions: The fishery manager's responsibilities 326 21. Sustainable Development of Fisheries in Southeast Asia 329 Aprilani Soegiarto The physical setting 329 Fishery production 330 Problems of sustainable development 331 Conclusions 332 vii Defining and Measuring Sustainability: The Biogeophysical Foundations 22. Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement 335 Henry A. Regier, Stephen A. Bocking, and H. Francis Henderson Six kinds of fisheries, with an emphasis on sustainability and equity 337 Sustainability concerns in marine fisheries: A case study 343 The fisheries of the Great Lakes: Another case study 346 Sustainability, science, and the environment 348 23. Sustainability of Managed Temperate Forest Ecosystems 355 Jerry F. Franklin Definition of sustainability 355 Status of knowledge of major ecosystem processes 356 Biophysical measurements for temperate forests 369 Alternative management options 371 Existing materials for assessing sustained productivity 376 Conclusions and recommendations 380 Comments, Ian 1. Payton 386 24. Sustainable Management of Temperate Wildlife: A Conceptual Model 389 Richard L. Knight Managing for sustainable wildlife 390 Use of models in managing for sustainable wildlife 395 Data sets used for monitoring and predicting sustainability of temperate wildlife 395 Measures of sustainability 397 The challenge 397 25. Sustainability of Wildlife and Natural Areas 401 Kent H. Redford and John G. Robinson Biodiversity 402 The sustainability of tropical biodiversity 403 26. Tropical Water Resource Management: The Biophysical Basis 407 Jeffrey Edward Richey, Eneas Salati, Reynaldo Luiz Victoria, and Luiz Antonio Martinelli A definition of the biophysical basis of sustainability of the Amazon 408 The Amazon hydrologic cycle: The biophysical basis of the intact system 412 Future measurement protocols for maintaining sustainability 423 Management options for achieving sustainability of production of goods and services and of ecosystem integrity 425 Comments, Jose G. Tundisi 428 27. Limitations in Measuring Sustainability 431 Richard A. Carpenter Relevance to policy and management 431 Health of ecosystems 434 Agriculture 435 Soil 436 Forcstry 437 Water 438 viii Foreword The Report of the World Commission on Environment agreed to join us in assembling a group of out- and Development, also known as the Brundtland standing scientists in an International Confer- Commission, has directed the attention of the ence on the Definition and Measurement of international development community and the Sustainability: the Biological and Physical Foun- ecologically concerned to the goal of sustainable dations, held at the World Bank in June 1992. The development. This goal was the focus in June primary goal of the conferenceŽ was to explore the 1992 of the United Nations Conference on Envi- prospects of establishing a scientifically rigorous ronment and Development, which resulted in the definition and set of measures for sustainability. Rio Declaration, Agenda 21, and the UN Com- The sustainability of human economies in so- mission on Sustainable Development. cial, economic, and environmental termsrequires The true significance of this term presents dif- the continuous and uninterrupted production of ficulties because it involves two concepts that a minimum amount of natural biological prod- may appear incompatible at first sight: ucts for food. In addition, we have come to rely sustainability and development. Some havecome heavily on a wide var'ety of organic products for to associate various aspects of development with medicines, fibre, energy, and construction mate- destruction and degradation of the environment, rials. Thus, the net primary production generated while forotherssustainabilityconnoteseconomic by photosynthesis is a major pillar that sustains stagnation. If the goal of sustainable develop- human beings as well as almost all other forms of ment is to be achieved, a rethinking and reorgani- life on Earth. Maintaining biological productiv- zation of some widely-used concepts of develop- ity is a key to sustainability. ment will be needed, as well as a rethinking of the A major problem that we are now facing for views that equate sustainability exclusively with havingneglected thiscardinal principleemerged protecting flora and fauna. in the 1980s when it was estimated that a signifi- Before a new paradigm for development cant proportion of the net primary production of emerges, a better scientific understanding of the all terrestrial ecosystems was being diverted di- requirements for sustainability is essential. We rectlyor indirectly to human use. With theantici- must be able to compare development and pated doubling of the human populationby 2050 sustainability at different times and in different and the urgent need for development, if present places to ensure that both are being achieved. To patterns of consumption persist, humanity will carry this out, cleardefinitions and practical mea- soon face the problems arising from diverting suringscalesarerequired.Theconceptofdevel- for its own use an ever increasing and opment has been studied in great detail, and its unsustainablyhigh volume of natural resources. definition, measurement scales, and the prob- In the face of such over-exploitation, the qual- lems of the indicators in current use have been ity of the environment and of human life is debated at length. However, comparatively little likely to decline rapidly, accompanied by wide- effort has been devoted to how sustainability is to spread suffering. be defined and measured. The factors necessary for the maintenance of As an international community of scholars biologicalproductionincludefertilityornutrient working on the problems of human survival, availability, energy, adequate moisture, proper development, and welfare, the United Nations substrates, subcritical levels of toxic substances, University(UNU)wasestablished, withitshead- and an adequate and genetically varied stock of quarters in Tokyo, Japan, to bring the best minds biological organisms. Such factors form the to bear on global problems of this nature. The biogeophysical foundationofsustainability.Only World Bank has been pursuing the basic goals of through biogeophysical measurements can the development and poverty eradication for five status and trends of natural and managed ecosys- decades. It was for this reason that we were tems be established and the true effectiveness of pleased when the other sponsoring organizations sustainable management practices be assessed. ix Defining and Measuring Sustainability: The Biogeophysical Foundations The tasks before the conference participants flect net primary productivity, biological diver- weredifficultones-to make significant progress sity, and perhaps other factors integrated over a in: (a) agreeing on a scientific definition of mosaic of different ecosystems. Recent work on biogeophysical sustainability; (b) providing the environmental indicators at the World Bank and biogeophysical framework to complement the elsewhereunderlinestheformidableproblemsof social and economic dimensions of an overall defining, measuring, analyzing, and interpreting measure for sustainability; and (c) recommend- such indicators. ing indicators and biogeophysical measures for This volume, which presents the highlights of monitoring and predicting the sustainability of the conference, shows that the participants have the maior managed ecosystems. indeed taken a major step forward toward realiz- The participants were presented with an even ing the tasks set before them. At the same time, greater and more important challenge-to rec- the ten conference recommendations indicate ommend, after careful consideration, a scientifi- the formidable problems that still remain in cally sound and practical set of indices of attempting to agree on a practical scientific biogeophysical sustainability. The ecological definition of and set of measures for equivalent of the system of national accounts and sustainability. The volume contains a wealth of the gross national product employed in econom- thought, discussion, and debate that will have ics would be of immense use in identifying over- to be taken into account in the final formulation stressed regions, comparing management prac- of the practical definition and set of measures. tices, and monitoring the effects of policy deci- For taking time from their important responsi- sions. To be practical, such indices would have to bili ties to focus for a few days on this important be based on data that can be readily and fre- aspect of sustainable development, the United quently obtained and be applicable to regions Nations University and the World Bank extend with landscape, national and continental dimen- their appreciation to all the distinguished par- siouis. Yet the indices would have to be transpar- ticipants who took part. ent enough for policymakers and citizens alike to As this volume is the result of a cooperative comprehend their fundamental significance in effort, we would like to conclude by thanking all the same way that people intuitively understand the co-sponsors and collaborators listed in the the meaning of the GNP. Thus, it appears that acknowledgments who made the conference suc- these measurements would somehow have to re- cessful and this publication possible. Heitor Gurgulino de Souza Ismail Serageldin Rector, United Nations University Vice President Tokyo Environmentally Sustainable Development The World Bank, Washington, D.C. x Acknowledgments As this volume is the result of a collaborative valuable support and assistance in convening effort, a special debt of gratitude is owed to the the conference; and following co-sponsors who made the conference * the governments of Norway and Sweden for and this publication possible: partial financial support. * the United States Environmental Protection Sincere appreciation is extended to all the Agency's Environmental Monitoringand As- distinguished participants who took part in the sessment Program and, in particular, Dan conference for taking time from their important McKenzie, who generously supported the responsibilities to focus for a few days on this conference; important aspect of sustainable development. * the Smithsonian Institution and especially Gratitude is also extended to Courtesy Associ- Assistant Secretary Thomas Lovejoy for his ates, particularly to Stacey Sickels, whose assis- contribution to the scientific organization and tance was crucial to the professional way in which leadership of the conference; the conference was managed. The editors extend a special note of thanks to * the East-West Center in Honolulu, where the the authors of the valuable contributions that first planning meeting for the conference was made this volume possible and who cooperated held and which also played a role in support- so agreeably during the editing process, and to ing the conference, and especially Richard Adelaida Schwab for expertly shepherding the Carpenter, without whose experience, insight, manuscript, to Maria Theresa Camilleri for dili- and perseverance the conference could not gently assembling the manuscript and transcrib- have achieved so much; ing the discussions, to Ramon Ray for preparing * the Ecological Society of America and its the graphics, and to Elizabeth Forsyth, Didier Sustainable Biosphere Initiative, and in par- Godat, Connie Eysenck, and Stephanie Gerard ticular its former president, Jane Lubchenco, for invaluable assistance during the editing and as well as Jim Gosz and Stephanie Cirillo, for production process. xi Contributors Authors Darwin W. Anderson is with the Department of Soil Science at the University of Saskatchewan, Saskatoon, Saskatchewan, Canada. Meinrat 0. Andreae is director of the Biogeochemistry Department of the Max Planck Institute for Chemistry, Mainz, Germany. Peter B. Bayley is with the Illinois Natural Survey in Champaign, Illinois. Gordon Beanlands is associate professor in the School for Resource and Environmental Studies at Dalhousie University in Halifax, Nova Scotia, Canada. Stephen A. Bocking is assistant professor in the Department of Environment and Resource Studies at Trent University in Peterborough, Ontario, Canada. Enrique H. Bucher is with the Centro de Zoologia Aplicada of the Universidad de C6rdoba in C6rdoba, Argentina. C. Lee Campbell is with the Department of Plant Pathology (USDA Air Quality) at North Carolina State University in Raleigh, North Carolina. Richard A. Carpenter is a consultant who was formerly with the East-West Center of the Environment and Policy Institute in Honolulu, Hawaii. Juan Carlos Castilla is a full professor in the Facultad de Ciencias Biol6gicas (Ecologia) at the Universidad Cat6lica de Chile, in Santiago, Chile. Gretchen C. Daily is the Winslow and Heinz Foundation Postdoctoral Fellow jointly at the Energy and Resources Group of the University of California in Berkeley, California and the Center for Conserva- tion Biology at Stanford University in Stanford, California. V. H. Dale is with the Environmental Sciences Division at the Oak Ridge National Laboratory in Oak Ridge, Tennessee. Christopher F. D'Elia is provost, University of Maryland Biotechnology Institute in College Park, Maryland. Robert E. Dickinson is a professor in the Institute of Atmospheric Physics at the University of Arizona in Tucson, Arizona. Paul R. Ehrlich is the Bing Professor of Population Studies at Stanford University in Stanford, California. Jerry F. Franklin is with the College of Forest Resources at the University of Washington in Seattle, Washington. Eduardo R. Fuentes is with the Global Environment Facility unit of UNDP in New York. Keechichiro Fuwa is Professor Emeritus of the University of Tokyo and President of the Society of Environmental Science in Tokyo, Japan. R. H. Gardneris with the Environmental Sciences Division at theOak RidgeNational Laboratoryin Oak Ridge, Tennessee. R. Graham is with the Environmental Sciences Division at the Oak Ridge National Laboratory in Oak Ridge, Tennessee. Walter W. Heck is with the USDA's Agricultural Research Service in Raleigh, North Carolina. H. Francis Henderson is retired from the Fisheries Division of the United Nations Food and Agriculture Organization in Rome, Italy. Dana L. Hoag is a professor in the Department of Agricultural Economics at North Carolina State University in Raleigh, North Carolina. xiii Defining and Measuring Sustainability: The Biogeophysical Foundations John P. Holdren is the Class of 1935 Professor of Energy at the University of California, Berkeley, and visiting distinguished scholar at the Woods Hole Research Center in Woods Hole, Massachusetts. C. S. Holling holds the Arthur R. Marshall, Jr. chair in ecological sciences at the University of Florida in Gainesville, Florida. C. T. Hunsaker is with the Environmental Sciences Division at theOak Ridge National Laboratory in Oak Ridge, Tennessee. D. Jones is with the Energy Division of the Oak Ridge National Laboratory in Oak Ridge, Tennessee. J.M. Klopatek is a professor in the Biology Department at Arizona State University, Tempe, Arizona. Richard L. Knight is associate professor in the Department of Fishery and Wildlife Biology at Colorado State University in Fort Collins, Colorado. Sharad Lele is at the Harvard Institute for International Development in Cambridge, Mass. Simon A. Levin is the George M. Moffett Professor of Biology in the Department of Ecology and Evolutionary Biology at Princeton University in Princeton, New Jersey. Jane Lubchenco is professor of zoology at Oregon State University in Corvallis, Oregon. Walter J. Lusigi is a senior ecologist in the Environmentally Sustainable Development Division of the Africa Technical Department, World Bank. Jeffrey A. McNeely is chief conservation officer of the World Conservation Union (IUCN) in Gland, Switzerland. Luiz Antonio Martinelli is with the Centre for Nuclear Energy in Agriculture (CENA) at the University of Sao Paulo, in Piracicaba, Brazil. S. R. Morton is senior principal research scientist at the Centre for Arid Zone Research in Alice Springs, N.T., Australia. Mohan Munasinghe is division chief for Environmental Policy and Research at the World Bank. Michael J. Munster is with the Agrosystem Resource Group of the Environmental Monitoring and Assessment Program in Raleigh, North Carolina. Deborah A. Neher is a professor in the Department of Plant Pathology at North Carolina State Uni- versity in Raleigh, North Carolina. R. V. O'Neill is with the Environmental Sciences Division at the Oak Ridge National Laboratory in Oak Ridge, Tennessee. Ian J. Payton is with the Forest Research Institute in Christ Church, New Zealand. Charles H. Peterson is with the Institute of MarineSciences at the Universityof North Carolina in Chapel Hill, North Carolina. Geoffrey Pickup is a program leader in the Division of Wildlife and Ecology at the Centre for Arid Zone Research in Alice Springs, N.T., Australia. Donald L. Plucknett isa consultant in international agricultural research and a former scientificadviser with the Consultative Group on International Agricultural Research at the World Bank. H. Ronald Pulliam is professor and director of the Institute of Ecology at the University of Georgia in Athens, Georgia. P. S. Ramakrishnan is professor of ecology and dean of the School of Environmental Sciences at Jawaharlal Nehru University in New Delhi, India. Kent H. Redford is associate professor and director of the Program for Studies in Tropical Conservation in the Center for Latin American Studies at the University of Florida in Gainesville, Florida. Henry A. Regier is professor of environmental studies and zoology at the University of Toronto's Ramsay Wright Laboratories in Toronto, Ontario, Canada. Jeffrey Edward Richey is with the School of Oceanography at the University of Washington in Seattle, Washington. Paul G. Risser, a botany specialist, is the President of Miami University in Oxford, Ohio. xiv Contributors John G. Robinson is vice president for international conservation of the Wildlife Conservation Society in New York. Eneas Salati is with the Fundacao Brasileira para Desenvolvimento Sustentavel in Rio de janeiro, Brazil. R. Maria Saleth is with the M.S. Swaminathan Research Foundation in Madras, India. Walter Shearer is the Advisor on New and Renewable Sources of Energy at the United Nations University office in New York. Kenneth Sherman is chief of the Branch of Ecosystems Dynamics of the Northeast Fisheries Science Center in Narragansett, Rhode Island, and adjunct professor of oceanography at the University of Rhode Island in Kingston, Rhode Island. Nigel J. H. Smith is a professor of geography at the University of Florida in Gainesville, Florida, and a former long-term consultant for the World Bank. Aprilani Soegiarto is Deputy Chairman for Natural Sciences at Lembaga llmu Pengetahuan Indonesia (Indonesian Institute of Sciences) in Jakarta, Indonesia. Lee M. Talbot is an environmental consultant at the World Bank. Jose G. Tundisi is a professor at the Centre for Water Resources and Applied Ecology at the University of Sao Paulo in San Carlos, Brazil. M. G. Turner is with the Environmental Sciences Division at the Oak Ridge National Laboratory in Oak Ridge, Tennessee. Reynaldo Luiz Victoria is with the Department of Physics and Meteorology at the Escola Superior de Agricultura "Luis de Queiroz" in Piracicaba, Brazil. Peter M. Vitousek is a professor in the Department of Biological Sciences at Stanford University in Stanford, California. Other Participants Stephanie Cirillio is Program Assistant at the Sustainable Biosphere Initiative in Washington, D.C. Ronald Carrol is with Sustainable Agriculture and Natural Resource Management (USAID) at the University of Georgia in Athens, Georgia. Ned Cyr is with the National Oceanic and Atmospheric Administration in Washington, D.C. Della Dennis is with the Sustainable Biosphere Initiative in Washington, D.C. Jerrold Dodd is at the University of Wyoming's Range Management Departmentin Laramie,Wyoming. Jerry Filbin is with the United States Environmental Protection Agency in Washington, D.C. Steven Gasteyer is with the Committee on Agricultural Sustainability for Developing Countries in Washington, D.C. Jerome Glenn is Executive Director for the American Council for the United Nations University in Washington, D.C. Robert Goodland is Environmental Assessment Advisor at the World Bank James Gosz is Director of the Division of Environmental Biology at the National Science Foundation in Washington, D.C. Judy Gradwohl is Director of the Office of Environmental Awareness at the Smithsonian Institution in Washington, D.C. Heitor Gurgulino de Souza is Rector of the United Nations University in Tokyo, Japan. Richard Haeuber is with the Sustainable Biosphere Initiative in Washington, D.C. Carole Hamilton is with the U.S. Bureau of Land Management in Washington, D.C. Gary Hartshorn is with the World Wildlife Fund in Washington, D.C. Eric Hyman is with Appropriate Technology International in Washington, D.C. xv Defining and Measuring Sustainability: The Biogeophysical Foundations Peter R. Jutro is with the United States Environmental Protection Agency in Washington, D.C. Thomas Lovejoy is Assistant Secretary for External Affairs at the Smithsonian Institution in Washington, D.C. Elizabeth McCance is with the Office of Environmental Awareness at the Smithsonian Institution in Washington, D.C. Dan McKenzie is with the U.S. Environmental Protection Agency's Environmental Monitoring and Assessment Program at Research Triangle Park, North Carolina. James McNair is with the Academy of Natural Sciences, Washington, D.C. Katy Moran is with the Office for External Affairs at the Smithsonian Institution in Washington, D.C. Henri Nsanjama is with the World Wildlife Fund in Washington, D.C. Gordon Orians is a professor in the Department of Zoology at the University of Washington in Seattle, Washington. Marissa Perrone is with the Office of the U.S. Trade Representative, Washington, D.C. Kathryn Saterson is Director of the Biodiversity Support Program at the World Wildlife Fund in Washington, D.C. Ingrid Schultz is with the Environmental Protection Agency's Environmental Statistics Initiative in Washington, DC. Roger Soles is with the U.S. Man and Biosphere Program at the Department of State in Washington, D.C. Silvia Tognetti is with the National Academy of Sciences in Washington, DC. Dan Tunstall is with the World Resources Institute in Washington, D.C. Susan Ware is an Internation Affairs Specialist at the National Oceanic and Atmospheric Administration in Washington, D.C. Naomi Weisman is with Technology Resources, Inc. in Laurel, Maryland. xvi An Introduction to the Definition and Measurement of Biogeophysical Sustainability Mohan Munasinghe and Walter Shearer This volume is based on papers prepared for the 40 percent of the net primary production of all International Conference on the Definition and terrestrialecosystemswasbehgdiverteddirectly Measurement of Sustainability: The Biophysical or indirectly for human use. Thus, if the global Foundations, which was convened at the World population were to triple while production and Bank in Washington, D.C. from June 22 to 25, consumption patterns remained the same, hu- 1992. manity would be faced with trying to co-opt for The conference was conceived as an exercise to human use more than the total global net annual advance thinking on the biogeophysical founda- production of organic matter. This in turn would tions for defining and measuring sustainability rapidly lead to diminishing production because and to provide useful policy considerations for of areductionofthere3ourcebase.lnthiscontext, the international development community. In it appears that net primary production offers one addition to the formal papers and written re- basisforunderstandingthelimitsofsustainability, sponses in the volume, this introductory chapter as well as some hope for developing from among is an attempt to describe the general context in all the variables a set of indicators (and perhaps which the conference took place and to capture an index) of one fundamental aspect of the highlights and salient points of the many sustainability-its biogeophysical component. papers presented and the discussions that tran- In 1987 the Report of the World Commission on spired. The papers themselves have been up- Environment and Development (the Brundtland dated and revised in light of the conference delib- Commission) and the popular version,Our Com- erations and subsequent developments. mon Future, proposed the concept of sustainable development. The adoption of this idea by the international development community and its Origins of the conference promotion as the theme of the UN Conference on Environment and Development (the Rio Confer- The motivation for the conference derives from ence)inl992furtherinspiredanattempttodefine the age-old concern about humanity's impact on and measure sustainability. The basic challenge theenvironmentand theultimatecarryingcapac- presented by sustainable development is in finding ity of the earth. Many thinkers, from Thomas ways to define, measure, and operationalize it. The Robert Malthus onwards, have expressed con- purpose of the conference was to contribute to- cern about the material limits to the magnitude of ward the development of a practical way of mea- human consumption and the support system re- suring movement, if any, toward greater quired for its sustenance. Others, most notably sustainability. the so-called technological optimists, have ar- As defined by the Brundtland Commission, gued that scientific progress and the increasing sustainable development must meet the needs of efficiency of resource use would help overcome the present without compromising the ability of this problem. This debate naturally led to the idea future generations to meet their own needs. This of sustainability and a search for ways to measure implies using resources today without limiting and monitor it. One example of the importance of the options for succeeding generations; of neces- such indicators emerged in 1986 when Peter sity, this must involve environment and natural Vitousek, Paul and Anne Ehrlich,and R.A. Matson resources as well as social dimensions. However, (in their well-known article in BioScience) esti- this relatively new and attractive concept is im- mated that the organic matter equivalent to about possible to satisfy in literal terms because, on the Defining and Measuring Sustainability: The Biogeophysical Foundations one hand, the current generation has to continue Organization of this volume altering the biosphere in order to develop, and on the other hand, each time a history-dependent The contributions to this volume are arranged system such as an ecosystem is modified, future along the lines of the conference. Part A covers the options change as well. This makes consideration major issues that affect all ecosystems in relation to of such impacts one of the most important issues biogeophysical sustainability.Theissuestreatedby in sustainable development. Therefore, a com- experts include key concepts and terminology, spa- promise is required between current uses and tialandtemporalscales,limitstothesustainableuse future options. Furthermore, it is important that of resources, cumulative effects, source/sink mod- sustainable development be understood not as eling of landscapes, and atmosphere and climate. decreeing a subsistence economy for Third World Several chapters are accompanied by reviews from countries, nor as freezing landscapes in a particu- other participants. In view of the key role played by lar configuration, but as conscious development the overview papers, we elaborate on them briefly. within bounds determined through the best sci- The first chapter, by Holdren, Daily and Ehrlich, entific evidence. lays out the issuesand concept of sustainability,sets Such discussions must avoid confusing the a specific time frame of 500-1,000 years for evaluat- terms sustainabledevelopmentand sustainability. ing sustainability, and notes that the crucial change The former is about promoting development and rate is about ten percent per century. ensuring that it issustained. It involves two seem- Thecontributionby Munasingheand McNeely ingly incompatible concepts: sustainability and provides three approaches to looking at development. Sustainability means maintenance sustainability: the economic, that is, growth and or even improvement, without degradation, over development; the environmental and natural re- the very long term. After decades of research, source; and the social and cultural. It compares intervention, and observation, profound scrutiny the policy relationships of these three approaches of many aspects of development is occurring. and points out the challenges for the world in However, the current state of understanding re- seeking to integrate them with sustainability. garding sustainability is poor, and it is this con- Because the challenges are great, the authors cept that needs the urgent attention of the intel- emphasize the need to clarify the issues not only lectual community. for decisionmakers but for the public who must In preparing for the 1992 meeting, it was enact the appropriate changes. recognized that sustainability involves complex The third contribution, by Vitousek and interactions-biogeophysical, economic, social, Lubchenco, brings to the fore three major issues for cultural, and political-and that sustaining the the elucidation of sustainability: the issue of global global life-support system is a prerequisite for processes that can sometimes manifest themselves sustaining human societies. This is why defining as global change; the idea of biological diversity; and measuring the biogeophysical foundations and the land-use issue, which becomes global by of sustainability became the theme of the confer- aggregation but can be important both locally ence. It isalso important to point out to the reader and regionally. that during the conference and in preparing the Hollingcontributesapersuasivechapteronadap- chapters in this volume, the term "biophysical" tive systems and furnishes an eloquent discussion was used. Subsequently it was decided that to of paradoxes-notonlytheparadoxofsustainability avoidpossibleconfusionwithbiophysics,atradi- and development, but also the paradox that all tional and well-established subdisciplineof phys- systems, no matter which ones are examined, ap- ics, the term used in the title of the present publi- pear to be governed primarily by a small number of cation and throughout this introduction would keyvariables.Althoughthesevariableshavediffer- be "biogeophysical." Although it has some draw- ent time scales, they aggregate into clusters, which backs, the new term is recommended for future provides an indication that the systems are hierar- use when referring to the natural life-support chical, have a definite structure, and should be system, the comprehensive study of which relies accessible by examining key variables. The author on the disciplines of biology, geology, chemistry, also looks at the notion of policy changesand man- and physics. Therefore, for the sake of consis- agement strategies, noting that in many cases man- tency, "biogeophysical" should be read in place agement appears to bequite clearand productive at of "biophysical" wherever it is encountered in the the outset but, over time, becomes essentially dys- remainder of the text. functional if not pathological. xviii An Introduction to the Definition and Measurernent of Utogeophyslcal Sustainability Beanlands' chapter takes on the difficult chal- threats to sustainability involve changes in land lenge of addressing cumulative impacts, point- use at large scales. The authors indicate that the ing out that there are indeed some relationships combination of remote sensing, geographic i nfor- between sustainable development and cumula- mation systems, and landscape theory offers a tive impacts but that they are difficult to measure framework for defining a significant number of and generally still beyond the reach of science. new indicators that are practical and interpret- Mostof thesecumulativeimpactsare slow changes able. However, they warn that continued progress on the order of 1-3 percent, however, and the involves difficult new areas of research. challenge is to find tnresholds that become criti- Part B contains reports of locations on the cal and to use them to drive decisions and man- planet where the environment and society's agement actions. reactions have teen studied in great depth by Pulliam's contribution looks at source-sink specialists. These locations include a rocky inter- modeling, focusing primarily on ecosystem sta- tidal marine coast in Chile, the Chesapeake Bay in bility measured in terms of productivity, diver- North America, arid zones in Australia, high- sity, stability, and adaptability. Using these mod- altitude forests in Asia, and large marine ecosys- els, the authorextends these interpretations to the tems and fisheries. landscape level, in this way providing models for Part C offers a series of reports on a variety of predicting population behaviour in terms of het- managed ecosystems for which leading scientists erogeneouslandscapes. The author then proposes provide theirexpert opinion on the current status some management strategies at this level on the of biogeophysical indicators of sustainability. basis of the monitoring of a single species. These reportscover agriculture,rangelands, fish- In the seventh chapter, Levin raises several eries, forests, wildlife, and natural areas and wa- important points, including the value of focusing ter resources in the tropical and temperate zones. on flexibility rather than on constancy, the need Reviews of several of these reports immediately in terms of species and populations to focus on follow the principal contributions. processes for the persistence of those species' Below wereview themain points that emerged survival and extinction, their spread in variabil- during the conference. ity, and the importance of spatial and temporal relationships between disturbances and those population processes. Sustainability means dif- Dimensions of sustainability ferent things at different scales, and the issue is howtoaggregatemeasurementsintheabsenceof Before considering the matter of defining and a correct level of aggregation. measuringbiogeophysical sustainability, thepar- Andreae and Dickinson assess changes in the ticipants wrestled with the concern that some- chemical and physical climate in light of the chang- thing important might be lost in trying to define ing composition of the atmosphere in relation to the biogeophysical aspects of sustainability inde- sustainability. The increasing concentration of pendent of its human and social dimensions. greenhouse gases in the atmosphere, the weaken- Concern was voiced that sustainability ap- ing ozone shield, increased local and regional pears to be essentially a social construct, and one pollution, and the decreasing capability of the way to determine the contribution of biologists, atmosphere to oxidize biogenic and anthropo- ecologists,andgeoscientists to thebiogeophysical genic emissions are symptoms of progressive basis of human life is to understand where social and unsustainable deterioration of the life-sup- issues come into play. First, it was proposed that portsystem provided by theglobal atmosphere. social issues become important when determin- The authors point out that because of multiple ing what is to be sustained, for how long, and in sources and sinks for many gases and aerosols, what manner, whether sustaining refers only to an effort to reduce one source can lead to in- nondecreasing average production or to resilience creased emissions of the same or another gas and adaptability, who is to benefit from what is from a different source. being sustained, and how the benefits will be The final chapter of Part A, from O'Neill and distributed. It was suggested that these determi- others, considersbiogeophysical sustainabilityat nations can be made only in the social arena, with the landscape and regional levels. It emphasizes the setting of the objectives or attributes of sustai- that landscapeecology is important because most nability emerging from a sociopolitical process. xix Defining and Measuring Sustainability: The Biogeophysical Foundations Second, it was suggested that social issues provide a way of identifying unsustainable prac- come into play when trying to understand why tices, more understanding of the human causes unsustainable practices and behaviour are so fre- would be valuable in this endeavour. quently observed among shortsighted people who Response to this view was that development thus undermine the ecological basis of their own and use of an index or set of indicators do not lives or the lives of future generations. These replace the need for disaggregated information considerations are extremely complex and in- and better understanding of the social causes of volve individual behaviour and social and politi- unsustainable practices. In fact, both approaches cal structures and organizations. Such informa- are needed and have complementary applica- tion is helpful in developing policies that would tions. The index is valuable for its monitoring and change unsustainable practices and promote the predictive powers, whereas more detailed stud- maintenance of the ecological basis of human life. ies and more understanding of human manage- It was also observed that with social and political ment practices are useful for developing and aspects determining the objectives of implementing policies to reduce unsustainable sustainabil i ty, there is bound to be conflict among practices. theseobjectives.Undersuchcircumstancesitwas A somewhat different view presented at the suggested that the task of the scientist is to pro- conferenceisthatbeyondsuchmattersaswhether vide information to help people choose among thereissufficientgenerationof financialresources conflicting objectives by assessing the trade-offs to ensure economic sustainability, whether there among these objectives and the consequences of is appropriate social organization and sufficient theirapplication. Thisinformationwouldbepar- motivation for social sustainability, or whether ticularly important where the trade-offs and con- there is sufficient power or collective will for sequences behave nonlinearly. This view sug- political sustainability, what is sustainable and gests that having recognized such concerns, it unsustainable is ultimately a biogeophysical, and should be easier to deal with sustainability with- not a social, matter. For example, it is a social out always having to address the social and cul- decision tochoosea land-usepattern fromamong tural issues. several options, but the sustainability of each A pragmatic and policy-oriented concern was pattern will be determined by whether the expressed in support of this view-focusing on biogeophysicalconditionscanbesustainedwhen developing one index or a set of indicators of the land is used. The only exception to this would sustainability runs the risk of aggregating to- appear to be in deciding to accept something less gether many distinct elements and thereby fail- than indefinite sustainability and adopt a par- ing to understand the human causes and mecha- ticular time frame, or in defining the spatial area nisms of unsustainable management. It is ex- to be sustained. Otherwise the concern raised pected that examination of what is known about about social issues does not appear to require sustainability in all its different aspects-such as consideration in dealing with biogeophysical renewable resources, nonrenewable resources, sustainability. Assuming a very long time hori- ecosystem services, and biological diversity- zon for sustainability, no new social issue ap- will reveal that these elements are distinct in the pears. Similarly, the manner in which an ecosys- way they absorb impacts from their use and in tem is sustained is not relevant as long as what they contribute to human well-being. Keep- sustainability actually occurs. And the concerns ing them separate, not insisting that theybemixed about who is to benefit and how the benefits will into one single indicator, and focusing instead on be distributed are not relevant because understanding why in specific cases there is sustainability can occur even if no one benefits at unsustainable use of a particular resource, eco- all. According to this view, the reasons for system service, orbiologicaldiversity mightactu- unsustainable practices and how they can be ally contribute to understanding (and thereby changed belong to the social and political realm obviating) the social causes of environmentally and are irrelevant to defining and measuring unsustainable practices. Although sustainability biogeophysical sustainability. Those holding this islargelyanaturalstateofaffairs,unsustainability view insist that sustainability in fact is the very is the result of social actions, and, therefore, re- simple concept of maintaining natural resource ducing these impacts involves a social process. stocks and that it should not be unnecessarily Thus, it would seem that if the goal of devising an complicated by encumbering it with social fea- index or set of indicators of sustainability is to turesand desired expectations. This would make xx An Introduction to the Definition and Measurement of Biogeophysical Sustainability it possible to identify a small number of indica- An innovative model was proposed for view- tors for describing biogeophysical sustainability ing this broader picture of sustainability, employ- thatwouldnotbesociallyandculturallysensitive ing an equilateral triangle with the ecological, and thus applicable in any location. economic, and sociocultural objectives of A simple analogy elucidates this position. To sustainabilityat itsapexes(seeMunasinghe 1993, keep the world running, if one could control it, and Munasinghe and McNeely this volume).This the first requirement would be to keep the offers an elegant way of reflecting the compro- biogeophysical system and its cycles working mises that have to be made in developing policy and to do so with maximum biological diversity. optionsandmakingdecisions.Theremaybesome These are the constraints that human beingsmust "win-win" policies that allow all of these objec- respect. Only then is it possible to deal with the tivestobeaddressed simultaneously,eventhough constellation of social and human concerns. theconceptualapproachcannotintegratethethree This approach is admittedly somewvhat purist objectives explicity. For example, there are more and reductionist in that the biogeophysical foun- than a billion people without access to safe drink- dations of sustainability are to be studied first ingwaterandsanitation.Fromaneconomicpoint and largely independent of the economic, social, of view, providing safe drinking water is one of cultural, and political aspects. The argument is the most cost-effective projects that can be under- often made that, because the term biosphere tra- taken because incidence of disease is reduced ditionally includes only animals, plants, bacteria, and, thus, productivity increased along with eco- and their life-support system, to the exclusion of nomic benefits. From an ecological and environ- human beings, environmental problems have mental point of view, it is also desirable because developed because some human societies con- water quality is improved, and waste material ceive of people as removed from the natural unity that pollutes water courses is reduced. From a of life and placed in opposition to the rest of the social point of view, such a project primarily biosphere. Thus, a cogent criticism of this ap- assists the poor, thereby reducing social tensions proach is that it could further add to the environ- and improving social sustainability. men tal problems caused by many people's feel- The problem is that beyond such simple indica- ing of detaclhment from, even superiority to, the tors there are necessary trade-offs between objec- environment and biosphere. tives-which is where the sustainable development Insummary,the finalsynthesisemergingfrom triangle becomes important. In addition, the tri- the conference places the biosphere first as the angle directs attention to the capacity to adapt and necessary basis for human life and activity. More- take advantage of new kinds of opportunities such over, there appears to be justification for this as come with the inevitable change in environmen- approach because of: (a) the extreme complexity tal conditions. It is because of such changes that of sustainability; (b) the fact that human life and analytical tools need to be developed that try to all human activitiesrelyonabiogeophysical foun- reconcile the conceptual tensions. The task of the dation that existed prior to (and therefore can be scientific community is to develop analytical tools viewed as independent of) human existence but and indicators and to offer policy options to the that can be perturbed and even radically altered decisionmakers. Thisinvolves takingcomplexcon- by human activity; and (c) the serious commit- cepts and models, boiling them down to the sim- ment to include in the approach those occasions plest idea possible to illustrate the extent of current where human (social and cultural) dimensions understanding of the problem, and providing a set do necessarily enter into considerations of of policy options to the decisionmakers. biogeophysical sustainability. Thisisparticularly However, the conference participants were also so in the case of defining the temporal and spatial reminded that individual human beings are a sig- boundaries for theecosystems tobe sustained.There- nificant and critically important factor in sustaining fore, this approach requires that care be taken to (i) their environment. Their involvement begins in consider the human dimensions where appropriate theirhomes (which theywould not pollute), spreads and (ii) accept a commitment to undertake the totheirneighborhoods,andextendstotheirregion. subseqtuent task of integrating the biogeophysical In eventually developing a sense of identity with foundations with theeconomic, social, cultural, and their region, individuals can put pressure on politi- political aspects that will lead to the development of cians. Thus, communicating directly with citizens a complete picture of sustainability and ultimately can often be as important, if not more so, than of sustainable development. communicating with politicians. xxi Defining and Measuring Sustainability: The Biogeophysical Foundations Concepts of sustainability Definition of sustainability A convenient, if not comprehensive, taxonomy Important considerations that were identified for of views regarding sustainability evolved in developing a definition of sustainability are: the course of discussions. First, the "input-out- the number of variables it contains put" view assumes that the internal dynamics of the ecosystem are more or less in a steady e ease in measuring these variables state-that is, they are not degrading over time. * its capacity for generalization Beyond that, the primary focus is on inputs and * its applicability to different situations, and outputs and whether they are sustainable. .the flebit it allows. Second, the related "state" definition requires exiiiy simply that a sustainable ecosystem be one in A definition should not attempt to freeze the which a state can be maintained indefinitely. currentstatebutinstead should define thebound- The "capital" or "stock" view requires the main- aries within which there can be flexibility-be- tenance of natural capital or stock at or above cause allowance must be made for the evolution current levels and, thus, that the products of the of some components of the system. Such flexibil- ecosystem be used at a rate within that ity could be permitted within an input-output ecosystem's capacity for renewal. Sustainability scheme, which would not require explicit adher- is thereby ensured by living off the income encetoaparticularculturalorsociologicalframe- rather than the ecological capital. The "ecosys- work but would allow any scheme to operate tem-characteristics-not-degrading-through- withincertainbiogeophysicalbounds.Thus,defi- time" view requires the perpetuation of the nitions of sustainability might differ in terms of character and natural processes of the ecosys- how they specify borders, envelopes, or states. tem and indefinite maintenance of its integrity The capacity for generalization refers to (productivity, diversity, stability, and adapt- transsystem or translandscape measurements, ability) withoutdegrading the integrity of other whereas flexibility refers to conditions within ecosystems. systems. Flexibility can exist within the system in Finally, there is the "potential throughput" terms of cultural and social structures as long as view, emphasizing the use of resources within the biogeophysical requirements are met. Space the capacity of those resources to renew them- is another important variable that ranges from selves. According to this view, sustainability is plots through landscapes, ecosystems, regions, defined on the basis of maintenance of poten- nations, and continents to the entire biosphere. tial, so that ecosystems can provide the same The conference participants strove for a defini- quantity and quality of goods and services as in tion of biogeophysical sustainability that would be the past. Potential is emphasized rather than broad and general, all encompassing, not particu- stocks, biomass, or energy levels. To maintain larlycontroversial,andnotinherentlyvalue-laden: this potential, which amounts to future op- "Biogeophysical sustainability is the mainte- tions, there are two areas of concern: the degra- nance and/or improvement of the integrity of dation of the physical productive capabilities the life-support system on Earth. Sustaining of the land and water and the loss of genetic the biosphere with adequate provisions for diversity. This might mean sacrificing 90 per- maximizingfutureoptionsincludesproviding cent of the stocks of a species while maintaining for human economic and social improvement a viable population of that species so that in the for current and future human generations future a society could rebuild the habitat for withinaframeworkofculturaldiversitywhile: that species. Thus, maintaining stocks or en- (a) making adequate provisions for the main- ergy levels is not as important as retaining the tenance of biological diversity and (b) main- productive abilities of the land, specifically of taining thebiogeochemicalintegrityof thebio- the soil and the biotic components. Obviously, sphere by conservation and proper use of its waste production can be tolerated only within air, water and land resources. Achieving these the capacity of the system to assimilate those goals requires planning and action at local, wastes. This basis for sustainability permits regionalandglobalscalesandspecifyingshort- many alternative sustainable systems that in- and long-term objectives that allow for the volve various mixes of goods and services. transition to sustainability." xxii An Introduction to the Definition and Measurement of Biogeophysical Sustainability This definition contains an element of vague- local scale allows because of the complex feed- ness because it refers to integrity, a term not backs that occur on regional and global levels. further defined or explained. It also recognizes This is an example of the differences that exist that some options might be closed by attempts to between ecological systems and economic sys- achieve sustainability that allow for an improve- tems, for which there is rarely a good one-to-one ment in human well-being. Furthermore, the par- match. ticipantswereunabletoresolvethephilosophical Furthermore, an important difference in per- issue of whether sustainability can be defined, in ception was found in the developing world and in its simplest and purest form, independent of sus- the industrialized world in defining ecological taining human welfare. capital. Ecological capital in the developing world In considering time scales, it was felt impor- is largely center2d on how to provide enough tant to include in the definition a consideration of food, fodder, and fuelwood for the sustenance of a transition phase during which there might be the community. Thus, there is an urgent need to some unsustainable practices, eventually leading expand economic activity, and the kind of criteria to a state that could be sustainable indefinitely. that would be required to measure the The general feeling was that somewhere on the sustainability of these activities would be very order of 50 to 100 years was a reasonable length different. In contrast, in the industrialized world for the transition phase but that this would be the need is to scale back some economic activities dependent on the resources and social and cul- in order to decrease their impact on the environ- tural considerations of each society. This transi- ment; therefore, the definition of ecological capi- tion period would be characterized by economic tal and the measurements of it that might be used growth in the developing countries, stabilization globally are likely to differ. Developing a unified of the world population, and the use of available definition of ecological capital requires the recon- nonrenewable resources to capitalize the estab- ciliation of these differences. lishment of a state that could be sustained. One approach to defining ecological capital is However, the choices made and activities un- to try to link the concept to risk aversion and, dertaken during the transition period would con- particularly, to culturally-specific definitions of strain the kind of state that can be maintained ecological capital. Various components contrib- indefinitely thereafter. Obviously, while draw- ute to ecological capital-air, soil, forests, and ing down on renewable and nonrenewable stocks biological diversity. Two important characteris- during the transition period in order to build an tics are how renewable they are and, if degraded, infrastructure that allows for a sustainable state how long it would take to recover them. From this at the end, some guidance would be needed to perspective, the most important element is bio- indicate whether the goals were actually being logical diversity because it probably requires the achieved during this period. It is important that longest recovery timeonce lost. From thepoint of during the transition, the capacity to renew is not view of risk, there is another way in which the lost.Speciesthatarelostorsoilsthataredegraded elements can be ordered. Take, for example, air, to the point where forests will not regrow, deter- the stratospheric ozone shield, soil, and biologi- mine the location of the relevant boundaries. cal diversity as four of the components of ecologi- cal capital. Air has the shortest-term impact be- cause the effects from its pollution will be noticed Ecological capital quickly. Stratospheric ozone depletion hasa some- what longer-term impact-that is, cancer rates The concept of ecological capital was examined will rise and food productivity will decline over a as a basis for better defining sustainability. How- longer time period than people can survive with- ever, ecological capital itself proved difficult to out good air to breathe. The consequences of soil define. It was observed that in any attempt at a degradation will appear over an even longer pe- definition, it is insufficient to simply focus on the riod of time because soil deteriorates slowly. Fi- local scale, that is on stocks of trees or fishes, for nally, biological diversity will take the longest to example. Consideration of ecological capital re- show ill effects-for example, in terms of undis- quires a look at much broader scales and an coveredchemicalsandpharmaceuticalproducts. understanding of ecosystems and the interaction Therefore, in one system of ranking, biological among ecosystems in much broader terms than a diversity is the most important elementbecause it xxiii Defining and Measuring Sustainability: The Biogeophysical Foundations takes the longest to recover (perhaps never), physical processes, all at different time scales, whereas in the other system it is the least impor- cannot be reduced to the simple metaphors of tant in terms of short-term impact. Risk modeling stocks, flows, and interest rates. Such metaphors (used in studies of economics and behavioral cannot capture the differences and the very dra- ecology) and analysis of risk aversion explore the matic changes that can occur, depending on the way people respond to risk and can help deter- level at which activities are directed. Thus, one mine whether they are risk prone or averse. Cul- proposal at the conference rejected the notion of tural and social factors may strongly influence stocks, flows, and ecological capital as a useful the outcome. paradigm for approaching the biogeophysical The foregoing examples complicate attempts foundations of sustainability. to aggregate diverse ecological components into In a similar vein, concern was expressed about a single index. Thus, the risk-averse strategy of the metaphor of ecosystem health. Although sci- somecultures would dictate that most concern be entists are expected to develop an index, like given to the shorter-term results: air quality, wa- temperature,which tells a lotaboutecosystems, it ter quality, and the stratospheric ozone shield, was pointed out that the analogy with health is whereas the longer-term impacts associated with not entirely correct. There is a major difference loss of soil and biological diversity would not between a tightly coupled system like the human receive as much weight in the aggregated index. body and an ecological system much more loosely Conversely, if the concern of the culture is about coupled. Such differences must be reflected in the the long-term consequences and the time it takes health construct used for ecological systems. No for recovery, the components will be weighted doubt the metaphor is inevitable and inescap- differently. One of the major difficulties in trying able, and although scientists prefer to use the to develop an aggregated index that covers all term integrity, the public will continue to refer to these components arises from the difference in ecosystems as healthy or unhealthy. Therefore, these culture-specific approaches. scientists must make a special effort to educate An important problem noted by conference the public about the real significance of the anal- participants was how to go about determining an ogy between human and ecological health. acceptable level of environmental damage, which in turn impliesa carrying capacity. The concept of maximum sustainable use invites the notion of Indicators maximum survivable abuse. One way the accept- abilityofdamagecouldbedeterminedisbyiden- The management of the global life-support sys- tifying the pointat whichan increase in theactiv- tem has been compared to piloting an airplane ity in question produces greater marginal cost without instruments. Basically, humanity does thanmarginalbenefits.Thisrathersatisfyingtheo- not have a complete set of indicators (instru- retical definition is, unfortunately, difficult to ments) or monitors of the global life-support sys- implement in practice because of measurement tem, and this situation will prevail for some time. problems. This would not be the case if all of the This shortcoming constitutes one of the major damages and all of the benefits could be quanti- challenges for those trying to develop indicators fied in commensurable units, say in monetary of biogeophysical sustainability. Another prob- units. Inpracticethisbecomesapoliticaldecision lem that was noted at the conference involves having to do with people's values and prefer- scale,whichcomplicatesenormouslyanyattempt ences about the risks and uncertainty of future to develop sustainability indicators. Thus, as a costs versus the shorter-term benefits they are practical matter, indicators are needed that are offered from the expanded activity in question. It usefuloverdifferenttimeandspacescalesandfor may be argued that society will generally prove differentlevels,fromthecommunitytothebiome. to be risk averse and will want to leave itself some This requires a test set of indicators. margin of error. In looking for an organizing principle for mea- However, concern was expressed that in con- suringbiogeophysical sustainability, threelevels sidering the concepts of stocks, flows, and eco- of activity can be identified. The first level in- logical capital, there is a danger that such simple volves determining measurernents of analogies do not adequately reflect the real com- sustainability. It appears unlikely that a set of plexity of ecosystems. It would appear that eco- universal measurements tocoverfreshwaterlakes, logical processes, species interactions, and geo- marine fisheries, grasslands, and mountain tops xxiv An Introduction to the Definition and Measurement of Biogeophysical Sustainability can be found. By defining the basis for measuring From an operational point of view, this list and determining biogeophysical sustainability at represents a checklist or framework. For each the ecosystem or landscape level, the relevant item listed, specific variables and an associated measurements can then be aggregated into indi- range of acceptable values must be identified, as cators at the second level. Therefore, it is impor- well as a way of measuring each by an inexpen- tant to assemble a generic list of indicators of sive technique. Moreover, measurements or val- biogeophysical sustainability to serve as a check- ues are needed for the different ecosystems. Be- list for determining the correct ground for each of cause the indicators relevant to biogeophysical the ecosystems. Specific measurements must then sustainabilitywillbedifferentforeachecosystem be identified for each indicator. However, it is im- type, it was decided to assemble the results into a portant to identify the status of this generic list of matrix with the relevant indicators given for each indicators at the regional and global levels,and this parameter category. Such a matrix, from which requires a composite index, which is the third level. specific measurements could be derived, would In summary, the three levels must be devel- have the added advantage of providing a frame- oped separately, but they are intimately work with which to identify any important over- interlinked . The first two levels, which arestraight- looked components for specific managed ecosys- forward from a biogeophysical point of view, tems. Using the Delphi technique, with contribu- involve selecting indicators and trends that are tions solicited from participants over a period applicable to specific systems at different spatial that extended well beyond the actual conference, and temporal scalesindifferentpartsof the world. a unified list of candidate indicators of The third level, involving the development of an biogeophysical sustainability evolved for the fol- index (rather like GNP) for the biogeophysical lowing eight managed and natural ecosystems: health of the globe, requires theassembling of the indicators and their trends inlto some type of * agriculture index that takes into account how those trends * forests actually make an impact on human welfare. The * rangelands ultimate goal might be to assemble all the pieces * wildlife and wild lands into an even broader index that will reflect all the * freshwater fisheries elements (including economic, social, and politi- * wetlands and ground water cal) of sustainability that determine long-term * coastal resources, and human welfare. * marine fisheries. To avoid the slow process of identifying and reachingagreementon therelevantmeasurements In addition, the following human dimensions of biogeophysical sustainability for the first level were considered to have an important impact on and later aggregating them into indicators for the biogeophysical sustainability: second level, an attempt was made at the confer- * human influences (such as land-use ence to "leap frog" by starting with the develop- change or subsidies to landscapes-for ment of the second level, based on the practical example, fertilizer) requirements of policy analysts and * human demography, and decisionmakers. Thus, the participants prepared * human well-being. a selection of parameters from which candidate indicators could be developed to monitor and It was considered important to complement predict biogeophysical sustainability (see table). these results with a list of issues that should be The parameters are as follows: considered when developinga set of indicators of biogeophysical sustainability. This list covers: o landscape composition and patterns * production of goods and services * scale (local, national, regional, global) * biological diversity * purpose and use of the indicators (policy * water quality and quantity orbroad resource allocation, specific man * soil properties agement), and • energy and nutrient flows * values (health, ecology, rules-of-thumb * atmospheric composition, and are to be used to specify the boundaries * climate. of sustainability. xxv Defining and Measuring Sustainability: The Biogeophysical Foundations Each selected parameter covers a number of such as fertilizers, that are being provided to the issues that may not be immediately obvious in managed ecosystems and how the managed sys- name or in ways that link it to other parameters. tems can be designed to replicate the internal Therefore, some discussion appears to be war- controls that govern natural ecosystems. Thus, ranted concerning the intended coverage of each comparative measurements of soil fertility status parameter and the types of measurements that (both in physical and chemical terms) between might be considered. managed ecosystems and natural ecosystems Landscape composition and pattern is intended to would be appropriate indicators of the include topography, geology and substrate con- sustainability of the system. ditions in combination with soil properties. It Energy and nutrient flows should include the requires periodic measurement of the various flow of materials, including both toxic and haz- types of ecosystems in a landscape and the kinds ardous materials. This is one way to handle toxic of changes taking place. Such measurements give materials and hazardous wastes, although they an indication of whether the system is sustainable could also appear throughout the list under such andinwhatwaythesystemisundergoingchanges parameters as water quality and quantity, atmo- in terms of species composition, invasion of spe- spheric composition, and soil properties. By subsum- cies, and degradation. ing pollutants and toxic materials in this way, One indicator included under production of there is no need to add a specific parameter for goods and services is land productivity, which is an pollution. Furthermore, energy and nutrient flows indicator of the biogeophysical health. Produc- could include an energy input/output ratio that tivity decline signals that something is wrong in would indirectly reflect CO2 emissions. This pa- the system. However, if the aim is to maintain rameter should also cover the kinds of subsi- productivity, it may be too late to wait for produc- dies that go into managed ecosystems in terms of tivity to drop. Thus, predictive indicators are economics as well as energy. essential for measuring productivity. Further- The climate parameter should reflect the agricul- more, it should be borne in mind that measuring ture-climate interaction at the regional level. Agri- net primary productivity for forest, marine, and culture, especially through its influence on the hy- someotherecosystemsturnsouttobeachalleng- drological cycle through land-use change, has an ing task that is only rarely accomplished. When influence on regional climate. If that interaction considering a mosaic landscapecontaining many leads to a decrease in hydrological recycling-as it ecosystems, urban areas would likelybe included. easilycan for land uses that go from rainforest at one Of course, although they have negative values for extreme todegraded agricultural land at theother- productivity, theyare important regions that cover theassociated decrease in rainfall or water retention a great part of the world and thus should not be capability can change the temperature cycles. This overlooked. can lead to a costly loss of resources. Biological diversity is an important parameter. Though not definitive, and despite needing It is sometimes possible to identify in a natural some refinement and precision, the proposed ecosystem a few key species that likely reflect the indicators constitute a significant step forward. system's overall biological diversity. Much bio- And although thissetdoes not introduceanynew logicaldiversityisavailableinmanagedsystems, indicators or measurements, its novelty rests in such asagriculture, in traditional societies. There- theselection process. Theconference participants fore, the biodiversity indicator should reflect the were confidenit that this set of indicators is valid biological diversity that is available in managed forscales fromonehectare to theentireglobe.They systems as well. were confident that it could be based on measure- If the landscape is part of a watershed, an mentsthatusetime-testedtechniquesandwhose important indicator of water quality would in- significance is well understood. volve the measurement of suspended particles The set of indicators proposed for temperate and chemicals that are lost to the stream or river. rangeland (see Risser, this volume) was estab- Among the important indicators of soil proper- lished as the model for other ecosystems. This ties are soil fertility changes-physical as well as set consists of five meaningful indicators that chemical. Comparative measurements can be specify the health of the rangeland. The indica- made from time to time between managed eco- tors for the different parameters (which are systems and natural ecosystems. This would add italicized, with the appropriate thresholds in to better understanding of the kinds of subsidies, parenthesis) are: xxvi An Introduction to the Definition and Measurement of Biogeophysical Sustainability The first example concerns the biogeophysical • landscape composiftion: range condition basis of sustaining tropical water resources using rating (good to excellent) the model of the Amazon region, which is repre- • aboveground primary production: peak sentative of what an intact hydrological cycle standing crop (>300 g/m2) ought to be. The goal is to monitor the integrity of a plant species diversity: eH' (>5.0) the hydrological cycle for the relevant time and * soil properties: soil organic carbon content space scales. Because of the need to consider in top 20 cm of soil (> 3-5 kg/M2), and biological diversity, and due to the extremely * nutrientflows: nitrogen content of vege- tight coupling between hydrology and ecosys- tation (0.6% on a dry weight basis). tems, the central issue of thebiogeophysical basis for maintaining integrity should be considered in This set of indicators can be used to identify the some detail in the development of a useful indica- biogeophysical sustainability of temperate range- tor. Thus, from the pointof view of the water cycle land ecosystems. A sixth indicator would reflect in the Amazon and i ts interaction with the vegeta- the scale of thelossof habitat. For this,a surrogate tion, the efficiency of the water cycle can be ex- i ndicator might be a change in the number of bird pressed as the river discharge per unit basin area species, which are very sensitive to changes in the divided by the precipitation ratio. Both the pre- extent and juxtaposition of rangeland areas. Of cipitation and the run-off or discharge have the course, rangeland ecosystems have much more advantage of being measurable. On the basis of temporal variability than other habitats, and thus the accepted view that in the Amazon basin temporal scale dynamics are important for un- roughly 50 percent of the precipitation is derived derstanding the variation and the applicability of from recycled water, the precipitation ratio be- different indicators. In particular, the measure- comes 0.5. An increase in this ratio would indicate ment of the size of the soil organic carbon pool has that more of the precipitation is going to run-off some universality as an integrative indicator of and that thereislessrecycling, givingstraightfor- the general health of rangelands. ward implications for the energy and water bal- This w;1s viewed as an excellen t example of the ances and for the vegetation. In this tropical forest type of sets of indicators that should be devel- environment the ecological niches are very fine. oped for all other ecosystems. Based on the kinds Two or three Celsius degrees of change can mean of conditionstobemaintained,itisexpected that the difference between life and death for that there is enough experience and expertise in the system, as opposed to a fifty-degree change (for ecological community to assembleasetof indica- example, seasonal variations) in a temperate or tors such as this for all ecosystems. Nevertheless, boreal coniferous forest. But other factors can some participants held the view that although change thedischarge rate per unitbasin area, and good indicators of biogeophysical sustainability to distinguish among them requires additional can be proposed, developing practical measure- parameters-such as those based on the stable ments for them can be quite a difficult task. isotopic composition of the water involved. Tak- It was noted that World Resources Report al- ingthe180concentrationoftherainwaternormal- ready publishes statistics for a number of impor- ized for area and time, changes in the recycling tant global indicators, six or eight of which ap- rate could be easily monitored. pear to be good trend spotters. These include This is an example of a physical indicator that atmospheric CO2 concentration, stratospheric contains a great deal of information about the ozone concentration, temporal changes in soil ecosystem,anditdemonstratestheneedformak- fertility, biological diversity loss, and changes in ing a trade-off between comprehensiveness and natural habitat. scrutability. It is illustrative of the type of indica- Integrative indicators are important because tor needed across a range of ecosystems and they provide greater information and are often conditions that incorporates as much of the im- moresensitiveto theiinteraction amongvariables portant information as possible into one or two and to critical thresholds than are single-variable numbers readily displayed and understood. indicators. Such integrative indicators are them- There may be instances where a particular set selves good candidates for componentsof an index of indicators is intelligible to the lay public and of biogeophysical sustainability. Important cx- also has scienti fic utility. One such example is the amples of individual indicators with significant use of the lake trout as an multivariable indicator integrative characteristics were also identified. of the state of the Lake Superior ecosystem in xxvii Defining and Measuring Sustainability: The Biogeophysical Foundations North America. The population of lake trout has Another concern is the relevant spatial scale on been proposed as an indicator that captures much which the measurements are to be made for each informationabout the state of integrity of the lake of the indicators. The issue of averagingover time basin. A particular population level has been is relevant. All natural and managed ecosystems identified that indicates a high state of integrity of are characterized by fluctuations on various time the lake ecosystem. Thus, this indicator not only scales. It is important to ascertain the magnitude integrates a variety of components but also is a of fluctuations that is tolerable-that is, before piecemeal indicatorforthedifferenttypesof stress they start threatening the sustainability of the on the ecosystem. For example, it can be inferred ecosystem. Of course, an imposed constraint that from the trout population dynamics whether the acceptsnochangesisnotlikelytobeacceptableto lakes are being overfished; from the number of most societies. In manycases, any human activity scars on the skin of adult fish whether the sea of a particular kind will produce a change. Thus, lamprey is being controlled; from the concentra- the argument is not whether there is going to be tion of contaminants in the flesh whether the a change, but how big a change is tolerable. contaminant loadings are being controlled; and A number of other useful contributions to the from the relative health of different stocks of lake development of indicators were also made, some trout whether their spawning grounds are free of of which are summarized here. silt. Thus, the well-being of the lake trout popula- tion is itself an integrative indicator of the ecosys- * Off-site effects and subsidies. Under any given tem. The point is not that this alone is a sufficient definition of the geographical scale of interest, indicatorofthesustainabilityoftheLakeSuperior it is possible to imagine a situation where ecosystem but that it includes both the qualities within a boundary everything is fine wh,ile sought in an indicator-that is, scientific validity adverse external impacts occur outside the and communicative facility to the general public. boundary. One of the primary differences be- Anumberofotherinterestingindicatorswith tween the situations faced by the Sahelian integrativecapabilitieswerealsodiscussedatthe pastoralist with impoverished soil and the conference, but their significance would require Saskatchewan farmer on the rich plains of more study before they could be included in the Canada is the inputs that the industrial world list of indicators. They include: (a) ecosystem relies on to mask the degradation of soils. integrity; (b) production efficiency, which relates Thus, a useful indicator of sustainability is the to energy input versus output as well as soil and size of the inputs used to mask the loss of biologicalsinksandreservoirs;and(c)acompari- resilience of these systems. A problem with son of human versus natural flows of energy and output measurements is that it is often possible materials. This last would be an attempt to deal to maintain high yields for a long time, al- with the scale of human activity in terms of rates though nutrients and energy have to be im- of mobilization of energy, soil, rock, and such ported from some other ecosystem. Therefore, elements as mercury, cadmium, sulfur, nitrogen, space and time boundaries of the system to be and phosphorous. sustained must be defined at the outset. Then for a predefined ecological management unit, primary production, biological diversity, and Practical considerations for applying natural rates of recycling become important indicators indicators within that unit. * Appropriate degree of aggregation in indicators. Mention should be made of some of the concerns For example, with regard to atmospheric com- raised about the results, particularly as regards position, it must be decided whether it is the indicators. First, it can be expected that indicators individual atmospheric components or the net and measurements will not be exactly the same, effect or the greenhousc potential of all of them nor will they measure exactly the same thing for together that is important when looking at different natural or managed ecosystem, includ- greenhouse gases. ing urban ecosystems. Also, much effort is still * Distribution of values. When looking for mea- required to select indicators that are the most sured values, it is sometimes not the mean operational but of sufficient generality to ensure value over a region or an area that should be global validity. sought, but rather the cumulative distribution xxviii An Introduction to the Definition and Measurement of Biogeophysical Sustainability function of values, including the tails of the indicator in the matrix. It will be necessary to distribution. identify-for each indicator and in the context of * Butterflyeffect. If there can be largeeffects from each region or ecosystem-the following: small causes (analogous to the butterfly effect * the appropriate spatial scale for the averaging, of chaos theory), one of the implications is that integration, or aggregation of the indicator there are some uses of certain ecosystems that * the appropriate averaging time simply must be foregone. Thus, requiring * the level of fluctuations consistent with cur- biogeophysical sustainability implies forbid- rent notions of sustainability ding some kinds of uses in some places be- * the magnitude of secular trends in the vari- cause large adverse effects may follow. ables that are consistent with these notions of o Surrogates. Surrogates can be used to monitor sustainability, and diversity. A great deal of the diversity of forest tainabieity, and *the needed or a Dro nlate dezree of a re ra- systems consists of invertebrates, microbial tion of the particular indicator; for example orgniss,and fungal organisms that are not patcua iniatr foxape organisms, and fungal organisms that are not.regarding energy flows, whether net primary easily monitored. However, such conditions productivity (NPP) is sufficient or whether as structure and stage of development of a informationaboutenergy flowsateverytrophic forest may indicate that at least the habitat . . conditions exist for those kinds of organisms. level IS required). Once this step has been achieved, there are two * Urban areas. Urban environments, with their . . . '. rapid growth, the consumptive lifestyles of to apply the selected indicators. Although the their inhabitants, and dependence on external health of the biosphere and of individual ecosys- support systems, might be seen at first as offer- health ofthetioshe aindivida ecosys- ' . . ~~~~~~tems is Important, the sustainabili ty concept sug- ing one of the better ways of preserving bio- gests the need to introduce an accounting prin- logical diversity-typically by concentrating ciple. The idea is to produce a system of measure- the population in urban areas and taking them ments that would be general enough to permit out of direct conflict with the natural ecosys- discussionofindicatorsofsustainabilitywithina tems elsewhere. However, further reflection framework of global environmental accounting. shows that people who live in cities actually The global accounting needs become obvious consume much more than people in the coun- because of the complications created by too many tryside. Furthermore, the evidence of past civi- indicators for different regions.Using a business lizations shows that urban consumption pat- analogy, the balance sheet at the end of the year terns are often not sustainable and ultimately shows whether the operation was sustainable begin to exceed the carrying capacity of the (profitable)orunsustainable(unprofitable). This surrounding countryside (for the prevailing requires the ability to convert all the assets of the technology).Thus,movingpeopletocitiesmay ecosystem (business) to a common unit so that be just a short-term expedient unless there is a accounting principles can be applied. It was sug- transition to long-term sustainable adaptation. gested that the most important accounting prin- ciple would be production-generally the har- vest, which is readily convertible to monetary Next steps terms, since there are markets. Two other aspects of importancearebiological diversity and ecosys- The conference demonstrated that although there tem processes and functions. is a focus on specific ecosystem types or specific This set of indicators provides an opportunity locations, it is not possible to reach a general for integrating several of them into a general consensus on what should be measured. Thus, it index that would cover broader scales. Therefore, is necessary to get down to specifics for each another approach would involve developing given ecosystem, for which more specific indica- within each of the parameters a very specific, and tors and measurements must be produced based probably region-specific, list of measurements to on the above considerations. see if these could have a universal value. If so, Therefore, the next logical step is to hand over further research would show the way to assemble to specialist groups the task of deciding which theseindividualindicatorsintooneormoreindi- specific measurements should be used for each ces aggregated across temporal and spatial scales. xxix Defining and Measuring Sustainability: The Biogeophysical Foundations The basic task is to determine what to do with the ments as well as the scientific goals and products infornmationobtained from specific indicatorsand needed to support those commitments. Further how the indicators can be used constructivelym scientific requirements can also be derived from to emphasize the concerns of sustainability. In the goals and policies expressed therein. doing this there must be some trade-off between However, it was suggested that fordeveloping comprehensiveness, on the one hand, and sim- indices of sustainability, it is necessary to go plicity, scrutability, and usability on the other. beyond the above-mentioned documents and Obviously thenumberofindicatorscanbeampli- consider the manner in which they would be fied ad infinitum, thus inviting the danger of used.Thepolicy challenge is frequentlyrelated to making an index complicated. It is harder to information that policymakers feel they need to assess the relative importance of different indica- make appropriate decisions. Discussions about tors-in particular, the ones that capture most of declining biological diversity or a change in hy- thepropertiesconsideredimportantwithoutadd- drological pattern in the Amazon basin do not ing so much complexity to the indicator that it usuallydrawtheattentionofpolicymakers.How- approaches the degree of complexity of the real ever, polic makers do want to know the conse- world. quence of such changes and how they effect the In addition to reducing the number of indica- sustainability of natural resource systems that tors by careful selection, another technique that support people and life. Providing such informa- cain simplify the processof formulating an indica- tion requires an ambitious agenda that will prob- tor, and thus an index, is to simply establish a ably test the current state of knowledge. norm and divide the actual measurement by that norm. This produces a dimensionless indicator or index that becomes a common measure be- Recommendations cause it is always a comparison. The measure- ment of a base year or a healthy ecosystem could The conference participants made the following be selected as the norm, and departures from it ten general recommendations: could be monitored. It was pointed out at the conference that the 1. More work is needed to refine the definitions indicators fall into three broad groups: chemical, and propose the indicators for biogeophysical biological and physical. The chemical indicators sustainability. The emphasis must be on start- are among the parameters for water quality and ing at the smaller scales because it is not pos- atmospheric composition, although such indica- sible to synthesize results that have not been tors would fall with the physical indicators. The obtained at the scale of the individual unit. biological indicators involve thebiological diver- 2. For the purposes of communication, a very sity parametersatdifferentlevels. Finally, there is simple index is urgently needed-one that a set of physical indicators in terms of energy would permit and facilitate communication balance and land-use change as well as the quan- between biologists, sociologists, and econo- titative aspects of some of the chemical and bio- mists. It is anticipated that the formulation of logical indicators. This method of organization such an index will require an interdisciplinary mighit be of use in the development of an aggre- effort. Such an index, or elements of such an gate index of sustainability. index, might includeatmospheric C 2concen- tration, atmospheric methane concentration, atmospheric oxygen concentration, net primary Toward policy considerations productivity,or biological diversity. Such an index should be useful overmany scales, popu- The conference participants felt that the most up- lations, and different types of ecosystem. to-date and perhaps the best general statement of 3. As thedistinction between temperateand tropi- what policymakersperceiveas their requirements cal agriculture is more one of measurement were in the documents that came out of the Rio and knowledge about the ecosystem than of Conference, namely Agenda 21 and the Rio Dec- actual biogeophysical differences, many of the laration. These provide an organizing principle measurements made in temperate agriculture for addressing sustainability because ihey spell systems could be useful in tropical agricultural out the existing policy commitments and agree- systems. This has not happened yet because xxx An Introduction to the Definition and Measurement of Biogeophysical Sustainability agricultural systems are much more diverse in 7. Integrating information is difficult and chal- the tropical areas as a result of a much greater lenging aspect of many disciplines of science. number of crops,cropping sequences, and cul- Even remote sensing is limited to a definite tural diversity. Thus, obtaining relevant indi- scale because of its limited resolution. Thus, cators becomes much more difficult, and it is landscape imagery provides a scale-depen- necessary to concentrate research and data dent pattern for different composites or mosa- gathering activities on obtaining the informa- ics. Research needs to be conducted on the best tion that will be needed for indicators of use of information, its interpretation, and its sustainability. aggregation to the larger scales. Such larger 4. For natural ecosystems more knowledge is scales have smoothed and reduced variances. needed about thedriving forcesthat keep those Although this is sometimes real, it often is an ecosystems in equilibrium, the natural pro- artifact of the aggregation process. Since aver- cesses involved, and the location of critical aging throws out a lot of information, there is thresholds. These last are important because if a need to develop decision-based rules for they are exceeded they can lead to discontinu- aggregating information so as to maintain the ity-that is, a rapid transition to new states information base present at small scales after with very differentconditions. To predict this, extrapolation to larger scales. In some cases, indicators are needed that can monitor the wherelinearaveragingisnotadequate,fractals proximity of the threshold. Priority should can be used to maintain information as scale is probably be given to those phenomena that increased. This would be a valuable tool in lead to major and dramatic shifts. Thus, it is developing the definitive set of indicators of important to be aware of the factors that reflect biogeophysical sustainability. the mai ntenance of the curren t state. It would 8. There appears to be a need to assess the state of also be valuable to know how change to a the science of sustainability to determine, for better state might be achieved. For managed example, how well sustainability can be pre- ecosystems there should be a similar list of dicted, measured, and understood. There are critical thresholds and critical capacities. It perhaps a dozen myths about sustainability would be helpful to identify the driving forces that can be, and have been, invoked. The time that have led to collapses or major clhanges in appears right for an independent review. the past. Thus, a series of case studies dealing 9. There must be a reorientation and refocus of with these shifts and critical thresholds should research in the ecological sciences. With lim- be conducted. ited human and financial resources and most 5. Ecosystems are high-order nonlinear systems. of the funds not being used on research that Much valuable information might be derived directlyaddresseslocalsustainability,themost from a study of ccosystems, using the tech- importantrecommendationisthattheresearch niques from physics, for example. Finding base and research funding be reoriented to equivalencies between kInown physical sys- focus on important questions about tems and ecological systems could be a practi- sustainability. cal prospect. Physicists have learned that it is 10. To better deal with all of this, a new discipline not always possible to predict the new state to dubbed "econology" is proposed. Econology which a nonlinear system will shift after a goes beyond combining the two older disci- transition. But they have managed to begin plines of ecology and economics and requires predicting whether there will be a state transi- research that brings more externalities into tion and when it will occur, on the basis of this approach. Such research is also needed to signs that manifest themselves as the state- overcome the automatic constraint inherent in transitioni threshold is approached. This is an dealing with steady-state models. important area for exploration in ecology. In addition, the participants proposed the fol- 6. The significance of large system change is still low-up activity to create high-level global indices uncertain, as when all indicators start chang- of biogeophysical sustainabilitv. The process be- ing in the "wrong direction." This is a topic on gins by developing specific sets of indicators for which much more research in needed. a series of ecosystems of different types and in xxxi Defining and Measuring Sustainability: The Biogeophysical Foundations different parts of the globe. This would involve References convening a series of working groups, with each one focusing on a particular ecosystem to de- Munasinghe, Mohan. 1993. Environmental Eco- velop theindicatorsand associated measurements nomics and Sustainable Development, Washing- of biogeophysical sustainability. Undertaking this ton, D.C.: The World Bank. task for a representative series of ecosystems, Vitousek, Peter, Paul Ehrlich, Anne Ehrlich, and natural and managed, that can thus provide the R.A. Matson. 1986. "Human Appropriationsof foundation for work on one or several indices of the Productsof Photosynthesis. " BioScience 36: biogeophysical sustainability is required for get- 368-73. ting down to more detailed scales. This should be followed by a meeting of representatives from each group to work out the commonalities among the sets of indicators and the formulation of one or several indices of sustainability. xxxii An Introduction to the Definition and Measurement of Biogeophysical Sustainability * m~~~~~~~~~~~~~~~I I _ _ _ _ ~~~~~ W zI I I ~~~Iii I I11 I~~~~Ii I 'Ipt A 11 KITI ii~~~~~~~~Qxxii PJart A Background Papers The Meaning of Sustainability: Biogeophysical Aspects John P. Holdren, Gretchen C. Daily, and Paul R. Ehrlich This paper benefited greatlyfrom interactions with R. Cicerone, A. Coale, T. Dietz, P. Gleick, R. Healy, R. Lenski, M. McDonnell, J. Lubchenco, T. Malone, B. McCay, N. Myers, D. Pimentel, G. Rabb, D. Skole, and M. Soule (U.S. National Academy of Sciences Planning Group for a study on ecological effects of human activities); Partha Dasgupta (Cambridge University); A. Ehrlich (Department of Biological Sciences, Stanford University); W. Falcon, L. Goulder, and R. Naylor (Institute for International Studies, Stanford University); R. Howarth, A. Kinzig, S. Lele, and R. Norgaard (Energy and Resources Group, University of California at Berkeley); G. Woodwell, R. Houghton, R. Ramakrishna, J. Amthor, and E. Davidson (Woods Hole Research Center);and M. Weitzman (Department of Economics, Harvard University). The responsibilityfor errors and infelicities, however, rests solely uith the authors. Our work on this topic was supported in part by grants from the Winslow and Heinz Foundations. J. Holdren also gratefully acknowledges the hospitality of the Woods Hole Research Center during a 1992 sabbatical in which much of his part of this work was done. A sustainable process or condition is one that can in literature going all the way back to the ancient be maintained indefinitely without progressive Greeks and somewhat more frequently and diminution of valued qualities inside or outside sweepingly in the two hundred years since the the system in which the process operates or the work of Malthus, above all in the period since condition prevails. (We exclude from consider- World War 11.3 Only in the past five years, how- ation, in applying this definition, the depletion of ever, has sustainability become a catchword ca- available energy from the sun on a time scale of pable of capturing the attention not only of envi- several billion years!)1 Such a definition may be ronmental scientists and activists but also of logically appealing, but it is hardly sufficient for (some) mainstream economists, other social sci- addressing the meaning of sustainability in the entists, and policymakers. context of practical choices about how to rnain- This enhanced salience presumably resulted tain or improve the well-being of humans on this from a suite of coincident factors. For one, the planet.2 What kinds of processes and conditions world community is no longer transfixed by the need to be sustained in the interest of maintaining Cold War. A second factor is the reluctant appre- or improving well-being? What are the sources ciation of the severity of the debt crisis in the and dimensions of the main threats to the developing world. A third is the substantial ad- sustainability of these? What places should be vancementinscientificunderstandingof thermag- investigated and what should be measured to nitude and consequences of ongoing global envi- find out? Can sustainability be made compatible ronmental transformations, including the deple- with-or traded off against-otherdesideratarelat- tion of stratospheric ozone, the buildup of green- ing to policy choices? (Consider, for example, sus- house gases, and the destruction of biodiversity. tainable development versus rapid development). Also very important has been the attention given The proposition that particular human prac- to the notion of sustainable development in the tices would prove unsustainable has cropped up report of the World Commission on Environment Defining and Measuring Sustainability: The Biogeophysical Foundations and Development (WCED 1987, also known as living within the carrying capacity of support- "the Brundtland report") and the avalanche of ing ecosystems (IUCN 1991) related studies that has followed. * Economic growth that provides fairness and Notwithstanding the extraordinary growth of opportunity for all the world's people, not just the "sustainability" literature in the past few years the privileged few, without further destroying (an unsustainable process, to be sure!), much of the world's finite natural resources and carry- the analysis and discussion of this topic remains ing capacity (Pronk and Haq 1992). mired in terminological and conceptual ambigu- ities, as well as in disagreements about facts and These definitions have the appeal of appearing practical implications.' These problems arise in to reconcile the concerns of diverse constituen- part because the sustainability of the human en- cies-above all the development and environ- terpriseinthebroadestsensedependsontechno- mental communities (Lele 1991)-but they raise logical, economic, political, and cultural factors at least as many questions as they answer. Is it as well as on environmental ones and in part possible to meet the needs of the present without because practitioners in the different relevant compromising the capacity of future generations fields see different parts of the picture, typically to meet their needs? How does one define needs think in terms of different time scales, and often anyway?Whatdeterminescarryingcapacity,and use the same words to mean different things. how does it vary from place to place and over It is therefore appropriate, even though this time? What is the relation between economic introductory chapter and the conference of which growth and development? What constitutes fair- it was originally a part are supposed to focus on ness? Let us sketch out tentative answers to some the biogeophysical aspects of sustainability, to of these broad questions-since those answers begin by locating the biogeophysical aspects will partly shape our understanding of the envi- within thecontext of the widerdebateabout what ronmental issues we want to address shortly in sustainability means and implies. We then ad- more detail-starting with the meaning of devel- dress, in turn, some problems with defining opment. We think development ought to be un- biogeophysical sustainability in practical terms, derstood to mean progress toward alleviating the the connection between biogeophysical mainillsthatunderminehumanwell-being.These sustainability and related concepts such as carry- ills are outlined in table 1-1 in terms of perverse ing capacity and the distinction between renew- conditions, driving forces, and underlying hu- able and nonrenewable resources, the state of man frailties. (The problems at each of these lev- knowledge and debate about the character and els are themselves diversely and often tightly origins of threats to biogeophysical sustainability, interconnected.) The development process isthen and some implications of the current state of seen to entail improving the perverse circum- knowledge and ignorance of these matters. We stances by altering the driving forces, which in undertakeall of this with a pronounced emphasis turn requires overcoming, to some extent, the on the global level of analysis, leaving to the underlying frailties. Sustainable development chapters that follow the task of addressing the then means accomplishing this in ways that do character and measure of sustainability in par- not compromise the capacity to maintain the im- ticular regions and ecosystems. proved conditions indefinitely. Development by this definition should by no meansbe considered synonymous with economic Biogeophysical sustainability in context growth, since growth by itself does not assure progress toward alleviating any of the indicated Much of the current salience of concepts of ills. (Economicgrowth may bea necessary condi- sustainability has come from a wide-ranging in- tion for alleviating some of them, but it is cer- ternational discussion about sustainable devel- tainly not a sufficient condition.) Note also that opment, which has been defined variously as, for we have placed sustainable in front of develop- example: ment to mean not that the development is of a form thatcan be continued indefinitelybutrather * Meeting the needs of the present without com- that the choice of processes and end states for promising the ability of future generations to development are compatible with maintaining meet their own needs (WCED 1987) the improved conditions indefinitely. Under this * Improving the quality of human life while sort of interpretation, even the much-maligned 4 The Meaning of Sustainability: Biogeophysical Aspects Table 1-1: Ills That Development Must Address Condition Meaning Perverse conditions Poverty 1.1 billion-20 percent-of the 5.5 billion people on the planet live in absolute poverty and perhaps 2billion people do not receivea sufficiently nutritiousdiet to alleviate disease Impoverishment of environement Disruption and erosion of environmental conditions and processes on which the well-being of those 5.5 billion people depend even more directly than on economic conditions and processes Possibility of war Civil, international, global, nuclear, or conventional wars manifest in the more than 100 instances of organized armed conflict since World War 11, nearly all of them in the south, with a total loss of life in the tens of millions Oppression of human rights In forms beyond the three already listed, which deny human beings their dignity, liberty, personal security, and possibilities for shaping their own destinies Wastage of human potential Resul ting from all of the foregoing and the despair and apathy that accompany them and from the loss of cultural diversity (Ehrlich 1980) Driving forces Excessive population growth Where excessive means growth that closes more options than it opens (Holdren 1973), a condition now prevailing almost everywhere Maldistribution of consumption Where the maldistribution is of three kinds: between rich and and investment and investment poor as the beneficiaries of both consumption and investment, between military and civilian forms of consumption and investment, and between the two activities themselves, that is, between too much consumption and too litle investment Misuse of technology Which occurs in forms both intentional (as in weapons of mass destruction) and inadvertent (as in the side effects of a broad spectrum of herbicides and pestiddes) Corruption and mismanagement Which are pervasive in industrial and developing countries Powerlessness of the victims Who lack the knowledge and the resources but above all the political power to change the conditions that afflict them Underlying human frailties Greed, selfishness, intolerance, Which collectively have been elevated by conservative political doctrine and and shortsightedness practice (above all in the United States in 1980-92) to the status of a credo Ignorance, stupidity, apathy, and denial The first consisting of lack of exposure to information, the second of lack of capacity to absorb it, and the third and fourth of having the information but lacking the conviction or optimism or fortitude to act on it term sustainable growth need not be an oxymo- As the human enterprise expands, interdependen- ron; it can be taken simply to mean growth in cies mediated through the world economy, the forms-and to end points-compatible with global environmental commons, and international sustainabili ty of the improved conditions it helps political and military relationslinkand intensify the bring about. threats posed byeach of these ills. Thus therequire- If improvementsin the human condition are to ments for sustainability include not only the envi- be not only achieved but also sustained, all of the ronmental factors to which we will shortly turn in ills will need to be addressed; this is so because detail but also military, political, and economic failure to address any one of them can eventually ones. The minimum requirements in each of these undermine the progress made on all the others. categories are presented in table 1-2. 5 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 1-2: Requirements for Sustainable Improvements in Well-being Area and requiremenf Rationale Military No weapons of mass destruction No one can be secure as long as these exist anywhere, and as long as any country insists on retaining them, others will have an incentive to acquire them Limited capabilities of national Security would be served by attaining a condition in which no nation's military forces military forces were strong enough to threaten the existence of other states; this can be facilitated by "defense dominance," in which national forces are structured to be much stronger in defense than in offense. If stronger peacekeeping forces are needed, they should be placed under international control Political Self-determination Smaller political units should coalesce into orbe absorbed bylargerones onlyby mutual consent, based on mutual advantage Participation/empowerment Societies are not stable-and hence not sustainable-unless their citizens have an effective voice in decisions that affect their lives The rule of law The rule of the strongest, the most devious, or the most unscrupulous is a prescription for perverse and destabilizing forms of competition Guarantees for human rights Majority rule does not include the privilege of abusing minorities; sustainability requires respect for cultural diversity as well as biotic diversity Economic Reduced disparities within and The large gaps between rich and poor that characterize income distribution within and between countries between countries today are incompatible with social stability and with cooperative approaches to achieving environmental sustainability Internalization of environmental E.conomic markets will lead to overconsumption of environmental resources and costs Ultimately to unsustainability if these resources are not priced or are underpriced Assignment of property rights to 1This approach seems essential to avoid the outcome in which high discount rates of future generations economic actors allow actions that undermine long-term sustainability to appear economically attractive (lowarth and Norgaard 1990) Environmental Preservation of the environmental What this requirement consists of and the way it might be attained are the topics of the bassofpresentandfuturewell-being rest of this chapter With that wider array of considerations as circular or unsatisfying in other ways. Consider context, we now take a closer look at the environ- the following capsule definitions: mental dimensions of sustainability that are the * Sustainability refers to a process or state that main focus of this volume. can be maintained indefinitely (IUCN 1991) * Natural resources must be used in ways that Definitions of environmental sustainability do not create ecological debts by overexploiting the carrying and productive capacity of the The environmental aspect of sustainability has Earth (Pronk and Haq 1992) been the subject of a rich literature, albeit only * Aminimumnecessaryconditionforsustainability recently with the term sustainability appearing is the maintenance of the total natural capital explicitly.5 As with the concept of sustainable stock at or above the current level (Costanza development, however, the definitions of envi- 1991). ronmental sustainability tobe found in the litera- The first statement is essentially a dictionary ture recent enough to use that term are often definition of sustainability; it tells us only what 6 The Meaning of Sustainability: Biogeophysical Aspects we already knew sustainability to mean. The tal sustainability go beyond the sorts of capsule secondstatementintroducestheinterestingterm definitions cited above and elaborate what "ecological debt" to describe an element of sustainability might entail and require (see, for unsustainability, but the elaboration in terms of example, boxes 1-1 and 1-2). The 1980 World overexploiting carrying capacity and productive Conservation Strategy of the International Union capacity is not much help, insofar as it merely for the Conservation of Nature, the United Na- transfers the definitional burden to over-exploi- tions Environment Program, and the World Wild- tation and carrying capacity. The third statement life Fund (IUCN 1980) concludes, for example, offers an actual specification of at least one ele- that sustainability requires "maintenance of es- ment of sustainability, but there is still buried sential ecological processes and life-support sys- within it a definitional problem: How is "total tems; preservation of genetic diversity; and sus- natural stock" tobedefined and measured?Assum- tainable utilization of species and resources." ing this hurdle can be overcome, the further ques- This three-part prescription seems to consist of tion will surely arise: What is inviolable about the different facets of the same thing: preservation of current level? Can environmental scientists give a genetic diversity and sustainableuse are essential good answer? We shall return to this issue below. to maintain essential ecological processes and life Of course, all serious writers on environmen- support systems. Box 1-1: Definition and Measurement of Sustainability: The Biophysical Foundations Keiichiro Fuwa Environmental issues have become so popular that politicians around the world no longer need to be persuaded of their importance. Natural scientists have been using the word sustainability for a fairly long time, and recently social scientists as well as politicians have started to use it quite frequently. However, it has yet to be defined clearly. Recommendations have been made for the definition of measurements and indicatorsof sustainability. Although by no means final, the following working definition of biophysical sustainability is satisfac- tory for the time being: Biophysical sustainability means maintaining or improving the integrity of the life support system of Earth. Sustaining the biosphere with adequate provisions for maximizing future options includes enabling current and future generations to achieve economic and social improvement within a framework of cultural diversity while maintaining (a) biological diversity and (b) the biogeochemical integrity of the biosphere by means of conservation and proper use of air, water, and land resources. Achieving these goals requires planning and action at local, regional, and global levels and specifying short- and long-term objectives that allow for the transition to sustainability. Biophysical refers not only to biology and physics but also to geology and chemistry. This is expressed in the definition, particularly through mention of biogeochemical integrity. Natural science has become so interdisciplinary that it is often confusing; nevertheless physics, chemistry, geology, and biology remain the most basic disciplines. Biogeophysicochemistry expresses them all in one word, albeit an exceptionally long one. Defining terms such as sustainability and sustainable development with reference to the global environment is, to my mind, complicated by the fact that humanity has been considered special and separate from other animals and plants. This has not always been the case. The Earth is divided into three spheres: atmosphere, hydrosphere, and lithosphere. The biosphere was added later as the fourth sphere but, unlike the others, includes those parts of the atmosphere, hydrosphere, and lithosphere in which lifeexists. Plants and animals are, of course, part of thebiosphere, but moreimportantly, humans are included as just one species of animal and are not treated specially. In recent years, particularly when serious environmental problems were recognized, human activity was so intense and pervasive that it came to be considered-for example, by the Man and the Biosphere Programme-as separate from the activity of other forms of life. Biophysical sustainability must, therefore, mean the sustainability of thebiosphere minus humanity. Humanity's role has to be considered separately as economic or social sustainability. Likewise, sustainable development should mean both sustainability of the biophysical medium or environment and sustainability of human development, with the latter sustaining the former. 7 Defining and Measuring Sustainability: The Biogeophysical Foundations Box 1-2. Coming to Grips with the Biogeophysical Issues in a Social Construct, or How to Talk about Sustainability without Being a Social Scientist Shara_ Lele "You cannot talk about sustainability without talking about people, about politics, about power and control." Comment by a sociologist at a seminar on sustainability University of California, Berkeley, 1988 "Sustainability is maintaining the ecological basis of economic well-being, so any discussion of sustainability must incorporate economic considerations." World Bank economist Comments such as these threaten to create a gridlock in our discussions of the biophysical foundations of environmental sustainability. But we are clearly not (and probably nobody is) capable of conducting such an all-encompassing discussion. How then do we discuss the biophysical founda- tions of environmental sustainability, however defined? Social, political, and cultural issues come into play in a number of ways at two critical stages in any discussion of environmental sustainability. Stage 1. In deciding, * What is to be sustained? That is, what relative ranking is to be given to, say, current resource productivity, productive potential, or genetic diversity? * What attributes, or combinations of attributes, of a particular system are to be maintained non- decreasing: average productivity, stability, resilience, or adaptability? * Over what time scale is this sustenance desired? * Who istobenefit? If a tradeoff is necessarybetween current and futureconsumption and well-being, or between the well-being of one community and that of another, who is to decide and how? * Should it be economic value of any resource flow or stock that is maintained non-decreasing, or should it be the physical quantity of that flow? Stage 2. In understanding, * Why is there environmental unsustainability, however defined, in the world today? * How would one achieve or move toward whatever notion of an environmentally sustainable society that is decided on in stage I? Stage 1 requires an explication of differing individual and cultural values, preferences, as well as beliefs about and approaches to a highly uncertain and unknowable future and then the resolution of such differences through some social process. Stage 2 requires an understanding of the complex array of social, political, and cultural factors in today's world that lead to environmentally unsustainable behavior. The 1991 "Strategy for Sustainable Living" by The economist Herman Daly, who has been a thesametriadoforganizations(IUCN1991)says pioneer in thinking systematically about these that "sustainable use means use of an organism, matters,6 recentlyofferedamorehelpful three-part ecosystem, or other renewable resource at a rate specification of the ingredients of sustainability within itscapacity for renewal ." Operating within (Daly 1991): the capacity for renewal clearly is one of the key * Rates of use of renewable resources do not elements of sustainability, but this formulation doesnot deal with either nonrenewable resources exceed regeneration rates or the possible off-site, out-of-ecosystem impacts * Rates of use of nonrenewable resources do not through which exploitationof oneresource within exceed rates of development of renewable sub- its capacity for renewal might adversely affect the sti tu tes renewability of other resources or the * Rates of pollution emission do not exceed as- sustainability of other ecosystems. similative capacities of the environment. 8 The Meaning of Sustainability: Biogeophysical Aspects Once this is clearly realized, it is easier to understand where our contributions as biophysicists and ecologists can be and ought to be in informing the process of reaching some societal consensus on the issues in stage 1. At the same time, we realize that, in our work, we have often made implicit decisions about the issues raised in Stage 1. We should therefore proceed as follows: 1. Clearly state the assumptions we are making about reality in a particular case, examine whether some assumptions are commonly shared, and determine the extent to which these may be justified. For instance, perhaps most ecologists believe that whatever is to be maintained nondecreasing in an ecosystem should be measured in physical, not economic, terms. This follows from their rejection of the belief commonly held and vigorously promoted by most economists: that technological change can continuously compensate for reduction in physical resource flows, thus preventing utility from decreasing. 2. Clearly state what value-based choices of objectives, of their ranking, of time horizons, and of users are being implicitly made in any particular case. 3. Identify a few scenarios corresponding to choices different from those that we might want to make. Having done this, we can then proceed with our basic tasks: 4. Synthesize the current state of knowledge about the relationships between biophysical processes that affect different objectives at different temporal and spatial scales. That is, what intensity of harvesting under what techniqueof loggingcanbe maintained in a tropical forest ata nondecreasing level for what time period? What would the implications of a nondecreasing resilience requirement be? 5. Identify a sparse set of indicators that best relate to each combination of objective, scale, and so forth and possibly identify threshold values for them. For instance, what would be the best indicator of stable harvests in the above-mentioned forest? What would be the indicator of resilience in the same system? What scales (spatial and temporal) may be most appropriate or sensible for measuring what attribute or type of sustainability? 6. Explore the ways in which the different scenarios interact; that is, the synergisms and contradictions among objectives, attributes, and indicators and between sustainability in general and other societal objectives. What are the tradeoffs between, say, maintaining timber productivity and maintaining biodiversity in a forest, or between average production and resilience? What are the tradeoffs between different levels of these attributes of sustainability and between the net yield or human well- being produced and the manner in which it is distributed within society? If we are able to do this in a self-aware and socially sensitive manner, we will be able to overcome the paralysis of analysis and make a major contribution to the sustainability debate. Thefirstoftheseconditions,bybeingstatedin the is met, by earmarking part of the proceeds from aggregate,partlyaddressestheproblemof off-site the exploitation of nonrenewable resources for impacts associated with the exploitation of indi- the development of renewable alternatives. vidualrenewableresources: theregenerationrates Daly's third condition, on rates of pollution constraints presumably reflect cross-resource or emission, does not seem as satisfying. If assimila- cross-ecosystem impacts occurring within the tive capacity of the environment means the ca- overall pattern of resource exploitation. pacity to assimilate the pollution without any The second condition, the rate of use of nonre- adverse effect on human health or welfare (in- newable resources, offers a clever solution to the cluding through diminution of ecosystem ser- question of how any use of nonrenewable re- vices), the difficulty is that there are many kinds sources can be contemplated within a of pollution for which the assimilative capacity, sustainability framework. Daly offers a detailed so defined, is probably zero (including, for ex- formulation on how to ensure that this condition ample, ionizing radiation, chlorofluorocarbons, 9 Defining and Measuring Sustainability: The Biogeophysical Foundations lead, and more). It does not seem to insist on no ment-not to mention the potentially continu- harm from pollution as a condition of ously derivable benefit flows-are partly un- sustainability; the question is rather what level of known (indeed, partly unknowable) and also harm is tolerable on a steady-state basis, in ex- partly incommensurable. (Without commensu- change for the benefits of the activity that pro- rability, one is stuck with trying to sustain the duces the harm.7 individual, incommensurablebenefit flowsrather We would also add to Daly's three-part formu- than-more sensibly-an aggregated total ben- lation that the first condition applies to resources efit flow within which tradeoffs among different for which substitution at the required scale is types of benefits could be contemplated.) currently and foreseeably impossible (essential Second, insisting that potential benefit flows resources). It is useful to distinguish those from remain constant over very long periods of time is resources for which substitutes are currently or problematic because environmental conditions foreseeably available (substitutable resources). and processes-climate, topography, the biota- Renewable substitutable resources could be are occurring all the time even in the absence of sustainably exhausted on the same basis as non- human interventions. The potential magnitudes renewable substitutable resources (Daily and of such changes over the very long term make the Ehrlich 1992). concept of foreveressentiallymeaningless,at least in relation to the sustainability of conditions that Biogeophysical sustainability humans of today care about. Third, it is conceivable that technological im- in theory and practice provements will permit well-being to be main- tained despite diminished benefit flows from the The two most important questions relating to a environment. This argument is probably the one definition of biogeophysical sustainability are most heavily relied upon by those not convinced "What is to be sustained?" and "For how long?" of the need to maintain the stream of environ- It is useful to distinguish, with respect to these mental services undiminished. But attempts to questions, between what one would like the an- substitute technology for diminishing or other- swers to be in theory and what one might have to wise inadequate environmental services invari- settle for in practice (see table 1-3). ably entail monetary costs and often generate Saying that what is to be sustained, in theory, significant new environmental impacts. In some is the magnitude and quality of benefit flows cases, these additional costs and impacts may continuouslyderivablefromtheenvironmentcap- more than offset the (presumed) benefits of the tures the idea that potential benefits are impor- activities that necessitated augmentation of the tant, not merely the benefits that society happens natural environmental services in the first place; to be deriving now. And, of course, saying that and even if it is supposed that this will not be the the time scale is forever takes the definition of case, it strikes us as imprudent in the extreme to sustainability seriously. assume that suitable technology for replacing Alas, several practical preblems intrude on the whatever environmental services are lost will attractiveness of this theoretical approach. First, become available in a timely manner and on the even the existing benefit flows from the environ- requisite scale. Table 1-3: Biogeophysical Sustainability in Theory and Practice What is to be sustained? For how long? In theory, the magnitude and quality of benefit flows that are Forever continuously derivable from the environment In practice, the magnitude and quality of stocks of cnvironmental Half-life of 500 to 1,000 years resources 10 The Meaning of Sustainability: Biogeophysical Aspects In any case, in light of the difficulties of mea- Contrasting views about the sustainability suring actual and potential environmental ben- of human activities efit flows, and in light of the conceptual and practical problems of insisting on no degradation Given the above (or any other) definition of forever, it may be necessary in practice to settle sustainability, some obvious questions present for trying to sustain the magnitude and quality of themselves: environmental stocks. The time scale on which this ought to be ensured might be defined in * Are current practices for transforming natural practical terms by a resource or stock half-life of resources into flows of economic goods and 500 to 1,000 years, a period much longer than services sustainable according to the indicated current planning horizons, but much shorter than definition? If not, in what respects and by what geologic time. A tentative rule for prudent prac- margins is sustainability violated? tice, then, would be to constrain the degradation of * Can the larger flows of goods and services monitorableenvironmentalstocks to not more than 10 required to shrink the gap between rich and percent per century. poor, or the still larger flows required to meet Note that degradation of 10 percent a century the needs of a doubled or tripled population, produces, strictly speaking (that is, with Q = Qo bedeliveredsustainablybyexpandingcurrent expl-0.lOtl), a half-life of about 700 years for the practices or by using improved practices that resource. Degradation of 20 percent a century arealreadyknown?Or would sustaininglarger would mean a half-life of 350 years, leaving a flows require improvements over the best quarter of the resource remaining after 700 years. practices now known? We focus on stocks in this prudent-practice To environmental scientists, the answer to the approach, because that is what can most easily be first question is clearly no. Current rates of degra- measured (albeit still not all that easily). Although dation of essential resources are typically an or- our approach is similar in this respect to the der of magnitude too high (in the range of 100 Costanza prescription quoted earlier, an impor- tant~~~~~ ~ difrec is th spcfcto ofa finiert percent a century or more) for them to qualify as tant difference iS the specfication of a finite rate sustainable. The margins by which sustainability of degradation as opposed to insistence on maTh- is exceeded by various types and combinations of taining the stocks at just their current level. This human activity are very difficult to ascertain, sidesteps slightly the argument with the econo- however. It follows from the first answer, in any mists and technologists over what is so special case, that current practices could not possibly about the current levels; putting the argument in sustain even larger flows of goods and services, termsof degradation rates relies on the presumed but whether best-known practices could do so circumstance that there is some degradation rate requires further careful analysis. that is too high to be regarded as sustainable, even Although environmental scientists would be allowing for economic substitution and techno- in nearly unanimous agreement on the answers logical change. just given, many members of other academic Of course, it would not really be acceptable to disciplines and numerous policymakers would run down environmental stocks at 10 percent a dispute not only these answers but also the centuryindefinitely.Thepointisratherthatarate relevance of the questions. It is worth looking of 10 percent a century (which after all means more closely at the origins of these differences about 0.1 percent a year) is slow enough to give in viewpoint. They undoubtedly arise in part society a reasonable chance of figuring out what from ambiguities in and disagreements about thisdegradation iscosting, which forms of degra- the meaning of sustainability. A more impor- dation can be compensated for, how those forms tant source of disagreement, however, are the can be stopped that cannot be compensated for or differing assumptions, perceptions, and knowl- tolerated, and so on, before it is too late. At edge about (a) the importance of environmen- current degradation rates, by contrast, which are tal conditions and processes in supporting hu- typically an order of magnitude or so higher (that man well-being, (b) the sensitivity of those con- is, in the range of 100 percent a century or more), ditions and processes to disruption, and (c) the natural services will be devastated before society character and amenability of society to remedy even understands what is happening, let alone the anthropogenic impacts now threatening finds time to takeevasiveactionon the needed scale. such disruption. 11 Defining and Measuring Sustainability: The Biogeophysical Foundations Confusion about the sensitivity of those condi- are often as ignorant about economic principles, tions and processes to disruption is evident in the and their relevance to environmental protection, comment attributed to economist William as economists are about ecological principles. Nordhaus that only 3 percent of gross national Approachingconsensusaboutbiogeophysical product (GNP) in the United States depends on sustainability clearly will require more research, the environment. In fact, the entire GNP in the morecornmunicationacrossdisciplines,andmore U.S. depends, ultimately, on maintaining the bio- education of the public and policymakers about a physical requisitesof sustainability. Furthermore, multitude of issues, notably: the importance of agriculture (the economic sector * The character and dimensions of the ways to which Nordhaus apparently was referring) is envir altr and futio afe vastly underestimated by its present share of GNP. human well-being, include ienctio fca Thegreatestdisparitiesininterpretationof the human well-being, scluesing the identifica- relationships between the human enterprise and tion of environmental services and the quanti- Earth's life support systems seem, in fact, to be g those between ecologists and economists. Mem- * The ways human impacts imperil environ- bers of both groups tend to be highly self-selected mental services, involving identification of en- and to differ in fundamental worldviews. Most vironmental systems at risk and the causes, ecologists have a passion for the natural world, extent, time scales, and degree of irreversibility where the existence of limits to growth and the of anthropogenic threats to these systems. consequences of exceeding those limits are ap- * The amenability of the threats to remedy, parent. Ecologistsrecognize that a uniquecombi- including potential improvements in nation of highly developed manual dexterity, technology and management, use of economic language, and intelligence has allowed humanity incentives to induce appropriate changes, and to increase vastly the capacity of the planet to the social, political, and economic barriers to support Homo sapiens (Diamond 1991); nonethe- implementation of the remedies. less, they perceive humans as being ultimately subject to the same sorts of biophysical constraints that apply to other organisms. The causes and character Economists, in contrast, tend to receive littleor no training in the physical and natural sciences of environmental damage (Colander and Klamer 1987). Few explore the natural world on their own, and few appreciate Understanding the amenability of the threats to the extreme sensitivity of organisms-including remedy requires a closer look at the factors and those upon which humanity depends for food, trends thatareat therootof theproblem. Anearly materials, pharmaceuticals, and free ecosystem approach to illuminate this issue was the "I = services-toseeminglysmallchangesinenviron- PAT" formula (Ehrlich and Holdren 1971,1972): mental conditions. Most treat economic systems as though they were completely disconnected (environmental) impact = population x from the planet's basic life support systems. The consumption per person (affluence) x impact narrow education and inclinations of economists per consumption (technology). in these respects are thus a major source of dis- agreements about sustainability. Today, a bit of further disaggregation seems Someof the responsibility for these continuing useful, so as not to confuse affluence with re- disagreements also rests, however, on the failure source use (the two being separable by means of of ecologists and other environmental scientists the inverse efficiency factor, resource use per to make a case for the importance of environmen- economic activity) and so as to separate measures tal conditions and processes and for the magni- of what technology does to the environment tude of anthropogenic threats to these, in terms (stress) from measures of actual damage, which understandablebyand persuasive toothers. This depends not only on stress but on susceptibility problemispartlyamatteroftoofewenvironmen- (itself a function of cumulative damage from tal scientists having made the effort to articulate previous stresses, as well as other factors). Thus, a coherent case, but also partly a matter of the great gaps in the environmental science itself. Damage = population Nor has it helped that environmental scientists x economic activity per person (affluence) 12 The Meaning of Sustainability: Biogeophysical Aspects x resource use per economic activity (resources) stresses on ecosystems; and of course it is largely x stress on the environment per resource use through institutions (economic, political, legal, (technology) and so on), through beliefs and values, and x damage per stress (susceptibility) through changes in these that damage feeds back to population, economic activity, and technology. Note that this expanded relation (like the previ- The relative importance of the different caus- ous I = PAT) is no more and no less than an ative and modulating factors and the nature and identity. It is true by definition. People are free to intensity of their interactions clearly vary drasti- argue about whether it is informative and useful- cally with the social and ecological contexts, hence and we thinkitis-but toargueabout whetheritis with location as well as with time. The situation right is foolishness. is further complicated by the wide array of Identities of this sort are instructive because mechanisms by which phenomena in one loca- they remind us that increases in population, af- tionandtime-bethesephenomenademographic, fluence, and the ratio of environmental stress to economic, technological,ecological,political,cul- economic activity (itself clearly a function of the tural, or other-propagate to and influence other composition of that activity and the technology locations and times. with which it is accomplished) are multiplicative Beyond these elaborations about the various in their effect on damage, so that the impact of contributing factors, it is important to be clear each factor is a matter not only of its own magni- about what we mean by damage. Damage means tude but also of the magnitudes of the others.' At reduced length or quality of life for the present the same time, such identities are deceptive, and generation or future generations. Damage may above all deceptively simple, in that they fail to result from short-term alteration of environmen- make explicit (a) the ways in which the variables tal conditions, long-term degradation of environ- on the right-hand side of the equation are not mental capital, and costs of attempts to avoid independent, (b) the ways in which institutions, reductions in length and quality of life with com- beliefs, and values can influence all of the vari- pensating technological and social interventions. ables and the nature of the interactions among This is of course an explicitly and them, or (c) the ways in which the relative impor- self-consciously anthropocentric definition, con- tance of the variables and the nature of the inter- sistent with the anthropocentric definitions of actions among them vary with location and time. sustainable development that provide the con- With respect to the lack of independence of the text for this debate. The anthropocentric approach variables, the magni tude and composi tion of eco- to environmental problems is not the only valid nomic activity per person, and their rates of one, but (a) it is the one most likely to succeed in change, are likely to depend in complicated ways the policy arena and (b) the difficulties in agree- on themagnitudeand composition of thepopula- ing on definitions, problems, and solutions are tion and their rates of change. The nature of the even greater if human well-being is not at the technology used to generate economic activity center of attention. (and thus the kind and magnitude of stresses Of course, any economic activity will lead to exerted on the environment by that technology someenvironmentaldamageexceptincaseswhere per unit of economic activity) will depend on the the susceptibility factor-damage per unit of magnitude and composition of all economic ac- stress-is zero. Such cases exist when environ- tivity (hence on population and economic activ- mental processes are capable of completely ab- ity per person) as well as on their rates of change. sorbing or buffering the imposed stress such that The damage to ecosystem services per unit of there is no short-term alteration of environmental imposed environmental stress-a form of conditions or long-term degradation of environ- dose-response relation-will generally be a func- mental capital of a magnitude sufficient to pro- tion both of the magnitude and composition of duce an impact on length or quality of life for any the stress and of their rates of change. members of the present or future generations of With respect to the role of institutions, beliefs, humans. But would many types and levels of and values, it is clear, on reflection, that these economic activity in real-world conditions actu- underlie as well as modulate changes in popula- ally meet this condition? tion, economic activity per person, and the tech- The critical issue is to specify a level of damage nological variables through which the combina- that is acceptable to society. An economist might tion of population and per capita activity exert argue, for example, that we should not refrain 13 Defining and Measuring Sustainability: The Biogeophysical Foundations from activities that cause any damage, but only Ignorance, knowledge, and uncertainty from those whose marginal costs (the sum of the internal costs plus the damages as here defined) Assuggested earlier, the listof what isnot known exceed their marginal benefits. That is, if one and what needs to be known in order to address could measure all of the costs and all of the "sustainability" with comprehensiveness and benefits in a single currency (such as 1992 dol- rigor is a very long one. Table 14 illustrates this lars), one would define the rational limit on the point by presenting in abbreviated form the re- scale of any economic activity as the level at search agenda on ecological aspects of the issue which the slopes of the cost and benefit curves that was developed recently as part of the Sus- were equal. Then maximum sustainable abuse tainable Biosphere Initiative of the Ecological (Daily and Ehrlich 1992) would mean the level of Society of America (Lubchenco and others 1991). abuse (stress) that pushes the total marginal cost Anothercompact survey of research requirements (slope of total cost curve) to just equal the total related to sustainability isavailable in the agenda marginal benefit (slope of total benefit curve). of the International Geosphere-Biosphere Alas, there is no hope of quantifying and mon- Programme of the International Council of Scien- etizing all the diverse kinds of damages associ- tific Unions (ICSU 1992). The most important ated with economic activity (even the damages and demanding research sub-agenda of all may occurringinthepresent,nottomentiontheproblem be one embedded in the environment-society el- of bringing future damages into our common cur- ements of these lists: namely, the question of how rency, which requires agreeing on a discount rate). to formulate and implement economic and social In practice, then, cost-benefit-type approaches incentives for preserving the essential character- to determining maximum sustainable abuse are istics and functions of environmental systems. stuck with the problem of apples-and-oranges At the same time, there is a great danger in aggregation of qualitatively different damages, falling into the scientist's trap of calling for more current and future damages, and damages and researchwithout sufficiently emphasizing what benefits. Additional daunting problems include we already know and the implications of that dealing with stochasticity and establishing an knowledge. We know forcertain, forexample, that: appropriate margin of safety in the face of uncer- * No form of material growth (including popula- tainty. All these difficulties mean that tastes and tion growth) other than asymptotic growth, is preferences about the proper weighting of differ- sustain able; ent categories become relevant and that the issue is political as much as technical. (A huge litera- * Many of the practices inadequately support- ture about risk perception and risk acceptance is ing today's population of 5.5 billion people are relevant in some respects to these issues of maxi- unsustainable; and mum sustainable-ormaximum tolerableormaxi- * At the sustainability limit, there will bea trade- mum prudent-abuse.) off betweenl population and energy-matter Table 1-4: Research Needed in Ecological Science on Sustainability Research area Need Ecological causes and consequences Changes in climate. Changes in atmosphere, soil, and freshwater and manne chemistry. Ecology of conservation and biodiversity Global distribution of species and changc factors. Biology of rare and declining species. Effects of global and regional change on diversit\ Strategies for sustainable ecological systems Patterns and indicators of responses to stress. Guidelines and techniques for restoration. Theory for the management of ecological systems. Introduced species, pests, and pathogens. Integration of ecology with economics and other social sciences. Source: Lubchenco and others 1991. 14 The Meaning of Sustainability: Biogeophysical Aspects throughput per person, hence, ultimately, that make the steps worthwhile even if the between economic activity per person and uncertain hazards later turn out to be small. well-being per person. . Insurance strategies entail paying to minimize This is enough to say quite a lot about what vulnerability,butwithoutexpectingthatother needstooeventually (a world of zero benefits (besides minimizing vulnerability) jus- needs to be faced up to tifyull the invstmnts the payent beon the net physical growth), what should be done now tify the investments; the payments beyond the (change unsustainable practices, reduce exces- expectation of other benefits are the premium, sive material consumption, slow down popula- and the issue becomes how big an insurance tion growth), and what the penalty will be for premium should be paid.' postponing attention to population limitation * Avoiding the biggest downside risks, finally, (lower well-being per person). means trying to leave the biggest margin of Of course there are implications of what is not safety against the hazards with the biggest negative consequences (largest areas and known as well as implications of what is known. gaqug numbers of people affected, highest degrees of The holes n soioety's knowledge should moti- irreversibility),evenif theprobabilitiesof these vate development of strategies for minimizing downside outcomes are unknown or appear to the dangers associated with uncertainty. Any be small. sensible prescription for dealing with the kinds of uncertainty we face will include adopting There is, of course, much more to be said about no-regrets strategies,buying insurance, and avoid- the meaning and measurement of biogeophysical ing the biggest downside risks: sustainability and about what human societies should be doing about it. But since this chapter is No-regrets strategies entail taking steps that intended only to set the stage for the more de- minimize vulnerability to the uncertain haz- tailed treatments to follow, we happily leave the ards while at the same time conferring benefits rest to them. Notes 7. Harm that would qualify as tolerable, in this context, could not be cumulative, else continu- 1. A billion is 1,000 million. ing additions to it would necessarily add up to 2. Althoughconcernsotherthanthemaintenance unsustainable damage eventually. Thus, for or enhancement of human well-being can be example, a form and level of pollution that posited as principles for guiding human be- subtract a month from the life expectancy of havior (see, for example, Ehrenfeld 1978), we the average memberof the human population, shall accept for the purposes of this chapter or that reduce the net primary productivity of that the perspective focuses on the well-being forests on the planet by 1 percent, might be of humans. deemed tolerable in exchange for very large benefits and would certainly be sustainable as 3.teSome landmarks inMthisary1864 sustin9 long as the loss of life expectancy or reduction literature include Marsh 1864; Vogt 1948; ipoctiyddtgwihim.of Osborn 1948- Brown 1954; Carson 1962; Ehrlich in productivity did not grow with time. Two of 1968; Cloud 1969; SCEP 1970; and Meadows us have coined the term "maximum sustain- and others 1972. able abuse" in the course of grappling with and others 1972. ~~~~~such Ideas (Daily and Ehrlich 1992). 4. A particularly helpful review calling attention 8.ch ing Discsin was aptdfo to ths difclisi.ha yLl 91 8. The following discussion was adapted from to these dificltisstatbyethe unpublished report of a National Academy 5. In addition to references cited in note 3 above, of Sciences study group, chaired by Holdren in some major works include Ehrlich and 1991, on human impacts of ecosystems. See Ehrlich 1970; Institute of Ecology 1971; also the acknowledgments to this paper. Ehrlich and others 1977; CEQ 1980; IUCN . Theideaof society'sbuyinginsuranceishardly 1980,1991; Myers 1984; Mungall and McLaren unprecedented: much of the $300 billion a year 1990; Woodwell 1990; Turner and others 1991; that the U.S. spends on defense, for example, Dooge and others 1992; Meadows and others represents an insurance policy against contin- 1992. gencies considerably lcss likely to come about 6. See, for example, Daly 1973, 1977, 1991; Daly than are some of the environmental disasters and Cobb 1989. one could mention. 15 Defining and Measuring Sustainability: The Biogeophysical Foundations References .1980. "Variety Is the Key to Life." Technol- ogy Review 82 (March-April), pp. 58-68. Brown, Harrison. 1954. The Challenge of Man's Ehrlich, Paul R., and Anne H. Ehrlich. 1970. Popu- Future. New York: Viking. lation, Resources, Environment. San Francisco: Carson, Rachel. 1962. Silent Spring. Boston: W. H. Freeman. Houghton-Mifflin. Ehrlich, Paul R., Anne H. Ehrlich, and John P. CEQ (Council on Environmental Quality). 1980. Holdren. 1977. Ecoscience: Population,Resources, The Global 2000 Report to the President. With the Environment. San Francisco: W. H. Freeman. U.S. Department of State. Washington, D.C.: Ehrlich, Paul R., and John P. Holdren. 1971. "Im- U.S. Government Printing Office. pact of Population Growth." Science 171, pp. Cloud, Preston. 1969. Resources and Man. A study 1212-17. and recommendations by the Committee on . 1972. "One-Dimensional Ecology." Bul- Resources and Man of the Division of Earth letin of theAtomic Scientists 28:5 (May), pp. 16, Sciences, National Academy of Sciences, Na- 18-27. tional Research Council, with the cooperation Holdren,JohnP.1973. "Populationand theAmeri- of the Division of Biology and Agriculture. San can Predicament." Daedalus (Fall), pp. 31-34. Francisco: W. H. Freeman. Ca rT Making Howarth, R. B., and R. B. Norgaard. 1990. Colander, D., and A. Klamer.s1987. "The Makin"IntergenerationalResource Rights, Efficiency, of an Economist." Economic Perspectives 1, pp. and Social Optimality." Land Economics 66, pp. 95-111. ~~~~~~~~~~1-11. Costanza, Robert, ed. 1991. Ecological Economics: Institute of Ecology. 1971. Man in the Living Envi- The Science and Management of Sustainability. ronment. Madison, Wisc.: Institute of Ecology. New York: Columbia University Press. ICSU (International Council of Scientific Unions). Daily, Gretchen C., and Paul R. Ehrlich. 1992. 1992. Global Change: Reducing Uncertainties. "Population, Sustainability, and Earth's Car- Stockholm: Royal Swedish Academy of Sci- rying Capacity." BioScience 42:10, pp. 761-71. ences, International Geosphere-Biosphere Daly, Herman E. 1977. Steady-StateEconomics. San Programme. Francisco: W. H. Freeman. IUCN (International Union for the Conservation ---. 1991. "Elements of Environmental of Nature). 1980. World Conservation Strategy: Macroeconomics." In R. Costanza,ed.,Ecologi- Living Resource Conservation. With the United cal Economics. New York: Columbia University Nations Environment Program and the World Press. Wildlife Fund. Gland, Switzerland. Daly, Herman E., ed. 1973. Toward a Steady-State . 1991. Caring for the Earth: A Strategy for Economy. San Francisco: W. H. Freeman. Sustainable Living. With the United Nations Daly, Herman E., and J. B. Cobb, Jr. 1989. For the Environment Program and the World Wildlife Common Good: Redirecting the Economy toward Fund. Gland, Switzerland. Community, the Environment, and a Sustainable Lele, Sharachchandra M. 1991. "Sustainable De- Future. Boston: Beacon. velopment: A Critical Review." World Develop- Diamond, Jared. 1991. The Rise and Fall of the Third ment 19:6, pp. 607-21. Chimpanzee. London: Radius. Lubchenco, Jane, A. Olson, L. Brubaker, S. Car- Dooge,J.C.l.,G.T.Goodman,J.W.M.LaRiviere, penter, M. Holland, S. Hubbell, S. Levin, J. J. Marton-Lefevre, T. O'Riordan, F. Praderie, MacMahon, P. Matson, J. Melillo, H. Mooney, and M. Brennan,eds. 1992. AnAgenda of Science C. Peterson, H. Pulliam, L. Real, P. Regal, and for Environment and Development into the 21st P. Risser. 1991. "The Sustainable Biomass Ini- Century. Cambridge, England: Cambridge tiative: An Ecological Research Agenda." Ecol- University Press. ogy 72, pp. 371-412. Ehrenfeld, David. 1978. The Arrogance of Human- Marsh, Gerge Perkins. 1864. Man and Nature, or ism. New York: Oxford University Press. Physical Geography as Modified by Human Ac- Ehrlich, Paul R. 1968. The Population Bomb. New tion. Reprinted in 1965 by Cambridge, Mass.: York: Ballantine. Belknap Press of Harvard University Press. 16 The Meaning of Sustainability: Biogeophysical Aspects Meadows, Donella H., Dennis L. Meadows, and SCEP(Studyof Critical Environmental Problems). Jorgen Randers. 1992. Beyond the Limits. Post 1970. Man's Impact on the Global Environment. Mills, Vt.: Chelsea Hills Publishing Co. Cambridge, Mass.: M.I.T. Press. Meadows,DonellaH.,DennisL.Meadows,Jorgen Turner, B. L., William C. Clark, Robert W. Kates, Randers, and W. W. Behrens III. 1972. The John F. Richards, Jessica T. Mathews, and Wil- Limits to Growth. New York: Universe Books. liam B. Meyer, eds. 1991. The Earth as Trans- Mungall, Constance, and Digby J. McLaren, eds. formed by Human Action. Cambridge, England: 1990. Planet under Stress. New York: Oxford Cambridge University Press. University Press. Vogt, William. 1948. Road to Survival. New York: Myers, Norman, ed. 1984. GAIA: An Atlas of Plan- Sloane. etary Management. New York: Doubleday. Woodwell, George M., ed. 1990. The Earth in Osborn, Fairfield. 1948. Our Plundered Planet. Bos- Transition: Patterns and Processes of Biotic Im- ton: Little, Brown & Co. poverishment. New York: Cambridge Univer- Pronk, Jan, and Mahbubul Haq. 1992. Sustainable sity Press. Development: From Concept to Action. The Hague WCED (World Commission on Environment and Report. New York: United Nations Develop- Development). 1987. Our Common Future. Ox- ment Program. ford, England: Oxford University Press. 17 Key Concepts and Terminology of Sustainable Development Mohan Munasinghe and Jeffrey McNeely The authors gratefully acknowledge assistance in preparing this chapter from Noreen Beg and Marten Jenkins, as well as useful inputs from Shelton Davis, Alfred Duda, John English, Genedy Golubev, Kenneth King, Alcira Kreimer, Anil Markandya, Ranjiva Munasinghe, Norman Myers, and lona Sebastian. The expression sustainable development was given development activity will inevitably ad- coined to demonstrate that economic growth and vahce some interests while prejudicing others. In environmental protection can be compatible. orderforinformedchoicestobemade,economic, Many definitions have been provided for this ecological,political,social,andculturalfactorsall phrase (Pearce, Barbier, and Markandya 1989; need to be considered and presented to Pezzey 1989). At the most basic level, dictionaries decisionmakers in an unambiguous (and unbi- define the verb sustain as "to give support to, ased) fashion. This is a formidable task, and this nourish, keep up/prolong," among others, and chapter attempts to make a small contribution to development as the improvement of human wel- the debate. fare and the quality of life. Development is there- Different disciplines place varying interpreta- fore not entirely synonymous with economic tionsontheconceptofsustainability.Whileecono- growth, which focuses mainly on real incomes. A mists emphasize the maintenance and improve- narrow definition of sustainable development ment of the living standards of humans, ecolo- would indicate that per capita income or well- gists and scientists have broadened the meaning being is constant or increasing over time. The to express concerns about preserving the adapt- wider concept of sustainable development is less ability and function of entire ecological and bio- precise and embraces a set of indicators of well- physical systems. At the same time, geographers being (including income) that could be main- andanthropologistshavefocusedontheviability tained or increase over time. The World Bank, in of social and cultural systems (Toman 1992). its World Development Report 1992, states that sus- Understanding sustainable development in tainable development meansbasing developmen- turn requires that the competition for resources tal and environmental policies on a comparison be placed in a historical context, in order to iden- of costs and benefits and on careful economic tify and describe the social and economic under- analysis that will strengthen environmental pro- pinnings of environmental degradation. By ex- tection and lead to rising and sustainable levels of amining how underlying processeshave evolved welfare (World Bank 1992). in the past, it becomes easier to understand the The practice of sustainable development in- goals of various types of development activities volves making choicesbetween alternatives. Any and institutional arrangements. Defining and Measuring Sustainability: The Biogeophysical Foundations A historical perspective ety, around 2370 B.C. In the Indus Valley of India and the Mayan lowland tropical jungles of Throughout the course of human evolution, the Mesoamerica, large-scale deforestation and the populations that survived wereby definition those resultant soil erosion precipitated a similar col- that had a sustainable relationship with their lapse of society, caused by the inability of fragile environment; that is, unsustainable behavior led ecosystems to support a massive, complex infra- todisplacementorextinctionofthepopulationor structure. In a somewhat different manner, the to a change in behavior. This does not mean that demands of rapid population growth on the envi- early human populations did not have significant ronment during the heyday of the Roman Empire ecological impacts or modify their environments led to long-term environmental decline in the to suit their needs better. Indeed, coincident with Mediterranean (caused by deforestation and soil the first arrival of Homo sapiens in North America, erosion from overgrazing). A similar picture some thirty-four genera of large mammals be- comes from many, if not most, individual civili- came extinct, and the first arrivals of humans into zations of the past (Darlington 1969). The process Australia, New Zealand, and Madagascar were of civilization could be more broadly defined to accompanied by significant losses of species of include not only the rise and fall of individual large animals that were easily harvested by the societies but also their progressively increasing new and sophisticated predator (Martin 1984). levels of organization and complexity. The future Presumably, huntersmissed at least some of these sustainability of this broader evolutionary pro- easily hunted species once they were gone, while cess will depend on the ongoing search for paths local cultures based on the harvesting of large of long-run sustainable development. mammals necessarily adopted other means of The increased demands that industrial coun- earning a living or themselves became extinct. tries place on natural resources, and growing The unsustainable pressures of human activi- poverty in developing nations, threaten the pros- ties on the environment greatly increased as the pects for achieving a level of ecological domestication of animals and the cultivation of sustainability while protecting human well-be- cropsbecamecommon.Thustraditionalnomadic ing. A major issue today is the highly resource- pastoralism is generally accepted as being more intensive per capita consumption in industrial environmentally benign than agriculturalism, nations, but population growth will also add to given that agriculture deliberately transforms the pressure on natural resources in the future. nature and ecosystems by altering soils and Todayhumansuseapproximatelyl2,000timesas growth patterns and through deforestation much energy (mainly in the form of fossil fuels) as (Goudie 1990; Ponting 1990). they did 400 generations ago when farming was Boyden and Dovers (1992) describe how the first introduced. Nearly 80 percent of this energy human aptitude for culture gave rise to is used in industrial nations, which constitute "technometabolism," involving new inputs and only 25 percent of the world's population. The outputs of materials and energy through various same imbalance is also seen in the per capita kindsof technological processes. They outline the generation of waste between the two groups of phases of this process. Phase 3, the early urban countries. The great intensification of phase, led to several biologically and socially technometabolism in the industrial countries has important changes, including the occupational resulted in the rapid increase of gaseous waste and social stratification of human society, the emitted into the atmosphere. In particular, scien- institutionalization of warfare, and the increased tists have noted an increase of 25 percent in the role of contagious disease as a cause of mortality. carbon dioxide content of the atmosphere since Ponting (1990) discusses how sociological the preindustrial age (see table 2-1 and figure 2-1 changes acted as a catalyst for the first known forexamplesofcarbondioxideemissionsbycoun- large-scale anthropogenic disruptions to the bio- try and average annual consumption per capita of physical environment. In Mesopotamia, the need energy, metals, and so forth for selected countries). for food surpluses to support a growing The increase in greenhouse gases is expected nonproducer class of bureaucrats and soldiers to lead to an increase in the mean global tempera- led to an intensive, irrigated agricultural system; ture of from2 'to 5 over the next fiftyyears. Some the consequent waterlogging and salinization of scientists believe that global warming will result the fields destroyed the basis for Sumerian soci- in a rise in sea level, an increased rate of deserti- 20 Key Concepts and Terminology of Sustainable Development Table 2-1: Average Annual Consumption per Capita, Various Years Product and United United year of data World Total Canada Germany France Italy Japan Kingdom States Energy (gigajoules) Total fossil fuels, 1989 54.71 19032 270.95 160.98 94.98 106.56 109.80 142.11 282.93 Solids, 1989 18.71 53.13 44.65 74.41 14.48 10.01 26.96 47.94 80.15 Liquids, 1989 22.48 90.27 127.06 57.75 60.24 69.35 67.51 56.97 127.21 Gas, 1989 13.53 46.92 99.25 28.82 20.26 27.19 15.33 37.20 75.57 Metals (kilograms) Crude steel, 1989 15320 489.09 529.72 563.08 312.77 486.71 757.64 304.00 411.44 Aluminun,refined, 1990 3.39 15.94 15.66 17.83 12.77 11.32 19.55 7.89 17.24 Copper, refined, 1990 2.04 9.54 6.95 13.29 8.46 8.24 12.76 5.52 8.54 Lead, refined, 1990 1.05 4.71 3.44 5.79 4.51 4.48 3.38 5.25 5.13 Nickel, refined, 1990 0.16 0.76 0.46 1.21 0.79 0.47 1.29 0.57 0.50 Tin, refined, 1990 0.04 0.19 0.11 0.28 0.15 0.11 0.28 0.18 0.15 Zinc, slab, 1990 1.32 4.93 4.75 6.85 5.03 4.69 6.59 3.29 3.95 Industral materials (0*grns) Cement, 1983-85 197.72 416.06 239.86 502.6P 376.17 670.11 550.66 242.85 327.23 Fertilizer, 1989-90 27.63 58.18 82.67 58.39 108.67 31.52 15.74 41.22 75.21 Forest products Roundwood (cubic meters) 1989 0.67 1.35 6.71 0.56 0.70 0.27 0.68 0.12 2.04 Paper and paperboard (kilograms), 1989 44.39 229.61 236.09 181.62 148.17 116.05 221.84 168.15 306.71 a. World consumption is assumed to be equal to world production. b. Consumption in the Federal Republic of Germany plus production in the German Democratic Republic. Figure 2-1: Cumulative Emissions of Carbon Dioxide from Fossil Fuels for Twenty-five Countries with the Highest Emissions, 1950-89 fication, and species extinction. Also, in the rnid- (billions of metric tons of carbon dioxide) 1980s more than 85 percent of the chlorofluoro- 160 -carbons (CFCs) released into the atmosphere came 140 from the industrial countries (see figure 2-2 for use of CFCs and halons by region in 1986). Under 120 the best outcome of the Montreal Protocol, the 100 concentration of CFCs in the stratosphere will increase to three times the present level in the 80 next thirty years, resulting in increased ultravio- 60 - let radiation from the sun due to depletion of the 40 - _ . - i t . ozone layer (Boyden and Dovers 1992). The con- 40 -, _ . 171 ofsequences are likely to include greater incidence 20 of skin cancer, cataracts, and so forth in humans olu . . . . ~~~~~~~~lethal effects of increased ultraviolet radiation on O - _T T s E nnn n , I as well as disruption of ecosystems due to the 'P ~~~~~~~~many organisms. Poverty and environmental degradation Note: The European Community comprises twelve Any discussion of sustainable development would countries: Belgium, Denmark, France, Federal Republic of be meaningless without recognition of the close Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, and the United Kingdom. relationship between poverty and environmental Source: Unpublished data from the Carbon Dioxide Information Analysis Center and the Oak Ridge National Laboratory. 21 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 2-2: Current and Projected Use of * Up to 20 metric hectares a year of primary Chlorofluorocarbons and Halons, by tropical forests (about 1 percent of total) are Region, 1986 and 2000 being lost. * In one of many examples, 5,000 Philippine Actual 1986: Projected 2000: villagers were killed by recent flooding caused 1.3 million tons 2.2 million tons in part by the presence of deforested hillsides. 70% 62% For water, / Approximately l billion peoplein thedevelop- ing world are still without access to clean wa- ter for drinking and bathing.' . Aboutl.7billionmustcontend withinadequate 15% sanitation facilities, resulting in 900 million ~~ ~-- - 22% cases of diarrheal diseases annually and 3 15% 16% million deaths (mostly infant mortalities); F] Industrial East Europe and i Developing another 500 million people suffer from countries former Soviet Union countries trachoma, 200 million from schistosomiasis or bilharzia, and 900 million from hookworm. Source: Munasinghe and King 1992, pp.24-25. For air, * About 1.2 billion people live in urban areas in developing countries that do not meet World degradation. It is clear that the poor suffer the Health Organization standards on dust and consequences of environmental degradation most, smoke; it is estimated that the reduction of especially since they are the most vulnerable and such pollutants would save 300,000 to 700,000 least able to avoid or mitigate the consequences. lives annually. TheHagueReport(PronkandHaql992)points . Firewood, charcoal, and dung, the primary out that the ratio of per capita incomes is 150:1 fuels of developing countries, endanger the between the top and bottom 20th percentiles of health of 400 million to 700 million people the world population. It estimates that most poor (especially women and children) with health people live in areas of high biodiversity and frag- consequences equivalent to smoking several ile ecosystems:.80 percent of Latin America, 60 packs of cigarettes a day. Automobile emis- percent of Asia, and 50 percent of Africa. Basi- sions, primarily lead, also contribute to health cally, the greatest proportion of the poor live in problems related to air pollution. rural areas: 69 percent in Sub-Saharan Africa, 74 . In the long term, even global warming is likely percent in South Asia, and 60 percent in Latin to have the most severe consequences for low- America (World Resources Institute 1992a). A income countries and the poor, as they will be contribu ting factor to poverty and environmental the least able to cope with the range of potential degradation is that rural areas lag behind urban impacts. areas in human development, with infant mortal- ity in some countries 30 to 50 percent higher and In developing countries, rapid population malnutrition as much as 50 percent higher. The growth, agricultural modernization, and inad- difficulties in determining the priorities of sus- equatelandtenuresystemsarecreatingeverlarger tainable development are highlighted by the fol- populations with little or no access to productive lowing statistics (World Bank 1992). land. This results in rural to urban migration or the increased use of marginal lands. As more and For land, more people exploit open-access resources in or- * Therateofsoildegradationanddesertification der to survive, the environment is further de- is increasing, mainly affecting the rural poor. graded.Thisdegradation occurs through soil ero- Countries like Costa Rica, Malawi, Mali, and sion, loss of soil fertility, desertification, defores- MexicomaybelosingO.5tol.5percentofgross tation, depleted fish and game stocks, loss of national product (GNP) annually in terms of biodiversity and natural habitats, depletion of farm productivity. groundwater, pollution, siltation of rivers, and so 22 Key Concepts and Terninology of Sustainable Development on. The end result is a reduction in the carrying sibilities, environmental regeneration rates, and capacity and productivity of the land and a loss of the existing resource base." Other researchers absorptive carbon sinks, as in the Amazon. This share the view that we can avoid the prospect of has both intragenerational and intergenerational Malthusian scarcity by resource substitution and consequences, exacerbating existing poverty and technological innovations (Toman 1992). threatening theeconomicprospectsof futuregen- However, many also share the view that the erations. scale of human pressure on natural systems is In developing countries, it is not so much the already well beyond a sustainable level (as dis- quality of life that is at risk because of environ- cussed inToman 1992). IntheformerSoviet Union, mental degradation, but life itself. Although eco- as Gerasimov (1974, as cited in Goudie 1990) has nomic growth is crucial, these poor nations must pointed out, up to the industrial revolution, "the adopt models of development that are less mate- natural environment taken as a whole was able, rial- and energy-intensive and more environmen- up to a point, to withstand anthropogenic distur- tally sound than in the past. The industrial coun- bances, although there were also local irrevers- tries can assist in this effort by facilitating the ible changes. Since the industrial revolution, the transfer of technology and financing environmen- general intensity of human impact on the envi- tally sustainable projects in developing countries. ronment has exceeded its potential for restora- tion in many large areas of the earth's surface, leading to irreversible changes not only on a local General ideas about sustainable but also on a regional scale." Examplesof human- development induced environmental degradation are illus- trated in figures 2-3 and 24, which provide ex- Probably the best-known and most frequently amplesof water degradationand thelocationand quoted definition of sustainable development is types of soil degradation. provided in the Brundtland Report as "develop- Sustainability has also been defined as "a rela- ment that meets the needs of the present without tionship between dynamic human economic sys- compromising the ability of future generations to tems and larger dynamic, but normally slower- meet their own needs" (World Commission on changing ecological systems, in which (a) human Environment and Development 1987, p. 8). This life can continue indefinitely, (b) human indi- definition is anthropocentric and based on the viduals can flourish, land] (c) human cultures can concept of intergenerational equity. develop, but in which (d) effects of human activities An economist's working definition of sustain- remain within bounds, so as not to destroy the able development could be "the maximization of net benefits of economic development, subject to Figure 2-3: Consumption of Irrigation Water and maintaining the services from and quality of natu- Volume of the Aral Sea, 1960-87 ral resources over time." This implies that renew- able resources (especially scarce ones) should be Volume Consumption used at rates less than or equal to the natural rate (cubic (cubzc kilometers of regeneration and that the efficiency with which kilometers) per year) nonrenewable resources are used should be opti- 1,200 - 100 mized, subject to how effectively technological 1,000 - V .._......_ progress can substitute for resources as they be- come scarce (Pearce and Turner 1990). To this 800 -60 could be added the requirement that waste be 600 - generated at rates less than or equal to the assimi- Consumption - 40 lative capacity of the environment (Barbier 1991). 400 - Dasgupta and Maler (1990) point out that a 200 - 20 decline in resource stocks per se is not a reason for concern. They state that "whether or not policy o -o should be directed at expanding environmental 1960 1965 1970 1975 1980 1985 resource bases is something we should try and deduce from considerationsof population change, Source: Micklin 1988. intergenerational well-being, technological pos- 23 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 2-4: Types of Soil Degradation, by Region diversity,complexity,andfunctionoftheecological life support system" (Costanza 1991; see figure 2-5). World Essentially, if economic growth is an increase 56°io%~ - 0 in quantity, then logically it cannot be sustainable Oceania indefinitely on a planet with a finite amount of 81% 2% resources. Economic development, however, is \ 4% 1 %7 an improvement in the quality of life and does not necessarily imply an increase in the quantity of 16% resources consumed. In this context, measures 28% such as the Overseas Development Council's Europe Africa Asia Physical Quality of Life Indicator and the United Nations Development Program's Human Devel- 52% ~ ~= 46%;j_ 58%i;_ opment Index have been proposed. Such qualita- 38^1 4% 2% tive rather than quantitative development may 17% 38 2' i Y10% be sustainable and could become the desirable 12% 300io long-range goal of humanity. North America Central America South America 63% 74° 51° Concepts and definitions of sustainable development J360 °-.o1% 1 As indicated earlier, wemightidentifythreebroad 7% 29% approaches to sustainability, as shown in Figure Wate=Chemnical degradation 2-6. Economists relate sustainability to the pres- = Wind erosion _ Phvsical degradation ervation of the productive capital stock. Physical scientists relate sustainability to the resilience or integrity of biological and physical systems Note: Categories not shown for a region represent (Perrings 1991). A third view relates sustainability to a concern about the adaptability and preserva- Source: Oldeman, van Engelen, and Pulles 1990, fig 5. tion of diverse social and cultural systems (Toman 1992). This section provides a brief overview of each in tum. Economic approach Figure 2-5: The Ecological System The economic approach to sustainability origi- nates in the Hicks-Lindahl definition of income as the maximum flow of benefits possible from a given set of assets, without compromising the flow of future benefits. This requires the preser- vation or increase of the base of assets over time. Solar Rat SOC10 Waste Radiated Solow (1986) describes the sustainability con- Re..-- Econo0mic -.-* EnergEn energy j Sub- Poll- Energy dition as follows: "A society that invests in systemw ution reproducible capital, the competitive rents on its curient extraction of exhaustible resources, will enjoy a consumption stream constant in time,. . . \ Ecological / an appropriately defined stock of capital-in- cluding the initial endowment of resources-is being maintained intact, and . . . consumption can Note: The capacity of the ecosystem may become over- being i coo loaded by the growing socioeconomic subsystem (the broken be Interpreted as the interest on patrimony." (The lines), constant stream of consumption is one interpre- Source: Munasinghe 1993. tation of intergenerational equity.) As discussed 24 Key Concepts and Terminology of Sustainable Development Figure 2-6: Approaches to Sustainable Development ECONOMIC 0 / Effidency * Growth * Stability /Poverty Biodiversity/Resilience * Consultation/Empowerment Natural Resoufces * * Culture/Heritage Pollution - * < * inter-generational equity > SOCIAL * popular participation ENVIRONMENTAL Source: Munasinghe 1993. below, the measurement of a country's base of ral capital fulfills life support functions that are assets might be broadened to include natural not met by man-made capital (for example, the capital, in addition to man-made capital and hu- ozone layer). Finally, the neoclassical interpreta- man-resource capital. tion of substitutability between inputs cannot be It has been argued that maintaining the stock easily applied to natural capital, given their of natural capital is not essential to the develop- multifunctionality and difficulties of physical ment of a sustainable economy, that technologi- quantification and economic valuation. For ex- cal change improves the efficiency of resource ample, table 2-2 presents a table of nonmarket use, and that more productive man-made capital goods available from a forest ecosystem. can be substituted for natural capital (electricity Cleveland (1991) points out that, while the forfuelwood, fertilizers formanure, and so forth). neoclassical production model assumes that capi- Livestock can replace wild animals in the diet, tal and labor are primary inputs to production without necessarily reducing species diversity (ignoringthesubstantialquantitiesofenergyused (the Masai in the savannas of Africa are an ex- in the process of harvesting resources itself), a ample); modern pharmaceuticals can replace biophysical model of the economic process as- medicinal plants; and domesticated plants can sumes that capital and labor are intermediate replace harvests from the wild. However, there inputs and that the underlying primary factors of are certain caveats to this argument (Pearce and production are low-entropy energy and matter. Turner 1990). According to the First Law of Ther- Cleveland re-analyzed the seminal work of Barnett modynamics, energy (and matter) cannot be cre- and Morse on scarcity (1963) from a biophysical ated or destroyed. Therefore, man-made capital perspective and found that energy use increases and natural capital are not independent: the latter with resource depletion, because lower-quality is often needed to make the former. Second, natu- deposits require more energy to locate and upgrade. 25 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 2-2: Environmental Functions of Forests Source of materials Sink General and and services for wastes life support Timber Absorption of waste Genetic pool Fuelwood Recycling nutrients Climate regulation Other business products Watershed protection Carbon fixing Non-wood products Protection of soil quality Habitat for people, flora, and fauna Genetic resource and resistance to erosion Aesthetic, cultural, and spiritual source Recreation and tourism Scientific data Source: Munasinghe 1992a. Additionally, natural and man-made capital external shocks could have severe economic and are only substitutable to a limited extent. Natural social consequences. Already, it is reported that capital is subject to irreversibilities. Natural capi- 10 percent of the Earth's potentially fertile land tal can be depleted but not increased, if previous hasbeen turned into desert or wasteland, while a decrementshave led to extinction (Pearce, Barbier, further25 percent may be endangered. Each year, and Markandya 1989). Furthermnore, the species 8.5 million hectares are lost through erosion and of plants and animals driven to extinction may siltation (Pronk and Haq 1992). Given such dire haveprovided significantbutpresentlyunknown conditions, natural capital would probably be benefits in the future. For example, a new chemi- preferable to man-made capital, in the sense that cal isolated from the back of the Ecuadorian poi- it may reduce vulnerability and increase resil- sonous frog Epipedobates tricolor is a painkiller 200 ience (Pearce and Turner 1990). Self-reliance is an timesaspotentasmorphine.Called "Epibatidine," important related quality that should be main- it is a member of an entirely new class of alka- tained (Lovelock 1979). loids: an organo-chlorine compound, which is A key concern in wishing to maintain a constant rarely found in animals. Thus important discover- stock of capital is the essentially socio-ethical issue ies remain to be made in the natural world, pro- of intragenerational and intergenerational equity. vided that species are allowed to survive and evolve. The concept of intragenera tional equity calls for a The issue of irreversibility is closely linked to spatial universality requirement: the current the role of uncertainty. There is considerable sci- highly skewed distribution of income and pov- entific uncertainty about the adverse conse- erty is recognized as unacceptable. This concept quencesofglobalwarming.Thechemistryofacid is also linked to the issue of resilience. rain is still beingdeveloped. We do not completely Intergenerational equity has a temporal universal- understand how to relate the role of ocean cur- ity requirement: the rights of future generations to rents to the determination of climate, and the an acceptable level of welfare must be protected. ways in which stands of natural forest affect Another consideration is that, when natural microclimates are still being researched. We are capital is destroyed, the habitats of other species not even certain about the loss of assimilative are damaged. As scientists discover more com- capacity that could be sustainable. Even changes plex interconnections in biophysical systems, of 5 or 10 percent in the carrying capacitv of the which in turn underlie the productive basis of Earth can be viewed as having enormous social human society, then the preservation of whole and political consequences on a global scale ecosystems (natural capital) may be viewed in (Schneider 1990). Given such uncertainty about economic terms. An extreme version of this ap- environmental benefits and costs, it becomes in- proach is centered around the so-called Gaia hy- creasingly difficult to make tradeoff decisions pothesis, which states that the totality of life on between man-made capital and natural capital. Earth is responsible for controlling the tempera- The issues of vulnerability and resilience also ture, chemical composition, oxidizing ability, and merit attention. In many developing countries, acidity of the Earth's atmosphere (Potvin 1992). the margin of flexibility is so low between a In addition, there is the purely ethical issue of sustainable and nonsustainable society that any existence rights for other species. 26 Key Concepts and Terminology of Sustainable Development It is clear that deciding what is the relevant the stress to sustainable levels on ecosystems that stock of capital and how it should be nieasured are central to the stability of the global system depends on the substitutability in productioni, (Perrings 1991). (This idea is related to the protec- and indispensability,of the different components tion of the resilience of fragile ecosystems through of capital: machines, techlnical knowledge, and the maintenance of natural capital.) renewable and nonrenewable resources (Pezzey From a thermodynamics perspective, 1989). The challenge for economists is to expand sustainability is related to the fact that in ecosys- the analysis of resource values to consider the tems where energy flows are open, the system function and value of ecological systems, making tends to organize itself into stable or quasi-stable greater use of ecological information, and to ex- states, within the constraints imposed by its envi- tend economic theory and analysis to examine ronment. In ecosystems where resource values more fully the implications of biophysical re- are incompatible with that self-organization, the source limits. Valuation techniques must be fur- systemwillswitchfromonethermodynamicpath ther developed, in particular to improve method- to another, or from one self-organization to an- ologiesforassessinghowfuturegenerationsmight other (Perrings 1991). Threshold values exist for value different attributes of natural environments. the diversity of species within an ecosystem. If The technique of en vironmental accounting seeks any one population in an ecosystem falls below to adjust the presen1t system of national accounts itscritical threshold level, theself-organization of and commonly used measures like gross (oines- the whole is altered. Theintegrity of an ecosystem tic product (GDP), by explicitly considering deg- is measured by its ability to maintain its self- radation of environmental assets and expendi- organization through the selection of an opti- tures to remove pollution (Lutz and Munasinghe mum operating point along the same thermody- 1991). One key question is how to distinguish namic path-that is, without undergoing the ir- between efficient allocation of resources and so- reversible change that occurs from switching cially oriented concerns such as issues of paths. Economic activity that imposes intergenerational and intragenerational equity unsustainable levels of stress on the natural envi- (Toman 1992). ronment may generate negative feedback effects. The impact of trade on sustainable develop- Using reasoning somewhat similar to the neo- ment isbeingexamined moreanc' more. Norgaard classical argunment for the substitutability of capi- (1987) has pointed out that development occur- tal, Potvin (1992) states that, "at the end of a ring as a result of exchange has encouraged agri- period during which depletable inventories are cultural specialization, with the resultant reduc- drawn down for use, if structural and chemical tion in the diversity of crops and supporting energy newly embodied in things of human de- species. He also notes that, although neoclassical sign and manufacture exceeds the energy lost to trade theory .issumes that the factors of produc- reserves themselves, then exploitation of these tion are mo;bile, the environmental services pro- inventories is consistent with ecosystem self-or- vided by biodiversity, which give land its value, ganization." This enables, for example, humanity cannot freely shift from one product to another. to use a finite minleral and fossil inventory in McRobert (1988) observes that traditional eco- order to generate a perpetual stream of income, nomics research on trade policy does not account assuming that society invested all the rents from for hidden environmental externalities (through the resource. If capital goods are acquired rapidly pollution and change in climate) that are implicit enough to make up for the continually declining in the transportation of goods. use of resources, resources can continue to be used and reserves remain positive. A vital ele- ment of the Cobb-Douglas production function, B iophysical approach Q = K-R'-, is that each input is essential. If either capital stocks or resources are run down to zero, Sustainability from a biophysical perspective is then output is zero (Hartwick and Olewiler 1986). linked to the idea that the dynamic processes of Ecological sustainability basically implies the the natural environment can become unstable as preservation of biodiversityat a sustainable level. a result of stresses imposed by hunman activity. Biodiversity, a term that las entered widespread Sustainability in this scenario refers to maintain- usage since the late 1980s, has arisen because ing a system's stability, which implies limiting other terms, suchas "living natural resources" or 27 Defining and Measuring Sustainability: The Biogeophysical Foundations "the living environment," were inadequate to are just a token sample of the benefits of preserv- deal with the complexities of conserving the liv- ing biodiversity, the importance of ecological ing systems on which human welfare depends. sustainability is evident. This new approach involves local communities, Anydiscussion of sustainability must examine scientists, indigenous peoples, and many parts of the question of the sustainable use of renewable government working together to ensure that bio- resources (especially biodiversity), in addition to logical resources-land, forests, oceans, and so the more easily quantifiable area of nonrenew- forth-are used in ways that are sustainable and able mineral resources. There is no direct rela- contribute to intergenerational equity. tionship between the market value of biological Biodiversity (as defined by the United Nations resources and their prospects for preservation Environment Program, Intergovernmental Ne- (Orians 1990). An example of this is deforestation gotiating Committee for a Convention on Bio- caused by strip mining of coal. The extractive logical Diversity 1992) means "the genetic, taxo- resource is valued more highly than the species nomic, and ecological variability among living that are being destroyed due to loss of habitat. organisms; this includes the variety and variabil- Livingorganisms also differ from minerals in that ity within species, between species, and of biotic they reproduce and can increase, whereas chemi- components of ecosystems." cal resources may be converted into other forms Conserving biodiversity is the foundation of of matter but do not generate more of their kind sustainable development: (Orians 1990). 1. It supports current productive systems. Sustainable use implies gaining benefits from 2. Future practical values and needs are unpre- a resource on a continuing basis, either through dictable. direct consumption (such as logging or huntiing) or through nonconsumptive means (for example, 3.obruecerstadingofth rosy a td imsasnsffcint tourism). For use to be sustainable in the long tobecertaonoftheroleand impactofremovilng term, consumption cannot exceed the natural in- any component (especially irreversible and crement. Acceptable levels of use are primarily a matter of judgment, based on reproductive rates, In addition, variety is inherently interesting and habitat condition, market demand, and so forth, more attractive. For example, both high agricul- but when the size of a stock is considered to have tural productivity and human health depend on reached a level so low as to threaten the continued the preservation of a diverse biota consisting of existence of the resource, then strict preservation the estimated 10 million species of plants and may be the only available option. animals that inhabit the planet. However, ap- Another issue to be considered in the discus- proximately 90 percent of the food for humans sion of sustainability is the definition of natural. comes from just fifteen plant species and eight Thisword isunderstood bymostpeople tobetthat animal species (Pimentel and others 1992). which is determined by wild nature and is often The continued productivity of agriculture and perceived as the opposite of human. In fact, natu- forestry depends on the preservation of ralness is a relative term coverinig a range of biodiversity. Microbes fix nitrogen from the at- human influences from the most pristine (for mosphere for use by crops and forests: an esti- example, Antarctica, where no permanent hu- mated 90 million tons of nitrogen is fixed for use man settlements are found, while major human by agriculture worldwide with a value of almost influences are felt primarily throughairpollution $50 billion annually (Hardy and others 1975, as and the harvesting of resources from the sur- quoted in Pimental and others 1992). Cross-polli- rounding seas) to the most human-influenced nation is essential for crop reproduction in many end of theecological scale (forexample, a primary cases. More than forty U.S. crops, valued at ap- man-made urban environment). proximately$30billion, require insect pollination Similarly, nativeness is not really a definition for production (Robinson and others 1989, as but rather a descriptive continuum in time that quoted in Pimental and others 1992). Although an ranges from the most ancient, continuous, species estimated $20billion a year is spent on pesticides, in an area (such as oaks in Europe or elephants in parasites and predators in natural ecosystems are Africa) to the most recent immigrants into an area providing about five to ten times this amount of (such asstarlings in Europe). Theacacia trees that value in terms of natural pest control (Piniental are so distinctive in the African savannas, for and others 1992). When one considers that these example, have been there less than I(00 years and 28 Key Concepts and Terminology of Sustainable Development are the result of human introduction of a devas- form of traditional agriculture have to be contin- tating exotic pathogen; just as they have become ued to provide the habitat that it requires. With- the defining characteristic of the African savanna out all the human cultural practices that go with for many people, so too will successful immi- the habitat, the species will be lost forever. Yet grants eventually become "natives" as natural this dimension of species maintenance has been selection creates new ecological communities. neglected in our own tradition of natural re- To most early ecologists the natural ecosystem source management. was the community that would be reached after a The general state of the physical environment long period without large-scale disturbance (fire, is to a large extent determined by living organ- wind storm, and so forth). This was the climax isms. For example, the presence of free oxygen in community, which would be in equilibrium, in the environment is the result of biological photo- which some parameter of interest (species com- synthesis,and livingorganisms playa majorrole in position, biomass, or net primary productivity) is the biogeochemical cycles of such elements as sul- roughly constant from year to year when aver- fur, calcium, nitrogen, and phosphorus (Orians aged over the whole landscape or in which op- 1990). Limits on the sustainable use of physical posing processes are approximately balanced on processes in the environment are related to both a landscape scale. More recently, ecologists have additions and subtractions of materials. The most learned that most areas are subject to various important mechanism whereby materials are re- typesoflarge-scaledisturbance,includingstorms, moved from the environment is the alteration of fires, changes in climate, and outbreaks of pests. ecosystems: deforestation, drainage of wetlands, Therefore, in many areas it may be unrealistic to and the conversion of diverse grasslands into de- try to define the natural vegetation for a site; graded pastures. These activities decrease the size several communities could be the natural veg- of theoriginalhabitat,alteringtheratiobetweenthe etation for any given site at any given time habitat's edge (with altered microclimates) and in- (Sprugel 1991). terior, thereby resulting in higher rates of species As ecologists learn more about the history of extinction.Second,suchactivitiesincreasedistances ecosystems, it becomes increasingly obvious that between patches of habitat, thereby lowering although the general idea of trying to preserve recolonization rates of species (Orians 1990). vegetation in a natural state might be desirable, The spatial nature of additions and subtrac- identifyinga specific point in time asepitomizing tions leading to overreach of a system's limits, the natural state is ill-advised. At any given mo- depends on the general type of environment. In ment, the vegetation of any area has some special the atmosphere, for example, because of the lack characteristic that makesitdifferent from another of physical structure and the high rates of move- time that might equally well have been chosen. ment and mixing of the medium, most problems Therefore, sustainability does not necessarily are global and regional rather than local (for imply maintaining some static natural state, but example, global warming and acid rain). In con- rather maintaining the resilience and capacity of trast, thecumulativeeffectsofhumanactivitiesat the ecosystem to adapt to change. the terrestrial level are primarily local, the most Gomez-Pompa and Kaus (1992) have pointed critical concern being alteration and fragmenta- out that manyof the treespeciesnow dominant in tion of the habitat (Orians 1990). Here, the mature vegetation of tropical areas were, and sustainability issuesbasicallyrevolvearound land still are, the same species protected, spared, or (and water) management. When land is privately planted in the land cleared for crops as part of the owned, land use management can become a com- practice of shifting agriculture. The current com- plex issue. Furthermore, publicly held lands are position of mature vegetation is therefore at least just as difficult to manage due to the pressure that partly the legacy of past civilizations: theheritage public interest groups exert on governments. of cultivated fields and managed forests aban- Decisions about land use management tend to be doned hundreds of years ago. made on small spatial scales, whereas the prob- Oneexampleof thecomplexityof sustainability lems of habitat loss call for large-scale solutions. issues is the now well-known perennial corn, Zea This is compounded by the fact that political diploperennis, which is a secondary species that boundariesdo not coincide with ecological bound- grows in abandoned cornfields. To protect the aries, rendering agreements on efficient land species, the slash and burn techniques of this management solutions more difficult. 29 Defining and Measuring Sustainability: The Biogeophysical Foundations A very recent example of unsustainable use 1991). It was followed up in 1992 when United and the need for clear management responsibility Nations Environment Program and International and ownership is the caviar-producing sturgeon, Union for the Conservation of Nature joined the which is rapidly becoming the latest victim of the World Resources Institute in publishing the Glo- collapse of the former Soviet Union. Over the past bal Biodiversity Strategy, a document outlining the six months, four independent states and two au- policies that need to be followed if biodiversity tonomous regions have appeared around the is to be conserved. Such an approach might be Caspian Sea, which contains more than 90 per- based on the following guidelines: cent of the world's sturgeon stocks. According to Economic incentives for sustainability should Dobbs (1992), "As a result, the tightly regulated be a foundation Of Tesource management (in- caviar-producing cartel formed by the former cluding the correct pricing of resources and Soviet Union and Iran has collapsed, leading to a internalization of environmental costs). free-for-all in which Russian poachers, Azerbaijani mafia bosses, and Turkmen bureaucrats muscle * Policies and programs must be responsive to their way into the lucrative business." Already the needs of the people who live closest to the threatened by a string of ecological disasters in resources being managed. and around the Caspian Sea, sturgeon stocks may * Policies and programs must be adapted to the becompletelydepleted within threeorfouryears. specific characteristics of the resources being Prior to the collapse of the former Soviet Union, managed. strict quotas were established by the Ministry of * Approaches must be flexible and able to adapt Fisheries, and a powerful inspectorate cracked to changes. down on poachers and dealers in illegal caviar. Sturgeonno longerswimup Azerbaijan'spoisoned Freeman (1991) shows that the economic value and dammed-up rivers, so local fishermen catch of resources depends in part on the management immature sturgeon in the Caspian Sea, a practice regime. Value is influenced notonlybybiological that will accelerate the demise of the fishery. Such a and economic factors but also by institutions that free-for-all may earn short-term profits for a few, at manage the resources and, ultimately, by values the cost of long-term welfare for many. embedded in the underlying sociocultural matrix. Economists may contend on theoretical grounds that environmental degradation should Sociocultural approach take place so long as the gains from the activities causing the degradation (such as clearing a forest Crucial, but often overlooked, factors in sustain- for agriculture) are greater than the benefits of able development are the social and cultural as- preserving the area in its existing form. The idea pects. Ethical values, beliefs, and institutions de- of an optimum stock of natural assets is based on velop within sociocultural systems to meet hu- this comparison of costs and benefits, but it as- man needs. The world is in a period of very rapid sumes that the full forgone benefits of preserving change, when new institutions are being created the area in its original form (opportunity costs) to manage natural resources, based on socioethical can be assessed accuratelyand that the gains from values about the environment. Few countries, for the activities are also accurately estimated. Some example, had national parks departments until economists question how well ecological pro- the 1960s, and most ministries of the environment cesses-or capital flows such as contributions to were created after the Stockholm Conference in geochemical cycles-can be captured by tradi- 1972. Many of these government decisions and tional cost-benefit analysis, suggesting instead structureswerebasedonexpectationsofsubstan- that in the face of uncertainty, irreversibility, tial budgets, but recent developments suggest the discontinuities, and catastrophic collapse of natu- inevitability of severe cutbacks in government ral systems, conserving what remains could be a expenditureand the unsustainability of the origi- sound risk-averse strategy (Pearce, Barbier, and nal schemes. Markandya 1990). Developing sustainable social and cultural A growing consensus in today's society is practices to help manage renewable resources is making it increasingly difficult for policymakers oneofthemajorchallengesofthecomingdecade, to ignore the issue of intragenerational equity. as numerous models are attempted. One such Large disparities of income-with the accompa- model was laid out in Caring for the Earth (IUCN nying risk of wars, conflicts over diminishing 30 Key Concepts and Terminology of Sustainable Development resources, migrations, and other destabilizing For the same reasons that we seek to maintain effects-are clearly not socially desirable nor sus- biodiversity, we could also seek to preserve social tainable. The recent conflict in the Middle East and cultural diversity (especially in indigenous (partly attributed to concerns over access to oil or tribal cultures). There may be hidden knowl- reserves) serves as a reminder of the enormous edgeconcerning,forexample,cooperativemodes waste of resources and the potential for environ- of behavior, social stability, and so forth that mental disasters. In general, conflicts can affect could improve overall sustainability and effi- numerous nationsdue to the flow of refugeesand ciency of resource use (especially common prop- pollutants acrossboundaries. Trade, finance,com- erty). In particular, ethnobotanists and agricul- munications, and ecological processes tie the turalists have shown us that indigenous (or world system into a tight web. The continuation "primitive") cultures may have much to teach so- of inequitabledevelopment isinherently unstable, called modern societies about alternative medici- as demonstrated by recent world events. The nal and agricultural practices. In our haste to increasing access of the poor to radio and televi- modernize, we may lose valuable information sion has created new expectations that govern- embedded in traditional cultures and value sys- ments may not be able to meet with their present tems that could improve our understanding of base of resources and unsustainable practices. practical steps to achieve greater sustainability. This may result in social discord (and often an Finally, given the need to change the dominant increasing rate of environmental degradation). paradigm in industrial nations (which empha- One remedy is to convert "have-nots" into stake- sizes material-intensive growth), the diversity of holders and managers of open-access resources. human societies and cultures and their embed- This involves rebuilding institutions and social ded wisdom could be used more effectively. systems, as well as redistributing assets and in- Since the industrial revolution, and especially come in some way. in the past few generations or so, a fundamental Likewise, the issue of intergenerational equity ecological, economic, and cultural shift has oc- needs to be addressed. One of the steps required curred. The world's collection of highly diverse to guarantee the continued presence of the hu- cultures adapted to local environmental condi- man species would be to arrest the creation of tions is now being replaced by a world culture intergenerational externality that results from the characterized by high levels of material consump- unsustainable management of renewable and tion. Economic growth based on the conversion nonrenewable resources. Fu ture generations will of fossil fuels to energy greatly expanded interna- have to bear the cost of any reduction in the flow tional trade, and improved public health mea- of capital caused by the reduction ordegradation sures have spurred such a rapid expansion of of the present stock of renewable resources. Is- human consumption that new approaches to re- sues that surround the present use of natural source management are urgently required. These resources-such as contamination of groundwa- approaches have overwhelmed the conservation ter, modification of climate, disposal of radioac- measures of local communities, often bringing tive waste, and harvesting of marine fisheries at overexploitation and poverty to many rural com- the appropriate level-need to be considered munities and great wealth to cities and certain while keeping the welfare of future generations social classes. in mind. This does not necessarily mean that we Overexploitation is to be expected in times of ignore today's problems of intragenerational eq- very rapid cultural change, as traditional controls uity in favor of future generations. For example, breakdownandhumanslearntoexploitresources "If a particular project being considered maxi- in new ways. The movement of Europeans into mizes the present value, but confers some unac- the Americas is only onedramatic exampleof this ceptably low or negative net benefits on future process.Technologicalinnovations-suchasplan- generations ... current gains could be set aside as tation agriculture or industrial logging-tend to a trust fund. The interest would serve to balance favor exploitation of biological resources and the the distribution of the netbenefits among genera- weakening of traditional approaches to conser- tions. Compensation does not necessarily have to vation, especially when a technologically supe- be monetary-whatever the form, the compen- rior group moves into a region occupied by tech- sation mechanism provides a way of sharing nically less-sophisticated groups. The dominant maximum net benefits among generations" or invading society has the option of movingon to (Tietenberg 1988). fresh resources when an area is exhausted and 31 Defining and Measuring Sustainability: The Biogeophysical Foundations wouldderivenoparticularadvantagefromadopt- Development-action that alters the environ- ingthetraditionsofsustainable,conservativeuse ment so that it caters more effectively to human practiced by the indigenous society. The domi- needs-is essential if the world is to be free from nant society is able to eam virtually all the cash poverty and squalor, but such development must benefitsof the forest, whilepayingalmost noneof be based on naturally regenerating resources that the long-term environmental costs imposed on can meet our needs indefinitely and on prudent use the indigenous society. of depletable resources. At the same time, the subordinated groups Within the process of development, more room lose any advantage from traditions of conserva- must be found for wild nature. The processes of tive use that might have been favored when they wild nature renew the oxygen in the air, maintain could exclude other groups from their territory. the cycles of essential elements, sustain the fertility These traditions evolved when costs and benefits of the land, and regulate the flow of rivers. We turn were internalized in the decisions made by com- to wild nature for new crops and new drugs as well munities. However, since the local people are as for the beauty that enriches life. Environmental now paying far higher environmental costs of protectionanddevelopmentarenotopponentsbut resource degradation, theironly rational response are inseparably one, having interlinked ecological, is to join the exploiters in trying to seek greater economic, and cultural components. short-run benefits as well. Thus, traditional man- The economic interdependence among nations agement systems that were effective for thou- is often viewed as basically desirable, and indeed sands of years have become obsolete in a few the World Commission on Environment and De- decades, replaced by systems of exploitation that velopment has called for greatly expanded interde- may yield short-term profits for a few but impose pendence through enhanced flows of energy, trade, long-term costs on many who are often poor. and finance. Each day several hundred billion dol- Meredith(1992)pointsoutthatculturepersists lars flow across national boundaries, because of only where it meets, at the minimum, the biologi- stock market and trading transactions. This is twice cal needs of a community or where it fits the base as large as the GNP of Sub-Saharan Africa, for of resources. It is therefore crucial that a certain example. Thisglobal interdependence hasbrought adaptive capacity exist in the dynamic relations veryconsiderablematerialbenefitstomanypartsof among resources. For example, technology trans- the world, greatlyincreasing percapitaGNP world- fer is better described as a stimulant to the devel- wide. However, some observers have suggested opment of local technology, since it can beadapted that such interdependence-making the world a to a particular base of resources or culture but single global system-is one source of the global cannot be transplanted in any viable form. depletion of resources. As the distinguished ecolo- One of the dominant economic needs in the gist Ray Dasmann pointed out over a decade ago, world today is the earning of foreign exchange whenweareallpartofasinglesystemconnectedby and expansion of international trade. Such forces powerful economic forces, itbecomeseasierto over- have contributed to the more complete exploita- exploit one part of the global system because other tion of biological resources. As an inevitable re- parts will soon compensate for such sult, cultural diversity is also reduced, for two overexploitation. The localized damage may not main reasons. First, a significant component of even be noticed until it is too late to do anything to cultural diversity that enables people to earn a avoid permanent degradation. living from the local biological environment is Thesystemoftradenowlinkingtheentireglobe, becoming less functional, and second, subordi- primarily for the benefit of urban populations, has nated groups begin to imitate the culture of the led to great prosperity for those who havebeen able dominant group, thereby losing a major portion to benefit from the expanded productivity. How- of theircultural uniqueness. As just one indicator ever, it has often resulted in the devastation of local of the loss of cultural diversity, about half of the ecosystems, tothedetrimentof thelocal,and mainly world's 6,000 main languages are moribund and poor, people who remain dependent on the now- spoken only by middle-aged or elderly people. depleted natural resources. Are all traditional systems doomed to failure, Social discrimination, cultural barriers, and ex- falling victim to state and private ownership? Or clusion from national political processes make in- do traditional systems of community-based re- digenouspeoplevulnerableand subject toexploita- source management still have something tc con- tion. Many groups become dispossessed and tribute? Let us examine a few of the issues. marginalized, and their traditional practices disap- 32 Key Concepts and Terminology of Sustainable Development pear. They become the victims of what could justi- that "the optimal control problem involves the fiablybedescribed ascultural extinction.TheWorld maximization (or minimization) of some index of Commission on Environment and Development performance as a function of a set of state vari- (1987) has recognized this problem, stating that "it ables and control inputs, subject to the constraint isa terribleironythatasformaldevelopmentreaches posed by the natural dynamic of those state vari- more deeply into rain forests, deserts, and other ables." There is a clear correlation between this isolated environments, it tends to destroy the only approach and the economic problem of maximiz- cultures that have proved able to thrive in these ing welfare, and minimizing social cost, through environments." the appropriation of environmental goods and Various major impacts of exploiting the environ- services. ment (such as the greenhouse effect and possible Perrings and others have demonstrated that changes in climate) suggest the inevitability of pro- the workability of the Hicks-Lindahl concept of found changes in the way humans relate to the sustainability in an economic sense is limited, environment. Theexact direction of these changes is because it is dependent on the controllability and unpredictable: the ecological practices of human (stochastic) predictability of the global system, communities could take any of a large number of an area of considerable uncertainty. Ecological formsinthecomingyears.Onepossibilitywouldbe sustainability, however, is more suitable for a a series of local adaptations to locally available control policy that does not depend on the resources, with distant resources being consumed controllability or stochastic predictability of only to the extent that such use is sustainable. In the global system; it is able to assure the particular, correct economic valuation of natural sustainability of uncontrollable systems sub- resources(withthemaximumpracticalinternaliza- ject to basic uncertainty by employing an eco- tion of environmental costs) would be needed to logical sustainability constraint (minimum ensure efficient management of these assets. This safety standards). This constraint imposes di- need notnecessarilymeana radical reduction in the rect restrictions on resource-using economic quality of life, but social, economic, and cnviron- activities that will, in theory, protect the stability, mental conditions will surely be fundamentally integrity, and resilience of the environment. different than they are in today's consumer society Economic modeling allows us to study rigor- and perhaps will come to resemble more and more ously issues that are interrelated and global in the sustainable approaches of traditional cultures. scale. However, what such models omit (by Cultural diversity, which is provided above all choice and by current shortcomings in scien- by the great variety of indigenous cultures in all tific understanding of environmental issues) parts of the world, contributes the human intellec- may turn out, in retrospect, to be crucial to tual "gene pool," the basic raw material for adapt- understanding a particular issue. Such models ing to the local environment. Indigenous people are therefore abstractions, albeit useful ones, who live in intimate contact with their major re- that should always be tempered by judgment. sources could therefore provide practical knowl- One of the most promising approaches involves edgetoguideashifttosustainablesocieties(atleast valuation of the environment and incorpora- in their own local context). The challenge is in tion of such monetary measures into conven- applying this knowledge and, where appropriate, tional economic deci sionmaking. The strengths transferring associated techniques and thinking to and limitations of economic approaches are resource management systems appropriate to explained below. today's circumstances. Finally, the sustainability of highly interconnected (but often diverse) modem societies will depend on how cultural pluralism is Economic approaches at the local encouraged but also on whether it is managed or project level effectively. Over the past few decades, economists have developed and presented several models that Reconciling different approaches to have attempted to reconcile traditional eco- operationalize sustainable development nomic theory, on which decisions are currently made, witlh efficient natural resource manage- The primarygoal of environmental management ment options that facilitate sustainable devel- is to use resources better. Perrings (1991) states opment. Environmental economic models that 33 Defining and Measuring Sustainability: The Biogeophysical Foundations seek to incorporate ecological concerns into Cost-benefit analysis and valuation methods neoclassical macro- and micro-economic theory seek to estimate the monetary and nonmonetary face difficulties. Complications arise because costs and benefits of a given project in mon- natural systems tend to cut across the etaryterms.Unfortunately,theyarenotalways decisionmaking structure of human society. For successful. Thus when projects and policies example, a forest ecosystem (like the Amazon) and their impacts are to be embedded in a could span several countries and interact with system of broader (national) objectives (for ex- many different economic sectors (such as en- ample, preservation of biodiversity)-some of ergy, transport, and agriculture) within the which cannot be easily valued in monetary country. Also, many externalities (for example, terms-multicriteria decisionmaking methods global warming, acid rain, or groundwater con- offer an alternative approach. These methods tamination) are not only difficult to measure in facilitate tradeoffs among different objectives. physical terms, but even more difficult to con- Often both cost-benefit and multicriteria analy- vert into monetary equivalents (to measure the ses are used jointly in a complementary way. willingness to pay of the parties affected by the externalities). Quite often the approach taken is to impose regulations and standards, expressed Safe minimum standards in physical measurements only, that try to elimi- nate the perceived external damages. However, Concernsover iintergencrational equity issuesand this approach may not be effective, because no recognitionof the limitationsof economic models attempt is made to compare the costs of compli- that have tried to address environmental issues ance with the real benefits provided (damages led to calls for the establishment of safe minimum avoided). Furthermore, in many developing standards, as formulated by Ciriacy-Wantrupand countries, the regulations and standards are developed by Bishop (Norton and Ulanowicz established without a realistic implementation 1991). Figure 2-7 illustrates the safe minimum mechanism or appropriate institutional struc- standard for balancing natural resource tradeoffs tures. At the same time, traditional regulations and imperatives for preservation. Toman (1992) and standards are undermined through inap- defines the safe minimum standard as a "socially propriate state policies, resulting in a higher determined dividing linebetween moral impera- rate of environmental degradation. tives to preserve and enhance natural resource With the adherence to a predetermined set of systems and the free play of resource tradeoffs." environmental limits (safe minimum stan- This requires the current generation to desist dards), traditional decisionmaking procedures from actions that could result in environmental that rely on technoengineering, financial and impacts with high-cost or irreversible damage. economic analyses of projects and policies, or a Examples of such resources include wetlands, the multicriteria method may be used to manage global climate, wilderness areas, Antarctica, and natural resources in a sustainable manner. The other areas with unique functional or even aes- primary role of such constraints is to ensure thetic values. that the trajectory generated through the para- Thisapproach differsfrom standard economic metricoptimizationof an imperfectlycontrolled approaches that require valuations of resources and imperfectly understood system does not and the use of economic incentives to achieve also threaten global stability. In other words, efficient resourceallocation (see thediscussion of ecological sustainability constraints can be ecological sustainabilityconstraint asdetailed by thought of as precautionary constraints; the Perrings 1991). Themethod placesgreatersignifi- limits they impose on economic activity will cance on irreversible damage to the ecosystem as depend on the local stability of the ecosystems opposed to short-run economic sacrifices experi- involved and on the projected losses if these encedasaresultofmeasurestocurbenvironmcn- ecosystems become unstable due to the effects tal impacts. The approach could be relatively of economic activity. Intragenerational issues equitable, if environmental safeguards are deter- such as poverty also will continue to be given mined by judgments that reflect societal values. high priorityin the sustainabledevelopment pro- More specifically, a better understanding of the cess. More generally, multicriteria tradeoffs will economic costs and benefits of environmental be required among economic efficiency, environ- impacts would help to establish better safe mini- mental degradation, and poverty reduction. mum standards. 34 Key Concepts and Terminology of Sustainable Development Figure 2-7: Safe Minimum Standard for Balancing Natural Resource Tradeoffs and Imperatives for Preservation Ecological and human catastrophe Moral inperatives for resource and ecosystem protection Increasing cost Free play of of ecological individual incenhves damages and resource tradeoffs < Increasing irreversibility Low cost, easily of ecological damages reversed effects Source: Bryan Norton, Georgia Institute of Technology. How such standards would be established and This point and others will have tobe dealt with as enforced, howevcr, has been debated at length. the process of implementation proceeds and our The anticipated agreement on global warming base of scientific knowledge expands. The role of will provide an opportunity to observe the appli- ecologists and economists is to present the maxi- cability of the standards approach in a practical mum possible information to the general public context. In view of the uncertainty underlying and to decisionmakers, so that costs and benefits global warming models, the "precautionary prin- are carefully weighed before a decision is made. ciple" is invoked whereby relatively costless steps In order to establish a valid decisionmnaking mightbetakenimmediatelyasaninsurancepolicy framework within which constraints could be to avoid very large (but uncertain) costs in the determined, it is necessary to put a value on future. Guidelines and objectives that might func- resources (Munasinghe 1992a, 1992b). Conceptu- tion as safe minimum standards are discussed in ally, the total economic value (TEV) of a resource appendix 2.1 aoconsists of its (a) use value (WV) and (b) nonuse appendix 2-1, along with examples of snterna- value (NWJV). Use values may be broken down tional proposals for the implementation of sus- further into the direct use value (DUV), the indi- tainable resource management. rect use value (IUV), and the option value (OV, potential use value). One needs to be careful not to double-count both the value of indirect sup- Valuation of environmental assets porting functions and the value of the resulting and impacts direct use (for a discusswon and example of this, see Aylward and Barbier 1992). The categories of Numerous questions concern different ap- nonuse value are existence value (EV) and bequest proaches to implementing sustainable develop- value (BV). Therefore, we may write: ment. One such question is the meaning of opti- mization (see Schneider 1990). Does optimization TEV = UV + NWV or entail the maintenance of maximum biomass or TEV = (DUV + IWV + OV) + (EV + BV) the maintenance of maximum diversity of spe- cies? Does optimization mean themaintenanceof Figure 2-8 shows this disaggregation of TEV in stability for the longest period of time or the schematicform.2 Beloweachvaluationconcept,a maintenance of maximum productivity of extant shortdescriptionof itsmeaning,and a fewtypical species? Since evolutionary change is a dynamic examples of the environmental resources under- process, at what point is the optimum achieved? lying the perceived value are provided. Option 35 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 2-8: Components of Total Economic Value Total economic value Use value Nonuse value Direct Indirect Option Bequest Existence use value use value value value value I I i I I Value of leaving Value frm Outputmed Functional and direct use and nonuse knowledge be consused benefits values for ofcontnued directly use values offspring existence * Food * Ecological * Biodiversity * Habitats * Habitats * Biomass fuinctons * Conserved * Irreverstble - Endangered * Receation * Flood control habitats changes speces Decresirng tangibilih of value to individuaLs Source: Munasinghe 1992a values, bequest values, and existence values are table 2-3, valuation methods can be categorized, shaded, to caution the analyst about some of the on the one hand, according to which type of ambiguities associated with defining these con- market they rely on and, on the other hand, by cepts; as shown in the examples, they can spring how they make use of actual or potential behav- from similar or identical resources, while their ior. These valuation techniques and other issues estimation could be interlinked also. However, relating to cost-benefit analysis are discussed in these concepts of value are generally quite dis- greater detail in appendix 2-2. tinct. Option value is based on how much indi- Only a small fraction of living species have viduals are willing to pay today for the option of direct economic importance in terms of commer- preserving the asset for future (personal) direct cial value. Other species are used for hunting and and indirect use. Bequest value, while excluding fishing,andsomespeciessuchasthegiantpanda individuals' own use values, is the value that are valued for their aesthetic value alone. Eco- people derive from knowing that others (perhaps nomic techniques havebeen developed that place their own offspring) will be able to benefit from a monetary value on use and nonuse values for the resource in the future. Finally, existence value independent species. Orians (1990) suggests a is the perceived value of the environmental asset tradeoff mechanism involving the use of "valued unrelated either to current or to optional use, that ecosystem components" (VECs). A VEC could be is, simply because it exists (see, for example, a single species of economic or aesthetic value, Randall and Stoll 1983). systems of interacting species, or an entire eco- A variety of valuation techniques may be used system. The theory and practice of the exploita- to quantify these concepts of value. The basic tion of single species is the most intensively stud- concept of economic valuation underlying all ied aspect of applied ecology (Orians 1990). The these techniques is the willingness to pay of indi- determnination of optimal harvesting rates can viduals for an environmental service or resource, usually be accomplished if basic demographic that is, thearea under the compensated orHicksian requirements and habitat requirements are demand curve (for an up-to-date exposition, see known. However, populations that have been Braden and Kolstad 1991, chap. 2). As shown in seriously reduced in number may fall below a 36 Key Concepts and Terminology of Sustainable Development Table 2-3: Taxonomy of Relevant Valuation Techniques Basis Conventional market Implicit market Constructed market Based on actual behavior Effect on production Travel cost Artifidal market Effect on health Wage differences Defensive or preventive cost Property values Surrogate goods Based on potential behavior Replacement cost Contingent valuation Shadow project Source: Munasinghe 1993. threshold of sustainability, which will not allow unsustainable rate. Therefore, the sustainable use the population to regenerate to pre-exploitation of a control agent (whether it is a toxic chemical, levels. This is because the interactions among a biological control agent, or a hunter) is the use species can be altered to the extent that a new at which the agent can continue to achieve its alternative stable state of the ecosystem is estab- desired effect (Orians 1990). lished in which the interrelationships are so al- Another instance where control agents may tered as to make returning to the original state have a counterproductive effect is related to the impossible (see Orians 1990; Perrings 1991). This mutualistic interactions between species. In cases is the case with commercial overfishing. where such interdependent relationships exist, In somecasestradeoffsbetweendifferenttypes sustainable use requires maintenance of such re- of value must be assessed to determine lationships. Thus, excessive use of pesticides on sustainabilityofuse.Forexample,dolphinscaught fruit trees is incompatible with pollination and, in fish nets have a purely aesthetic value, but the hence, the production of fruit. tuna harvest has a purely economic value. Thus, Speciesrichness-thesumtotalofspeciesinan relative weights would have to be assigned to the ecosystem-can be a VEC and is recognized as values in order to determine policy. such in the U.S. Endangered Species Act, which Managers of natural resources cannot ignore states that maintaining species richness is a soci- relationships between predator and prey or is- etal goal (Orians 1990). The protection of species sues related to appropriate scale when they aim richnesscan generally (theoretically) beachieved forsustainedyields.Forexample,harvestingprac- through efficient management of land use. But tices of balsam fir in New Brunswick have created this is difficult because, as mentioned earlier, age distributions and stand densities that are decisions about land use are usually made on a favorable for rapid growth of the spruce bud- smaller scale than that which is required to maxi- worm population, which defoliatesand kills trees mize survival of the species. Moreover, protected on a massive scale. Managing and making deci- areas only cover approximately 5 percent of the sions on scales that are too small does not allow terrestrial landscape, while the demand for land for the maximum sustainable yield to be pro- continues to grow; it is unlikely that society will duced if conditions suitable for outbreaks of bud- be prepared to preserve all species at the cost of worms persist (Orians 1990). Furthermore, one alternative land uses. must not forget that interactions between preda- According to the 1992 World Resources Report, tor and prey also have important evolutionary cropland covers 11.2 percent (having grown 2.2 componentsaffectingsustainability. Forexample, percent in a decade, while the world population attempts to reduce populations of certain prey increased 20 percent); permanent pasture covers (such as insects that are pests) usually create 25 percent, having increased 0.1 percent in the last mortality patterns that do not mimic the stochas- decade; and forest and woodland covers 31 per- ticbehaviorofthetargetpopulation,inparticular cent, having increased 1.8 percent in a decade with regard to prey genotypes. An example of (World Resources Institute 1992b). "Other land" this is the evolution in North America of agricul- totals 32 percent, having increased 1 percent in tural pests better able to withstand a given mor- the last decade, and wilderness areas cover 26 tality agent because farmers use pesticides at an percent (including the 5 percent that is legally 37 Defining and Measuring Sustainability: The Biogeophysical Foundations protected). Many of these land uses are entirely Multicriteria analysis consistent with the preservation of biodiversity, and the implied conflict between cropland and In the multicriteria (or multiobjective) approach, other uses is perhaps not really a struggle be- desirable objectives need to be specified. These tween agriculture and alternative uses. First, ag- often exhibit a hierarchical structure. The highest ricultural land also contains important biological level, representing the broad overall objectives diversity; second, many forest lands harbor sig- (such as improving the quality of life), is often nificant biodiversity even when they are used vaguely stated and therefore has limited opera- primarily to produce timber; and third, most tional function. Some of these, however, can be increases in agricultural productivity are, in any broken down into more operational lower-level case, going to come from intensification rather objectives (such as increasing income), so that the than expansion onto marginal lands. Finally, as extent to which the latter are met may be practi- many ecologists have pointed out, the optimal cally assessed. Sometimes only proxies are avail- use of areas not suitable for permanent agricul- able (for example, if the objective is to enhance ture is to ensure that sufficient land exists to recreational opportunities, the attribute number enable considerable benefits to be earned for so- of recreation days can be used). Although value ciety, in terms of harvested goods, specific ser- judgments may be required to choose the proper vices(suchaswatershed protectionand tourism), attribute (especially if proxies are involved), in and more abstract services (such as conserving contrast to thesingle-criterionmethodologiesused biodiversity and maintaining evolutionary po- in economic cost-benefit analysis, measurement tential). But if conflict over land use does occur, does not have to be in monetary terms. More improved economic valuation and multicriteria explicit recognition is given to the fact that a analysis must be used to help decisionmakers variety of concerns may be associated with plan- determine which species are more highly valued, ning decisions. given a set of criteria deemed important. An intuitive understanding of the fundamen- A key consideration to keep in mind is that tals of multiobjective decisionmaking can be pro- species may have divergent values for different vided by a two-dimensional graphical exposition groups, depending on income level and whether such as that presented in figure 2-9. Assume that private or social costs and benefits are being a project has two noncommensurable and con- assessed. For example, a living snow leopard may flicting objectives, Z, and Z2. Assume further that have high aesthetic value to individuals in an alternative projects or solutions to the problem affluent society, even though they may never be (A, B, and C) have been identified. Clearly, point able to see the animal. To an individual poacher survivingat subsistence level, however, thevalue of the pelt far outweighs social and ethical consid- erations. The solution to this dilemma is to en- Figure 2-9: Pareto Optimal Curve and force strict regulations to preserve such species, Isopreference Curves while providing poachers alternate and viable ways of generating income. Information cam- paigns to alert consumers to the true costs in- Increasing curred by their purchases (potential extinction of E preference the species) have also proven successful to a certain extent (Pearce 1991). > The methods described in appendix 2-2 seek to , estimate costs and benefits of a given project in g Equi- monetary terms. As previously stated, when Pareto prfcurVen projects or policies and their impacts are to be optimal integrated into a system of broader (national) curve objectives, some of which cannot be easily quan- tified in monetary terms, multicriteria decisionmakingoffersa supplementaryapproach Objective 2 that may facilitate the optimal choice among in- vestment options or available policies. 38 Key Concepts and Terminology of Sustainable Development B is superior to (or dominates) A in terms of both has limited validity. The applications (studied in Z1 and Z2- Thus, alternative A may be discarded. developing and industrial countries) pay insuffi- However, we cannot make such a simple choice cient attention to key methodological require- between solutions B and C, since the former is ments and are generally inconsistent. Problems better than the latter with respect to objective Z2 also exist in the definition of attributes (for ex- but worse with respect to Z,. In general, more ample, the environmental consequences of land points (or solutions) such as B and C may be use are not clearly expressed). identified to define the set of all nondominated The authors suggest that it would be prefer- feasible solution points that form a Pareto opti- able to work with value or utility functions rather mal curve (or curve of best options). This line is than purely physical scales for measuring at- also called a transformation curve or efficient tributes. Moreover, one of the most useful tools of frontier. multiattribute analysis is the tradeoff curve, which For an unconstrained problem, further rank- displays available options and their impacts in ing of alternatives cannot be conducted without terms of two objectives. introducing value judgments. Specific informa- Orians (1990) outlinesa multicriteria approach tion has to be elicited from the decisionmaker to to decisionmaking, whereby the sustainable use determinethepreferredsolution.Initsmostcom- of a particular ecosystem is defined by how the pleteform, such informationmaybe summarized differentVECsareranked and weighted. AsVECs by a family of equi-preference curves that indi- conflict, an optimal solution requires that how cate the way in which the decisionmaker trades specific systems are used is balanced with pat- off one objective against the other. The preferred terns of how uses are allocated among different alternative is that which results in the greatest systems. Such values are not constant and can utility; that is, which occurs (for continuous deci- change significantly over time. As such, some sion variables as shown here) at the point of nonsustainable development is unavoidable-it tangency D of the highest equi-preference curve, is, in fact, more likely to occur when rates of with the Pareto optimal curve. In this case, point critical processes remain unknown, whenbound- E (on an even higher equi-preference curve) is not aries between domains of a system's stability are attainable. not appreciated, or when ecosystem components Several multicriteria methods have been de- are undervalued. Such nonsustainable develop- veloped (for an introductory overview relevant mentisalso favored by short-termplanninghori- tonaturalresourceanalysis,seeMunasinghe1993; zons, small spatial frames, and economic incen- for an extensive survey including references to tives for overuse. Such uncertainties and short- about 150 applications, see Romero and Rehman term perspectives call for a precautionary ap- 1987; for a shorter but more recent survey, see proach, in which allowance is made for the limi- Petry 1990). Which practical method in particular tations of current knowledge, for risk, and for is suitable to determine the best alternative avail- changing values. A principal cause of able depends on the nature of the decision. For nonsustainable development is that private ben- instance, interactive involvement of the efits accrued promote overexploitation at a high decisionmaker has proved useful in the case of social cost. In such situations, thesolutionmay lie problems characterized by a large number of in combining a set of regulations governing re- decision variables and complex causal interrela- source use with an alteration of economic incen- tionships. Some objectives can be dealt with tives to internalize societal costs. throughdirectoptimization, whileothersrequire The major accomplishment of multiobjective the satisfaction of a certain standard (such as a decision models is that they allow for more accu- level of biological oxygen demand no lower than rate representation of decision problems, in the 5 milligrams per liter). sense that several objectives can be accounted for. A recent applied research study in developing However, a key question concerns whose prefer- countries, performed by Meier and Munasinghe ences are to be considered. The model only aids a (1992), examines the practicality of using single decisionmaker (or a homogeneous group). multiattributedecisionanalysisin thesettingofa Various interested groups often assign different developing country. The study determined that priorities to the respective objectives, and nor- the use of multiattribute scoring systems, in which mally it may not be possible to determine a single nonprice attributes are accounted for separately, best solution via the multiobjective model. Also, 39 Defining and Measuring Sustainability: The Biogeophysical Foundations the mathematical framework imposes constraints record the depreciation of natural capital, and (3) on the ability to represent the planning problem excluding environmental damages while includ- effectively. Nonlinear, stochastic, and dynamic ing clean-up costs in the national income. formulationscanassistinbetterdefiningtheprob- There has been some discussion in the litera- lembutimposecostsbecauseofthecomplexityof ture as to how user costs (revenues generated by formulating the problem and then solving the nonrenewable resources) should be invested in model (Cocklin 1989). an asset to produce a future stream of income Nevertheless, in constructing the model, the equal to that which is lost through consumption. analyst communicates information about the na- The more rigorous proponents of sustainable ture of the problem, specifying why factors are development would argue that the user cost important and how they interact. Liebman (1976) should be reinvested in an asset that not only observes that modeling is thinking made public replaces income lost but also acts as a physical and considers the transfer of knowledge to repre- substitute for the nonrenewable resource. Re- sent perhaps the most important contribution of gardlessof whichmethod isemployed, theimple- modeling. With respect to the second point of mentation of a user cost adjustment in national criticism (diverse preferences), Liebman suggests accounts would generally result in GNP being that there is value in constructing models from revised downward (Moseley 1992). differing perspectives and comparing the results. As previously stated, ecologists argue that the major threat to biodiversity is not direct human exploitation of species, but the alteration and Economic approaches at the macro level destruction of habitat that result from changesin land use, urbanization, infrastructure develop- Aspolicymakersstarttousesomeoftheconcepts ment, discharge of varfous pollutants into the of sustainabledevelopment, theyhavefound that environment, and so forth. Human population the traditional use of GNP as a macroeconomic growth is the most frequently cited cause of the indicator of growth does not guarantee environ- anthropogenic stressors on the environment, but mentally benign growth. Thus economists have the stability of ecological systems-of which hu- increasingly attempted to include a variety of mans are an integral part and on which our con- natural resources in their calculations of national tinued existence depends-does not imply a cer- products and incomes. Current national account- tain equilibrium level of human population ing systems do not capture the value of natural growth as some groups would advocate. How- resources adequately, and, therefore, develop- ever, the continued ability of Earth's ecosystems ment strategies that rely on standard income ac- to maintain stability will depend on the level of counting techniques do not result in sustainable stress to which they are subjected. Therefore, development. A significant first step in the imple- reductionists, in the interest of limiting the fur- mentation of sustainable development is to meld ther degradation of physical and biological re- thetraditional accounting techniques into a frame- source systems, need to work toward understand- work that permits the computation of measures ing and addressing the incentives that underlie such as an environmentally adjusted net domes- human population trends. This will involve the tic product (EDP) and an environmentally ad- accelerated implementation of projects that ad- justed net income (ENI). Some of the effects of dress rural poverty, health, education, land ten- unsustainable development such as resource ure, and the unequal distribution of income and degradation, pollution, and poor waste-disposal assets that is prevalent in most developing coun- practices, together with their repercussions on tries. If policymakers and social scientists con- society, cannot be captured by market-based in- tinue to believe that social ignorance and institu- formation and standard accounting techniques. tional irresponsibility are solely to blame for the Correction of three primary shortcomings of cur- current trend of world population growth, we rent accountingwill move the implementation of will not come any closer to resolving the funda- sustainable resource management practices for- mental issues (Perrings, Folke, and Maler 1992). ward because the new approach stresses the long- The successful implementation of sustainable run maximization of EDP. These shortcomings development will require price reform (at the are (1) excluding natural and environmental re- globallevel)ofenvironmentalresourcesforwhich sources from the balance sheets, (2) failing to effective markets can be created. Conducted in 40 Key Concepts and Terminology of Sustainable Development the name of self-sufficiency and national defense, * Design and adopt accounting systems that give government intervention, particularly in mar- appropriate economic value to natural re- kets for agricultural products and other tradable sources and establish economic pricing based natural resources such as timber and minerals, on the polluter-pays principle. has driven prices at the national level below those * Promote conservation action through intema- ruling in the world market. This widens the gap tional cooperation and national planning. between private and social value, which perpetu- *Maintain representative examples of the full ates unsustainable development at the local and global level. The reform (ideally, the elimination) spectrum of tems biological cemmeni of fiscal policies such as subsidies (for clearing ties, habitats, and their ecological processes. land in agriculture and forestry, unsustainable * Increase scientific understanding of natural systemsof agricultural pricesupport,and so forth) resources and apply that understanding to that encourage environmental degradation their more efficient management. through the farming of marginal land should, one * Expand research into thebiophysical resources hopes, result in the more socially optimal and as a basis for improving management. sustainable use of natural resources (Perrings, * Develop indicators of social sustainability and Folke, and Maler 1992). incorporatethemintooverall indicatorsof sus- tainable development. * Give full consideration to issues of cultural and local levels diversity when designing and implementing projects. The following is a more detailed set of guidelines and objectives that will move the world closer to Regional and national objectives its goal of sustainable resource management. These are by no means the only approach to * Implement management procedures at the sustainable practice, and individual approaches landscape level in order to integrate human will depend on one's values, beliefs, and current activities with conserving natural resources. and long-term goals. As previously stated, the United Nations Environment Program, Intema- * Reduce economic incentives that promote the tional Union for the Conservation of Nature, wastefuluseofnaturalresourcesandestablish World Wildlife Fund, and World Resources Insti- prices that reflect full economic costs (includ- tute have recently published lists of objectives inginternalizationofenvironmentalimpacts). and management regimes for responsible devel- * Strengthen the management tools for conserv- opriient and the protection of biodiversity; the ing natural resources and apply them more following list includes some of their conclusions. broadly. As numerous scientists have pointed out, interac- * Greatly strengthen the humanand institutional tions among the different goals change as the scale of the system is extended from the local, to capaciy foreconservin deuoingbiodei the regional, to the national level, and so forth. strinaby y in devloveingcn- The behavior of more complex systems in such a tries, by using the full range of government hierarchy cannot be studied in isolation. Thechoice agencies, nongovernmental organizations,and of sustainable development goals to be pursued the private sector. will differ from level to level. The followingbreak- * Establish sufficient protected areas containing down is therefore presented in the form of global, viable populations of the target species and regional, and national objectives (adapted from design activities so that they take place prima- Holmberg 1992). rily outside the protected areas. * Maintain viable populations of the nation's G lobalI objectives native plants and animals, well distributed . Develop national and international policy throughout their geographic range. frameworks to foster the sustainable use of * Maintaingeneticvariabilitywithinandamong natural resources and the maintenance of populationsof nativespecies,tomaintaintheir biodiversity. evolutionary potential. 41 Defining and Measuring Sustainability: The Biogeophysical Foundations Local objectives Sustainability indicators * Promote awareness and understanding of the Thedevelopmentof sustainability indicatorsrec- importance of natural resources and ognizes the need to monitor human impacts on biodiversity among the public (especially ur- theenvironmentandalsoattemptstorelatelevels ban dwellers), decisionmakers, and politicians. of human use to a reproducible indicator. Both * Implementpolicies,incentives,andconditions the analytical definition and practical measure- that will enable rural populations, including ment of such indicators are complicated. It is those living in and around protected areas, to conceptually easier to deal with sustainability continue using their natural resources in sus- indicators at larger levels of aggregation. The rest tainable ways. of this section briefly reviews broadly defined * Introduce conservative levels of harvesting indicators, as an introduction to the topic. biologicadon evaluation of Some recent attempts have been made to de- biologi r 'esrcies a ed vise ways to measure and monitor sustainability. thetargetspeces andcontinuedmonitoringto Shearer (1992) proposes a Biophysical assess the impact of use. Sustainability Index, on the basis that all life on * Establish management structures and legal Earth is sustained by the net primary biological frameworks-within the local communities production generated byphotosynthesis. Moseley whenever possible-with the capacity to con- (1992) provides examples of sustainability indi- trol levels of use. cators for five sectors: greenhouse gases, agricul- * Institute education programs that improve ture, freshwater resources, forestry, and energy. local-level management of natural resources. Shearer observes that the factors necessary to Governments are already starting to follow maintain biological production include fertility and provide backing for some of these guidelines or availability of nutrients, energy, adequate and management regimes. A new Green Fund- moisture, proper substrates, minimum level of the Global Environment Facility-has been cre- toxi. substances, and anadequateand genetically ated under the management of the World Bank, vaned stock of biological organisms. He uses United Nations Environment Program, and In- these factstodevelopaBiophysicalSustainability ternationalUnionforthe ConservationofNature Index (BSI), which he suggests can be used to and is channeling several hundred million dol- monitor policy rather than to replace scientific lars into biod iversi ty projects in developi ng coun- measures of sustainability. tries. A new Biodiversity Convention, after sev- .The Biophysical Sustainability Index is com- eral years of negotiation, was adopted at the Earth prised of a netprimary productivity factor,NPPF Summit in Rio de Janeiro, Brazil, in June 1992. (which reflects more of the economic concerns) If we focus on the biodiversity issues, several and a biological diversity factor, BDF (which agreements have been, or are in the process of represents the ecological aspects); such that, being, completed at the intemational level. These BSI = NPPF x BDF. include the Global Biodiversity Strategy, pre- The NPPF is defined as the ratio of theannual net pared by theWorld Resourceslnstitute, the Inter- primaryproduction, ANPP, of the region over a national Union for the Conservation of Nature, given year, y, to that of the same region over the and the United Nations Environment Program, previous year, y-. The ANPP of a given year is in consultation with the Food and Agriculture defined as Organization and UNESCO (the United Nations ANPP = PPA + PPP - HPCP Education,Scientific,and Cultural Organization); where PPA is the primary production of annuals, the Draft Convention on Biological Diversity cur- PPP is the primary production of perennials, and rently being prepared by United Nations Envi- HPCP is the harvested primary capital of perenni- ronment Program; and the guidelines prepared als. The data needed to calculate the ANPP is ob- by the Scientific and Technical Advisory Panel tainable through satellite imagery and aerial pho- (STAP) for the Global Environmental Facility tography, using a geographic information system. (GEF) regarding criteria for eligibility and priori- The BDF of a region is defined as the ratio of the ties for selection of GEF projects (many of which current selectedbiological diversity,CSBD,of the are directly linked to the preservation of region (the number of species of a set of taxa biodiversity).Detailsareprovidedinappendix2-1. thriving in the region) to the natural selected 42 Key Concepts and Terminology of Sustainable Development biological diversity, NSBD, of the same region the highest level of emissions in that year. Moseley (the number of species of the same set of taxa arguesthatthemostequitablewayofapportioning thriving in the region prior to human interven- the sink would be through a combination of popu- tion). The set of taxa selected is chosen on the lation and land area criteria. basis of those that are regularly monitored, those The Intergovernmental Panel on QimateChange that are reasonably well known, those that are hasdevisedaconceptualunit-theglobal warming mostly immobile (for ease of monitoring), and potential-that allows for comparisons between those that are conspicuous. different greenhouse gasesbased on theircontribu- Essentially, the proposed BSI has the following tion to global warming. The figure isdetermined by form: integrating an expression for the removal rate times BSI = (ANPP / ANPP ) anexpressionfortheinfraredabsorptionpotencyof (CSBD / NSBD) the gas. A global warming potential output quota is Moseley (C992 present N somewhatdiffere based on (1) a quota for the world derived from the Moseley (1992) presentsuaisomewhat different sinks for CO2 and methane and (2) a quota for approach to measuring sustainability, using the individual countries dependent on the average of Bru nd tland Commission's definition of sustain- their percentageof world populationand land mass. able developmentoas-arguidingaprinciple touexam- The difficulty inherent in this method is that it ine five sectors-greenhouse gases, agriculture, assumes a substitutability between the sinks of freshwater resources, forestry, and energy-in differentgreenhousegases. However, unless thresh- terms of their potential as sustainability indica- old enission levels are reached, a certain amount of tors at the national level. substitution at the abstract level is perrmssible. GREENHOUSE GASES AGRICULTURE Globalsinkcapacity(forincreasedernissionsdueto A viable indicator of sustainable agriculture is the human activity) is defined as the difference be- preservation of the integrity of agricultural soils. A tween the massof anthropogenic gas ernlssionsand key measure of productivity that can be examnined the annual change in atmospheric mass of that gas. is the ratio of crop production to fertilizerconsump- Moseley discusses different means of determrining tion,whichindicateshowlandrespondsto fertilizer sink capacity rights on a national basis. subsidies over time. From a sustainability point of If population is the determninant, then nations view, the ratio should remiain constant or increase that encourage population growth would be re- over time. Decreasing ratios point to (1) fertilizer warded. Another method of apportionrnent would use that has increased to the point of dimninishing be allocation based on a country's relative land retumsor(2)theinsufficientreplacementoforganic mass. However, since land has different levels of matterthrough fertilizer inputs, leading to soil deple- carrying capacity, this would also be an inequitable tion. As shown in table 2-4, the ratio of crop produc- form of distribution. Apportionment based on 1989 tion to fertilizer use has declined drastically over levels of emissions wvould reward countries with the last decade. Table 2-4: Ratio of Crop Production to Fertilizer Use Changes in irrigated land as a % of cropland between 1978 and 1988 Year 1980 1985 1986 1987 1988 (percent) United States 27 43 43 42 33 0 Sweden 21 26 28 27 26 0 Japan 13 13 13 12 12 -5 Thailand 119 100 86 59 67 7 Nepal 169 87 73 84 74 19 Kenya 43 32 33 32 30 0 Mali 106 71 55 60 115 3 Mexico 43 35 31 30 30 1 Ecuador 55 61 47 51 42 2 Source: World Agricultural Trends and Indicators, 1970-88, USDA, 1989. 43 Defining and Measuring Sustainability: The Biogeophysical Foundations FRESHWATER RESOURCES able development may be generally interpreted Moseley identifies two potential indicators of to mean the maintenance or improvement of the sustainability for freshwater reources. The first quality of life on a continuing basis. There are measuresthelevel of humanexploitation of fresh- three main interpretations of sustainabilitybased waterresources(domesticand externalriverflows on different disciplinary approaches: economic, and groundwater generated from endogenous ecological/biophysical, and sociocultural. precipitation). A measure that better examines Historically, the rise and decline of past cul- the sustainability of water use practices is to tures may often be traced to unsustainable prac- identify the percentage of population served by tices,especiallydegradationofsomevitalnatural wastewater treatment facilities. Countries with resource. Nevertheless, human society as a whole high levels of water use should ideally have cor- has increased in complexity and organization respondingly high levels of wastewater treat- over the ages, and the future sustainabilityof this ment capacity. Although these are useful mea- evolutionary process will depend on the search sures of sustainable practice, data for developing for paths of long-run sustainabledevelopment. In countries are not easily available. modem times, the pressure of human activities has caused environmental damage on a scale not FORESTRY experienced before. High levels of consumption Theconceptofsustainableyield (theamountofa per capita in some parts of the world are one resource that can be harvested without reducing reason. its long-term stock; see Pearceand Turner 1990) is The rapidly growing populations of the poorer more readily applicable to fisheries than forestry, countries will continue to add to the environmen- because tree populations have a longer period of tal burden as development proceeds, especially if regeneration. However, environmental indica- the material- and pollution-intensive patterns of tors for temperate forests have been prepared for past growth (experienced by the industrial na- member nations of the Organization for Eco- tions) continue to be followed. Poverty and envi- nomic Cooperation and Development (Moseley ronmental degradation are closely linked, be- 1992). Annual harvest was compared to annual cause the majority of the global population who increment, in the form of a ratio. However, aggre- are pooralso tend to be the most severely affected gate figures at the national level can be mislead- by pollution and the degradation of natural re- ing. Although the United States is losing overall sources.Therefore,sustainabledevelopmentsug- forestcover, annual increment is increasingbecause gests that future progress will have to focus more old forestsarebeingreplaced by younger ones with on the quality of human life and less on the higher growth rates. Although the production of quantity of resources consumed, although the wood increases, wildlife habitats and recreational two aspects are related, especially for low-in- areas are lost as a consequence of this policy. come groups. The economic approach to sustainability is ENERGY based on the Hicks-Lindahl concept of the maxi- By examining consumption of commercial en- mum flow of income that can be generated while ergyperconstant U.S. dollarof GNP, theproduc- at least maintaining (or increasing) the stock of tion of renewable energy and the exploitation of assets (or capital) that yield these benefits. Prob- renewable resources, it is possible to observe if a lems of interpretation arise in identifying the country is moving towards sustainable use of kindsof capital tobemaintained and theirsubsti- energy (Moseley 1992). Measures to increase en- tutability (man-made, natural, human resources, ergy efficiency are another key indicator of sus- and so forth) as well as in valuing these assets, tainable use and are particularly important be- particularly ecological resources. The issues of causetheyallowforeconomicgrowthwhilekeep- uncertainty, irreversibility, and catastrophic col- ing the quantity of throughputs constant. lapse also pose difficulties. The ecological view of sustainability focuses on the stability of the biophysical system. Of Conclusions particular importance is the viability of sub- s,ystems(species,biotic components) that arecriti- The dictionary definitions of the words sustain cal to theglobal stabilityof theoverall 2cosystem. and develop suggest that the expression sustain- Protection of biological diversity is a key aspect. 44 Key Concepts and Terminology of Sustainable Development Furthermore, natural systems maybe interpreted ability between resources and technological to include all aspects of the biosphere (including progress. Another requirement is that waste be primarily man-made environments like cities), generated at rates less than or equal to the assimi- with emphasis placed on preserving their resil- lative capacity of the environment (to preserve ience and dynamic ability to adapt to change, resiliency) and that efforts be made to protect rather than conservation of some ideal static state. equity within and between generations. Finally, The sociocultural concept of sustainability seeks the implementation of sustainable development to maintain the stability of social and cultural will require a pluralistic and consultative social systems, including the reduction of destructive framework that, inter alia, facilitates theexchange conflicts.Bothintragenerationalequity(especially of information between dominant and hitherto elimination of poverty) and intergenerational disregarded groups-to identify less material- equity (protection of the rights of future genera- and pollution-intensive (qualitative rather than tions) are important elements of this approach. quantitative) paths for human progress. Preservationofcultural diversity across theglobe, In addition to thebroad recommendations that and the better use of knowledge concerning sus- would be helpful at the global, national, and local tainable practices inherent in many indigenous levels, certain specific short-term actions will meet communities, should be pursued. Modem soci- allcriteriasimultaneously.Forexample,improve- ety must encourage and harness pluralism and ments in access to clean drinking water and sani- grass-roots participation in order to engender a tation will increase productivity (economic), re- more effective framework for making decisions. duce environmental pollution (ecological), and Reconciling these concepts and operationalizing benefit the poor (social). In the medium and long them as a means to achieve sustainable develop- term, other project and policy options will re- ment are formidable tasks. The diversity of imme- quire tradeoffs among the different objectives. diate needs and concerns, as well as long-term Better techniques of valuing the environmental goals throughout the world, suggests that there is and social impactsof humanactivities will helpto no universal right or wrong approach to sustain- improve decisionmaking. In certain cases, the able development. To a village-level official in the safeminimum standardsapproach could be help- developing world, sustainable development may ful, especially when uncertainty and irreversibil- mean, first and foremost, dealing with poverty ity are involved. Multicriteria analysis is another and human misery through health and education supplementary technique to help policymakers services. Meanwhile, to a government regulator make difficult tradeoffs. in an industrial country, it may mean protecting Since approaching sustainable development the physical environment from emissions of both from the perspective of only one discipline has greenhouse gases and other effluents that cause significant shortcomings, it will be necessary for acid rain by developing the right economic incen- biologists, physical and social scientists, and tives and monitoring compliance with point dis- economists to work together. This will permit charge regulations. The existence of diverse per- decisionmakers and society to makebetterchoices ceptions and the existing structure of betweenalternativecourses,basedonthepresen- decisionmaking do not readily allow for consen- tationofrelevantandunbiasedinformation.There sus building, and the variety of problems at the is urgent need for continued research by microlevel does not facilitate constructive syn- multidisciplinary teams to improve our under- thesis at the level of macro policy. standing of economic valuation, the functions of One useful practical approach to sustainable ecosystems,theflowofenergy,andthedynamics development that may be more comprehensible of social systems. At the same time, society has to to policymakers and the public might be to maxi- continue with its present evaluation of its funda- mizenetbenefitsof economicand social develop- mental values and to recognize and define the ment, subject to maintaining the services from values that it seeks to promote through different and quality of natural resources over time. This types of development activities and institutional implies that renewableresources (especially scarce arrangements. ones) should be used at rates less than or equal to To conclude, even though public awareness of the natural rate of regeneration and that the effi- the relationship between the environment and ciency with which nonrenewable resources are development has greatly increased, and govern- used should be optimized, subject to substitut- ments are now devoting considerable energy and 45 Defining and Measuring Sustainability: The Biogeophysical Foundations resources to the issue, much more needs to be they willing to alter the terms of trade and done. No country in the world uses its natural reverse economic flows, so that development in environment as well as it could and should. The the south is stimulated rather than continually natural and biophysical assets of the Earth are undermined? Will the south confront its own being wasted, partly because our economic sys- inadequacies in implementing programs to re- tems value only what human societies have cre- duce poverty, internal income disparities, and ated. The recently concluded United Nations population growth? Are governments every- Conference on Environment and Development where willing to permit local communities to was an important watershed and fresh starting have greater influence in planning and manag- point that has generated much momentum. But ing their own resources, where doing so will do our leaders yet have the political courage reduceinequity,exploitation,andwasteofnatu- and will to implement the solutions that are ral resources? And are the leaders of the conser- required? Areconsumers in thenorth willing to vation movement willing to put human needs curb their overuse of energy and emission of at the forefront of their concerns and to insist greenhouse gases that already overtax the resil- that while development must be sustainable, it ience of the global atmospheric commons? Are must also be centered on people? Notes 3. Work is ongoing to identify defensive ex- penditures. Such expenditures by firms are 1. A billion is 1,000 million. treated in the current System of National 2. The various terms in the equation for TEV Accountsasintermediatecostsandarethere- maybe grouped in somewhat different ways fore not part of value added or final output. for convenience; see, for example, Walsh Defensive expenditures by households and Loomis, and Gillman 1984. In order to mea- governments, in contrast, are treated as final sure willingness to pay for wilderness pro- expenditures and included in GDP. Present tection, they sought to separate (a future- research seeks to address this and other is- oriented) preservation value from recreational sues and inconsistencies in the System of use value (in current use). Accordingly, these National Accounts (see Lutz and Munasinghe authors defined preservation value (PV) as 1991). option value plus existence value plus be- quest value, that is, PV = (OV + EV> + BV). 46 Key Concepts and Terminology of Sustainable Development A ppendix 2-1. UNCED convention on PP ~~~~~~~~~~biological diversity (1992) Current Strategies for Key elements of the convention of the United Nations Conference on Environment and Devel- Ensuring Biodiversity opment are that each contracting party shall un- dertake the following: * Develop national strategies (or adapt existing Global biodiversity strategy (1992) strategies) for the conservation and sustain- able use of biological diversity and shall inte- At the international level, the Global Biodiversity grate concerns for biodiversity into relevant Strategy calls for conservingbiodiversity through sectoral or cross-sectoral plans. international cooperation,integratingbiodiversity . Identify and monitor components of biological conservation with international economic policy, diversity over which it exercises sovereign strengthening the international legal framework rights. for conservation to complement the Convention gts on Biological Diversity, integrating biodiversity om the establishentiand srengen conservation constraints into development assis- of national inventory, regulation, or manage- tance, and increasing the funding for biod iversi ty ment and control systems related to biological conservation, with the development of innova- resources. tive, decentralized, and accountable ways to raise a Develop methodologies to undertake sampling funds and spend them effectively. and evaluation on a national basis of the com- At the local level, the strategy recommends ponents of biological diversity and the status reform of land tenure systems and development of ecosystems. of new resource management partnerships be- * Establish a system of in-situ conservation for tween government and local communities, cre- ecosystems and natural habitats. ation of institutional conditions for bioregional conservation and development, provision of in- * Adoptmeasuresfortheex-situconservationof centives to encourage the sustainable use of prod- components of biological diversity where pro- ucts and services from the wild for local benefits; viding in-situ facilities is impracticable or not and adequate use (with appropriate benefits pro- feasible. vided) of local knowledge of genetic resources. * Integrateconsiderationof theconservationand The strategy outlines appropriate actions to sustainable use of biological resources into strengthen the tools and technologies of national decisionmaking, adopt measures re- biodiversity conservation, including identifying lating to the use of biological resources to national and international priorities for strength- avoid or minimize adverse impacts on biologi- ening protected areas and enhancing their role in cal diversity, protect and encourage custom- biodiversity conservation; enhancing the value ary use of biological resources in accordance and improving management plans for protected with traditional cultural practicesthatarecom- areas; strengthening the capacity to conserve spe- cis pouatos an geei diest inntua patible with the requirements of sustainable cisppltosadgntciesti natural use, and support local populations to develop habitats; and strengthening off-site conservation uand support local p ations t develop facilities to conserve biodiversity, educate the and implement remedlal acdion mi degraded public, and contribute to sustainable develop- areas where biological diversity has been ment. reduced. The strategy explicitly recognizes the impor- * Take appropriate measures for the fair and tance of increasing the appreciation and aware- equitable sharing of benefits derived from re- ness of the value of biodiversity. It suggests that search, development,and useof biological and the dissemination of informnation by relevant in- genetic resources between the sourcesof those stitutions be encouraged and calls for the promo- resources and the persons who use them. tion of basic and applied research on biodiversity * Develop economic (including pricing) and so- conservation as well as the development of human g i v g resources capacity for biodiversity conservation. 47 Defining and Measuring Sustainability: The Biogeophysical Foundations biodiversity and the sustainable use of biologi- of the ozone layer. In terms of the protection of cal and genetic resources on private lands. biodiversity, STAP suggests that GEF should * Establish and maintain programs for scientific support projects that are linked to the manage- and technical education, training, and research ment and protection of various ecosystems- inte .dntfcain,cnsrato,and sus- including in-situ as well as ex-situ conserva- tainathe identfbicatgion, cnerv an . .tion, where necessary, in protecting the tamable~~ us,fbooia iest n t on biodiversity of species and of genetic mate- ponents; provide support in such areas to de- riad-andsth Gf should s o regio eao- vlpnconrie.ral-and tha t GEF should support regions, eco- systems, and species that, for reasons of degra- * Promote public education and awareness of dation of habitat, pressure, or threat of extinc- biodiversity conservation and its importance. tion, require immediate attention. STAP places * Introduce appropriate procedures for envi- priority on the allocation of resources for the ronmental assessments of proposed projects protection of biodiversity as follows (and sug- likely to have significant impacts on biological gests that useful techniques and instruments to diversity. implement such programs include legislative * Establish global lists of biogeographic areas techniques, tax policies, subsidy approaches, * Esablsh goba lits o bigeoraphc aeas and land use planning): of importance and species threatened with extinction. * Comprehensive approach to an entire ecosys- * Facilitate the transfer of technology and en- tem throughout its area, including human courage participation in biotechnological re- populations, taking into account species rich- search activities. ness, diversity, and degree of threat. * Produce regularly updated world reports on * Establishment and consolidation of protection biodiversity based on national assessments in areas. all countries. . Promotion of sustainable use of biota. * Education, training, and research. . Inventories. STAP guidelines (1991) Institution-strengthening, including the sup- Projccts eligible for GEF funding accomplish port of scientific communities and the devel- one or a combination of the following aims: the opment of national mechanisms to coordinate reduction of greenhouse gas emissions, the pro- programs for the conservation and sustainable tection of biodiversity, the protection of inter- use of biodiversity. national waters, and the reduction of depletion * Public awareness programs. 48 Key Concepts and Terminology of Sustainable Development death. The loss of potential net earnings (called the Appendix 2-2. human capital technique) is one proxy for foregone output, to which the costs of health care or preven- Cost-Benefit Analysis tion may be added (as a form of replacement/ preventive expenditure). The above measure as- sumes that earnings reflect the value of marginal Given the importance of sustainable develop- product and that medical treatment costs are well ment, it is evident that environmental externali- defined. The method also encounters difficulties ties should be incorporated into the framework of when the cause-effect link between environmental traditional cost-benefit analysis. The following quality and ill-health is unclear, or the sickness is section (mostly derived from Munasinghe 1993) chronic (that is, of long duration). summarizes ways to improve the integration of This technique seeks to avoid ethical controver- natural resource and environmental issues into siesassociated withvaluinga singlelife,attempting economic analyses of projects and policies. The instead to placea valueon thestatistical probability main emphasis is given to methods and ap- of ill-health or death (akin to the actuarial values proaches for valuing environmental effects, is- used by life insurance companies). Moreover, gov- sues relating to the discount rate, and the prob- ernmentsandpublichealthauthoritiesroutinelyset lems raised by risk and uncertainty. priorities and allocate health expenditures that af- fect human well-being. This in turn provides a baseline for determining implicit values placed by Valuation of environmental costs society on various health risks. and benefits DEFENSIVE OR PREVENTIVE COSTS The basic concepts involving both use and nonuse Often, costs may havebeen voluntarily incurred by values of environmental assets were described in communities or individuals to mitigate or undo the the main text. The various valuation techniques damage caused by an adverse environmental im- surnmarzed in table 2-3 are discussed below in pact. For example, if the drinking water is polluted, greater detail. extra filtration and/or purifying chemicals may need to be used. Then, such additional defensiveor Direct effects valued on conventionial preventive expenditures (ex-post) could be taken as markets a minimum estimate of the benefits of mitigation. markets The assumption is that the benefits of avoided Methods considered in this section are based on environmentaldegradationatleastexceed thecosts how a change in environmental quality directly of avoidance. Theadvantageof thetechniqueisthat affects actual market-related production. defensive or preventive outlays (already made) are easier to determine than the value of the original EFFECT ON PRODUCInON environmental damage. One weakness is that the An investment decision often has environmental defensive actionsare sometimesdecided upon quite impacts, which in turn affect the quantity, quality, arbitrarily with little reliance on market forces, so orproductioncostsofarangeofproductiveoutputs that the costs bear little relation to the potential that may be valued readily in economic terms. In a environmental benefit. Recently, Harrington and case studyon soil conservation in Lesotho, increased others (1989) evaluated the economic damages of a production from conserved land is estimated (Bojo waterborne disease outbreak, emphasizing that the 1991). In the valuation of a Peruvian rainforest, the valuation of averting behavior requires the estab- values of different production schemes are com- lishment of a relationship between observable de- pared (Peters and others 1989). Another example fensive expenditures, and non-observable willing- includes impacts on tropical wetlands (Barbier and ness to pay. others 1991). Potential expenditure valued on EFFECT ON HEALTH conventional markets Thisapproachisbased on health impactscausedby pollution and environmental degradation. One This section summarizes techniques in which po- practical measure that is relevant is the value of tential or future actions could be valued in conven- human output lost due to ill health or premature tional markets to provide a measure of environ- 49 Defining and Measuring Sustainability: The Biogeophysical Foundations mental degradation,provided there isa highdegree its per year to a park), as a function of variables of certainty that such actions will be undertaken. such as consumer income, price, and various socioeconomic characteristics. The price is usu- REPLACEMENT COST AND SHADOW PROJECT ally the sum of observed cost elements like (a) If an environmental resource that has been im- entry price to the site; (b) costs of travelling to paired is likely to be replaced in the future by the site; and (c) foregone earnings or opportu- another asset that provides equivalent services, nity cost of time spent. The consumer surplus then the costs of replacement may be used as a associated with the demand curve provides an proxy for the environmental damage. This is an estimate of the value of the recreational site in ex-ante measure similar to the (ex-post) defensive question. More sophisticated versions include costs approach. It maybe argued that thebenefits comparisons (using regression analysis) across from the environmental resource should be at sites, where environmental quality is also in- least as valuable as the replacement expenses. cluded as a variable that affects demand (for a The replacement cost approach has been applied detailed survey, see Mendelsohn 1987). Until a to protecting groundwater resources in the Phil- few years ago, most applications of this tech- ippines, by determining the cost of developing nique were to be found in the market econo- alternative water sources (Munasinghe 1990b). mies, but quite recently several examples have A shadow project is usually designed specifi- emerged involving developing world applica- cally to offset the environmental damage caused tions. The travel cost for domestic trips to a byanotherproject.Thecostoftheshadowproject forest is estimated in a Costa Rica case study reflects an institutional judgment on the value of (Tobiasand Mendelsohn 1991). Inanother study environmental assets that are thereby restored. on the value of elephants in Kenya, the travel Theapproach hasbeen discussed in the context of cost of tourists from Europe and North America project-level sustainability. The original project is used to estimate consumer surplus (Brown and shadow project together form a sustainable and Henry 1989). package that helps to maintain undiminished some vital stock of environmental resources. For PROPERTY VALUE example, if the original project was a dam that In areas where relatively competitive markets inundated some forest land, then the shadow exist for land, it is possible to decompose real project might involve the replanting of an equiva- estate prices into components attributable to dif- lent area of forest elsewhere. Often, the equiva- ferent characteristics such as house and lot size, lency criterion is hard to satisfy exactly-in the proximity to schools, shops and parks, etc. above example, the two tracts of forest may have (Cropper and Oates 1992). To value an environ- the same volume of biomass, but could differ mental variable like air or water quality, the widely in terms of biodiversity. method seeks to determine that component of the property value attributable to the relevant Valuation using implicit (or surrogate) environmental variable. Thus, the marginal markets willing-to-pay (WTP) for improved local envi- ronmental quality is reflected in the increased Often, relevant market data are not available in priceof housingin cleaner neighborhoods. This directly usable form to value environmental re- method has limited applicability in developing sources. In many such cases, analysis of indirect countries because it requires a well-functioning market data (for example, using statistical and housing market as well as sophisticated informa- econometric methods) permits the valuation to tion and tools of statistical analysis. Jimenez (1983) be carried out implicitly. A variety of such surro- used this technique to explain changes in housing gate market-based methods-including travel pricesin a Manila slum area,upgraded partly due to cost, the "hedonic" methods (property value and water and sanitation service improvements. wage differential), and proxy goods-as well as their applicability under different circumstances, WAGE DIFFERENCES are described below. As in the case of property values, the wage differ- ential method attempts to relate changes in an TRAVEL COST economic price variable (that is, the wage rate) to This method seeks to determine the denmand for environmental conditions. The underlying as- a recreational site (for example, number of vis- sumption is that there is some component of the 50 Key Concepts and Terminology of Sustainable Development wage that is determined by the environmental compensation they would be willing-to-accept pollution or hazard associated with the job or (WTA) if they were deprived of the same re- work site. The technique is relevant when com- source. The contingent valuation method is more petitive labor markets exist, where wages (that effectivewhentherespondentsarefamiliarwiththe reflect the marginal product of labor) equilibrate environmental good or service (for example, water the supply and demand for labor. One concern is quality) and have adequate information on which that the approach relies on private valuations of tobasetheirpreferences. CVMislikelytobefarless health risks, rather than social ones. In this con- reliable when the object of the valuation exercise is text, the level of information on occupational a more abstract aspect, such as existence value. hazards must be high, for private individuals to Generally, declared WTA tends to be signifi- make meaningful tradeoffs between health-risk cantlygreater than the corresponding WTP. This and remuneration. Finally, the effects of all fac- maybepartlyattributableto"strategicbias"where tors other than environment (for example, age, respondents feel they would be better off inflat- skill level, job responsibility) that might influence ing the amounts they would receive rather than wages must be accounted for, to eliminate bias the sums to be paid out, if the hypothetical ques- and isolate the impacts of environment. tions posed were somehow to become a reality in the future. In the case of poorer individuals, WTP This MARKEED is D useful when an environmentalgood may be limited by the ability-to-pay, whereas This method is useful when an environmental good WTA is not. The questionnaires have to be care- or service has no readily determined market value, WTA isin e quentes ave to but a close substitute exists that does have a com- fully designed, implemented, and interpreted to petitively determined price. In such a case, the overcome the above mentioned difficulties, as market price ofthesbstitutemaybewell as other types of bias (for details, see The market pnce of the substitute may be used as a Energy Journal 1988). Munasinghe (1990a) pro- proxy for thevaluerof theenvironmental exaurce vides several early examples of the application of Barbler and others (1991) provide an example CVM to value the quality of electricity services in involving marketed and non-marketed fish sub- developing countries. stitutes. A review by Pearce and Markandya (1989) compared valuation estimates obtained from market-based techniques and CVM, using results In cases where market information cannot be from seven studies carried out in industrial na- used directly or indirectly, market-like behavior tions. They found that the corresponding esti- needs to be deduced through construction or mates overlapped within an accuracy range of simulation. The methods summarized below plus or minus 60 percent. The conclusion is that depend on direct questions, surveys, or market- CVM, cautiously and rigorously applied, could ing experiments. provide rough estimates of value that would be helpful in economic decisionmaking, especially ARTIFICIAL MARKET when other valuation methodsareunavailable. A Such markets are constructed for experimental case study using the CVM for estimating the purposes,todetermineconsumerWTPforagood value of elephants in Kenya (Brown and Henry or service. For example, a home water purifica- 1989) shows that it is possible to achieve an un- tion kit might be marketed at various price levels, derstanding of the order of magnitude of the or access to a game reserve may be offered on the benefits through modestmethods. Anotherstudy, basis of different admission fees, thereby facilitat- on WTP for water services in southern Haiti, tests ing the respective estimation of values placed by the CVM for different biases, indicating the limits individuals on water purity or on the use of a of its reliability (Whittington and others 1990) . recreational facility. CONTINGENT VALUATION The discount rate When relevant market behavior is not observ- able, the contingent valuation method (CVM) Economists typically use a forward-looking ap- puts direct questions to individuals to determine proach in which past (or sunk) costs and benefits how much they might be willing-to-pay (WTP) are ignored, while a discount rate is applied to for an environmental resource, or how much future costs and benefits to yield their present 51 Defining and Measuring Sustainability: The Biogeophysical Foundations values. Standard criteria for cost-benefit analysis term relative to the case where discount rates (CBA), such as the net present value (NPV) and were higher. Further, using a very low discount internal rate of return (IRR) are derived in this rate to protect future generations is inequitable, way. The issue of choosing an appropriate dis- since this would penalize the present generation count rate has been discussed in the context of and increase inequalities across time periods- general CBA for many years (Dasgupta and oth- especially when the present contains widespread ers 1972, Harberger 1976, Littleand Mirrlees 1974). poverty. Two concepts mainly help to shape the dis- In order to facilitate such intergenerational count (or interest) rate in a market economy. transfers, one option is to impose a sustainability First, there is the rate of time preference of indi- constraint, whereby current well-being is maxi- viduals, which determines how they compare mized without reducing the welfare of future present-day with future consumption. Second, generations below that of the current generation. there is the rate of return on capital, which deter- In practice, this would entail monitoring and mines how an investment (made by foregoing measurement of capital stocks (man-made, hu- today's consumption) would yield a stream of man, and natural), anda broad investment policy future consumption (net of replacement). In an which seeks to ensure that compensating invest- ideally functioning market the interest rate, which ments offset depreciation of existing assets (Pearce equilibrates savings and investment, also equals 1991). Theoretically, the aim would be to ensure both the marginal rates of time preference and that the overall stock of assets is preserved or return on capital. In practice, government policy enhanced for future generations, but practical distortions and market failures lead to diver- application of this principle would be difficult. gences between the rates of time preference and return on capital. Furthermore, the social rate of time preference may be less than the individual Risk and uncertainty time preference rate, because long-lasting societ- ies are likely to have a bigger stake in the more Risk and uncertainty are an inherent part of eco- distant future than relatively short-lived indi- nomic decisions. Risk represents the likelihood viduals. of occurrence of an undesirable event such as an The long-term perspective required for sus- oil spill. In the case of uncertainty, the future tainable development suggests that the discount outcome is basically unknown. Therefore, the rate might play a critical role in intertemporal risk of an event may be estimated by its probabil- decisions concerning the use of environmental ity of occurrence, whereas no such quantification resources. The rate of capital productivity is very is possible for uncertainty since the future is high in many developing countries because of undefined. The risk probability and severity of capital scarcity, and the rate of time preference damage could be used to determine an expected also is elevated because of the urgency of satisfy- value of potential costs, which would be used in ing immediate food needs rather than ensuring the CBA. However, the use of a single number (or long-term food security (Pearce and Turner 1990). expected value of risk) does not indicate the de- Projects with social costs occurring in the long gree of variability or the range of probability term and net social benefits occurring in the near values that might be expected. Additionally, it term will be favored by higher discount rates. docs not allow for individual perceptions of risk. Conversely, projects withbenefitsaccruingin thc The risk probability may be used to devise an long run will be less likely tobe undertaken under insurance scheme to protect against the risk. high discount rates. Thus, some environmental- In the case of uncertainty, it is not possible to ists have argued that discount rates should be estimate the expected value of costs or insure lowered to facilitate environmentally sound against an unknown eventuality. The increasing projects meeting the CBA criteria. However, this scale of human activity, the complexity of envi- would lead to more investment projects of all ronmental and ecological systems, and the lack of types, thereby possibly threatening fragile envi- knowledge of how these systems might be af- ronmental resource bases. Norgaard (1991) ar- fected, all emphasize the need to deal with uncer- gues that lowering discount rates can in fact taintymoreexplicitly. Acautiousapproachisthe worsen environmental degradation-by lower- key to dealing with uncertainty. Global warming ing the cost of capital and thereby lowering the is an illustrative example. In the past, the green- cost of production, more is consumed in the near house effect of CO2 emissions was not known or 52 Key Concepts and Terminology of Sustainable Development recognized as a risk. At the present time, there is future generations (or, as he terms them, "miss- still considerable uncertainty about the future ing markets" because "future generations are impacts of global warming, but given the large unable to enter bids to protect their interests"). magnitude of potential consequences, caution is Without such a structure, decisionmakers may warranted. tend to ignore costs to future generations, and The traditional and simple way of incorporat- minimize costs to current generations at the ex- ing risk and uncertainty considerations in project pense of the future. If the entitlement structure is level CBA has been through sensitivity analysis. adjusted, the policymakercan thenexamine three Using optimistic and pessimistic values for dif- policies to protect the interests of future genera- ferent variables can indicate which variables will tions: (1) mandated pollution abatement; (2) full have the most pronounced effects onbenefits and compensation for future damages (for example, costs. Wenote that whilesensitivityanalysisneed by taxation); and (3) anannuity to compensate the not reflect the probability of occurrence of the future for costsimposed in the present. In the face upper or lower values, it is useful for determining of uncertainty, the first option might be the most which variables are most important to the success efficient. or failure of a project (Dixon and others 1988). Other important sources of uncertainty linked More sophisticated approaches to analyze risk with environmental issues include uncertainty anduncertaintyareavailable(BradenandKolstad over land tenure (which leads to deforestation 1991). and unsustainable agricultural practices), and The issue of uncertainty plays an important uncertainty of resource rights (which can acceler- role in environmental valuation and policy for- ate the rate of depletion of a nonrenewable re- mulation. Optionvaluesandquasi-optionvalues source). Policymakers can address these issues are based on theexistenceof uncertainty. Option by instituting land reforms, and by designing value (OV) isessentially the "premium" that con- appropriate taxation policies that return rents to sumers are willing to pay to avoid the risk of not public sources rather than to private agents. having something available in the future. The sign of option value depends upon the presence References of supply and/or demand uncertainty, and on whether the consumer is risk averse or risk lov- Aylward, B., and Edward Barbier. 1992. "Valu- ing. Quasi-option value (QOV) is the value of ing Environmental Functions in Developing preserving options for future use in the expecta- Countries." Biodiversity and Conservation 1, tion that knowledge will grow over time. If a pp 34_50 development takes place that causes irreversible pp. 34-50. environmental damage, the opportunity to gain Barbier, Edward. 1991. Economics, Natural Re- knowledge through study of flora and fauna is sourceScarcity,and Development: Conventional lost. Increased benefits to be derived through andAlternativeViews.London:EarthscanPub- future knowledge expansion (which is indepen- lications, Ltd. dent of exploitation) leads to a positive QOV. Barbier,Edward,W.M.Adams,andK.Kimmage. Thissuggeststhattheresourceexploitationshould 1991. "Economic Valuation of Wetland Ben- be postponed until increased knowledge facili- efits: The Hedejic-Jamiare Floodplain, Nige- tates a more informed decision. If information ria." LEEC Discussion Paper DP 91-02, April. growth depends on the use taking place, which is Bamett, Harold, and Chandler Morse. 1963. Scar- unlikely in an environmental context, then QOV city and Growth: The Economics of National Re- is positive (negative) when the uncertainty ap- source Availability. Baltimore, Md.: Johns plies to the benefits of preservation (exploitation) Hopkins University Press. (Pearce and Turner 1990, Fisher and Hanemann Botkin, D. B. 1990. Discordant Harmonies: A New 1987). B o r D. B 190 centu Harmonies: AxNew Bromley (1989) suggests that the way in which Ecologyfor the 21st Century. New York: Oxford policymakers address uncertainties depends on University Press. their perception of the existing entitlement struc- Bojo, J.1990. "Economic Analysis of Agricultural ture. The interests of the future areonly protected Development Projects. A Case Study from by an entitlement structure that imposes a duty Lesotho". EFI Research Report, Stockholm: on current generations to consider the rights of Stockholm School of Economics. 53 Defining and Measuring Sustainability: The Biogeophysical Foundations Boyden, S., and S. Dovers. 1992. "Natural Re- Dobbs, M. 1992. "Bootlegging Thrives, Sturgeon source Consumption and Its Environmental FlounderasCaviarCartelSplits."lnternational Impacts in the Western World. Impacts of In- Herald Tribune, June. creasingPerCapita Consumption." Ambio21:1. Energy Journal. 1988. Special issue on Electricity Braden, J. B., and C. D. Kolstad, eds.1991. Mea- Reliability, 9 (December). suring the Demand for Environmental Quality. Fisher, Anthony C., and W. Michael Hanemann. New York: Elsevier. 1987. "Quasi Option Value: Some Misconcep- Bromley, Daniel W. 1989. "Property Relations tions Dispelled." Journal of Environmental Eco- and Economic Development: The Other Land nomics and Management 14, pp. 183-90. Reform." World Development 17:6, pp. 867-77. Freeman,A. Myrick. 1991. "ValuingEnvironmental Brown Jr., G., and W. Henry. 1989. "The Economic Resources under Alternative Management Re- Valueof Elephants." London Environmental Eco- gimes." Ecological Economics 3, pp. 247-56. nomnics Centre, Discussion Paper 89-12. Gomez-Pompa, A., and A. Kaus. 1992. 'Taming the Cleveland, C. J. 1991. "Natural Resource Scarcity Wilderness Myth." Bioscience 42:4, pp. 271-79. and Economic Growth Revisited: Economic Goodland, R., H. Daly, S. L. Serafy, and B. von and Biophysical Perspectives." In R. Costanza, Droste, eds. 1991. Environmentally Sustainable ed., Ecological Economics: The Science and Man- Economic Development: Building on Brundtland. agement of Sustainability. New York: Columbia Paris: UNESCO. University Press. Goud ie, A. 1990. The Human Imnpact on the Natural Cocklin, C. 1989. "Mathematical Programming Environment. 3d ed. Cambridge, Mass.: M.I.T. and Resources Planning I: The Limitations of Press. Traditional Optimization." Journal of Environ- Grumbine, E. 1990. "Protected Biological Diver- mental Management 28, pp. 127-41. sity through the Greater Ecosystem Concept." Costanza, R., ed. 1991a. Ecological Economics: The Natural Areas Journal 10:3, pp. 114-20. Science and Management of Sustainability. New Hanley, Nick D. 1989. "Valuing Rural Recreation York: Columbia University Press. Benefits: An Empirical Comparison of Two . 1991b. "The Ecological Economics of Approaches." Journal of Agricultural Econotnics Sustainability: Investing in Natural Capital." 40:3 (September), pp. 361-74. In R. Goodland and others, eds., Environmnen- . P tally~~~~~~~ Sutial Ecnmc.eeopet ,Bid Harberger, A. C. 1976. Project Evaluation: Collected tally Sustainable Economic Development: Build- Papers. Universit of Chicago Press. ing on Brundtland, pp. 83-90. Paris: UNESCO. Harrs. A.e.y Krupnigc W. Croppr,M.L.,ad W.F. Oaes. 992."Envron- Harrington,W., A. J .Krupnick, and W. O.Spofford, Cropper, M. L., and W. E. Oates. 1992. "Environ- Jr. 1989. "TheEconomic Losses of a Waterborne mental Economics: A Survey." Journul of Eco- DiseaseOutbreak." Journal of UrbanEconomics, nomic Literature, Vol. XXX (June), pp. 675-740. 25:1, pp. 116-37. Daly, Herman. 1990. "TowardsSomeOperational Hartwick, J. M., and N. D. Olewilcr. 1986. The Principles of Sustainable Development." Eco- Economics of Natural Resource Use. New York: logical Economics 2:1, pp. 1-6. Darlington, C. D. 1969. The Evolution of Man and Hodgson Gregor and John Dixon.1988. Logging Society. New York: Simon and Schuster. Hogsn Grgr an JonDxn'98 ogn versus Fisheries and Tourism in Palawan. Occa- Dasgupta, Partha, and Karl-Goran Maler. 1990. sional Paper 7. Honolulu, Hawaii: East-West The Environment and Emerging Development Is- Environment and Policy Institute. sues. Proceedings of the World Bank Annual Con- . . . S feeco Deeomn Ecn .c 190 Holmberg,J. 1992. "Opera tionalizing Sustai nable ferenceon Development Economnics 1990, pp. 1u1- Deeomn nteWrl a'. ol ak 29. Washington, D.C.: World Bank. Development inthe World Bank." World Bank, Dasgupta, P., S. Marglin, and A. K. Sen. 1972. Hyman J.B. and.C. WeDrstedt.1991. eRoleof Guidelinesfor Project Evaluation. UNIDO, New H B York. Biological and Economic Analysis in the Listing Dixon, John, and others. 1988. Economic Analysis of of Endangered Species." Resources (Summer). the Environmental Impacts of Development Projects. IUCN (International Union for the Conservation London: Earthscan Publications, in association of Nature). 1991. Caringfor the Earth: A Strategy with the Asian Development Bank. for Sustainable Living. With the United Nationis 54 Key Concepts and Terminology of Sustainable Development Environment Program and the World Wildlife Munasinghe, Mohan. 1990a. Electric Power Eco- Fund. Gland, Switzerland. nomics. London: Butterworths Press. Jimenez, E. 1983. "The Magnitude and Determi- . 1990b. "Managing Water Resources to nants of Home Improvement in Self-Help AvoidEnvironmentalDegradation."Environ- Housing: Manila's Tondo Project." Land Eco- ment Department Working Paper 41. World nomics 59:1, pp. 70-83. Bank, Washington, D.C. Lele, Sharad M. 1991. "Sustainable Development: . 1992a. "Biodiversity Protection Policy: A Critical Review." World Development 19:6, Environmental Valuation and Distribution Is- pp. 607-21. sues." Ambio 21:3, pp. 227-36. Liebman, J. 1976. "Some Simple-Minded Ob- . 1992b. Water Supply and Environmental servations on the Role of Optimization in Management. Boulder, Colo.: Westview Press. Public Systems Decisionmaking." Interfaces . 1993. Environmental Economics and Sustain- 6, pp. 102-08. ableDevelopmnent.Washington,D.C.:WorldBank. Little, I. M. D. and 1. A. Mirrlees. 1974. Project Munasinghe, Mohan, and K. King. 1992. "Pro- Appraisal and Planning for Developing Countries. tecting the Ozone Layer." Finance and Develop- Basic Books, New York. ment (June), pp. 24-25. Lovelock,J. E. 1979. Gaia: A New Lookat Lifeon Earth. Nabhan, G. P., A. M. Rea, K. L. Hardt, E. Mellink, Oxford, England: Oxford University Press. and C. F. Hutchinson. 1982. "Papago Influences Lutz, Ernst, and Mohan Munasinghe. 1991. "Ac- on Habitat and Biotic Diversity: Quitovac Oasis counting for the Environment." Finance and Ethno-Ecology." Journal of Ecology 2, pp. 124-43. Development 28 (March), pp. 19-21. Norgaard, R. B. 1991. "Sustainability as Martin, P. 1984. "Prehistoric Overkill: The Global Intergenerational Equity." Asia Regional Series, Model." In P. Martin and R. Kline, eds., Quater- Report IDP-97, World Bank, Washington D.C. nary Extinctions: A Prehistoric Revolution, pp. Norton, B., and R. Ulanowicz. 1991. "Scale and 354-403. Tucson: University of Arizona Press. Biodiversity Policy: A Hierarchical Approach." McRobert, D. 1988. "Questionable Faith." Probe Ambio 20:1, pp. 1-6. Post 11:1. Oldeman, L. R., V. W. P. van Engelen, and J. H. M. Meier, P., and Mohan Munasinghe. 1993. Incorpo- Pulles. 1990. "The Extent of Human-induced rating Environmental Concerns into Power Sector Soil Degradation." In L. R. Oldeman, R. T. A. Decision-making: A Case Study of Sri Lanka. Hakkeling, and W. G. Sombroek, eds., World World Bank, Washington, D.C. Mapof theStatus ofHuman-induced SoilDegrada- Mendelsohn, M. 1987."Modelling the Demand tion: An Explanatory Note. 2d ed. rev., annex 5. for Outdoor Recreation." Water Resources Re- Wageningen, the Netherlands: International search 23:5, pp. 961-7. Soil Reference and Information Centre. Meredith, T. C. 1992. "Environmental Impact As- Orians, Gordon H. 1990. "Ecological Concepts of sessment, Cultural Diversity, and Sustainable Sustainability." Environment 32:9 (November), Rural Development." EnvironmentaliImpact As- pp. 10-15, 34-39. sessment Review 12:1/2 (March/June). Pearce, D. W. 1991. Development and the Natural Micklin,Philip.1988."DesiccationoftheAralSea:A World. World Bank, Washington D.C. WaterManagementDisasteroftheSovietUnion." Pearce, David, Edward Barbier, and Anil Science 241 (September 2), pp. 1170-76. Markandya. 1989. Blueprintfora Green Economy. Mitchell, Robert C., and Richard T. Carson. 1989. London: Earthscan Publications. Using Surveys to Value Public Goods: The Contin- . 1990. Sustainable Development: Economics gent Valuation Method. Washington, D.C.: Re- and Environment in the Third World. London: sources for the Future. Edward Elgar, Ltd. Moseley, W. 1992. "Measuring the Environmen- Pearce, David W., and Anil Markandya. 1989. The tal Sustainability of Human Economies: Some Benefits of Environmental Policy: Monetary Valu- Suggestions and Examples of Indicators at the ation. Paris: OECD. National Level." Environment Department, Pearce, David W., and R. Kerry Turner. 1990. World Bank, Washington, D.C. Draft. Economics of Natural Resources and the Environ- 55 Defining and Measuring Sustainability: The Biogeophysical Foundations ment. London: Harvester-Wheatsheaf. Schneider, S. 1990. "Debating Gaia." Environment Pearce, F. 1991. Green Warriors: The People and the 32:4, pp. 5-9, 29-32. Politics behind theEnvironmental Revolution. Lon- Shearer, W. 1992. "A Proposal for a Biophysical don: Bodley Head. Sustainability Index." United Nations Univer- Perrings, Charles. 1991. Ecological Sustainability sity, New York and Tokyo. Private communi- and Environmental Control. Centre for Resource cation. and Environmental Studies, Australian Na- Solow, Robert. 1986. "On the Intergenerational tional University. Allocation of Natural Resources." Scandina- Perrings,Charles,C.Folke,andKarl-GoranMaler. vian Journal of Economics 88:1, pp. 141-49. 1992. "The Ecology and Economics of Soule, M. E. 1991. "Conservation: Tactics for a Biodiversity Loss: The Research Agenda." Constant Crisis." Science 253, pp. 744-50. Ambio 21:3, pp. 201-11. Sprugel, D. G. 1991. "Disturbance, Equilibrium, Peters, C. M., A. H. Gentry, and R. O. Mendelsohn. and Environmental Variability: What Is "Natu- 1989. "Valuationof an Amazonian Rainforest." ral" Vegetation in a Changing Environment?" Nature 339 (June 29), pp. 655-56. Biological Conservation 58, pp. 1-18. Petry, F. 1990. "Who Is Afraid of Choices? A Tietenberg,Thomas.1988.EnvironmentalandNatu- Proposal forMulti-Criteria AnalysisasaTool ral Resource Econornics. 2d ed. Glenview, Ill.: for Decision-making Support in Develop- Scott Foresman and Company. ment Planning." Journal of International De- Tobias, D. and R. Mendelsohn. 1991. "Valuing velopment 2, pp. 209-31. EcotourisminaTropicalRain-ForestReserve." Pezzey, John. 1989. "Economic Analysis of Sus- Ambio 20:2, April. tainable Growth and Sustainable Develop- Toman, M. A. 1992. "'The Difficulty in Defining ment." Working Paper 15. Environment De- Sustainability." Resources (Winter), pp. 3-6. partment, World Bank, Washington, D.C. Pro- United Nations Conference on Environment and cessed. Development. 1992. "Conservation of Biological Pimental, D., and others. 1992. "Observing Bio- Diversity." In Agenda 21, chap. 15. Rio dejaneiro. logical Diversity in Agricultural/ Forestry Sys- Walsh, Richard G., John B. Loomis, and Richard tems." Bioscience 42:5, pp. 354-62. A. Gillman. 1984. "Valuing Option, Existence, Pollan, M. 1990. "Only Man's Presence Can and Bequest Demands for Wilderness." Land Save Nature." Journal of Forestry 88:7, pp. Economics 60:1 (February), pp. 14-29. 24-33. Westman, W. E. 1977. "How Much Are Nature's Ponting,C. 1990. "Historical Perspectives on Sus- Services Worth?" Science 197, pp. 960-64. tainable Development." Environment 32:9, pp. Whittington, D., J. Briscoe, X. Mu, and W. Barson. 4-9, 31-33. 1990. "Estimating the Willingness to Pay for Potvin, Joseph. 1992. "Classification and Appraisal Water Services in Developing Countries." Eco- Criteria for Conservation Investments." Glo- nomic Development and Cultural Change 38:2. bal Environment Facility, World Bank, Wash- o D.C.Environment Draf ilit. , World Bank,Wash- World Bank. 1992. World Development Report 1992: ington, D.C. Draft. Development and the Environmnent. New York: Pronk,J., and M. Haq. 1992. SustainableDevelopment: Oxford University Press. From Concept to Action. The Hague Report. New World Commission on Environment and Devel- York: United Nations Development Program. opment. 1987. Our Common Future. Oxford, Randall, Alan, and John Stoll. 1983. "Existence England: Oxford LJniversity Press. Value in a Total Valuation Framework." In g y Managing Air Quality and Scenic Resources at World Resources Institute. 1992a. Global National Parks and Wilderness Areas. Boulder, Biodiversity Strategy. Washington, D.C. Colo.: Westview Press. . 1992b. World Resources Report. New York: Romero, C., and T. Rehman. 1987. "Natural Re- Oxford University Press. source Management and the Use of Multiple Young, M. D. 1992. Sustainable Investment and Criteria Decision-making Techniques: A Re- Resource Use: Equity, Environmental Integrity, view." European Journal of Agricultural Econom- and Economic Efficiency. Paris: UNESCO; ics 14, pp. 61-89. Carnforth: Parthenon Publishing Group. 56 3 3 Limits to Sustainable Use of Resources: From Local Effects to Global Change Peter M. Vitousek and Jane Lubchenco The term sustainable has undergone a rapid mi- whether humanity can harvest a resource with- gration and radiation from its long-term home in out reducing its stock but also a question of forestry and fisheries into the wider ecological, whether humanity can use resources without at agricultural, and development communities. In the same time changing regional and global sys- its previous habitat, the term referred to manage- tems in deleterious, uncontrolled, or unpredict- ment strategies in which a small enough fraction able ways. of a resource (trees, fish) is harvested so that the 'The evolution of the term sustainable can be stock of the resource is not diminished; growth in illustrated straightforwardly with a set of ex- volume or population of the resource is sufficient amples. Consider first a region occupied by natu- to replace the amount harvested. In contrast, ral grassland ecosystems, say the mixed-grass nonsustainable management strategies take a prairie of central North America. Net primary larger fraction of the resource, thereby diminish- production in the grassland is supported by ni- ing the stock and (eventually) the amount har- trogencyclingfromsoilorganicmattertograsses, vested. In an economic analogy, the stock of a often through animals, and back to soil organic resource is equivalent to principal and its growth matter; occasional firesaccelerate thiscyclebriefly rate is equivalent to interest; use of interest alone (Ojima 1987; Parton and others 1987). A small is sustainable because it leaves the principal number of humans can use the region sustainably intact, but any higher rate of use diminishes by harvesting grazing animals, if they do so at a principal. low enough intensity to avoid depleting the stock The meaning of sustainable is clearas it applies of those animals. to forestry and fisheries, although long-term suc- Now consider the same region being used for cessinmanagingeitherforestsorfishina sustain- pioneer agriculture. The prairie soil is plowed, able way has been elusive (Ludwig, Hilborn, and crops are planted, and the prairie grasses are Walters 1993). However, in its expanded niche as suppressed. Net primary productivity remains partofsustainableagriculture,sustainabledevel- high fora time and in a form useful to humans, opment, and a sustainable biosphere, the term but it is now supported by net oxidation of soil has taken on additional meanings. Where once organicmatterandreleaseoftheorganicnitrogen only a single resource, the one being harvested, it contains (Bolin and others 1983). If this organic was considered, we are now concerned with the matter or nitrogen is not replaced, the system is integrated and cumulative effects of human ac- not sustainable; it achieves its productivity only tivities or management practices on a multitude by depleting an essential resource, and sooner or of resources and processes, from the local to the later yields will fall and the agricultural systems global scale. Now, it is not only a question of will fail. Defining and Measuring Sustainability: The Biogeophysical Foundations Pioneer agriculture of this sort can be made 1977). This productivity occurs despite the olig- sustainable through shiftingcultivation if enough otrophic waters surrounding most reefs and is land is available and if the human population thought to be intimately dependent on the effi- using the land is small enough. In this system, cient recycling of nutrients within the reef com- humans practice agriculture for a short period of munity. Reefs are thus unlike many other coastal time on any given plot, then abandon that plot marine ecosystems, which depend primarily on and use another, and so on until after some time nutrients from adjacent systems, for example, via they return to the first plot. Shifting cultivation upwelling or terrestrial runoff. The high inci- can be sustained if the interval between visits to a dence of symbiotic relationships between plants given site is long enough for soil fertility (and and animalson coral reefs is thought to reflect the other features of the land) to regenerate (Nye and advantage accrued to an efficient transfer of nu- Greenland 1960). However, such a system is highly trients among these components. vulnerable to any increase in population or eco- Despite being very productive, however, reefs nomic pressure. are highly susceptible to certain kinds of Finally, considera modemragricultural system overexploitation (Birkeland 1992; Munro and in the same area. Modem agriculture, of course, Williams 1985; Russ 1985). Traditional, even covers a multitude of practices, but common to heavy, subsistence fishing on coral reefs is usu- most of them is the fact that net primary produc- ally sustainable, especially when nutrients, for tion of the crop is supported by inputs that hu- example in the form of scraps and fecal material, mans control. For example, nitrogen from fertil- are returned to the reef. These fishing practices izergenerallybecomesarelativelylargesourceof appear to retain nutrients within the system. nitrogen for crops. Is such a system sustainable? Large-scale export fisheries, however, have dra- The question must now be asked on at least two matically different effects. Not only do they re- scales: locally and regionally or globally. On the move more biomass, but they now transport bio- local scale, the answer depends on the particular mass completely out of the system instead of agricultural practice examined, and it is arguable recycling it. The result of this combined depletion in many cases. Many observers are impressed of fish stocks and export of nutrients is an over- withtheincreasingamountandvarietyof inputs fished and impoverished reef and thus an required to maintain productivity and with con- unsustainable fishery (Birkeland 1992). current rates of erosion, salinization, and other As fish become scarce on these reefs, destruc- changes in soils. Many others are impressed that tive methods of fishing are often employed to increasing inputs have led to some quite spec- harvest the remaining fishes. Dynamiting is a tacular increases in yields and also that humans common, if illegal, method. The resulting de- demonstrablyaregood atadaptingland manage- struction of reefs leads to further impoverish- ment to match many alterations in soils. ment, now involving the stocks of other, nontar- On a regional or global level, the case for the get species as well as destruction of the habitat sustainability of many modem agricultural prac- itself. Practiced on a large scale, this cumulative tices is not so debatable. Inputs to modern agri- impoverishment has consequences for the entire cultural systems include energy from fossil fuels, basin. fertilizer, pest control, weed control, and often Thesetwoexamplesillustratesomeoftheways irrigation water; outputs include alterations in in which our thinking about sustainability has the composition of the atmosphere and in re- changed. The scale of the area under consider- gional air quality, changes in water quality and ation has increased as we appreciate the regional sometimesdepletion of groundwater,and changes and global impacts of local practices. Inclusion of in biological diversity both within and outside the larger scale often changes conclusions about the region. As widely practiced, such agricultural the sustainability of a practice, as feedbacks at systems are not sustainable in that they cannot regional and global scales impose different limits persistonalargescalewithoutalteringtheregion than those of the more narrow local focus. To and indeed the Earth system as a whole. determine sustainability, one must consider not Coral reefs provide a different example of the only the stock of a particular resource but also the relationship between spatial scale and the process required to maintain those stocks, other sustainability of harvesting practices. Coral reefs populations, and other functions. Maintenance of are widely recognized as being among the most these processes often involves larger scales and productiveecosystemson Earth (Crisp 1975; Lewis diffcrent constraints. 58 Limits to Sustainable Use of Resources: From Local Effects to Global Change In this chapter, we focus on some of these There is no doubt that the ongoing increase is integrated regional and global consequences of a by-product of human activity, primarily the human activity, not on other (very important) combustion of fossil fuels and, secondarily, questions, such as the extent to which high-input changes in land use. The amount of carbon re- systems can be maintained in the long term on leased from the burning of fossil fuel is more than local scales or how much biomass and nutrients sufficient to account for the global increase can be exported from a system before it collapses. (Schlesinger 1991). More convincingly, the rela- Weconsiderhumanactivitiestobenonsustainable tive atmospheric abundance of the carbon iso- to the extent that they alter features of the Earth topes 13C and 'IC have decreased over time in a system, such as the composition of the atmo- pattern and magnitude that demonstrate that sphere and its capacity to process pollutants, the their concentrations are being diluted by carbon stability of the dimate, the formation and mainte- released from fossil fuel combustion (which is nance of soil fertility, the ability of aquatic (fresh- 14C-free and 13C-depleted) and to a lesser extent waterand marine) systems to processand recycle loss of terrestrial biomass (13C-depleted; nutrients, and above all the maintenance of the Siegenthaler and Oeschger 1987; Stuiver 1978). diversity of organisms that carry out many of Thishuman-caused increasein carbondioxide these functions (Ehrlich and Mooney 1983). We isalreadysubstantial(morethan25percentof the first consider some human activities that have initial value), and it is the major factor driving global consequences, then focus on human re- anthropogenic enhancement of the greenhouse sponses to these changes. effect. Moreover, increased concentrations of car- bon dioxide are likely to affect terrestrial biota directly by increasing growth rates of some but Human activity and global change not all plants and by increasing the amount but decreasing the quality of food available to many That the current level of human activity alters animals and decomposers (Bazzaz 1990; Mooney many features of the Earth system on regional and others 1991). Elevated carbon dioxide could and global scales is beyond dispute. Change in also have direct effects in marine ecosystems climate (for example, global warming!) receives a (Smi th and Buddemeier 1992). large amount of attention, but it is not the best Anthropogenic increases in concentrations of documented, not currently the most important, a number of other stable gases also have been andnotthemostpermanentofthecomponentsof documented (Watson and others 1990). These global change caused by humans. include the industrially produced chlorofluoro- carbons (CFCs), methane, and nitrous oxide. The The changing atmosphere increase in methane is believed to be due to a combination of agricultural activities (particu- The atmosphere mixes more rapidly than the other larly the growing of paddy rice and the mainte- great spheres of the Earth system (oceans and ter- nance of domestic ruminants) and industrial pro- restrial ecosystems), and it is not surprising that cesses (Cicerone and Oremland 1988). The rea- global changes have been detected most readily in sons for increasing nitrous oxide are less certain the atmosphere. Measurements of carbon dioxide but are believed to relate to changes in tropical concentrations have been carried out since 1957; land use and the massive alteration of the global during that time, concentrations have increased nitrogen cycle brought about by intensive agri- more than 10 percent, from 315 to over355 parts per culture (Matson and Vitousek 1990; Vitousek and million, and the rate of increase has accelerated Matson, forthcoming). (Keelingandothers 1989).Could thisrapidincrease All of these gases can enhance the greenhouse result from a fortuitous interaction between our effect; in addition, CFCs and nitrous oxide break relatively brief record of measurements and a natu- down in the stratosphere and cause a breakdown ral fluctuation in carbon dioxide concentrations? of stratospheric ozone. The ability of CFCs to Theanswer is no, because measurements of carbon affect stratospheric ozone was identified in 1974 dioxide concentrations in air bubbles trapped in (Molina and Rowland 1974), and their impor- Greenland and Antarctic ice show that concentra- tance as an agent for global change was debated tions were stable near 280 parts per million for at actively at that time. Nevertheless, everyone was least 1,000 years before the ongoing exponential surprisedbythediscoveryofaspringtimeholein increase began (Watson and others 1990). the Antarctic ozone in the mid-1980s, though 59 Defining and Measuring Sustainability: The Biogeophysical Foundations perhaps not by the subsequent proof that CFCs try (Keller and others 1991; Shukla, Nobre, and cause this depletion as a consequence of a previ- Sellers 1990), to alter the chemistry of major river ously unsuspected set of interactions (Prather systems (Peieris and others 1991), and to be the and Watson 1990; Rowland 1989). most important cause of global change to coastal Not all of the human-caused changes to the marineecosystemsandcoralreefs(Howarthl988; atmosphere involve stable, globally distributed Smith and Buddemeier 1992). Nevertheless, its gases. Anthropogenic increases in concentrations most important effect probably is simply to alter of chemicallyreactive gases have led to decreased local systems; some types of major ecosystems tropospheric concentrations of the hydroxyl radi- have nearly disappeared (tall-grass prairie, tropi- cal, the major oxidizing agent in the atmosphere cal deciduous forest), and many others have been (Thompson 1992). The resulting decrease in the degraded or fragmented. ability of the atmosphere to cleanse itself leads to The global effects of changes in land use have an increased atmospheric lifetime, and hence in- been summarized in two ways: in terms of the creased concentration, of methane. amount of land altered by humanity and in terms At the same time, a syndrome of elevated of the fraction of terrestrial productivity that hu- tropospheric ozone concentrations, acidic pre- manity controls. Turner and others (1990) have cipitation, and elevated nitrogen deposition oc- estimated that nearly half of the land surface of curs over most of the economically developed Earth has been transformed by human activity, in regions of the Earth (Crutzen and Zimmerman that it has been converted to cropland or to im- 1991; Logan 1985), and similar changes are now proved pastures or been desertified. Much of the being observed seasonally in developing tropical rest of Earth has been affected through logging or regions (Fishman and others 1991; Keller and extensivegrazingonrangelandsbuthasnotbeen others 1991). High application rates of nitrogen transformed in character. Alternatively, Vitousek fertilizer, intensive animal husbandry, and the and others (1986) have calculated that nearly 40 production of nitrogen oxides and sulfur oxides percent of the terrestrial net primary productivity by internal combustion engines and other indus- of Earth is now being used, dominated, or de- trial processes all contribute to these changes; stroyed by human activity; again, much of the biomass burning is the most important source in remainder is affected, although not overwhelm- many developing tropical areas (Andreae and ingly so. others 1988; Crutzen and Andreae 1990). Eurnpe, These two estimates are in reasonably close where increases in agricultural production have agreement, and both imply that all of Earth's been most impressive (and most heavily subsi- terrestrial ecosystems and an overwhelming ma- dized), is affected particularly severely by these jority (certainly more than 99 percent) of its spe- by-products of human activity (Schulze 1989). cies must persist on little more than half of the area they once occupied. To the extent that these Changes in land use natural systems, species, and populations pro- vide goods or services that are essential to the Human-caused change in land use (land clearing, sustainability of human systems, their shrunken agricultural intensification, urbanization, and so base of operations must be a cause for concern. forth) is currently the most consequential compo- nent of global change, and its effects are already Loss of biological diversity with us. However, land use can be difficult to treatasaglobalchange,becauseunliketheatmo- The most permanent component of anthropo- sphere or oceans, terrestrial ecosystems do not genic global change is the extinction of species mix on any time scale that is relevant to human andgeneticallydistinctpopulations. Extinctionis sustainability. Consequently, it is impossible to a natural phenomenon; under normal circum- characterize global changes in land use by mea- stances, an average species lasts perhaps 10 mil- surements in one or a few locations. Any global lion years from appearance to extinction (Ehrlich effect must be the sum of many changes to local andWilson 1991)."Normal" conditionsarepunctu- ecosystems. In practice, change in land use alters ated by episodes of mass extinction, of which five enough local ecosystems substantially enough to are known in the past hundredsof millions of years. contribute directly to increased concentrations of Human activity is accelerating the process of greenhouse gases (Watson and others 1990), to extinction dramatically. Observations of well- affect regional climate and atmospheric chemis- studied groupssuch asbirds, together with calcu- 60 Limits to Sustainable Use of Resources: From Local Effects to Global Change lations of losses based on species per area and degrading or altering the system in now-unpre- species per energy relationships (Wilson and Pe- dictable ways (that is, the impact is not sustain- ter 1988), suggest that current rates of extinction able), many intelligent people argue that it is too areorders of magnitude abovebackground rates. soon to act because the changes are not yet to the Most of the extinctions that have occurred to date point that they really matter to humanity. (Argu- have been caused by changes in land use, al- ments that action is too expensive, that future thoughbiologicalinvasionsbyexoticspecieshave technological advances will mitigate any prob- also played a significant role (D'Antonio and lems, or that precipitate action in the face of Vitousek 1992). Other components of global uncertaintymightberegrettedlateraresubsetsof change are likely to contribute more and more in this point of view.) the next century. What is the point at which it really matters? If these human-caused extinctions continue, How would we know if it had been approached the next few decades will entrain a mass extinc- or exceeded? We cannot answer these questions tion of a magnitude greater than any since the definitively for several reasons. First, for many Cretaceous-Tertiary boundary 65 million years components of global change, the level of uncer- ago. This loss of diversity is by far the least tainty about the overall consequences of change reversiblecomponentof globalchange.Thegreen- is relatively high. We know that human activity house gases we are concerned with have atmo- alters soil fertility and water quality over large spheric lifetimes lasting from a decade to a little areas of Earth; we also know that human activity more than a century; their concentrations could is causing a substantial increase in rates of popu- return to background levels in at most a few lation and species extinction. We know that ex- centuries if anthropogenic forcing wereremoved. tinctionimpoverishes the genetic library that sup- Climate might be a little slower to respond due to ports agriculture and health care. However, the buffering by the oceans; the restoration of soil extent to which this loss of diversity itself affects fertility on severely degraded sites could take a aspects of how the ecosystem functions, such as millennium or two. In contrast, overall levels of the maintenance of soil fertility or water quality, species diversity (in terms of the number of spe- is not known (Lawton and Brown 1993; Vitousek cies on Earth) might recover from a catastrophic and Hooper 1993). Without this information, we mass extinction in a million years, and the loss of cannotpredictwithanyconfidencehowmany(or particular species and their genetic information which) species could be lost before ecosystem would be permanent. functions that support humanity aredegraded by direct consequences of the loss of diversity. Both the Ecological Society of America (Lubchenco Human response to global change and others 1991) and the international Scientific Committee on Problems of the Environment There is no doubt that human activity causes (Schulze and Mooney 1993) have identified this global change and no mystery about the ultimate as a high-priority area of research, but answers cause. The scale of human activity-the product will not come quickly. of our population and our effect on the rest of the Second, thereiseveryreason to believe that the Earth system-has become large relative not just Earth system will not respond gradually and to that of other species but also to the flow of evenly to global change. Rather, substantial lags, energy and materials on a global scale. Global nonlinearities, thresholds, and interactions can effects on the atmosphere, on land use, and on be anticipated even if the human-caused forcing other species are already clear; effects on climate functions themselves vary gradually and con- and other components of the Earth system are tinuously. The best current example of a nonlin- coming. There are no serious arguments against ear response to change is the Antarctic ozone these points. Where arguments exist, they con- hole. The effect of CFCs on ozone depletion had cern how bad the changes will be and how soon been identified, but no one predicted the rapid they will occur. The specific arguments are: Do development of the ozone hole, which was dis- we need this species? Will that change in climate covered more or less by accident (Farman, reduce our gross national product (GNP) signifi- Gardiner, and Shanklin 1985; Rowland 1989). cantly? More generally, while an informed and Climatic change in the past also involved sig- reasonable person must concede that the current nificant thresholds and nonlinearities. The cli- levels of human impact on the Earth system are matic oscillation between glacial and interglacial 61 Defining and Measuring Sustainability: The Biogeophysical Foundations conditions (COHMAP Project 1988; Imbrie 1985) (and should) be addressed by research programs. involves rapid changessuperimposed on gradual Many are now being studied more or less system- changes in forcing functions; positive feedbacks atically. This research will contribute to showing based on ice cover, atmospheric carbon dioxide how the world works and how it is altered by concentrations, and perhaps cloud condensation human activity. It may identify some surprises in nuclei produced by marine phytoplankton all advance, and it can suggest areas where change is could contribute to the oscillation (Houghton, occurring particularly rapidly. However, it can Jenkins, and Ephraums 1990). On a shorter time neverbeenough to direct policyunambiguously. scale, there is evidence that patterns of ocean By the time we determine the significance of each circulation can change quite rapidly from one speciesina tropical rain forestin termsofitseffect quasi-stable state to another, driving very rapid, on ecosystem function, very likely more than half substantial, and relatively persistent changes in of these species will be extinct. We will never the global climate (Broecker 1987). Human forc- identify (in advance) all of the surprises that will ing is now driving Earth's climatic system into occur as the Earth system changes, and monitor- conditionsthatdifferfromthoseatanytimeinthe ing will never be sufficient to detect early warn- Pleistocene; it is unlikely that we will be able to ing signs of all the important components of use the past record to guess what surprises global change. Therefore, research should not be (nonlinearities, thresholds, and so forth) these considered a substitute for action designed to newconditionswillbring,but thepastdoes tell us reduce human impacts on the biosphere. More- that some surprises are very likely. over, policies guiding this action must be based Finally, environmental monitoring is now in- on the most current scientific understanding of adequate to pick up many likely changes in the these impacts and must recognize the inherent global environment, including some of those that difficulty in making precise predictions about could affect humanity most directly. The mea- highly complex systems. surements of atmospheric carbon dioxide that Indeed,ourcurrentknowledgeabout thescope, have been made over the past thirty-five years are significance, and variety of global changes result- tremendously useful; it is difficult to imagine ingfromhumanactivityshouldcatalyzeimmedi- where we would be without them. They demon- ate action to reduce these impacts. Specifically, strate global change unambiguously. There have prompt, vigorous actions should be initiated to been a number of attempts to use such measure- reduce the rate of growth of the human popula- ments further to determine the sources and sinks tion, reduce the use of energy, reduce the con- of carbondioxideglobally(see,forexample,Tans, sumption of resources, and implement policies Fund, and Takahashi 1990), but the measure- and practices that are sustainable at local, re- ments are carried out in so few sites (and those gional,and global scales. Noneof thescobjectives sites are so removed from local sources of varia- is easy; each is essential. tion) that it is difficult to obtain information at a Despite many uncertainties about how to scale finer than an entire hemisphere. achieve theseobjectives, it is imperative that new, Our ability to detect global change in land use more responsible policies and practices be imple- is much worse, even though remote-sensing tech- mented. Programs such as the Sustainable Bio- nology is available and appropriate to the task. sphere Project of SCOPE (the Scientific Commit- Equally important, our knowledge of the distri- tee on Problems of the Environment) are de- bution and changing patterns of Earth's biologi- signed to help identify options for a more sustain- cal diversity remains haphazard. able use of resources. These options must be All of these concerns present impediments to based on the recognition that sustainability must ourabilitytoanalyzetheeffectsof human-caused be evaluated at not only the local, but also the changes in the Earth system, and all of them can regional and global scales. 62 Limits to Sustainable Use of Resources: From Local Effects to Global Change References Ehrlich, Paul R., and E. 0. Wilson. 1991. "Biodiversity Studies: Science and Policy." Andreae, M. O., E. V. Browell, M. Garstang, G. L. Science 253, pp. 758-62. Gregory,and R.C. Harris. 1988."BiomassBurn- Farman, J. C., B. G. Gardiner, and J. D. Shankliii. ing and Associated Haze Layers over 1985. "Large Losses of Total Ozone in Antarc- Amazonia." Journal of Geophysical Research 93, tica Reveal Seasonal C10/NO. Interaction." pp.1509-27. Nature 315, pp. 207-10. Bazzaz, F. A. 1990. "The Response of Natural Fishman, J., K. Fakhruzzaman, B. Cros, and D. Ecosystems to Rising Global CO2 Levels." An- Nganga. 1991. "Identification of Widespread nual Review of Ecological Systerns 21, pp. 167-96. Pollution in theSouthern Hemisphere Deduced Birkeland, C. 1992. "Differences among Coastal from Satellite Analyses." Science 252, pp. 1693-96. Systems: The Controlling Influences of Nutri- Howarth, R. W. 1988. "Nutrient Limitation of ent Input and the Practical Implicatioins for Primary Production in Marine Ecosystems." Management." In Coastal Systems Studies and Annual Review of Ecologyand Systematics 19, pp. Sustainable Development, Proceedings of the 89-10. COMAR interregional scientific conference. Houghton, J. T., G. J. Jenkins, and J. J. Ephraums, UNESCO Technical Papers in Marine Science eds. 1990. Climate Change: The IPCC Scientific 64. Paris: UNESCO. Assessment. Cambridge, England: Cambridge Bolin, B., P. J. Crutzen, P. M. Vitousek, R. G. University Press. Woodmansee, E. D. Goldberg, and R. B. Cook. Imbrie,J. 1985. "A Theoretical Framework for the 1983. "Interactions of Biogeochemical Cycles." Pleistocene Ice Ages." Journal of the Geological In B. Bolin and R. B. Cook, eds., The Biogeochemi- Society of London 142, pp. 417-32. cal Cycles and Their Interactions, pp 8-40. Sceyo odn12 p 1-2 Chichester, England: John Wiley and Sons. Keeling, C. D., R. B. Bacastow, A. F. Carter, S. C. Piper, and T. P. Whorf. 1989. "A Three Dimen- Broecker, W. S. 1987. "Unpleasant Surprises in sional Model for Atmospheric CO2 Transport the Greenhouse." Nature 328, pp. 123-26. Based on Observed Winds. 1: Analysis of Ob- Cicerone, R. J., and R. Oremland. 1988. "Bio- servational Data." Geophysical Monographs 55, geochemical Aspects of Atmospheric Meth- pp. 165-236. ane." Global Biogeochemical Cycles 2, pp.299-327. Keller, M., D. J. Jacob, S. C. Wofsy, and R. C. COHMAP Project. 1976. "Climatic Changes of Harris. 1991. "Effectsof Tropical Deforestation the Last 18,000 Years: Observations and Model on Global and Regional Atmospheric Chemis- Simulations." Science 241, pp. 1043-52. try." Climatic Change 19, pp. 145-58. Crisp, D. J. 1975. "Secondary Productivity in the Lawton,J.H.,and V.K.Brown.1993."Redundance Sea." In D. E. Reichle, J. E. Franklin, and D. W. in Ecosystems." In E.-D. Schulze and H. A. Goodall, eds., Proceedings of a Symnposium on Pro- Mooney, eds., Biodiversity and Ecosystem Func- ductivity of World Ecosystems, pp. 71-89. Wash- tion, pp. 255-70. Berlin: Springer-Verlag. ington, D.C.: National Academy of Sciences. Lewis, J. B. 1977. "Processes of Organic Produc- Crutzen, P. J., and M. 0. Andreae. 1990. "Biomass tion on Coral Reefs." Biology Review 52, pp. Burning in theTropics: Impact on Atmospheric 305-47. Chemistry and Biochemical Cycles." Science Logan,J. A. 1985. "Tropospheric Ozone: Seasonal 250, pp. 1669-78. Behavior, Trends, Anthropogenic Influence." Crutzen, P. J., and P. H. Zimmerman. 1991. "The Journal of Geophysical Research 90, pp. 10, 463- Changing Photochemistry of the Atmosphere." 10, 482. Tellus 43, pp. 136-51. Lubchenco, J., A. M. Olson, L. B. Brubaker, S. R. D'Antonio, C. A., and P. M. Vitousek. 1992. "Bio- Carpenter, M. M. Holland, S. P. Hubbell, S. A. logical Invasions by Exotic Grasses, the Grass- Levin, J. A. MacMahon, P. A. Matson, J. M. fire Cycle, and Global Change." Annual Review Melillo, H. A. Mooney, C. H. Peterson, H. R. of Ecology and Systematics 23, pp. 63-87. Pulliam, L. A. Real, P. J. Regal, and P. G. Risser. Ehrlich, Paul R., and H. A. Mooney. 1983. "Extinc- 1991. "The Sustainable Biosphere Initiative: tion, Substitution, and Ecosystem Services." An Ecological Research Agenda." Ecology 72:2, BioScience 33, pp. 248-54. pp. 371-412. 63 Defining and Measuring Sustainability: The Biogeophysical Foundations Ludwig, D., R. Hilborn, and C. Walters. 1993. DeclineinaSpruce(Piceaabies) Forest." Science "Uncertainty, Resource Exploitation, and Con- 244, pp. 776-83. servation: Lessons from History." Science 260, Schulze, E.-D., and H. A. Mooney, eds. 1993. pp. 17-18. Biodiversity and Ecosystem Function. Berlin: Matson, P. A., and P. M. Vitousek. 1990. "Ecosys- Springer-Verlag. tem Approach to a Global Nitrous Oxide Bud- Shukla, J., C. Nobre, and P. Sellers. 1990. "Amazo- get." BioScience 40, pp. 667-72. nian Deforestation and Climate Change." Sci- Molina, M. J., and F. S. Rowland. 1974. "Strato- ence 247, pp. 776-83. spheric Sink for Chlorofluoromethanes: Chlo- Siegenthaler, U., and H. Oeschger. 1987. "Bio- rine Atomic Catalysed Destruction of Ozone." spheric CO2 Emissions during the Past 200 Nature 249, pp. 810-12. Years Reconstructed by Deconvolution of Ice Mooney, H. A., B. C. Drake, R. J. Luxmoore, W. C. Core Data." Tellus 39B, pp. 140-54. Oechel, and L. F. Pitelka. 1991. "Predicting Smith,S.V.,andR.W.Buddemeier.1992."Global Ecosystem Responses to Elevated CO2 Con- Change and Coral Reef Ecosystems." Annual centrations." BioScience 41, pp. 96-104. Review of Ecologyand Systematics 23, pp.89-118. Munro, J. L., and D. M. Williams. 1985. "Assess- Stuiver, M. 1978. "Atmospheric Carbon Dioxide mentand Management of Coral Reef Fisheries: and Carbon Reservoir Changes." Science 199, Biological, Environmental, and Socio-economic pp. 253-58. Aspects." Proceedings of the Fifth International _ ,_ . . ~~~~~~~~Tans, P. 0., 1. Y. Fund, and T. Takahashi. 1990. Coral Reef Congress, Tahiti, vol.4, pp. 543-81. "Observational Constraints on the Global At- Nye, P. H., and D. J. Greenland. 1960. "The Soil mosphereCO Budget." Science 247, pp. 1431-38. under Shifting Cultivation." Technical Comnmu- Thompson, A. M. 1992. "The Oxidizing Capacity nications (Commonwealth Bureau of Soils, of the Earth's Atmosphere: Probable Past and Farnham Royal, England) 51. Future Changes." Science 256, pp. 1157-64. Ojima, D. S. 1987. "The Short-term and Long- Turner, B. L. II, W. C. Clark, R. W. Kates, J. F. term Effects of Burning on Tall-Grass Prairie Richards, and J. T. Matthews. 1990. The Earth as Ecosystem Properties and Dynamics." Ph.D. Transformed by Human Action. Cambridge, En- diss., Colorado State University. gland: Cambridge University Press. Parton, W. J., D. S. Schimel, C. V. Cole, and D. S. gland: Cambr Uvs P ress. Ojima. 1987. "Analysis of Factors Controlling Vitousek, Peter M., Paul R. Ehrlich, A. H. Ehrlich, Soil Organic Matter Levels in Great P'lains and P. A. Matson. 1986. "Human Appropria- Gras san il Sier Sety of Areca Iolrn tion of the Products of Photosynthesis." Grasslands." SoilScienceSocietyof rmericalour- BiSiec 346 p 6-3 nal 51, pp. 1173-79. BioScience 34:6, pp. 368-73. Peierls B. L. N.F. Caraco M. L. Pace, and S. . Vitousek, Peter M., and D. U. Hooper. 1993. "Bio- Peierls, B. L., N. F. Caraco' M. L. Pace, and logical Diversity and Terrestrial Ecosystem Bio- Cole. 1991. "Human Influence on River Nitro- geochemistry." In E.-D. Schulze and H. A. gen." Nature 350, pp. 386-87. Mooney, eds., Biological Diversity and Ecosys- Prather, M. J., and R. r. Watson. 1990. "Strato- tern Function, pp. 3-14. Berlin:Springer-Verlag. spheric Ozone Depletion and Future Levels of Vitousek, Peter M., and P. A. Matson. Forthcom- Atmospheric Chlorine and Bromine." Nature ing. "Agriculture, the Global Nitrogen Cycle, 344, pp. 729-34. and Trace Gas Flux." In R. Oremland, ed., Rowland, F. S. 1989. "Chlorofluorocarbons and Biogeochernistryof Global Change: RadiativeTrace the Depletion of Stratospheric Ozone." Ameri- Gases. New York: Chapman & Hall. can Science 77. pp. 42-44. Watson, R. T., H. Rodhe, H. Oeschger, and U. Russ,G. 1985. "Effects of Protective Management Siegenthaler. 1990. "Greenhouse Gases and on Coral Reef Fishes in the Central Philip- Aerosols." InJ.T.Houghton,G.J.Jenkins, and pines." Proceedingsof the Fifthinternational Coral J. J. Ephraums, eds., Climate Change: The IPCC Reef Congress,Tahiti, vol.4, pp.219-24. Scientific Assessment, pp. 1-40. Cambridge, Schliesinger, W. H. 199 1. Biogeochemistry: An Analy- England: Cambridge Uniiversity Press. sis of Global Change. San Diego, Calif.: Aca- Wilson, E. O., and F. M. Peter, eds. 1988. demic Press. Biodiversity. Washington, D.C.: National Acad- Schulze, E.-D. 1989. "Air Pollution and Forest emy of Sciences. 64 Sustainability: The Cross-Scale Dimension C. S. Holling A new class of problems is challenging the ability to achieve sustainable development: * These problems are more and morefrequently caused by slowly accumulated human influences on air, land, and oceans that trigger sudden changes that directly affect the health of people, the productivity of renewable resources, and the vitality of societies. * The spatial span of connections is intensifying so that the problems are nowfundamentally cross-scale in space as well as in time. * The problems are essentially nonlinear in causation and discontinuous in both their spatial structure and temporal behavior. * Both the ecological and social components of these problems have an evolutionary character. The problems are therefore not amenable to solutions based on knowledge of small parts of the whole nor on assumptions of constancy or stability offundamental relationships: ecological, economic, orsocial. Such assumptions produce policies and science that contribute to a pathology of rigid and unseeing institutions, increasingly brittle natural systems, and public dependencies. But recent advances in theory, method, and regional experience are leading to a truly cross-scale understanding and to the identification of the attributes of renewal capital that are the foundations for sustainable development in a world of surprises. In the most fundamental sense, the renewal capital for nature is the physical structure of the environment that sustains and is controlled by the biota at all scales. For people, it is social trust and accessible knowledge. The biophysical dimensions of sustainable devel- tion. Partial policies fail. Integrated policies may opment cannot be separated neatly from the eco- have a chance to succeed. nomic or the social dimensions. To attempt to do Lamentably, partial policies are more comfort- so would encourage piecemeal strategies of in- ably congruent with the disciplinary expertise vestments that have failed to improve the status thatisanimportantfoundationforeducationand of people. Those strategies have invested in parts research. But a biologist's or an ecologist's of the whole, typically investments in resource discipline-based design for sustainability cannot development, while ignoring the responses of be trusted any more than an economist's or an nature and the adaptive traditions of people. The engineer's. Doing so leads to a disciplinary and present recognition of the role of nature in issues policy myopia that generates the very problems of sustainability is certainly an advance, but not if and conflicts that sustainabledevelopment is sup- that appreciation simply encourages a policy lurch posed to address. away from narrow economic development and I argue here, however, that practical ways are toward equally narrow environmental protec- emerging to measure and invest in sustainable Defining and Measuring Sustainability: The Biogeophysical Foundations development that draw on a spectrum of disci- extremes of weather intersect with increasing plinary scholarship within a framework that leads vulnerability of the stand as treesbecome mature. to integrated understanding and integrated poli- For different tree species, in different regions of cies. This has become possible with recent ad- the boreal forest, the natural disturbance might vances in theory, method, and the sciences them- be an outbreak of insects or a forest fire. selves and from regional experience in the resto- Sustainability at that scale can be seen as the ration of ecosystems. This approach is leading to maintenance of successional cycles of stand-level truly cross-scale understanding and to the identi- boom-and-bust to produce a perpetuating mo- fication of the attributes of renewal capital that saic of standsof trees of different ages, each stand are the foundations for sustainable development covering 100 to 1,000 hectares. The resource capi- in a world of surprises. tal responsible for maintaining that pattern is the In the most fundamental sense, the renewal set of biotic and abiotic processes that perpetuate capital for terrestrial nature is the physical archi- thedynamic mosaic. It canbe measured by physical tecture of the biophysical environment that sus- attributes of vegetation patterns and by climate. tains and is controlled by the biota at all scales. By this time,and at thisgeographicextent,Iam For people, it is social trust and accessible and describing a good part of the present unlogged, usablc knowledge. unmanaged, high-latitude forest of North America, Europe, or Asia. Now I am at a scale where there are groups of ecosystems of conifer- A cross-scale journey ous and mixed forests and of lakes, bogs, and wetlands. They aggregate to form the boreal for- Metaphors can help clarify complex and appar- est biome whose existence is itself a passing and ently paradoxical notions such as sustainable transient thing, which emerged in its present development. Do ecosystems-their structure, form perhaps 8,000 years ago following the re- function, and behavior-provide a useful meta- treat of the ice sheets. Pollen records demonstrate phor? That depends on the scale of observation, that the aggregation of tree species following the both in time and space. retreat of the ice sheets was a highly individualis- If I observe a 400-year-old, 1,000 hectare stand tic process depending on an individual species' of Douglas fir trees in British Columbia from the response toweather,uniquedispersal properties, perspective of my three-score years and 3 or 4 and distance to the source of seeds. The processes kilometers of easy walk, I see a true model of defining the system at this scale now include sustainability. The resource capital that measures geophysical cycles thatareresponsible for rhythms thatparticularperspectiveofsustainabilitymight of glaciation, erosion, and land movement. be the standing biomass, just as the new genera- I could go on in this journey in time and space tion of resource economists and accountants now and only stop when I encounter the whole uni- propose. verse and the big bang of its origin. Or I could If I view the stand from the perspective of the proceedintosmallerscalesfromthestartingpoint tree's lifetime, not mine, however, I see of a stand of 400-year-old trees and pose ques- sustainability as perpetuation of a 400-year pe- tionsof persistenceofpatcheswithinwhichplants riod of tree growth, which was initiated by a compete for water, nutrients, and light, then per- major disturbance covering at least 1,000 hect- sistence of plants within those patches, persis- ares. The resource capital allowing that period of tence of branches, persistence of leaves and growth might be measured by the nutrient-hold- needles, and so on. Each set of questions would ing attributes of soil, not by the standing biomass. provide a different perspective on sustainability If I extend time still further to several lifetimes and a different way to measure it. of trees and expand my spatial perspective to a I am not dwelling on these different sub-continental scale, I realize that the originat- scale-dependent perspectives in order to claim ing disturbance events are periodic. Such distur- that sustainability is so relative as to be meaning- bances are not intrusions from outside but are an less. Rather, I do so to illustrate that a universal inherent part of ecosystem succession. In thecase feature occurs at all these scales, from needle to of a stand on the storm-swept west coast of planet.Eachdescriptionisacycleofbirth,growth, Vancouver Island, the disturbance could well be death, and renewal. What sustains such cycles? a windstorm capable of clear-cutting many hun- Oddly, the processes of death and renewal rather dreds of hectares as a normal process whenever than those of birth and growth lie at the heart of 66 Sustainability: The Cross-Scale Dimension sustainability. That is where we need to search for borrowed from the economist Schumpeter (as measures of sustainability: measures of distur- reviewed in Elliott 1980), in which the tightly bance and of the capacity to renew after distur- bound accumulation of biomass and nutrients bance. Consider the succession of ecosystems. becomes increasingly fragile (overconnected in systems terms) until it is suddenly released by agents such as forest fires, insect pests, or intense Ecosystem function pulses of grazing. The second is one of reorgani- zation, in which soil processes of mobilization Over the last decade, the literature on ecosystems and immobilization minimize nutrient loss and has led to major revisions in the original reorganize nutrients to become available for the Clementsian view of succession. That initial view next phase of exploitation. was one of a highly ordered sequence of species During this cycle, biological time flows un- assemblages moving toward a sustained climax evenly. The progression in the ecosystem's cycle whose characteristics are determined by climate proceeds from the exploitation phase (box 1 of andedaphicconditions.Thisrevisioncomesfrom figure 4-1) slowly to conservation (box 2), very extensive comparative field studies (West, rapidly to release (box 3), rapidly to reorganiza- Shugart, and Botkin 1981), from critical experi- tion (box 4), and rapidly back to exploitation. mental manipulations of watersheds (Bormann During the slow sequence from exploitation to and Likens 1981; Vitousek and Matson 1984), conservation, connectedness and stability in- frompaleoecological reconstructions(Davis 1986; crease, and a capital of nutrients and biomass is Delcourt, Delcourt, and Webb 1983), and from slowly accumulated. That capital becomes more studies that link models of systems and field and more tightly bound, preventing other com- research (West, Shugart, and Botkin 1981). petitors from using it until the system eventually The revisions include four principal points. becomes so overconnected that rapid change is First, the species that invade after disturbance triggered. The agents of disturbance might be and during succession can be highly variable and wind, fire, disease, insect outbreak, or a combina- determined by chance events. Second, both early tion of these. The stored capital is then suddenly and late successional species can be present con- released, and the tight organization is lost to tinuously. Third, large and small disturbances allow the released capital to be reorganized and triggered by events like fire, wind, and herbi- the cycle to begin again. vores are an inherent part of the internal dynam- ics and in many cases set the timing of succes- sionalcycles.Fourth,somedisturbancescancarry Figure 4-1: Flow of Events between Four Ecosystem the ecosystem into quite different stability do- Functions mains: for example, mixed grass and tree savan- 5 l nas turn into shrub-dominated semi-deserts g 4 Reorganization 2 Conservation (Walker 1981); that is, more than one climax state A is possible. I,- In summary, therefore, the notion of a sus- - tained climax is a useful but essentially static and incomplete equilibrium view. The combination Nf 4 * Accessible carbon C K-stratex of these advances in understanding ecosystems energy *Consolidaton by studying population systems has led to one trate A version of a synthesis that emphasizes four pri- . R-strateg / * Fire marystagesinanecosystem'scycle(Holling 1986). X * Opportunist * Pest The traditional view is that succession of an ,/ ecosystem is controlled by two functions: exploi- A tation, in which rapid colonization of recently disturbed areas is emphasized, and conservation, in which slow accumulation and storage of en- _. Exploitatton 3. Reaease ergy and material is emphasized. But the revi- 3 sions in understanding indicate that two addi- tional functions are needed (see figure 4-1). One is Weak t Connectedness - Strong that of release, or creative destruction, a term 67 Defining and Measuring Sustainability: The Biogeophysical Foundations The arrows show the speed of that flow in the structures to dynamic entities whose levels are cyde: arrows dose to each other indicate a rapidly vulnerable to small disturbances at certain critical changing situation, and arrows far from each other times in the cycle (Holling 1992). That represents a indicate a slowly changing situation. The cycle re- transient but important bottom-up asymmetry. flectschangesintwoattributes:(1)theYaxis,which There are two key states in which slower and is the amount of accumulated capital (nutrients, larger levels in ecosystems become briefly vul- carbon) stored in variables that are the dominant nerable to dramatic transformation because of keystonevariablesatthemoment,and (2)the Xaxis, small events and fast processes. One is when the which is the degree of connectedness among vari- system becomes overconnected and brittle as it ables.Theexit from thecyde, shown at the leftof the slowly moves toward maturity (box 2 of figure 4- figure, indicates the stage at which a flip is most 1). Atthisstage, relationsamong the plant species likely to lead into a less or more productive and aretightlycompetitive.Fromanequilibriumper- organized system, that is, devolution or evolution spective, the system is highly stable (that is, re- as revolution! turn times are fast in the face of small distur- That pattern is discontinuous and is depen- bances),butfromaresilienceperspective(see,for dent on the existence of changing multi-stable example, Holling 1987), the domain over which states that trigger and organize the release and stabilizing forces can operate becomes increas- reorganization functions. Instabilities and cha- ingly small. Brittleness comes from such a loss of otic behavior trigger the release phase, which resilience. Hence the system becomes an accident then proceeds in the reorganization phase, where waiting to happen. In the boreal forest, for ex- stabilitybeginstobereestablished.Inshort,chaos ample, the accident might be a contagious fire emerges from order, and order emerges from that becomes increasingly likely as the amount, chaos! Resilience and recovery are determined by extent, and flammability of fuel accumulate. Or it the fast release and reorganization sequence, could be an outbreak of insects that spreads as whereas stabilityand productivityaredeterrnined increasing amounts of foliage both increase food by the slow exploitation and conservation se- and habitat for defoliating insects and decrease quence. the efficiency with which their vertebrate preda- Moreover, there is a nested set of such cycles, tors search for them (Holling 1988). each with its own range of scales. In the typical boreal forest, for example, fresh needles cycle yearly, the crown of foliage cycles with a decadal Figure 42: Hierarchies of Space and Time for period, and trees, gaps, and stands cycle at peri- Forests and Atmospheres ods of close to a century or longer. The result is a Topography hierarchy in which each level has its own distinct Micro Meso Macro spatial and temporal attributes (see figure 4-2). 1 10 1 10 1000 cm m m km 100 Dynamics of hierarchies 4 -10,000 years A critical feature of such hierarchies is the asym- -2( _ Blome Century metric interactionsbetween levels (Allen and Starr 2 Patch-Forest 1982; O'Neill and others 1986). In particular, the S-tan El Nioc larger, slower levels maintain constraints within 0 o +- Year which faster levels operate. In that sense, there- e Cro4wn H fore, slower levels control faster ones. If that were -2 Needle Tropical stormnc5- onth the only asymmetry, however, it would be im- - Front C Day possible fororganisms to exert control over slower 4Conv. stonm (7) Hour environmental variables. Many geologists criti- 4 cize the Gaia theory on these grounds (Lovelock Breeze- -Minute 1988): How could slow geomorphic processes -6 microburst, possibly be affected by fast biological ones? How- lightnig stnke ever, it is not broadly recognized that the birth, -8 - II I I I I I growth, death, and renewal cycle, shown in fig- -6 -4 -2 0 2 4 ure 4-1, transforms hierarchies from fixed static Space - log kms 68 Sustainability: The Cross-Scale Dimension Small and fast variables can also dominate ture of soils that, if retained, allows a number of slow and large ones at the stage of reorganization different kinds of ecosystem to flourish, with (box 4 of figure 4-1). At this stage, the system is trees or not? Is it the mosaic of architectural underconnected, with weak organization and features that maintains the diversity of habitats weak regulation. As a consequence, this is the and species? I cannot comfortably answer these stage most affected by probabilistic events that questions, but I can identify a direction in which allow a diversity of entrained species, as well as to search for the answers. exotic invaders, to become established. On the one hand, it is the stage most vulnerable to ero- sion and to the loss of accumulated capital. On the Two puzzles of sustainability other hand, it is the stage from which jumps to unexpectedlydifferentand more productive sys- Recently, I resolved two difficult puzzles that had tems are possible. At this stage, instability comes emerged during a review of some twenty-three from the loss of regulation rather than the brittle- examples of managed ecosystems (Holling 1986). ness of reduced resilience. Those examples fell into four classes: forest in- The degree to which small, fast events influ- sect, forest fire, savanna grazing, and aquatic ence larger, slower ones is critically dependent on harvesting. One puzzle seemed to be a paradox of the accumulation, cycling, and conservation of theorganizationofecosystems.Theotherseemed accumulated capital. And thatin tum dependson to be a paradox of the management of ecosys- the meso-scale disturbance processes. Human tems. Both have turned out to be the consequence management of renewable resources or impacts of the natural workings of any complex, evolving of macroscale phenomena, such as change in system. climate, can release a pattern of disturbance that The first paradox suggested that the great di- destroys large amounts of accumulated renewal versity of life in ecosystems is traceable to the capital over large areas. If too much capital is function of a small set of variables, each operating destroyed over too large an area, the system can at a qualitatively different speed from the others. flip into a qualitatively different stable state that The second suggested that any attempt to man- persists unless there is explicit rehabilitation age ecological variables inexorably leads to more through management. As an example, that is why brittle ecosystems, more rigid management insti- grazing at sustained, extensive, but moderate tutions, and more dependent societies. I shall levels can transform productive savannas into deal with each in turn. less-productive systems dominated by woody shrubs (Walker and others 1969). Or why success- ful efforts at controlling forest fires can lead to so The ecosystem organization puzzle much accumulated fuel over such a large area that the inevitable runaway fire destroys accu- How could the great diversity within ecosystems mulated soil and the capacity of trees to regenerate. possibly be traced to the function of a small num- The question for issues of human transforma- ber of variables? The models that were developed tion from the scale of fields to that of the planet, and tested for these examples certainly generated therefore, is how much change it takes to release complex behavior in space and time. Moreover, disturbances whose intensity and extent are so those complexities could be traced to the actions great that the renewal capital is destroyed to the and interactions of only three to four sets of point where regeneration of plants is seriously variablesandassociatedprocesses,eachofwhich compromised. Thus, two sets of questions need to operated at distinctly different speeds. The speeds be addressed in order to assess the sustainability were therefore discontinuously distributed and of development. First, how much disturbance differed from their neighborsoftenby asmuch as will be generated, for how long, and over what an order of magnitude. A summary of the critical areas? Second, how much renewal capital is de- structuring variables and their speeds are pre- stroyed, for how long, and over what areas? sented in table 4-1. For the models, at least, this That still leaves open the question of what structure organizes the time and space behavior specific kinds of renewal capital are the measure of variables into a small number of cycles, pre- of sustainability. Is it the number of trees as the sumably abstracted from a larger set that contin- newresourceeconomistssuggest?Butwhyshould ues at smaller and larger scales than the range trees be the only future permitted? Is it the struc- selected. 69 Defining and Measuring Sustainability: The Biogeophysical Foundations But are those features simply the consequence (1) A small number of plant, animal, and abiotic of the way modelers make decisions rather than processes structure biomes over scales from the results of an ecosystem's organization? This days and centimeters to millennia and thou- uneasy feeling that such conclusions can be a sands of kilometers. Individual plant and bio- figment of the way we think, rather than of the geochemical processes dominate at fine, fast way ecosystems function, led to a series of tests scales; animal and abiotic processes of using field data to challenge the hypothesis that meso-scale disturbance dominate at interme- ecosystem dynamics are organized around the diatescales;andgeomorphologicalonesdomi- operationof a small numberof nested cycles, each nate at coarse, slow scales. driven by a few dominant variables. (2) These structuring processes produce a land- The critical argument is that, if there are, in scape that has lumpy geometry and lumpy fact, only a few structuring processes, their im- temporal frequencies or periodicities. That is, print should be expressed on most variables. That the physical architecture and the speed of vari- is, time-series data for fires, intensity of seeding, ables are organized into distinct clusters or number of insects, flow of water-indeed any quanta, each of which is controlled by one variable for which long-term, yearly records are small set of structuring processes. These pro- available-should show period icities that cluster cesses organize behavior as a nested hierarchy around a few dominant ones. In the case of the of cycles of slow production and growth alter- eastern maritime boreal forest of North America, nating with fast disturbance and renewal (as for example, those periodicities were predicted to shown in figure 4-1). be three to five years, ten to fifteen years, thirty- (3) Each quantum is contained to a particular five to forty years, and more than eighty years. Similarly,afewdominantspatial "footprint" sizes range of scales in space and time and has Its similalyd fewdsoninantwith one of the cycles of own distinct architecture of object sizes, dis- should be associated tance between objects, and fractal dimension disturbance and renewal in the nested set of such within that range. cycles. Finally, the animals living in specific land- g scapes should demonstrate the existence of this (4) All of the many remaining variables, other lumpyarchitecturebyshowinggapsinthedistri- than those involved in the structuring pro- bution of their sizes and gaps in the scales at cesses, become entrained by the critical struc- which decisions are made for location of region, turing variables, so that the great diversity of foragingarea,habitat,nests,protection,and food. species in ecosystems can be traced to the All the evidence we have so far confirms just function of a small set of variables and the those hypotheses for boreal forests, for boreal niches they provide. The structuring processes region prairies, for pelagic ecosystems (Holling are the ones that both form structure and are 1992), and for the Everglades of Florida affected by that structure. These structuring (Gunderson 1992). Various alternative hypoth- variables are, therefore, where the priority esesbasedondevelopmental,historical,ortrophic should be placed in investing to protect arguments were disproved in the fine traditions biodiversity. of Popperian science, leaving only the (5) The discontinuities that produce the lumpy world-is-lumpy hypothesis as resisting disproof. structure of vegetated landscapes impose Therefore, there is strong evidence for the fol- discontinuities on the behavior and morphol- lowing conclusions: ogy of animals. For example, gaps in the distri- Table 4-1: Key Variables and Speeds in Five Groups of Managed Ecosystems System Fast Intermediate Slow Forest insect lnscct, needles Foliage crown Trees Forcst fire Intensity Fuel Trees Savanna Annual grasses Perennial grasses Shrubs Aquatic Phvtoplankton Zooplankton Fish Source: The key references are McNamee, McLeod, and Holling 1981 for forest insects; Holling 1980 for forest fires; Walker and others 1969 for annual grasses, and Steele 1985 for phytoplankton. 70 Sustainability: The Cross-Scale Dimension bution of body mass of resident species of economic activity. In the case of management of animals correlate with scale-dependent eastern North American spruce and fir forests, discontinuities in the geometry of vegetated the target was an anticipated outbreak of a defo- landscapes. Thus these gaps, and the body liating insect, the spruce budworm (Clark and mass clumps they define, become a way to others 1979); for the forests of the Sierra Nevada develop a rapid bioassay of ecosystem struc- Mountains, the target was forest fires (Holling tures and of human impacts on that structure. 1980); for the savannas of South Africa, the target This opens the way to develop a comparative was the grazing of cattle (Walker and others ecology across scales that might provide the 1969); for the salmon of the Pacific northwest coast, same power for generalization that came when the target was salmon populations (Walters 1986). physiology became comparative rather than In each case t`ie goal was to control the vari- species specific. ability of the target: insects and fire at low levels, cattle grazing at intermediate stocking densities, (6)Conversely,changesinlandscapestructureat and salmon at high populations. The level de- defined ranges of scale caused by land use sired wasdifferentineachsituation,butthecom- practice or by changes in climate will have mon feature was to reduce variability of a target predictable impacts on the community whose normal fluctuations imposed problems structure of animals (that is, animals of some and periodic crises for pulp mill employment, body masses can disappear if an ecosystem's recreation,farmingincomes,or fishermen'scatch. structure at a predictable range of scale is The typical response to threats of fire or pesti- changed). Therefore predicted (using models) lence, flood, or drought is to narrow the purpose, or observed (using remote imagery) impacts of focus on it exclusively, and solve that problem as changing climate orland use on vegetation can so defined. Modern engineering, technological, also be used to infer the impacts on the diversity economic, and administrative experience can deal of animal communities. well with such narrowly defined problems. And in each example, the goal was successfully The lessons for both sustainable development achieved: insects were controlled with insecti- and biodiversity are clear: focus should be placed cide; the frequency and extent of fires were re- onthestructuringvariablesthatcontrolthelumpy duced with fire detection and suppression tech- geometry and lumpy time dynamics. They are niques; cattle grazing was managed with modem the ones that set the stage upon which other rangelandpractice;andsalmonpopulationswere variables play out their own dramas. That is, the augmented with hatchery production. health and viability of the physical and temporal At the same time, however, elements of the infrastructure of biomes at all scales are what system were slowly changing. First, reducing the sustains the theater; given that, the actors will variability of the ecological target produced a look after themselves! slow change in the spatial heterogeneity of the ecosystem. Forest architecture became more con- tiguous over landscape scales, so that if defoliat- The ecosystem management puzzle ing insects or fire were released, the outbreaks could cover larger areas with more intensive im- An even more surprising and puzzling feature pacts than before management. Rangeland gradu- Aemergedin more comparison and the twentyuree ally lost drought-resistant grasses because of a emerged in the comparison of the twenty-three shift in competition with more productive but examples. All the examples were associated with more drompetitive gre prouct but manaemet ofa rsoure werethe erysuccss oredrought-sensitive grasses. If drought oc- managementofaresourcewheretheverysuccess curred, the consequences were therefore more of management seemed to set the condition for extensive, more extreme, and more persistent: collapse. Is there some general property of grasslands turned irreversibly into shrub-domi- unsustainabilitythatistouchedintheseexamples, nated semi-deserts. Wild populations of salmon or is the observation simply the result of cases in the many streams along the coast gradually selected because of an unconscious attraction to becameextinctbecausefishingpressure increased catastrophic visions? Again, some independent in response to the increased populations achieved tests were necessary. by enhancement. That left the fishing industry Each of the examples represented policies of precariouslydependentonafewhatcherieswhose management whose goal was to control a target productivity declined with time. variable in order to achieve social objectives, typi- In short, the success in controlling an ecologi- cally maintaining or stimulatingemployment and cal variable that normally fluctuated led to more 71 Defining and Measuring Sustainability: The Biogeophysical Foundations spatiallyhomogenized ecosystemsoverlandscape with that conclusion, implying, as it does, that the scales. It led to systems more likely to flip into a only solution is a radical return of humanity to persistent degraded state, triggered by distur- being"childrenof nature."Theargumentisbased bances that previously could be absorbed. Man- on two critical points. One is that reduced vari- agement successfully froze each system in the ability of ecosystems inevitably leads to reduced conservation stage of the cycle of growth and resilience and increased vulnerability. The sec- production, disturbance, and renewal (box 2 of ond is that there is, in principle, no different way figure 4-1), thereby making each system a larger for agencies and people to manage and benefit and larger accident waiting to happen. from the development of resources. Those changes in the ecosystems could be Again some independent evidence is needed. managed if it were not for concomitant changes in Are there counter examples? Oddly, nature itself two other elements of the interrelationships: in provides counter examples of tightly regulated the management institution(s) and in the people yet viable systems in the many cases of physi- .n the society who reaped thebenefits or endured ological homeostasis. Consider temperature regu- the costs. Because of the initial success, in each lation of endotherms (warm-blooded animals), case the management agencies shifted their ob- for example, which represents a system where jectives from the original social and ecological internal body temperature is not only tightly ones to the laudable objective of improving op- regulated within a narrow band, but at an aver- erational efficiency of the agency itself: spraying age temperature perilously close to lethal. More- insects, fighting fires, producingbeef, and releas- over, the costs of achieving that regulation re- ing hatchery fish with as much efficiency and quire ten times the energy for metabolism than is with the least cost possible. Efforts to monitor the required by an ectotherm. That would seem to be ecosystem for surprises rather than for product a recipe for disaster, and a very inefficient one at therefore withered in competition with internal that. Yet evolution somehow led to the extraordi- organizational needs, and research funds were nary success of the animals having such an adap- shifted to moreoperational purposes. Whymoni- tation: the birds and mammals. tor or study a success? Thus the gradual reduc- In order to test the generality of the hypothesis tionofresilienceoftheecosystemswasunseenby of variability loss/resilience loss, I have been any but maverick and suspect academics! collecting data from the physiological literature Successbrought changesin the society,aswell. on the viable temperature range of the internal Dependencies developed for continuing, sus- body for organisms exposed to different classes tained flow of the food or fiber that no longer of variability. The data are organized into three fluctuatedasitoncehad.Moreinvestmentsthere- groups ranging from terrestrial ectotherms fore logically flowed to expanding pulp mills, (cold-blooded animals) exposed to the greatest recreational facilities, cattle ranches, and fishing variability of temperature from unbuffered am- technology. That is the development side of the bient conditions, to aquatic endotherms exposed equation, and its expansion can be rightly ap- to an intermediate level of variability because of plauded. Improving the efficiency of agencies themoderatingattributesofwater,toendotherms should also be applauded. But if, at thesame time, that regulate temperature within a narrow band. the ecosystem from which resources are garnered As predicted, the viable range of internal body becomes more and more brittle, more and more temperature decreases from about 40'C for the sensitive to large-scale transformation, then the most variable group to about 30 C for the inter- efficient but myopic agency and the productive mediate, to 20'C for the tightly regulated endot- but dependent industry simply become part of herms. Resilience, in this case the range of tem- the source of crisis and decision gridlock. peratures that separates life from death, clearly So there is the paradox: success in managing a does contract as experience with variability is target variable for sustained production of food reduced. Therefore, reducing the variability of or fiber apparently leads to an ultimate pathology living systems from organisms to ecosystems in- of more brittle and vulnerable ecosystems, more evitably leads to loss of resilience. rigid and unresponsive management agencies, But that seems to leave an even starker para- and more dependent societies. That seems to dox of control inevitably leading to collapse. But, confirm one opinion that sustainable develop- in fact, endothermy does persist and therefore ment is an oxymoron. serves as a revealing metaphor for sustainable But something seems to be inherently wrong development. This metaphor contains two fea- 72 Sustainability: The Cross-Scale Dimension tures that were not evident in my earlier descrip- grated nature of the ecological, industrial, and tions of examples of resource management. social interrelationships, led to an abrupt trans- First, the kind of regulation is different. Five formationofpolicywhoseattributesbecamemuch different mechanisms, from evaporative cooling like those just described (Baskerville, forthcom- to the generation of metabolic heat, control the ing). Itisa policy that functions forawholeregion temperature of endotherms. Each mechanism is by transformingand monitoringthesmaller-scale not notably efficient by itself. Each operates over stand architecture of the landscape and by focus- a somewhat different range of conditions and ing the productive capacities of industry. with different efficiencies of response. It is this There is even the suspicion, in these examples overlapping "soft" redundancy that seems to char- of regional resource management, that institu- acterize biological regulation of all kinds. It is not tions and society themselves achieve periodic notably efficient or elegant in the engineering advances in understanding and learning through sense, but it is robust and continually sensitive to the same four-box cycle of growth and produc- changes in intemal body temperature. That is tion, release, and renewal that shapes the spatial quite unlike the examples of regulation by man- and temporal dynamics of ecosystems. But each agement where goals of operational efficiency proceedsatitsown paceand in itsown space, and gradually isolate the regulating agency from the that creates extraordinary conflicts when ecosys- things it is regulating. tems, institutions, and societies function on scales Second, endothermy is a true innovation that that are extremely mismatched. If the scale of all explosively released opportunity for the organ- three becomes more congruent, it is likely that the isms evolving it. Maintaining high body tem- inevitable bursts of human learning can proceed perature-short of death-allows the greatest with less conflict and more creativity. range of external activity for an animal. Speed and stamina increase, and activity can be main- tained at both high and low external tempera- Conclusions tures. A range of habitats forbidden to an endot- herm is open to an ectotherm. The evolutionary This chapter has used metaphors and paradoxes consequence of temperature regulation was sud- to provide some insight into what sustainable denly toopenopportunityfordramaticorganiza- development is and what measures its proper- tional change and the adaptive radiation of new ties. The ecosystem metaphor led to the conclu- life-forms. Variability was therefore not elimi- sion that there is a cycle of slow growth and nated. It was reduced and transferred from the production that triggers fast disturbance and re- animal's internal environment to its external one, newal. The slow growth and production phase as a consequence of the continual probes of the accumulatesnatural capital. It is analogous to the whole animal for opportunity and change. Hence processes of what we call development. the price of reducing internal resilience and The fast disturbance and renewal phase re- maintaining high metabolic levels was more than leases bound and constrained capital and reorga- offsetby thatcreation of evolutionaryopportunity. nizes it to reestablish the ecosystem cycle. It is Surely the release of human opportunity is at analogous to the conditions of what we call the heart of sustainable development! It requires sustainability, and it is the phase where diversity flexible,diverse,and redundant regulation, moni- is maintained. Therefore sustainability is mea- toring that leads to corrective responses, and sured by some attributes of disturbance and re- experimental probing of the continually chang- newal, and development is measured by some ing reality of the external world. Those are the attributes of growth and production. features of adaptive environmental and resource A paradox helps clarify the specific attributes management (Holling 1978; Walters 1986). Those that determine sustainability of an ecosystem. are the features missing in the descriptions of The paradox is that a few simple processes seem traditional, piecemeal, exploitive resource man- to generate the great complexity and diversity agement and its ultimate pathology. withinecosystems.Ecosystemsarehierarchically In fact, that is what eventually happened in at structured into a number of levels. Relatively few least one of the examples quoted. In New processes determine this structure, and each im- Brunswick, the intensifying gridlock in forest poses distinct frequencies in space and time on management, combined with slowly accumulated the ecosystem over different ranges of scale. They and communicated understanding of the inte- entrain all other variables. 73 Defining and Measuring Sustainability: The Biogeophysical Foundations Hence both sustainability and biodiversity are versed by normal, internal recovery. That state is determinedbythestructuringvariablesofdistur- the condition defined as poverty, a condition of bance and renewal that control the lumpy geom- inability to cope. etry and lumpy time dynamics. To use another metaphor, they set the stage on which other variables play out their own dramas. The health References and viability of the physical and temporal infra- structure of biomes at all scales sustain the theater; Allen, T. F. H., and T. B. Starr. 1982. Hierarchy: given that, the actors will look after themselves! Perspectives for Ecological Complexity. Chi- A second paradox suggests that many existing cago, Ill.: University of Chicago Press. examplesofmanagementofrenewableresources Baskerville, G. L. Forthcoming. "The Forestry have led inexorably to more brittle ecosystems, Problem." In L. H. Gunderson, C. S. Holling, more rigid management institutions, and more and S. S. Light, eds., Barriers and Bridges for the dependent societies. Its resolution comes from Renewal of Regional Ecosystems. another biological metaphor of regulation: that of Bormann, F. H., and G. E. Likens. 1981. Patterns homeostatic regulation of body temperature in and Process in a Forested Ecosystem. New York: endotherms. Indeed, successful control of var- Springer-Verlag. ability there does reduce resilience within the system regulated. But, unlike the pathology of Clark,W.C.,and others. 1979. "Lessons for Eco- management noted, the regulation responds to logical Policy Design: A Case Study of Eco- internal change and is robust. It transfers internal system Management." Ecological Modelling 7, variability externally to release opportunity for pp. 1-53. probing, creative opportunities. Davis, M. B. 1986. "Climatic Instability, Time The release of human opportunity is at the Lags, a anCommunity Disequilibrium." In l. heartofsustainabledevelopment!ltrequiresflex- Diamond and T. Case, eds., Community Ecol- ible, diverse, and redundant regulation, monitor- g ing that leads to corrective responses, and experi- Delcourt, H. R., P. A. Delcourt, and T. I. Webb. mental probing of the continually changing real- 1983. "Dynamic Plant Ecology: The Spectrum ity of the external world. of Vegetational Change in Space and Time." Finally,sustainabledevelopment isnotaneco- Quaternary Science Reviews 1, pp. 153-75. logical problem nor a social problem nor an eco- Elliott, J. E. 1980. "Marx and Schumpeter on nomic problem. It is an integrated feature of all Capitalism'sCreativeDestruction:ACompara- three. Effective investments in sustainable devel- tive Restatement." Quarterly Journal of Eco- opment simultaneously retain and encourage the nomics 95, pp. 46-58. adaptive capabilities of people, of business (en- Gunderson, L. H. 1992. "Spatial and Temporal terprises), and of nature. The effectiveness of Hierarchies in the Everglades Ecosystem with those adaptive capabilities can turn the same Implications for Water Management." Ph.D. unexpected event (for exatnple, drought, change diss., University of Florida, Gainesville, Fla. in price, shifts in market) into an opportunity for Holling, C. S. 1980. "Forest Insects, Forest Fires, onesystemoracrisisforanother.Thoseadaptive and Resilience." In H. Mooney, J. M. capacities depend on processes that permit re- Bonnicksen, N. L. Christensen,J. E. Lotan, and newal in society, economies, and ecosystems. For W. A. Reiners, eds., Fire Regimes and Ecosystem nature, it is the structure of the biosphere; for Properties. General Technical Report WO-26. businesses and people, it is usable knowledge; Washington, D.C.: U.S. Department of Agri- and for society as a whole, it is trust. culture Forest Service. We may postulate that investments to increase . 1986. "Resilience of Ecosystems: Local productivity are only viable if all these sources of Surprise and Global Change." In W. C. Clark renewal capacity are maintained or enhanced. and R. E. Munn, eds., Sustainable Development Temporary erosion of these might be bearable so of the Biosphere, pp. 292-317. Cambridge, En- long as recovery is made within the critical time gland: Cambridge University Press. unit of half a human generation (note the relation .1987. "Simplifying the Complex: ThePara- to intergenerational equity and freedoms of digms of Ecological Function and Structure." choice). But continued erosion of any one ulti- European Journal of Operational Research 30, pp. mately reaches the point where it cannot be re- 139-46. 74 Sustainability: The Cross-Scale Dimension .1988. "TemperateForest lnsect Outbreaks, Steele, J. H. 1985. "A Comparison of Terrestrial Tropical Deforestation, and Migratory Birds." and Marine Systems." Nature 313, pp. 355-58. Memoirs of the Entomological Society of Canada Vitousek, Peter M., and P. A. Matson. 1984. 146, pp. 21-32. "Mechanisms of Nitrogen Retention in Forest - . 1992. "Cross-scale Morphology, Geom- Ecosystems: A Field Experiment." Science 225, etry, and Dynamics of Ecosystems." Ecological pp. 51-52. Monographs 62:4, pp.447-502. Walker, B. H. 1981. "Is Succession a Viable Con- Holling, C. S., ed. 1978. Adaptive Environmental cept in African Savanna Ecosystems?" In D. C. Assessment and Management. London: John West, H. H. Shugart, and D. B. Botkin, eds., Wiley and Sons. Forest Succession: Concepts and Application, pp. Lovelock, J. 1988. The Ages of Gaia. New York: 431-47. New York: Springer-Verlag. W. W. Norton and Co. Walker, B. H., D. Ludwig, C. S. Holling, and R. M. McNamee, P. J., J. M. McLeod, and C. S. Holling. Peterman. 1969. "Stabilityof Semi-arid Savanna 1981. "The Structure and Behavior of Defoliat- Grazing Systems." Journal of Ecology 69, inglnsect/Forest Systems." Research in Popula- pp. 473-98. tion Ecology 23, pp. 280-98. Walters,C.J.1986.AdaptiveManagementofRenew- O'Neill, R. V., D. L. DeAngelis, J. B. Waide, and able Resources. New York: McGraw Hill. T. F. H. Allen. 1986. A Hierarchical Concept of West, D.C., H. H. Shugart, and D. B. Botkin. 1981. Ecosystems. Princeton, N.J.: Princeton Uni- ForestSuccession:ConceptsandApplication.New versity Press. York: Springer-Verlag. 75 Cumulative Effects and Sustainable Development Gordon Beanlands Preventing the slow, persistent, and cumulative been published recently, containing in excess of degradation of natural systems resulting from 150 references (Davies 1991; Delcan Corporation human activity is the ultimate environmental 1988; Williamson and Hamilton 1989). challenge facing society. Although a concise defi- This increased attention is related in part to nition is open to debate, the term cumulative publicity given to rates of loss or deterioration of effects is generally considered to refer to the resources of the global commons. For example, long-term accumulation of residual environmen- the general public has a better understanding of tal changes resulting from all previous develop- how multiplehumanactivities, spread over space mental actions. Conceptually it is easier to con- and time, have resulted in cumulative problems sider cumulative effects as a second-order set of of change in climate, ozone depletion, acid rain, problems, or a problem syndrome. groundwater contamination, species extinction, The problem of cumulative effects has been and habitat fragmentation. recognized for many years and is, explicitly or Moreandmoreoftheprofessionalcommunity implicitly,attherootof mostconceptsof environ- is realizing that the cumulative aspects of the mental conservation, protection, and manage- more serious environmental problems account ment. For example, according to Orians (1986, p. for the difficulty in developing viable solutions. 2), "Management of renewable resources is basi- For example, two informal surveys of the most cally cumulative effects management." Over the pressing problems of cumulative effects facing last ten to fifteen years, however, the problem of society resulted in the following lists: cumulative effects has become a major focus of Williamson, Armor, and Johnston (1985) attention for applied natural scientists, environ- mental managers, and policymakers. This atten- * Multiple small hydro dams tion has been reflected in a number of scientific * Flood control projects and management conferences devoted specifi- * Coastal wetlands development cally to the problems of cumulative effects * Wetlands drainage (CEARC and National Research Council 1986; * Hazardous waste disposal Conservation Foundation 1990; Estevez and oth- ers 1986; Wiliamson, Armour, and Johnston 1985). * Nutrient loading (estuaries) The number of publications on cumulative * Dredging of waterways effects has also grown rapidly, increasing from * Urbanization of farmland less than five papers a year between 1975 and * Air pollution and acid rain 1980 to an average of twenty-four a year between 1985 and 1988 (Williamson and Hamilton 1989). * Eutrophication of lakes Several bibliographies on cumulative effects have * Nonpoint soil erosion Defining and Measuring Sustainability: The Biogeophysical Foundations Peterson and others (1987) Definitions, concepts, and approaches * Acidic precipitation * Urban air quality Numerous attempts have been made to define cumulative effects or impacts. One of the earliest * Bioaccumulation of toxins and most influential definitions was developed * Climatic change in 1978 by the U.S. Council on Environmental * Spread of infrastructure Quality (CEQ 1978,40 C.F.R., sect. 1508.7) under * Loss and fragmentation of habitat regulations made pursuant to the National Envi- ronmental Policy Act: * Erosion and degradation of soil ronmentawhich • Agricultural chemicals The impact on the environment which results from the incremental impact of * Loss and degradation of groundwater the action when added to other past, * Changes in freshwater present, and reasonably foreseeable fu- * Increased harvest rates ture actions regardless of what agency (Federal or non-Federal) or person un- This increased awareness of cumulative envi- dertakes such actions. Cumulative im- ronmental effects on the part of the general public pacts can result from individually minor and professional communities has been reflected but collectively significant actions tak- in a growing number of statutes requiring the ing place over a period of time. assessment of cumulative effects. For example, a Clark (1986, p. 114) offered the following com- recent survey of federal laws in the United States mentaryon cumulativeimpactassessment, which discovered that the terms cumulative impacts or incorporates a working definition of cumulative cumulative effects appear seventeen times in ten effects: statutes, and the term cumulative impacts ap- pears in sixty-one sets of regulations (Conserva- Cumulative impact assessment examines tion Foundation 1990). In addition, the recently the consequences of multiple sources of passed Canadian Environmental Assessment Act environmental disturbance that impinge requires that cumulative effects be considered in on the same valued environmental com- various stages of the environmental review pro- ponent. The characteristic "multiple" na- cess including screening, comprehensive study, ture of the sources of cumulative impacts assessment, and mitigation (Canadian House of may arse in three ways: the same kind of Commons 1992). source recurs sufficiently frequently Not surprisingly, consideration of the ulti- through time, the same kind of source mate implications of cumulative environmen- recurs sufficiently densely through space, tal changes at the global level was among the different kinds of sources impose similar factors that led the World Commission on Envi- consequences on a valued environmental ronment and Development to promote the con- component. ceptofsustainabledevelopment(WCED 1987). Peterson and others (1987, p. 5) considered In other words, the challenge of sustainable cumulative effects to occur when "at least one of development includes arresting or reversing two circumstances prevails: persistent addition the cumulative depletion and degradation of of material, a force, or an effect from a single the natural systems on which current and fu- source at a rate greater than can be dissipated; or ture generations depend. On a world scale, compounding effects as a result of the coming cumulative effects and sustainabledevelopment together of two or more materials, forces, or ef- are inextricably linked, reflecting the mega en- fects, which individually may not be cumula- vironmental problem and the mega environ- tive." mental solution, respectively. Perhaps the most succinct definition of cu- This chapter highlights the evolution of the mulative effects is offered by the Canadian concept and practice of cumulative effects as a Environmental Assessment Research Council specific subset of environmental change and (CEARC 1988, p. 2): draws comparisons with the policymaking and Cumulativeeffectsoccurwhenimpactson decisionmaking implications related to sustain- the natural and social environments take able development place so frequently in time or so densely in 78 Cumulative Effects and Sustainable Development space that the effects of individual "in- symptomaticofthecontinuuminenvironmental sults" cannot be assimilated, or the im- planning, which extends from project-based im- pacts of one activity combine with those of pact assessments, to policy and program assess- another in a synergistic manner. ments, to cumulative effects assessment, to re- Individually, these impacts may not be gional planning, to sustainable development. qualitatively different from environmen- A review of two recent bibliographies on cu- tal effects associated with single-project mulative effects (Davies 1991; Williamson and developments, but collectively they often Hamilton 1989) indicated that 20 to 30 percent of require different kinds of research and the publicationslisted offered some form of guid- management approaches if they are to be ance (procedure. approach, handbook, frame- dealt with effectively. work, technique, methodology, or model) on con- These definitions, and others, require that cu- ducting cumulativeeffects assessments. The gen- mulativechangesbe interpreted within specified eral utility of such guidance, however, may be time, space, and organizational scales. It is the open to question. For example, Granholm and expanded nature of these scales that separates others (1987), after screening more than ninety cumulative effects from the more limited time methods, concluded that most were good at de- periods and geographical boundaries used in scribing problems but performed poorly when it addressing the effects of single projects. Critical came to analysis and evaluation. scientific and management factors must be taken Nevertheless, the scientific and management into account to ensure that cumulative effects are communities have developed and tested a num- considered within theappropriatecontextof time ber of approaches designed specifically to in- and space. For example, ecological time lags may crease our understanding of and ability to man- require thatcumulativeeffectsbemonitored over age cumulative effects. These initiatives can be extendedperiods,whilejurisdictionalboundaries grouped according to whether they are (a) ge- may be the priority concern from a management neric in focus, (b) driven by regulatory require- perspective. ments, or (c) in response to threatened or vulner- In general, these considerations support the able resources. adoption of an approach to managing cumulative effects based on at least a regional level (water- shed, municipality, and so forth) and extending Generic focus over a period of years. Therefore, although scien- tists may be tempted to study the cumulative Bain and others (1986) propose a general method- changes over time in well-bounded ecosystems ology for evaluating the cumulative effects of (lakes, wetlands, estuaries), managing the mul- multiplehumandevelopments.Itconsistsofthree tiple sourcesof stresson those systems effectively phases: analysis, evaluation, and documentation. requires a much broader geographic coverage. A unique aspect of this approach is its use of a This need to link the scientific and management computer screening process in the evaluation aspects in efforts to resolve the problems of cu- phase to compare and select developmental mulative effects was the major theme of an inter- configurations. national symposium held in 1986 (CEARC and Sonntag and others (1986) have developed an National Research Council 1986). analytical framework for assessing cumulative Perhaps the complexity inherent in the con- impacts. The framework thvolves determineng cept of cumulative effects has been responsible activities (classified according to project charac- for theproliferation of methodologies, techniques, teristics), system structure and process (the na- and approaches developed to address the issues ture of the receiving system), and cumulative involved. Some of these are merely extensions of impacts (determined through a step-wise analysis). techniques to assess environmental impact de- Lane and others (1988) present a comprehen- veloped to meet legislative requirements. In other sive framework for assessing cumulative effects. cases, it is often difficult to distinguish methods Itbegins witha decision tree todistinguish among for identifying, evaluating, and managing cumu- four types of cumulative effects. This is followed lativeeffectsfromthebasicprinciplesinvolvedin by characterization of the type of causality in- regional planning, riverbasin planning, and inte- volved. The entire process is guided by a sequen- grated resource management. Such overlap is tial series of activities. 79 Defining and Measuring Sustainability: The Biogeophysical Foundations Regulatory focus lytical tool for evaluating alternative decisions relative to the conservation of grizzly bears. Under contract to the U.S. Army Corps of Engi- Even this brief sample shows that the method- neers, INTASA Inc. developed a six-step proce- ology for assessing cumulative effects has a short dure for assessing the regional cumulative im- history, both in its development and in its imple- pacts of developing hydropower. The procedure mentation. Federal government agencies, driven is based on the use of indexes that require few by the need to meet regulations, have been re- new data, are easily calculated, and can be used sponsible for most of the supportive research and for comparisons between regions. The procedure trial applications. It is a complex topic that crosses has been applied to a number of river basins over numerous fields of study in the natural, (INTASA Inc. 1981). social, and managerial sciences, and its bound- The U.S. Federal Energy Regulatory Commis- aries are difficult to set. For this reason, substan- sion has proposed a cluster impact assessment tive progress in a practical sense has been slow. procedure for assessing the cumulative impacts Bodies of water, particularly wetlands act as of multiple small-scale hydroelectric develop- n olleton wan d.ispoal sytems; in th ments on a single watershed (Emery 1986). The natural collechon and disposal systems; In the cumulative ewords of Preston and Bedard (1988, p. 577), they cumulative effects resulting from project cluster- "a evee sasre ffo-hog ec ing are first examined through sub-basin disag- i"can be viewed as a series of flow-through reac- gregation and then dispersed across sub-basins tion vessels." It is perhaps for this reason, along using linear algebra and principles of informa- with the factthat theyarerelativelyeasytobound, tion theory. that they have been the predominant focus of An analysis of potential cumulative effects is attention forcumulativeeffectsassessment. Such required before any oil or gas resources may be is the case from regulatory, management, and developed on wet tundra on the Alaskan North scientific perspectives. In fact, the professional Slope. In response to this requirement, the U.S. Journal of Environmental Management devoted an Fish and Wildlife Service has developed a method entire issue (vol. 12,1988) to all aspects of cumu - using an integrated geobotanical and historical lative impacts on wetland ecosystems. Thus, the disturbance map for predicting and evaluating evolution of our concepts and the growth of our cumulative impacts. This method is based on a empirical evidence concerning this complex phe- landscape approach and uses maps produced by nomenon are largely based on the study of rivers, the oil industry foroperational purposes (Walker lakes, wetlands, and estuaries. Since terrestrial and others 1986). systems are more heterogeneous, and arguably more vulnerable to direct human intervention, the experience with water ecosystems may not be Focus on vulnerable resources directly transferable. Based on this brief overview, some parallels Under an agreement with the U.S. Environmental can be drawn between cumulative effects and Protection Agency, the U.S. Fish and Wildlife sustainable development. First, they both lack Service conducted a series of workshops on bot- precise definitions, which has resulted in much tomland hardwood wetlands that produced a confusion over basic objectives and operating paradigms. Second, in both cases, researchers methodology for assessng cumulative Impact and managers need to think laterally, across a (Gosselink and Lee 1988). The methodology ap- number of disciplines, in order to gain a compre- plies the landscape approach of island biogeogra- hensive understanding of the concepts involved, phy to the fragmentation and loss of habitat. The which few are trained to do. Third, in both cases key hypothesis is that individual features are not it is difficult to establish practical operational as important as the overall pattern in the formu- boundaries, with the result that managers are lation of conservation measures. overwhelmed by the sense that "everything is Weaver, Escano, and Winn (1987) describe a connected to everything else." Finally, for both cumulative effects assessment model for applica- cumulative effects and sustainable development, tion in the management of grizzly bears. It is our intuitive understanding of the concepts in- designed to quantify individual and collective volved is much moreadvanced thanourability to effects of various land uses and activities in space apply that knowledge in a meaningful and prac- and time and to provide managers with an ana- tical manner. 80 Cumulative Effects and Sustainable Development Cumulative effects A specific type of change only that they be perceived to exist. Such is the case, for example, with the gradual depletion of Why is change associated with cumulative effects the ozone layer, where experts cannot give a different than other types of environmental quantitativethresholdofconcernbutareworried change? There are two main reasons: the direc- about the health implications of a continuing tion and rate of change. First, all environmental decline. variables, whether physical or biological, change The focus of attention in studies of cumulative naturally over time but tend to fluctuate within effects appears to be split between recording some long-term envelope of stability. In other gradual changes and deciding on appropriate words,theyappeartobeinsomeformofdynamic thresholds. In cases where thresholds are not a equilibrium. In the case of cumulative effects, concern, theresee!mstobeaninterestinrecording however, the implication is that change in the cumulative effects for some future reference. An variable of concern is unidirectional and that no example of this archival approach is the recent counterbalancing forces are at play. It may be, of report on the state of the environment in Canada course, that theobserved changeismerely a small (Canadian Ministry of Supply and Services 1991). part of a very long natural cycle, which is the This large volume consists almost exclusively of counterargument to the assertion of global warm- graphs and tables, with explanatory text, depicting ing. In other cases, however, such as the increas- cumulative changes in a wide variety of environ- ing acidity of precipitation and the gradual accu- mental and natural resources. Examples indude mulation of toxic chemicals in groundwater aqui- concentrations of mercury in fish, breeding popu- fers, the changes are clearly induced by humans. lations of ducks on the western prairies, growth Implicit in the concern over the unidirectional rates for maple trees, concentrations of various changeofcumulativeeffectsisthatthevariablein pollutants in air, concentrations of nitrite and question is moving in relation to some norm, nitrate in the Great Lakes, abundance of harbor limit, standard, or threshold value. If there is no seals, global mean surface air temperatures, and stated or implied threshold, then monitoring the global emissionsof carbon dioxide from thebum- change is of academic interest only, in the sense ing of fossil fuels (see figures 5-1 through 5-8). In that it will not precipitate any concern. In the all of these cases, a gradual but clearly established context of cumulative effects, it is not necessary trend is evident over a period of at least ten years or that thresholds be established, or even known, more and, in some cases, decades. Figure 5-1: Average Mercury Concentrations in Walleye Collected from Lake St. Claire, Canada, 1970-89 Concentration (ppm, wet weight) 2.5 2.0 _ 1.5 0.5 1970 72 74 76 78 80 82 84 86 88 Source: Canadian Ministry of Supply and Services 1991, p. 21-15. 81 Defining and Measuring Sustainabiltty: I he blogeopnyswal -ounuuriru Figure 5-2: Trends in the Size of Breeding Figure 5-3: Growth of Sugar Maples in Areas Populations of Mallard and Northern Pintail of High, Moderate, and Low Levels of Atmo- in Western Canada, 1966-89 spheric Pollution, 1900-90 Number of ducks (in millions) Ring width (millimeters) 12 - 0.5 - 1.0 1955 60 65 70 75 80 85 90 95 190010 2030405060708090 Year Year Low poLlution Moderate pollution - --- -- - Northern Pintail - Mallard - ---- High pollution Source: Canadian Ministry of Supply and Services 1991, p.6-18. p. 24-11. Figure 5-4: Trends in Canada's Air Quality, 1974-89 Percentage of maximum Percentage of acceptable level (for all 1974 average Figure 5-5: Abundance of Harbor Seals in varables except lead) valuefor lead British Columbia, Canada,1973-86 120 - 100 - --=----------Number of seals (thousands) 60 1 ------ ~~~~~~12 40- - --___,.__1--_-_ .... .. - .-~~z. .... 20 lo 1974 1977 1980 1983 1986 1989 _ Year ___ ----- - Total suspended Sulphur dioxide 4 - . . . . _ particulates (annual average) (annual average) 2- Ozone (peak Carbon monoxide one-hour average) (peak eight-hour average) ------ Nitrogen dioxide Lead 1974 76 78 80 82 84 86 (annual average) (annual average) Source: Canadian Ministry of Supply and Services 1991, Source: T Furmanczyk, Environment Canada, personal p.8-17. communication 82 Cumulative Effects and Sustainable Development Figure 5-6: Concentrations of Nitrate and Nitrate in Parts per Billion e in the Open Waters of the Great Lakes, 1970-88 Lake Erie (central basin) Lake Ontario 500 - 500 - 400 400 300 -300- 200 200 100- I 1 1 1 II1011,-0 100 1970 72 74 76 78 80 82 84 86 88 1968 70 72 74 76 78 80 82 84 86 88 Year Year Lake Michigan Lake Huron 500 - _500- 400 - 400 - 3002- 300- 200 - 200 - 100 ( 1983 85 87 89 1971 73 75 77 79 81 83 85 87 89 Year Year Lake Superior 500 400- 300- 200- 100 - o- I I , 1983 85 Year 87 89 Source: Provided by Dave Dolan, International joint Commission, Windsor. 83 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 5-7: Variation of Global Mean Surface Air olds have not been established or (b) current levels Temperature in Canada, 1861-89 are well separated from threshold values. The Canadian publication gives examples of Degrees (celcius) both positive cumulative effects, such as decreas- 0.4 - ing levels of pollution, and negative cumulative 0.2 - effects, such as declining populations of certain 0 A A ^ lDh e o species. The reference rate of change of 1 to 3 percent a year appears to be valid in both situa- -0.2- 19 ll L T tions. In other words, it may take as long for -o04 - 9 S v u l systems to recover following the removal of stres- sors as to be degraded following the onset of the -0.6 stressors. Recovery is also a cumulative effect. 1860 80 1900 20 40 60 80 Year Source: Canadian Ministry of Supply and Services 1991, The decisionmaking dimension p. 22-12. The success of all approaches to environmental protection and management ultimately depends Figure 5-8: Global Emissions of Carbon Dioxide on the degree to which they influence relevant from the Burning of Fossil Fuels, 1860-1988 decisionmakers. The management of cumulative effects is no different. Cumulative effects lead to Emissions (billions of tonnes procrastination in decisionmaking. Since the rate Of carbon per year) of change is small and constant, there is a high 6 X degree of ambivalence as to exactly when correc- 5 - tive action is warranted. In general, the point at 4- which decisions are taken-the decision flash 3 - point-is influenced by three main factors. The first is the nearness of the cumulative variable to 2 - a predetermined or implied threshold. As the I _, variable approaches the threshold, the probabil- 0- _ ity of remedial action beingundertaken increases. 1860 80 1900 20 40 60 80 The experience with declining populations of Year Atlantic salmon in eastern Canada is an illustra- tive example. During the 1950s and 1960s spawn- Source: Canadian Ministry of Supply and Services 1991, ing populationsof Atlantic salmon in many rivers p. 2-21. in eastern Canada exhibited a slow but constant The second characteristic that distinguishes decline. The change was the result of the cumula- cumulative effects from other types of environ- tive effects of hydroelectric developments on the mental variability is the rate of change. The term rivers, destruction of spawning habitat, wide- implies a slow rate, but how slow is slow? A spread water pollution, and overfishing on the cursoryexaminationofthetrendsinthegraphsof ocean feeding grounds. The declining trend had the Canadian State of the Environment Report (Ca- been known and monitored for many years by nadian Ministry of Supply and Services 1991) fisheries biologists, but no action had been taken provides some clues. In almost all cases, a fitted to conserve the remaining stocks. trend line represents an annual change of be- By the mid-1970s, the situation had, in the tween 1 and 3 percent. This reference rate seems minds of some officials, become quite critical. A to apply equally as well to cumulative changes in graph was prepared showing the cumulative de- human activities, for example, trends in air emis- cline in all east coast spawning stocks over the sions, as it does to cumulative effects in natural time period for which data were available. When variables. Such a rate of change appears to be the minister responsible for making a decision readily observable-measurable-but not of a was shown that the extrapolated trend line indi- sufficient magnitude to warrant immediate correc- cated a likelihood that local populations would tiveaction,especially insituationswhere (a) thresh- be totally extinct within a couple of decades, he 84 Cumulative Effects and Sustainable Development was motivated to take action. In other words, in of carbondioxideemissions increased to approxi- the mind of the key decisionmaker, the difference mately5 percent. The continuationof thispost-war between the cumulative effects variable (salmon rate raised concerns among atmospheric scien- population) and the threshold (extinction of local tists that have led to international negotiations at populations) had been reduced to a point where the political level aimed at reducing the rate of intervention was required. carbon dioxide emissions on a global basis. Un- In this case the reluctance to take action was doubtedly, the absolute concentration of carbon related to the political and financial costs in- dioxideintheatmosphere-aquantitativethresh- volved. The initial decision eventually resulted in old-influenced these developments; however, the government buying back salmon fishing li- theabruptchangein thecumulativerateof change censes from hundreds of commercial fishermen was probably also a contributing factor. In this at a cost of millions of dollars, as well as initiating case, an increase in the annual rate of change from a multi-faceted and long-term salmon enhance- 3 to 5 percent was sufficient to raise concerns and ment program. To date, the populations have lead to action. shown signs of recovery, but at a slower rate than expected by the scientists. This example illus- trates the comprehensiveness of mitigation mea- Comparing cumulative effects sures that are often required to reverse trends and sustainable development resulting from cumulative effects problems. It also demonstrates the need to make decisions at As indicated, the management of cumulative ef- the highest levels of authority since the actions fects and sustainable development share a number necessary to correct cumulative effects problems of similarities and linkages. These can be discussed often require politically sensitive negotiations. In under two headings: concepts and practice. this case the scientists had been aware of the cumulative decline for some time but lacked the Concepts authority to make the required decisions. The second factorinfluencingthedecisionpoint Neither cumulative effects or sustainable devel- is related to the length of time between observa- opment is well defined. Both are subject to vague tions of trends. Repetitive readings over short meanings and disagreements among profession- periods of time reveal only small changes that als over the nature and scope of the intended may lead to a certain degree of complacency and problematic. This lack of definition is both a acceptance. lnfrequentmeasurementsof thesame strength and a weakness. In spite of the vague- variable that show relatively large incremental ness of the terms, both evoke an intuitive under- changes are more likely to have a shock value. standingof the complexity of the issues involved. The third factor related to decisionmaking in Sincetheyrepresenthigherordersofaggregation the context of cumulative effects has to do with of cause-and-effect relationships, they are con- the rate of change of the variable in question. It ceptually robust enough to be applied to a wide was noted above that annual rates of change variety of societal problems. Both concepts are between 1 and 3 percent seem to be common to inherently interdisciplinary in nature, which re- many trends that represent cumulative impacts. flects the reality of dealing with the complex It was also postulated that, depending on thresh- linkages among environmental, economic, cul- old levels, such rates have a low probability of tural, and political issues. precipitating immediate corrective action. This It can be argued that both terms represent raises the question as to whether there is, in a obverse intellectual traps. The term cumulative general sense, a rate of cumulative environmen- effects is an intuitively obvious way to state a tal change beyond which immediate concern is problem, but it does not pose equally tractable more likely to be generated. solutions. The term sustainable development, in Figure 5-8 provides some clues to answer this contrast, represents an intuitively obvious solu- question. Before 1950 the average annual rate of tion without providing insights into the funda- increase in carbon dioxide emissions from the mental nature of the problems. The former cap- burning of fossil fuels was approximately 3 per- tures the essence of the ultimate environmental cent. As a result of increased industrialization problem, and the latter promotes the formulation followingtheendofWorldWarll,theannualrate of the ultimate social solution. 85 Defining and Measuring Sustainability: The Biogeophysical Foundations Both concepts are firmly rooted in expanded sheds,and coastal zones. Thesecanbeconsidered scales of time, space, and organization. At the as rudimentary experiments in managing devel- globallevel,theyposesimilarchallengesforman- opment at expanded scales of space and organi- agement and control. Many of the basic problems zation, although they usually lack the longer-term, being addressed by sustainable development, strategic,intergenerationalaspectsofsustainable such as climatic change, ozone depletion, loss of development and cumulative effects manage- biodiversity, and accumulation of toxins, are ex- ment. Given the complexity and comprehensive- amplesofcumulativeeffectsoperatingona world ness of the two approaches, it is not surprising scale. At thislevel of aggregation,bothapproaches that our record on implementation is so limited. have identical scale, boundary, and threshold characteristics and define problems and solu- tions in a similar way. The notion of References intergenerational equity is germane to the defini- tion of sustainable development but does not Bain,M.B.,J.S.Irving,R.D.Olson,E.A.Stull,and appear in the literature on cumulative effects. Yet G. W. Witmer. 1986. "Cumulative Impact As- the focusof studyison historical trends that often sessment: Evaluating the Environmental Ef- span more than one human generation. It is pos- fects of Multiple Human Developments." sible, therefore, to define the objective of cumula- ANL/EES-TM-309. Argonne National Labo- tive effects management as ensuring that the vari- ratory, Argonne, 111. able of concern remains within its natural enve- CEARC (Canadian Environmental Assessment lopeof stabilitybetween generations. When stated Research Council). 1988. "The Assessment of in this way, the similarity between the two con- Cumulative Effects: A Research Prospectus." cepts is clearly evident. Ottawa, Canada. Practice CEARC and National Research Council. 1986. Cumulative Environmental Effects: A Binational So far, the concepts behind sustainable develop- Perspective. Proceedings of a workshop ment and cumulative effects management have co-sponsored by the Canadian Environmental proven difficult to translate into practice. The Assessment Research Council and the National expanded scales of time and space that underlie Research Council, Ottawa, Canada. the concepts do not generally match the jurisdic- Canadian House of Commons. 1992. "Bill C-13, tional mandatesof existing institutions. Even more an Act to Establish a Federal Environmental important, theorganizational and decisionmaking Assessment Process Passed onMarch 19,1992." infrastructure required to manage and control Ottawa, Canada. human activities effectively at such scales has yet Canadian Ministr of Supply and Services. 1991. to be developed. The State of Canada's Environment. Ottawa: Goy- The legislative requirement to consider cumu- ernmeof Canada lative effects in many development applications ement of Canada. has spurred scientific and management studies CEQ (Council on Environmental Quality). 1978. on this topic. Progress, however, is slow and "Regulations under the U.S. National Environ- experience is limited to controlling cumulative mental Policy Act." Washington, D.C. changes in relatively small and well-defined eco- Clark, William C. 1986. "The Cumulative Impacts systems or management units. of Human Activities on the Atmosphere." In Most of the current effort going into sustain- Cumulative Environmental Effects: A Binational able development is intellectual in nature, focus- Perspective. Proceedings of a workshop ing on refining concepts, definitions, and termi- co-sponsored by the Canadian Environmental nology. The first operational experiments will Assessment ResearchCouncil and theNational likely revolve around attempts to integrate envi- Research Council. Ottawa, Canada. ronmental, economic, and social considerations Conservation Foundation. 1990. Making Decisions in policy formulation. These are complex issues, on Cumulativelmpacts. Proceedingsof a confer- and it will be some timebefore tangibleresultsare ence co-sponsored by the Council on Environ- available for review. mental Quality, the U.S. Environmental Pro- In the meantime, our experience in working at tection Agency, and the National Science Foun- larger scales islimited to planning regions, water- dation, Washington, D.C., June. 86 Cumulative Effects and Sustainable Development Davies, Katherine. 1991. "Cumulative Environ- and the National Research Council, Ottawa, mental Effects: A Compendium." Report pre- Canada. pared for the Federal Environmental Assess- Peterson, E. B., Y. H. Chan, N. M. Peterson, G. A. ment Review Office, March. Constable, R. B. Caton,C.S. Davis, R. R. Wallace, Delcan Corporation. 1988. "Annotated Literature and G. A. Yarranton. 1987. "Cumulative Ef- Review of Cumulative and Incremental Envi- fects Assessment in Canada: An Agenda for ronmental Impact." Prepared for Natural Re- Action and Research." Background paper pre- sources Branch, Canadian Parks Service, Ot- pared for the Canadian Environmental As- tawa, Canada, August. sessment Research Council, Victoria, Canada. Emery, R. M. 1986. "Impact Interaction Potential: Preston, E. M., and B. C. Bedard. 1988. "Evaluat- A Basin-wide Algorithm for Assessing Cumu- ing Cumulative Effects on Wetland Functions: lative Impacts from Hydropower Projects." A Conceptual Overview and Generic Frame- Journal of Environmental Management 23, pp. work." Journal of Environmental Managernent, 341-60. 12, pp. 561-64. Estevez, E. D., J. Miller, J. Morris, and R. Mannan. Sonntag, N. C., R. R. Everitt, L. Rattie, C. P. Wolf, 1986. Managing Cumulative Effects in Florida J. Truett, A. Dorcey, and C. S. Holling. 1986. Wetlands. Proceedings of a conference, held in "Cumulative Impact Assessment: Review of Sarasota, Florida, October 1985. New College State-of-the-Art and Research Recommenda- Environmental Studies Program Publication tions." Environmental and Social SystemsAna- 37. Madison, Wisc.: Omni Press. lysts, Ltd., Vancouver, Canada. Gosselink, J. G., and L. C. Lee. 1988. "Cumulative Walker, D. A., P. J. Webber, M. D. Walker, N. D. Impact Assessment Principles," In J. A. Kusler, Lederer, R. H. Meehen, and E. A. Nordstrand. M. L. Quammen, and G. Brooks, eds., Proceed- 1986. "Use of Geobotanical Maps and Auto- ings of the National Wetland Symposium: Mitiga- mated Mapping Techniques to Examine Cu- tion of Impacts and Losses, pp. 196-203. Techni- mulative Impacts in the Prudhoe Bay Oil Field, cal Report 3. Association of State Wetland Alaska." Environmental Conservation 13, pp. Managers, Chester, VT. 149-60. Granholm, S. L., E. Gerstler, R. R. Everitt, D. P. Weaver,J.L.,R.E.F.Escano,and D.S.Winn.1987. Bedard, and E. C. Vlachos. 1987. "Issues, Meth- "A Framework for Assessing Cumulative Ef- ods, and Institutional Processes for Assessing fects on Grizzly Bears." Transcripts of the North Cumulative Biological Impacts." Prepared for American Wildlife Natural Resource Conference Pacific Gas and Electric Company, San Ramon, 52, pp. 364-76. Calif. Williamson, Samuel C., C. L. Armour, and R. L. INTASA Inc. 1981. "National Hydroelectric Johnston. 1985. "Preparinga FWS Cumulative Power Resources Study: Environmental As- Impacts Program: January1985 Workshop Pro- sessment." Prepared for the U.S. Army Corps ceedings." Biological Report 85(11.2). U.S. Fish of Engineers, Institute for Water Resources, Ft. and Wildlife Services, Washington, D.C. Belvoir, VA., contract DACW72-80-C-002. Williamson,Samuel C., and Karen Hamilton. 1989. Lane, P. A., R. R. Wallace, R. L. Johnson, and D. "Annotated Bibliography of EcologicalCumu- Bernard. 1988. "Reference Guide to Cumulative lative Impacts Assessment." Biological Report Effects Assessment in Canada." P. Lane and 89(1). U.S. Fish and Wildlife Service, Washing- Associates, Ltd., Halifax, Canada, October. ton, D.C. Orians, Gordon H. 1986. "Cumulative Effects: WCED (World Commission on Environment and Setting the Stage." In Cumulative Environmen- Development). 1987. Our Common Future. tal Effects: A Binational Perspective. Proceedings United Nations, World Commission on Envi- of a workshop co-sponsored by the Canadian ronment and Development. Oxford, England: Environmental Assessment Research Council Oxford University Press. 87 Managing Landscapes for Sustainable Biodiversity H. Ronald Pulliam Many of the ideas presented in this chapter were developed and discussed by participants in a weekly meeting of students and postdoctoral associates working with me. I am particularly grateful to John (Barny) Dunning, Jianquo Liu, David J. Stewart, Brent Danielson, Scott Pearson, and Brian Watts for sharing their data, ideas, and criticisms. Portions of the researcn discussed here were supported by the National Science Foundation, the Department of Energy, and the U.S. Forest Service. Most animal and plant species live not just in nomic development. It considers first how popu- nature preserves but also in the matrix of man- lation models incorporating habitat-specific aged, human-dominated ecosystems in which demography can be used to predict how single- such preserves are embedded. In the long run, species populations might respond to changes in preservation of biological diversity will depend land use. An example is then given where a at least as much on how we manage the matrix as management plan developed primarily over con- on how much land we set aside in preserves. cern for an endangered species leads to the possi- Accordingly, a great challenge for sustainable bilitythatotherspeciesrnightalsobecomethreat- development is designing and managing land- ened or endangered. A discussion of problems scapestobalancetheneedsof economicdevelop- associated with managing for a single endan- mentandthepreservationofbiologicaldiversity. gered species leads to a consideration of how The traditional approach to preserving bio- changes in land use might influence patterns of logical diversity has been to set aside nature biological diversity in general. Finally, the chap- preserves in which neither economic develop- ter concludes with a discussion of using ecologi- ment nor resource extraction is allowed. Impor- cal models to explore how land management tant as this approach has been to the preservation plans might simultaneously influence economic of some species, biological diversity cannot be (market) profits, biological diversity, and other preserved solely by setting aside such preserves aspects of environmental quality. (Hansen and others 1991; Liu 1992; Wilcove 1989) because (1) thearea of such preserves is too small, (2) the rapid increase in human population and Landscapes of suitable patches the subsequent pressure for other uses of the land make it doubtful that much more land can be set All organisms live in heterogeneous environ- aside, and (3) species in preserves are greatly ments. Individuals of the same species living in affected by the changes in land use and other hu- relatively close proximity to one another may manactivitiesinthelandscapessurroundingthem. experience totally different physical and biotic This chapter explores how ecological theory environmentstotheextentthatsomemaybeable can be employed in schemes to preserve biologi- to survive and reproduce while others cannot. At cal diversity while permitting sustainable eco- spatial scales substantially larger than what one Defining and Measuring Sustainability: The Biogeophysical Foundations individualexperiences, thelandscapeexperienced therefore suitable breeding sites for this species. by a population represents a mosaic of good and Hardwood stands and pine stands between five bad places for the species. The growth, or lack and eighty years of age are not suitable for thereof, of the population is determined not only Bachman's sparrows,because theirunderstoryor by the quality of the individual microsites occu- canopy vegetation is too dense and does not piedbutalsobythespatialandtemporaldistribu- permit sufficient light to reach the forest floor. tion of suitable and unsuitable microsites or Figure 6-1 shows the actual distribution of suit- patches of habitat. able sites on the study area in 1990. Based on the A field or woodlot that is a single patch of known history of land use of the area and the relatively uniform quality for a bird or mammal proposed management plan for the site, the prob- may, at the same time, be a mosaic of patches of able distribution of suitable sites can be recon- quite different quality for individual nematodes structed for the past or projected into the future. or shrubs. For any species, the landscape contain- Several of the factors that influence the loca- ing a population may be mapped as a mosaic of tionofsuitablehabitatforBachman'ssparroware suitable and unsuitable patches. Each map is quitegeneralinthattheyinfluencethelocationof specific to the habitat requirementsof one species habitat for many terrestrial species. Factors like and must be done at a scale appropriate to that soil type, topography, and vegetative cover all organism. Ingeneral,thescalemustbefineenough give information on the suitability of a site for a to resolve the areas occupied by individuals over particular species, and all of these factors can be significant portions of their lifetimes. readily mapped. For Bachman's sparrow, soil In order to map the patches of suitable habitat type and topography influence the rate at which for a particular species, one must have a set of seedling trees grow and, therefore, the ages of criteria for drawing the habitat boundaries. Fol- pine stands that have vegetation profiles suitable lowing Elton (1949) and Andrewartha and Birch for the sparrow. In addition, time since distur- (1984), ecologists choose a habitat boundary to bance, successional status, and management his- have "certain homogeneity with respect to the tory may provide additional information on the sort of environments it might provide for ani- suitability of a site. For Bachman's sparrow, the mals" (Andrewartha and Birch 1984, p. 223). In age since planting of a young stand and the time managed landscapes, drawing habitat bound- since the last understory fire in an old stand both aries is usually simplified by the strong contrasts provide valuable information on suitability of the between habitats with different management his- stand. tories. For example, a pine plantation is clearly The combination of factors determining suit- discernible from a neighboring old field or de- ability of the site is different for every species. ciduous woodlot. In landscapes less dominated Such information alone is not enough to deter- by human activities, habitat boundaries are often mine unambiguously the presence or absence of "softer" and more arbitrary. a species, but it can usually be used to categorize Suitable sitesfora particularspeciesmayoften habitats as suitable or unsuitable and in some be distributed as isolated patches embedded in a cases to assign a probability of occupation. As matrix of unsuitable habitats. Spatially explicit discussed below, a map based on information models, developed to predict population trends, about the amount and location of suitable sites require that suitablehabitat be distinguished from under existing or proposed patterns of land use unsuitable habitat so that all suitable sites can be can be an invaluable tool for species manage- located on a map. To illustrate how this can be ment. done, figure 6-1 shows suitable habitat for Bachman's sparrow on a 5,000-hectare tract at the Savannah River Site, a U.S. Department of Energy Habitat-specific demography: Sources facility near Aiken, South Carolina. Dunning and and sinks Watts (1990) have shown that this species re- quires a dense layer of grasses and forbs in the The previous discussion of distribution does not first meter above ground and an open understory differentiate among habitats of different quality (2 to 4 meters above ground). At the Savannah except inasmuch as they might be suitable or River Site, both old-growth pine stands that are unsuitable for a given species. What do we mean frequently burned and very young pine stands by suitable habitat? Might a species occur in have the appropriate vegetative structure and are unsuitable habitat? 90 Managing Landscapes for Sustainable Biodiversity Figure 6-1: Distribution of Suitable Breeding Habitat for Bachman's Sparrow in 1990 on a 5,000-Hectare Tract at the Savannah River Site ~~~ t ~~~~~km 7 C IrilAl~~~~~~~~~~~~9 Pines 1-5 years 2 v F _~~~~~~~~~~~~~~ Pines >80 years Note: Bachian's sparrow breeds both in older-growth pine forests and young clear-cuts but not in middle-aged pine stands. Following Pulliam (1988), consider a popula- The finite rateof increasecan vary fromyearto tion that has n0individuals in the late spring just year as the survival rates and reproductive rates prior to the reproductive season. If none of the vary. The geometric mean of the rates for a se- adults dies during reproduction, and each adult quence of years characterizes the mean growth produces an average of b offspring, then at the rate of the population over that time. If the long- end of the breeding season there will be nT + bnT term mean I is less than 1.0, the population will individuals. Furthermore, if adults survive the decline, and if it exceeds 1.0, the population will nonbreeding season with probability PA and ju- grow. Obviously, the population cannot grow veniles survive with probability PP, then at the forever, so for a population that does not go end of theyear there will be P^nT+P PbnTindividu- extinct or become infinitely abundant, the als. The finite rate of increase for the population long-term I must be close to 1.0. (defined as I = PA + bP,) gives the number of The finite rate of increase can also be used to individuals at the beginning of year T + 1 per describe spatial variation in growth rates. In this individual at the beginning of year T. case, we refer to the habitat-specific rate of in- 91 Defining and Measuring Sustainability: The Biogeophysical Foundations crease and calculate I based on the birth and death Spatially explicit models rates that apply in a specific habitat or patch of habitat alone. This concept of habitat-specific Most population models are conceptual models growth rate is complicated by dispersal. If each that do not attempt to incorporate the complexi- habitat or patch of habitat is isolated from all ties of real landscapes. As Levins (1966) has others, then the value of I calculated for any one pointed out, general models are often neither habitat is the growth rate experienced by the precise nor realistic. One of the general themes of population in that habitat. However, if habitats bothconservationbiologyandlandscapeecology areconnected,thepopulationgrowthrateisgiven is that details, such as the geometry of habitat by the weighted average across all habitats, that patches in a landscape, can influence population is,different partsofthepopulationaregrowingat trends and extinction probabilities. Population different rates. models that are not spatially explicit make very Some habitats are clearly more suitable than unrealistic assumptions about the dispersal be- others. Consider the simple case where two habi- havior of individuals and do not reflect the com- tats are of different quality and migration occurs plexity of real landscapes. Whereas such models between them. Habitat 1 is the better habitat, are useful for gaining general insights into popu- called the source, and here reproduction exceeds lation dynamics, they are poorly suited for man- mortality, so that the habitat-specific growth rate, aging particular specieson particularlandscapes. 1,, is greater than 1.0. In habitat 2, the sink, mortal- On the other hand, spatially explicit models are ity exceeds reproduction, so 12 is less than 1.0. well suited for incorporating realistic details of Assume that the subpopulation in the source particular species and landscapes, but, because habitat grows at the l until it reaches a population they are so specific, the conclusions reached can- (n1*), which represents the maximum number of not always be generalized to other species and breeding individuals that can be accommodated landscapes. in the source habitat. Once the source has reached Mobile Animal Population-or MAP-is a class the maximum number of breeding individuals, of spatially explicit population simulation mod- there are 1ln I individuals at the end of each els that accounts for the actual spatial locations of nonbreeding season; of this total only n1' can habitat patches and simulates habitat-specific remain tobreed, and the remaining n (11 I-)nmust demography and the dispersal behavior of or- emigrate from the source habitat into the sink ganisms. The original version of the model, habitat. BACHMAP, wasdeveloped to simulate theabun- In thisexample, the sink subpopulation would dance and distribution of Bachman's sparrow in soon disappear in the absence of any immigra- managed pine plantations at the Savannah River tion, because at the end of each year there would Site. The model is, however, flexible and can be befewerindividualsthantherewereatthebegin- adapted to a variety of animal species on any ning. However, with the steady immigration of landscape, either hypothetical or real. n1 (11 - 1) individuals from the source habitat, MAP can be linked to a geographic informa- the sink population will grow to an equilibrium tion system description of a real landscape (Liu population of 1992). The model has mostly been run on land- scapes ranging from 1,000 to 5,000 hectares but n 2* =n (1 -1) / ( - 12). can be adapted to larger or smaller landscapes. The model represents a landscape as a grid of cells, each of which is the size of an individual Note that 11 - 1 is the per capita reproductive territory of the species being simulated (2.5 hect- surplus in the source habitat, and 1 - 12 is the per ares for Bachman's sparrow). Clustersof adjacent capita reproductive deficit in the sink habitat. cellsrepresentthesizeandlocationofforesttracts Clearly, if the reproductive surplus in the source in the landscape; these tracts are assumed to be is much larger than the reproductive deficit in the relatively uniform in their suitability for the spe- isink, tuchears habitathwi reproducntavedefcin f re d cies of interest. MAP models contain subroutines sink, thant habitat , even thoug the that specify forest management practices, succes- vidalsthat c h , ee tugh t sion, and, in some cases, the growth rates of tree sink subpopulation depends on emigration from species. Thus, the model can depict the structure the source for its very existence. In other words, of the current landscape and project the most of the individuals in a local population may landscape's structure in the future based on a exist in habitat unsuitable for them (Pulliam 1988). management plan specifying a harvest and re- 92 Managing Landscapes for Sustainable Biodiversity planting schedule. Other management activities Managing for diversity such as thinning or buning stands, which might influence suitability of the stand, can be easily Managing an ecosystem may be directed at a incorporated into the model. single species or at multiple species. Each ap- Bachman's sparrows breed in pine stands that proach addresses different concerns and faces have a dense ground layer of grasses and forbs different challenges. and an open understory (Dunning and Watts 1990). In managed pine plantations of the south- Problems with single-species approach east, these conditions occur in very young pine stands (less than five years after planting) and in Managing an ecosystem for a single endangered very old s'.ands (more than eighty years old). or threatened species may lead to the decline of Middle-age stands (between six and seventy-nine populations of other species and, in some cases, years old) have dense understories, which shade may lead to additional species becoming threat- out the ground-layer forbs and grasses that pro- ened or endangered themselves. A case in point is vide food and nesting sites to the species. the many salmonid species in rivers of the north- MAP models simulate structure and change of western United States. These species and subspe- the landscapeand explicitly incorporate dispersal cies, many of which are rare and threatened with behavior. Dispersal is assumed to occur in the extinction, have substantially different habitat spring just before reproduction. If an adult fe- ' male dies and leaves surviving female offspring, requirements, and changes i management that then one of those offspring inherits the natal .improv habitari territory. Juveniles that do not inherit their natal habItat for another. territory disperse in search of unoccupied, suit- In some cases, vast tracts of land may be man- able breeding sites. On each move to a new site, aged primarily for a single species, as is the case the dispersing female faces a fixed risk of mortal- for the spotted owl (Strix occidentalis) in the north- ity, so the probability of dying while dispersing is western United States. Such circumstances natu- higherwhentherearefewersuitablesitesormore rally result in conflict between economic and of the suitable sitesare already occupied. conservation goals and may result in conflict BACHMAP has been parameterized based on between the needs of the target species and those field studies of Bachman's sparrows and similar of many other species. A somewhat similar ex- species(DunningandWattsl990;Haggertyl986). ample exists in the southeastern United States, Pulliam, Dunning, and Liu (1992) provide a de- where the red-cockaded woodpecker (Picoides tailed discussion of how these parameters were burealis) is a species of major concern in managed determined and provide the results of an exten- pine forest. This species requires large tracts of sive sensitivity analysis. By running the model older-growth stands of longleaf and other pine with various combinations of parameters from species. Management for the red-cockaded wood- the set of feasible ranges for each parameter, they pecker interferes with the ability of humans to conclude that the population dynamics of the achieve high economic return from harvesting species is more sensitive to variation in demo- thepinesand,asdiscussedbelow,mayalsoresult graphic variables than to variation in variables in the decline of other species of concern. describing dispersal behavior. Changes in adult The Savannah River Site (SRS) is a 770-square- and juvenile survivorship have especially large kilometer reserve consisting mostly of managed impacts on populatien size and probability of exWinction. Accordingly, current field efforts to bottomlands along theuSavannah Riverdand other obtain better parameter values are focusing on these parameters. major streams. Natural resource management is MAP and other spatially explicit models pro- part of the mission of the SRS, and forest lands on vide land managers with a new tool for predict- the site are managed by the U.S. Forest Service. ing how management plans and changes in land Silvicultural practices at the site over the past use may affect species of concern. A few such fortyyearshaveresulted inamosaicof even-aged models are sufficiently well developed and pa- pine stands mostly ranging from 10 to 100 hect- rameterized so that they can be used immedi- ares in size. As part of a long-term forest manage- ately. For most species, however, the relevant mentprogram, theSavannahRiverForestStation parameters are not known, and several years of hasdeveloped the "Savannah River Site Wildlife, concentrated fieldwork would be required to Fisheries, and Botany Operation Plan" (SRFS determine them. 1992). One of the primary objectives of this plan is 93 Defining and Measuring Sustainability: The Biogeophysical Foundations to maintain a viable population of the point is simply that managing primarily for one red-cockaded woodpecker. endangered species can potentially threaten The operations plan considers the habitat re- others. This is not meant to be a criticism of the quirements of a variety of rare plant and animal SRS Wildlife, Fisheries, and Botany Operations species,includingthered-cockadedwoodpecker. Plan. In many ways, with its multispecies ap- The approach is to identify management indica- proach and emphasis on habitat management, tors, which are plant and animal species and the operations plan is a model approach. It plant communities chosen for management em- could be improved, however, by incorporating phasis and monitoring. The plan identifies about quantitative predictors of population trends as sixty such management units, the great majority the one discussed here. The basic problem is of which are vertebrate or plant species. For each that such quantitative tools are in their infancy, unit, a rninimum objective is set, which for most and management plans cannot wait for such species is a minimum number of individuals or models to be sufficiently parameterized and breeding pairs to be maintained at the site. validated. Over the next decade, such models The operations plan proposes a major change can be expected to develop rapidly and become in the management strategy for SRS forest over useful tools for resource managers. Manage- the next fifty years. It calls for a shift from pine ment plans such as the SRS Wildlife, Fisheries, rotations of 30 to 40 years to 80 to 100 years, in and Botany Operations Plan should remain flex- order to increase the number of older-growth ibletoolsthatarecapableofmakingmid-course pinestandsofthesortfavoredbythered-cockaded corrections as predictive tools are further de- woodpecker. Liu (1992) and Liu, Dunning, and veloped and refined. Pulliam (forthcoming) have used a MAP model to simulate the long-term impact of this proposed Figure 6-2: Predicted Population Trends for change on the viability of Bachman's sparrow. Bachman's Sparrow over the Next Fifty Years Based For the entire SRS, the management objective is to on Two Alternative Management Plans maintain 1,100 breeding pairs of sparrows (SRFS Harvesting the Stands in Clusters 1992). Our simulations have been run on a 5,924- Population size hectare portion of the site for which we have a 180 detailed geographic information system giving ] vegetativecoverand history of land use. Extrapo- 50 Minimum objective lating down from the objective of 1,100 pairs for 120 - the entire SRS gives a management objective of 90 -/ 115 pairs for the 5,924-hectare study site. 60 - BACHMAP was linked to this geographic in- 30 _- _ _= PI = 0_3 formation system to project future pattems of I land use under the proposed operations plan and 1990 2000 2010 202t) 2030 2040 2050 to simulate the impact of the changes in land use lime (year) on Bachman's sparrow. Figure 6-2 shows the predicted population trends for Bachman's spar- Harvesting the Oldest Stands First row for two harvest strategies. There are cur- Population size rently about 65 breeding pairs of the sparrow on 140- this 5,924-hectare area (Dunning, unpublished 120- data). If the proposed operations plan is em- 100- Minimum objective ployed, the population size of Bachman's spar- 80- .4 row is predicted to decline to 20 to 30 breeding 60 - pairs around the year 2000 and then to increase 40 slowly and reach the minimum management ob- 200 jective of 115 pairs sometime after the year 2030. ( I l l Based on replicated simulations and sensi- 1990 2000 2010 2020 2030 2040 2050 tivity analysis, Bachman's sparrow does not [ime (year) appear to be in danger of local extinction due to the proposed management plan, but, nonethe- Note: In both cases, the sparrow populations decline to a low less, over the short-term, it will probably de- level before gradually increasing to the minimum objective size. The parameter Pr refers to the survival probabilitv of juvenile cline significantly from its current level. The sparrows If the lower value (0.3) is assumed, the sparrow population never recovers to its original level 94 Managing Landscapes for Sustainable Biodiversity Fragmentation and species diversity Forest fragments in heterogeneous landscapes are very different from oceanic islands in that the The theory of island biogeography has been ap- surrounding matrix may be uninhabitable for plied to habitat fragments created by deforesta- some species, but suitable for others. This may be tion (Diamond 1975; Wilson and Willis 1975), and particularly true in human-dominated landscapes despitewarningstothecontrary(Margules,Higgs, where natural areas are embedded in a complex and Rafe 1982) may still provide the best general matrix of habitat types, some of which are suit- guide to the preservation of biological diversity able for some of the species in the natural area. in habitat fragments. The number of species (S) on Figure 6-3 illustrates this point with three hypo- anisland isapproximatelyrelated to island area(A) thetical bird species occupyinga forest fragment by the "species-area" equation S = cAz, where c is an and some of the habitats in the surrounding land- empirically determined constant that variesamong scape. Species A inhabits the forest understory taxa and z is a constant, usually in the range of 0.25 and adjacent old-field habitats that have similar to 0.35. The application of the theory of island ground cover, but species A does not inhabit the biogeography in general, and this equation in par- mowed ground layer in residential areas. Species ticular, assumes that the area surrounding a habitat B occurs only in the forest since the mid-story fragment is truly unsuitable to the species in ques- vegetation it requires is not found in either the tion, as would be water surrounding an oceanic residential areas or old fields. Species C, a canopy island. Furthermore, the equation is an equilibrium species, does well in the forest habitat and the relationship that is only reached after many of the surrounding residential areas, which also have original species have become extinct (see Brown mature forest trees, but this species does not 1978; Kadmon and Pulliam 1994). inhabit old fields. In this example, the effective Figure 6-3: Response of Different Species to Deforestation Habitat 1 .Habitat 2 Habitat 3 (Old field) (Forest) (Residential) 41 Species A Species B Species C (40% deforestation) (90% deforestation) (60% deforestation) H H3 Hl jH3 _ H i HI 1 HI HI H3 H"II -3 H'H2 Hi I` H 4 HHI H3 HH H3 IHI HI HI wHI H3HI Note: Different species may respond very differently to deforestation. Species B is a speciahst on old-growth forest, and when 90 percent of the forest is lost, 90 percent of the habitat for this species is lost. Species A and C, however, can also thrive in other habitats in the human-dominated landscape surrounding the forest preserve. Accordingly, the effective deforestation is less for these species. 95 Defining and Measuring Sustainability: The Biogeophysical Foundations degree of forest fragmentation is different for suitable habitat in the surrounding second each species. For species B, the forest fragment growth and managed habitats. If, for example, constitutes an island of suitable habitat. The is- 90 percent of the forest species can use 50 per- land is larger for species A and C, but the island cent of the managed landscape plus the 10 per- for species A is different from that for species C. cent of remaining forest, then only about 14 The species-area equation has been used to percent, 1 - (0.6)o3, of these species would be predict the consequences of habitat fragmenta- lost. If half of the remaining 10 percent of forest tion. According to the relationship between spe- specialist species were doomed to extinction, cies and area in the equation, the percentage of the total loss of species would be about 18 the original species remaining after a loss of suit- percent, instead of the 50 percent predicted by able habitat should be (AR / AO)z, where AR is the a 90 percent loss of habitat. area remaining and AO is the original area avail- able. Thus, with z = 0.3, a loss of 50 percent of the original habitat should result in a loss of about 20 Economic considerations percent of the original species, and a 90 percent loss of habitat should result in a loss of about 50 Landscape models of the sort discussed in this percent of the species. chapter can be used to explore how land use and The species-area curve has been used to try to land management practices influence both bio- predict the consequences of widespread defores- logical diversity and economic revenues. For ex- tation in the Amazonian Basin and other parts of ample, the MAP models, which predict patterns the world undergoing rapid changes in land use. of species abundance and biological diversity, Primary forests have long since been lost in other require information on patterns of land use and parts of the world, however, with apparently far forest management practices such as frequency of less loss of species than predicted by the simple burning, length of rotation, and size of stand, all applicationof thespecies-area curve. Forexample, of which have an important impact on economic more than 90 percent of the forests of eastern revenue (seebox6-1 for indicatorsofsustainability United States have been logged or otherwise at the landscape level). cleared, but far less than 50 percent of the native ECOLECON is a spatially explicit, biota has been lost. There are at least three reasons object-oriented model that simulates the popula- for this discrepancy. First, extinction across a tion dynamics of animals and economic perfor- large area may require decades or even centuries mance based on forest landscape management to proceed to completion. Accordingly, many practices (Liu 1992; Liu, Cubbage, and Pulliam, species currently considered threatened and en- forthcoming). ECOLECON incorporates a MAP dangered maybeevidenceofagradual processof model with detailed information about extinction still under way. Second, the extinction species-specific use of habitat and demography of some species has been prevented by direct with a forest growth and yield model containing intervention of humans, sometimes at very great economic information about management costs cost. Third, even after most of the forest has been and timber prices. Given current parameters, the cut, much of the resulting secondary growth, model simulates thegrowth rates of loblolly pines managedforests,andotherhuman-dominatedhabi- based on information about density of planting tat may providesuitablehabitat formanyspecies. In and typeof soil and calculates yield of pulpwood- other words, the effective loss of forest for many chipand saw-and sawtimber, dependingon the species may have been far less than 90 percent. length of rotation chosen. Net income is based on To evaluate the potential of human-dominated yield, the prices of pulp and timber, and the costs habitat to buffer the impact of deforestation on offorestregenerationand maintenance, plus prop- species loss,consider the followingexample meant erty tax and administrative costs. to represent the situation for forest birds in the ECOLECON predicts the impact of land use eastern United States. Suppose 90 percent of the and management decisions on animal popula- original forest is cu t, but that most of the original tions while calculating net income, net present forest inhabitants can use some portion of the value, and land expectation value. The param- human-dominated matrix surrounding the re- eters of the biological portion of ECOLECON are maining old-growth forests. In other words, only presently set for predicting the abundance and a fraction of the species originally inhabiting the distribution of Bachman's sparrow on managed forest are so specialized that they can not find pine plantations. Liu (1992) and Liu, Cubbage, 96 Managing Landscapes for Sustainable Biodiversity Box 6-1. Indicators of Sustainability at the Landscape Level Sharad Lele The following are some of the indicators of sustainability that need to be addressed at the landscape level: * Changes in land use. One of the major factors affecting sustainability is change in the pattern and level of land use. Therefore, monitoring changes in overall patterns of land use is essential. This involves periodic assessment of changes in the amount, distribution, and size of different types of patches (land uses) in the landscape, including both human-created and natural or semi-natural types. From this assessment, functional changes in the landscape can be inferred, assuming that there is basic under- standing of conditions and processes within the patch types and an understanding of patch by patch interactions (that is, edge effects). Changes in conditions within a patch type-successional changes in composition and structure-should also be a part of this assessment. * Biodiversity. Some periodic assessment of biodiversity is important. This could take a variety of forms but certainly should include attention to population levels of indicator species essential to the ecosystem's health. In addition, protocols can be developed to assess overall levels of diversity based on the occurrence of various patch types or structural features within the landscape. A basic assumption is that manyelementsofdiversityare to be found in managed landscapes and not simplywithin natural areas; this must be considered in any monitoring program and in any effort to develop inferences about various management alternatives. * System inputs and losses. Assuming that the basic landscape unit is a watershed or drainage basin, system inputs and losses should be assessed on that basis. Natural inputs would be expected to be primarily through the atmosphere; hence, they would need to be addressed in any monitoring program. System losses are probably best assessed in the river that drains the landscape. This requires hydrologic monitoring and a program of sampling and measuring particulate and dissolved materials leaving the watershed, for example, chemical species and particulate soil materials and sediments and chemicals, including nutrients. * Soil properties. Given the fundamental importance of soils in the production of food and fiber, some program for periodically assessing soil conditions-amount or depth, physical properties (macroporosity), chemical conditions (level of nutrients and organic matter), and biology (diversity of soil microflora and fauna)-should be conducted. Emphasis should be on conditions in patch types heavily managed forcommodities, but periodic comparisons with natural systems asbaselines are also essential. A part of the assessment should also be the periodic analysis of soil subsidies being used to maintain fertility, such as levels of fertilization. * Economic and energy balance. Periodically, some measure(s) of economic (including energy) inputs to and outputs from the drainage system are desirable. In economic terms, this would be an economic input-output analysis. * Human condition. Some monitoring of the human population and its status needs to be conducted; perhaps this should be done at the regional rather than the landscape level. In any case, some measures of the human condition are appropriate in order to determine whether it is stable, declining, or improving. Various measures are possible, includinghealth status, caloric intake, levels of satisfaction, income, and morbidity. and Pulliam (forthcoming) show that whereas We are currently developing versions of land expectationvalueisusually maximized with ECOLECON and related models to predict popu- short rotations (approximately twenty years), lation sizes and extinction probabilities for se- sparrow populations are largest with rotations lected bird, mammal, amphiban, and inverte- more than eighty years in length. Factors such as brate species and to predict trends in biological size of stand, fragmentation of forest, and amount diversity in general. Weare also workingon ways and location of mature pine stands in the land- to expand the model to include forestsother than scape all influence economic revenues as well as mono-specific stands of pines. Finally, we plan to size of the animal population and probability of expand the economic portion of the model to extinction. Whereas no management plan maxi- include other extractive and recreational land mizes both economic profit and population size, uses. We now have a working model that can be some management schemes result in both rela- used in an adaptive management framework to tively high net income and bird populations that, help forest managers choose landscape manage- while intermediate in size, are nonetheless suffi- ment schemes that yield reasonable economic cient to make extinction very unlikely. revenues without threatening plant and animal 97 Defining and Measuring Sustainability: The Biogeophysical Foundations species of concern to management. Although our logical diversity with models of agricultural and models to date are specific to species and condi- forest management, and even these must be tions in the southeastern United States, the ap- viewed as subsystems to be integrated into larger proach is quite general and could be a useful tool models of global economic and Earth system for managers in other locations wishing to bal- trends. ance economic and ecological goals. Ref erences Sustainable landscape management Andrewartha, H. G., and L. C. Birch. 1984. The Maintenance of biological diversity is only one Ecological Web. Chicago, Ill.: University of Chi- aspect of sustainable management of landscapes. cago Press. For example, landscapes can be characterized as Brown, J. H. 1978. "The Theory of Insular Bioge- sustainable only if the nutrient losses of their ography and the Distribution of Boreal Birds component parts (forests, agricultural fields, and and Mammals." Great Basin Naturalist Memoirs so forth) are balanced by nutrient inputs. Further- 2, pp. 209-27. more, sustainability implies that the manage- Diamond, J. M. 1975. "The Island Dilemma: Les- ment of one habitat or one landscape is not done s at the expense of another. Thus, the agricultural son of Mod ern Biog a SieConser- productivity of one landscape cannot be said to vation 7o Ntr 129-46P be sustainable if that productivity is had at the v pp expense of the nutrients of another system or, for Dunning, John B., and B. D. Watts. 1990. "Re- that matter, at the expense of supplies of fossil gional Differences in Habitat Occupancy by fuels. Also, a landscape cannot be viewed as Bachman'sSparrow."TheAuklO7,pp.463-72. sustainable if it exports (leaks) toxic materials Elton, C. 1949. "Population Interspersion: An Es- that degrade other systems. For example, forest say on Animal Community Patterns." Journal or agricultural productivity in one landscape is of Ecology 37, pp. 1-23. not sustainable if it requires the use of pesticides Haggerty, T. M. 1986. "Reproductive Ecology of that are transported to adjacent landscapes or Bachman's Sparrow (Aimophila aestivalis) in into groundwater. Central Arkansas." Ph.D. diss., University of The necessity to consider how management of Arkansas, Fayetteville. one habitat or landscape influences neighboring Hansen, A. J., T. A. Spies, F. J. Swanson, and J. L. habitats or landscapes suggests a hierarchical Ohmann. 1991. "Conserving Biodiversity in approach to sustainable systems. According to Managed Forests: Lessons from Natural For- this approach, components of a subsystem are ests." BioScience 41, pp. 382-92. said to be sustainable only if the practices within ene . the subsystem can be maintained indefinitely Kadmon, R., and H. Ronald Pulliam. 1994. "Is- without degrading other subsystems or the larger land Biogeography: Effect of Geographical Iso- system of which they are a part. For example, an lation on Species Composition." Ecology 74, agricultural field must be managed so as not to pp. 977-81. degrade either the soil within the field, on which Levins, R. 1966. "Strategy of Model Building in its own productivity depends, or the integrity of Population Biology." American Scientist 54, pp. adjacent subsystems (such as biodiversity of for- 421-31. est reserves or quality of groundwater). The agri- Liu, Jianquo. 1992. "ECOLECON: A Spatially Ex- cultural field must also be managed so as to '. . preserve the integrity of the larger human ecosys- CitMod fo Coloc Eonms ofnspecie , , ., .. . . . ~~~~Conservation in Cornplex Forest Landsca e." tem of which it is a part, including the economic P security of the humans who manage the system. h Maintaining biological diversity must thus be Liu, Jianquo, F. Cubbage, and H. Ronald Pulliam. viewed inthecontextof managing theintegrityof Forthcoming. "Ecological and Economic Ef- human-dominated landscapes and the regional fects of Forest Landscape Structure and Rota- economic systems on which they depend. This tion Length: Simulation Studies Using approach will require integrating models of bio- ECOLECON." Ecological Economics. 98 Managing Landscapes for Sustainable Biodiversity Liu, Jianquo, John B. Dunning, and H. Ronald tions 2, pp. 165-77. Pulliam. Forthcoming. "A Spatially Explicit (SRFS) Savannah River Forest Station. 1992. "Sa- Model of Animal Population Dynamics on a vannahRiverSiteWildlife,Fisheries,andBotany Changing Landscape: The Bachman's Sparrow Operations Plan." Savannah River Forest Sta- at theSavannah RiverSite." Conservation Biology. tion, Forest Service, U.S. Department of Agricul- Margules, C., A. J. Higgs, and R. W. Rafe. 1982. ture. "Modem Biogeographic Theory: Are There Any Wilcove, D. S. 1989. 'Protecting Biodiversity in Lessons for Nature Reserve Design?" Biological Multiple-useLands:LessonsfromtheU.S.Forest Conservation 24, pp. 115-28. Service." Trends in Ecology and Evolution 4, pp. Pulliam, H. Ronald. 1988. "Sources, Sinks, and 385-88. Population Regulation." TheAmerican Naturalist Wilson, E. O., and E. 0. Willis. 1975. "Applied 132, pp. 652-61. Biogeography:TheDesignof NaturePreserves." Pulliam,H.Ronald,JohnB.Dunning,Jr.,andJianquo In M. L. Cody and J. M. Diamond, eds., Ecology Liu. 1992. "Population Dynamics in a Complex and Evolution of Communities, pp. 522-34. Cam- Landscape: A Case Study." Ecological Applica- bridge, Mass.: Belknap Press. 99 Defining and Measuring Sustainability: The Biogeophysical Foundations would maintain as many future options as pos- sible and still allow for current developments. Eduardo R. Fuentes Because ecological systems are very complex and we understand them only to a very lirr ited extent and because we need freedom for human I will make my comments from the perspective of creativity, I suggest that rather than attempting a biologist who is interested in landscapes and is to define what is sustainable, we should con- living in a developing country. Ronald Pulliam's centrate on what we perceive as not sustainable very attractive presentation addresses the ques- and as clearly and unnecessarily eliminating tionofsustainabledevelopmentandbiodiversity future options. Finding the bounds of what at the landscape level, with particular emphasis does not seem sustainable today if continued on metapopulations, sources, and sinks. In my indefinitely is more realistic than attempting to comments, I refer to sustainable development specify conditions for sustainable situations. from theperspectiveof developingcountriesand Defining thes- bounds could be somewhat then address some emerging questions concern- easier and more coherent with the necessary ing landscape development, metapopulations, freedom and the inherent changes of a market sources, and sinks. economy than attempting to specify, and later plan, the exact future use of land for each par- ticular region. Therefore, rather than thinking Sustainable development of ideal landscapes with a rigidly stable physi- ognomy, we might think of the maximum Sustainable development has been defined as stresses that various subsystems can tolerate using resources today without affecting the andstillmaintainsomefutureoptions.Perhaps options for future generations. This is a rela- one of the mo>st important consequences of tively new and attractive concept. However, it presentdiscussionson sustainabledevelopment isa criterion that isalmostimpossible to satisfy, will be the greater scope that environmental because, on the one hand, the current genera- impact assessments could have in the future if tion must necessarily continue altering the bio- they consider not only current consequences sphere to develop and, on the other, frequently but also reductions in future options. when a history-dependent system, such an eco- With all the limitations of our current knowl- system or the biosphere, is modified, there are edge, at least five different axes can be used to changes in future options. A compromise is define a five-dimensional envelope that indi- thereforecalled upon between today's usesand cates where serious consequences for future future options. In the future, other, new op- options begin. tions will be available, but if they cannot be Development must consider the fate of the specified now, they should not concern the two most nonrenewable resources: soils (with present analyses. all their internal complexity) and biodiversity For developing countries that will have to (in its widest meaning). Hence development continue changing their landscapes in the fu- schemes ought to consider maintaining these ture, this is a very important point. For them at variables within accepted standards and moni- least, sustainable development should not be tor their changes. Development must also con- understood as freezing the landscape at a par- sider changes in climate, since climate defines ticular configuration, but rather as continuous the basic scenario for all organisms and human development within a setofrestrictionsassoci- activities. Overall temperatures might be in- ated with the maintenance of as many future creasing,andrainfallpatternsmightalsochange options as possible. in the near future. Change in climate should be For example, it would not be culturally sus- a prime target when considering, assessing, tainable to attempt to freeze growth in Latin and monitoring development. Human popula- AmericaorinacountrylikeChile,whereabout tion and especially per capita impacts should 40 percent of the people are below the poverty also be considered. It isof the highest interest to line. In such countries, land uses are expected quantify and monitor these impacts. The to change, and our definitions should consider globalization of the economy is likely to pose those changes. The question, then, is what de- difficulties for our attempts to evaluate the per fines the appropriate set of restrictions that capita impact of people in the large consump- 100 Managing Landscapes for Sustainable Biodiversity tion centers in the world. A final and crucial Sources and sinks bound for sustainable development concerns pollution and the accumulation of toxic and Sources and sinks are part of systems in which radioactive materials, which will eventually there is heterogeneity, for example, landscapes. make life on earth impossible. Within the rather In general, if portions (patches) of the landscape extreme bounds imposed by the limits of physi- have different potentials, there will be flows, ologically tolerable contamination, additional sources, and sinks. The examples given in the frames could deal with different levels of qual- chapter by Pulliam are examples of a particular ity of life. Monitoring sustainable development case in which thepotentials relateto habitatswith should also involve the amounts of materials different relative population densities. and residues accumulated in the different com- However,inalandscapethereareseveralkinds partments of the planet and how our well-being of heterogeneities, generating different types of depends on the state of those deposits. These sources and sinks: five dimensions can be used to define a hypervolume~~~~~~~~ tha deemnsbud o ad * Water moving along the slopes of a watershed hypervolume that determies bounns sor lanr- from the source areas distributed throughout scape transformationsit or in a portion of it and sinking by percola- limiting future options. tion or through the lower end of it Several questions can be posed at this point. Should each stage be sustainable or should only * Sediments behaving roughly parallel to water end states be? Should we attempt to manage . Individuals of a particular species, including landscapes so that all states can be reversed? beneficial (resource species) such as many What is the appropriate spatial scale in each coastal marine species (see, for example, case? Should all or a few areas in the globe Roughgarden and others 1988), wildlife spe- move within the five-dimensional envelope? cies (Lande 1988; Poiani and Fuentes 1989; Another question pertains to the particular sys- Roughgarden and Fuentes 1977), as well as tem that should be aimed at. A landscape, for pest species, as in the Norte Chico of Chile example covered with forests, can be used at (Fuentes and Campusano 1985), all respond- any of N different levels (states) depending on ing to permanentorsporadichabitat gradients energy and material inputs from surrounding and constituting sourcesand sinksof different areas. At one extreme are quasi-pristine states; kinds at the other are various types of agricultural * Sets of species, as in colonization events fields. All of them could be sustainable, and followingnaturaldisturbanceofcomplexforest most involve some kind of compromise with ecosystems (Veblen, Schlegel, and Oltremari future options. The greater the transformation 1983), competing fugitive species (Bengtson and the greater the transformed area, the lesser 1991), predator-prey systems requiring the options for the future. The question is what different patches(as in Huffaker's 1958famous inputs are needed and from where they will be experiment with mites and oranges). obtained so that the process in the area can be considered sustainable. For individuals both of the same and of differ- The area considered in sustainable develop- ent species, the question of sources and sinks ment as well as the transformations allowed are involves habitat quality with its intraspecific and very important, because phenomena are differ- interspecific connotations and specificities. Con- ent at different scales. For example, the area in sequently, different population processes are which to express the phenomenon of altitudi- likely to have different areas of importance. The nal bird migration is different from the area in previousexamplewiththerangesforthealtitudi- which to observe the dynamics of sustainable nal migration of birds, the salmon cycle, and the tree gaps, or thearea in which to observe the life tree-phasedynamics illustrates this point well. In cycle of salmons. Maintaining a spatially or all these cases, the area involved and its heteroge- functionally restricted version of a system, neity is crucial if the system is to be maintained, which naturally needs more area or a higher but the areas themselves are different or at most amplitude of oscillations for its expression than overlap only partially. the area we can afford to maintain, involves Within species, numbers as well as composi- energy and material inputs and can be very tion within a given area vary within years, be- costly and be nonsustainable in the long run. tween years, between decades, between centuries. 101 Defining and Measuring Sustainability: The Biogeophysical Foundations Sustainable use implies that some species will be Fuentes, Eduardo, and C. Campusano. 1985. 'Test "lost" anyway. What species composition should Outbreaksand Rainfall in theSemiaridRegionof concern us when thinking of the sustainable de- Chile." Journal of Arid Environments 8, pp. 67-72. velopment of a given area? Huffaker, C. B. 1958. "Experimental Studies on More generally, if spatial heterogeneity and Predation: Dispersion Factors and Predator- therefore geophysical as well as population Prey Oscillations." Hilgardia 27, pp. 343-83. sources and sinks are as common as they seem to be, their disruption should be an important con- lagda.l 18. " ,er ticiendem24rap.y1455- cern when defining the bounds for sustainable logical Conservation." Scence 241, pp. 145560. development. However, if the relevant areas for Poiani, A., and E. R. Fuentes. 1989. "Preferences thevariousgeophysical and populationprocesses of Native Rodents for Shrub Clumps of Vari- overlap only partially, there are difficult choices ous Sizes and Compositions: Implications for to be made. Which of these partially overlapping the Structure of the Chilean Matorral." Redia areas should be preserved? Obviously, all areas 72, pp. 133-48. cannot be preserved if we want to develop the Roughgarden, J., and Eduardo R. Fuentes. 1977. land, and necessarily some source-sink systems "The Environmental Determinants of Size in will have to be altered. Research on sources and Solitary Populations of West Indian Anolis sinks should therefore not only describe these Lizards." Oikos 29, pp. 44-51. systems but also consider the consequences of Roughgarden J., S. Gaines, and H. Possingham. transforming and even deleting some of them. 1988. "Recruitment Dynamics in Complex Life Cycles." Science 241, pp. 1460-66. References Veblen, T. T., F. M. Schlegel, and J. V. Oltremari. 1983. "Temperate Broadleaf Evergreen Forests Bengtson, J. 1991. InterspecificCoof South America." In J. D. Ovington, ed., Mentaon, tio99." "intrs ic Competiton in Ecosystems of the World, vol. 10: Temperate Broad Metapopulations." Bi1olgicalJournaloftheLin- Leaved Evergreen Forests. New York, N.Y.: nean Society 42, pp. 219-37. Elsevier. 102 Scale and Sustainability: A Population and Community Perspective Simon A. Levin It is a pleasure to acknowledge support of the National Science Foundation, the Andrew W. Mellon Foundation, the U.S. Department of Energy, the National Aeronautics and Space Administration, the U.S. Environmental Protection Agency, Mclntire-Stennis, and Hatch. Most important, I thank Colleen Martinforherindispensable helpwith this chapterandforten years of collegial interaction. The invitation to construct this chapter came just as I was completing the manuscript, "The Problem of Pattern and Scale in Ecology," which was published by the Ecological Societyof America (Levin 1992a). This chapter covers much of the same material and summarizes the appropriate points presented in the Ecology paper. Parts are reproduced with permission of the Ecological Society of America. Understanding the problem of sustainability re- basis for extrapolation; such understanding pro- quires understanding the scale of natural space videsthefoundationforunderstandingandman- and time of population dynamics and how the agement. pattem of habitat on various scales influences Addressing the problem of scale also has fun- survival of a population. It is often argued that a damental applied importance. Global and regional fluctuating population, because it is, at times, changesinbiologicaldiversity,inthedistribution smallerthanastablepopulationofthesamemean of greenhouse gases and pollutants, and in cli- size, faces higher probabilities of extinction and mate all have origins in and consequences for genetic bottlenecks that are associated with low fine-scale phenomena. The general circulation population size. But neither the notion of fluctua- models that provide the basis for predicting cli- tion nor even that of size makes sense without mate operate on spatial and temporal scales many reference to the spatial and temporal scales of orders of magnitude greater than the scales at interest; an understanding of how variability is which most ecological studies are carried out associated with area and time period is thus fun- (Hansen and others 1987; Schneider 1989); satel- damental to defining and effecting sustainability. lite imagery and other means of remote sensing Achieving sustainability requires characteriz- provide spatial information somewhere in be- ing the natural patterns of variability within an tween the two, overlapping both. General circu- ecosystem or landscape and understanding what lationmodelsand remote-sensing techniquesalso biotic and abiotic processes are essential for their must lump functional ecological classes, some- maintenance. Understandingpatternsin terms of times into very crude assemblages (such as the the processes that produce them is the essence of "big leaf" to represent regional vegetation), sup- science and the key to developing principles for pressing considerable ecological detail. To de- management. Without an understanding of velop the predictive models that are needed for mechanisms, each new stress on each new system management, or simply to allow us to respond to mustbeevaluateddenovo, without any scientific change, we must learn how to connect the dispar- Defining and Measuring Sustainability: The Biogeophysical Foundations ate scales of interest of scientists studying these down in unnecessary detail. The essence of mod- problems at different levels. eling is, in fact, to facilitate the acquisition of this Thecapability to make theseconnectionsacross understanding by abstracting and incorporating scales is fundamental to developing strategies for just enough detail to produce observed patterns. sustainability.Anymanagementstrategyinvolves The reference to particular scales of interest intervention, and intervention is perturbation. emphasizesa fundamental point: there is no single An understanding of how a population, a com- correct scale on which to describe populations or munity, or an ecosystem will respond to pertur- ecosystems (Allen and Starr 1982; Greig-Smith bation must involve some understandingof what 1964; Meentenmeyer and Box 1987; Steele 1978, mechanisms construct that system and mediate 1989; Wiens 1989). Indeed, the forces governing the patterns it exhibits on diverse scales of space, the history of the evolution of life, shaped by time, and organizational complexity. competitive pressures and co-evolutionary inter- Indeed, even the determination of manage- actions, are such that each species observes the ment objectives requires us to select scales of environment on its own unique suite of scales of interest.Whatsystemcomponentsareof primary space and time (see, for example, Wiens 1976). interest? Are individual species to be maintained, When we observe the environment, we neces- or should the focus be on functional groupings? sarily do so on only a limited range of scales; Microbial decomposition must be maintained, therefore, our perception of events provides us but do we care about the preservation of every with only a low-dimensional slice of a high-di- species within the microbial community? The mensionalcake.Insomecases,thescalesofobser- benthic community performs critical functions vation may be chosen deliberately to elucidate for the system, butmanyof thesecan be protected key featuresof thenatural system; moreoften, the even if the species composition of the community scales are imposed on us by our perceptual capa- is altered. bilitiesorby technological orlogistical constraints Similar considerations apply to the manage- (Steele 1978). In particular, the observed variabil- ment of multispecies fisheries, at least to the ity of the system depends on the scale of descrip- extent that consumer preferences can be shifted tion (Haury, McGowan,and Wiebe 1978;Stommel from one species to another. Terrestrial commu- 1963). nities provide other examples; the value of a In describing natural phenomena, we mimic forest for example, will be measured differently evolutionbyaveragingover uncertainty. At very by different individuals. If a forest is viewed fine spatial and temporal scales, stochastic phe- solely as a source of fiber, the relevant measures nomena (or deterministically driven chaos) may of performance will differ from those that would make the systems of interest unpredictable. Thus be guided by a recognition of the forest as wilder- we focusattention on largerspatial regions,longer ness, or as critical habitat for its denizens, or as time scales, or statistical ensembles, for which mediator of climate. Once again, scale and per- macroscopic statistical behaviors are more regu- spective are critical. lar. This is the principal technique of scientific In general, there is no sirgle way to measure inquiry: by changing the scale of description, we the functioning of an ecosystem or its ecological move from unpredictable, unrepeatable indi- or human value. One must impose some selective vidual cases to collections of cases whose behav- filter on the infinity of possible measures of a ior is regular enough to allow generalizations to system's performance and decide which features be made. In so doing, we trade the loss of detail or are most representative. This applies not only to heterogeneity within a group for the gain of pre- the level of functional detail but also to the spatial dictability; we thereby extract and abstract those and temporal scales of interest. Thus, it is essen- fine-scale features that have relevance for the tial to understand how the perspective associated phenomena observed on other scales. The impli- with a particular choice of scales biases the view cations for sustainability are profound. In preser- of asystemandhowtoextrapolatefromonescale vation, we must not become embroiled in the to another. details of how a system functions. Wemustdeter- To achieve these objectives, we must under- mine its essential features and assure that these are stand how information is transferred from fine maintained. Those essential features cannot be de- scales to broad scales, and vice versa. We must fined without reference to a set of external valua- learn how to aggregate and simplify, retaining tions of the system. Science can illuminate a deci- essential information without getting bogged sion making process, but it cannot substitute for it. 104 Scale and Sustainability: A Population and Community Perspective The concepts of scale and pattern are inelucta- other areas, indeed other nations, to provide the bly intertwined (Hutchinson 1953). The descrip- life support services that have been eliminated tion of pattern is the description of variation, and locally. That may work well on local scales, and the quantification of variation requires the deter- less well but adequately on broader scales, but it mination of scales. Thus, the identification of cannot go on indefinitely, and ultimately each patternisanentree into theidentification of scales region must be made as closed and self-sufficient (Denman and Powell 1984; Powell 1989). as possible for sustainability to be achieved. It is Our efforts to develop theories of how ecosys- well understood that larger systems are less open tems or communities are organized must revolve than smaller ones, because of elementary geomet- around attempts to discover patterns that can be ric principles. As the developing nations begin to quantified within systems and compared across achieve equity in terms of the drain they place on systems. Thus, considerable attention has been theenvironment, therewillbenoplaceleftto turn directed to techniques for describing ecological for sustenance. We mustunderstand how to man- or population patterns (Burrough 1981; Gardner age these fragmented systems, on scales from the and others 1987; Milne 1988; Sokal, Jacquez, and forest to the biosphere. Wooten 1989; Sugihara, Grenfell, and May 1990). The view of systems as mosaics of islands has Once patterns are detected and described, we can taken a number of interesting directions. The seek to discover the determinants of pattern and concept of patch dynamics (Levin and Paine 1974; themechanismsthatgenerateand maintain those Paine and Levin 1981; Pickett and White 1985; patterns. With an understanding of mechanisms, Watt 1947) has become a popular theme in both one has predictive capacity that is impossible the terrestrial and marine literatures and has led with correlations alone. to new views of community structure. The problem of ecological pattern is insepa- Metapopulation models, in which systems are rable from the problem of the generation and viewed as composed of interacting populations maintenanceof diversity (Levin 1981). Not only is of local demes, have been shown to be of impor- theheterogeneity of theenvironment often essen- tance in conservation biology (Armstrong 1988; tial to the coexistence of species, but the very Burkey 1989; Fahrig and Paloheimo 1988a, 1988b; description of the spatial and temporal distribu- Gilpin and Hanski 1991; Nuernberger 1991), evo- tions of species describes patterns of diversity. lutionary theory (Levene 1953), and epidemiol- Thus, understanding pattern, its causes, and its ogy (Levin and Pimentel 1981) and have become consequences is central to understanding evolu- the focus of considerable theoretical effort (for tionary processes such as speciation as well as example, Nee and May 1992), especially the role ecological processes such as succession, commu- that the structure of the metapopulation plays in nity development, and the spread and persis- facilitating the coexistence of species. tence of species. Food webs Mosaics and fragmented environments One of the most natural ways to describe a com- A fundamental challenge in achieving munity or an ecosystem is in terms of the trophic sustainability is to cope with the fragmented en- relationships among species and the tangled web vironments that human activity has imposed on that results (Elton 1958; Levin, Levin, and Paine the landscape. Not only are natural systems mo- 1977; Odum 1983; Paine 1966, 1980). Consider- saics of patches in various stages of successional able theoretical interest has been directed to regu- development, but the broader landscape itself is larities that can be detected in the topological a patchwork of urban, agricultural, natural, and structure of such webs (Cohen 1977, 1989; Pimm other pieces. This landscape is thus the result of a 1982; Sugihara 1982; Yodzis 1989). This seems all process that has defied sensible regional and glo- the more remarkable because such patterns seem bal patterns, satisfying limited local or exploit- to hold true regardless of the criteria used to ative objectives without thought to the define the elements of a web or the criteria for sustainability of the whole enterprise. Generally, deciding that a link exists between two species theproblemofsustainabilityhasbeenavoidedby (but see Cohen 1989; Schoener 1989). Indeed, viewing the exploited areas as being open sys- there clearly is no unequivocal way to charac- tems and drawing heavily on other systems and terize a web. Is a taxonomic subdivision most 105 Defining and Measuring Sustainability: The Biogeophysical Foundations appropriate, or would a functional one serve have provided data on spatial variations in the better? Should subdivision stop at the species distributions of phytoplankton and zooplankton level, consider different demographic classes, be over half a century (see, for example, Colebrook partitioned according to genotype, and so forth? 1982; McGowan 1990). The evidence from these However a class were defined, one could parti- studies has been that large spatial and temporal tion it further according to various kinds of crite- scales show the greatest variations and that these ria, reducing variability within a class while sac- correlate well with large-scale variations in cli- rificing the predictability that can be achieved for mate (Dickson and others 1988; McGowan 1990). largerassemblages. Thisis the same kind of prob- The approach taken (Radach 1984) is first to ask lem confronted when one deals with spatial and how much of the variation can be explained by temporal scale, but with added layers of com- variationinthephysicalenvironmentandthento plexity. look to autonomous biological factors to account for the balance. This mode of attack perhaps requires scrutiny, given the possibility that in- Global change in climate trinsicbioticfactorsmightaccountforsomevaria- and ecological models tion in climate. Only mechanistic approaches that examine the effectsof scale can address thispuzzle. Global changes in climate and in the concentra- Agren and others (1991) review models of the tions of greenhouse gases will have major effects linkage of production and decomposition and on the vegetational patterns at local and regional discuss the linkages of process at different scales. scales (Clark 1985; MacArthur 1972); in turn, Theproblemofscalingfromtheleaftotheecosys- changes that occur at very fine scales, such as tem and beyond fundamentally challenges pre- alterations in rates of stomatal opening and clos- dictionsof theeffectsof global change (Ehleringer ing,ultimatelywill haveimpactsatmuchbroader and Field 1993; Norman 1980). Agren and others scales (Jarvis and McNaughton 1986). General (1991) point out that ecosystem models that oper- circulation models, which form the basis of pre- ateatonlyonelevelofintegrationarenotlikelyto dictions of climate, operate on scales of hundreds incorporate mechanisms properly and that it is of kilometers on a side, treating as homogeneous essential to develop methods forintegratingfrom all of the ecological detail within (Schneider 1989; finer scales; this reiterates the central theme of Hansenandotherst987).Ontheotherhand,most this chapter. A related problem is the need to ecological studies are carried out on scales of connect processes operating at different levels of meters or tens of meters (Kareiva and Anderson integration, as for example, the linkages between 1988), and even ecosystem studies are at scales grassland biogeochemistry and atmospheric several orders of magnitude less than those rel- processes (Parton and others 1989; Schimel and evant to general circulation models. Thus, a fun- others 1990). damental problem in relating the large-scale pre- It is worth noting (Holling, personal commu- dictions of the climate models to processes at the nication) that separating climatic and biotic influ- scale of ecological information is to understand ences on changing patterns of ecosystem can be how information is transferred across scales (Jarvis extremely problematic. Extrinsic influences can and McNaughton 1986; Levin 1993). triggerqualitativechangesin thedynamicsof the To address this problem, both statistical and system (Levin 1978); cases in point may involve correlational studies are needed, as is modeling firesoroutbreaksthataretriggeredbyachangein designed toelucidate mechanisms. A useful place climatebut show very littlecorrelation withit(for to begin is the quantification of spatial and tem- example, Holling 1992b). poral variability as a function of scale (see, for example, Kratz, Frost, and Magnuson 1987; McGowan 1990). Long temporal and spatial se- Pattern and scale ries can be used to examine similar patterns in the variation of climate and components of the eco- The critical environmental problem facing soci- system; where scales of variation match, there is ety today issustainability: the maintenanceof our at least the basis for investigating mechanistic natural and managed systems so that they will be relationships. An example is thecontinuous plank- available, in roughly the form we found them or ton recorder surveys of the North Atlantic, which better, for later generations. To achieve 106 Scale ana bustainabitlity: A Population and Community Perspective sustainability, however, it is helpful if not essen- ment on a unique range of scales and thus re- tial to know what it is. What are the properties of sponds to variability individualistically. Thus, no these systems that we most wish to sustain? In- description of the variabilityand predictability of deed, does sustainabil ity mean preservation as is, the environment makes sense without reference or does it incorporate the natural patterns of to the particular range of scales that are relevant change and growth that occur and are perceived to the organisms or processes being examined. differently at different spatial, temporal, and or- Such issues are most clear for spatial and tem- ganizational scales? poral scales but apply as well to organizational Only one answer is possible to the last ques- complexity. The recognition in marine fisheries tion. Change and renewal are essential features of that total yield in multispecies fisheries remains any system, and attempts to suppress such change fairly constant os er long periods of time, though can undermine sustainability. Modern views of the composition of species may change dramati- optimal strategies for managing forest fires are cally (May 1984), is a consequence of broadening evidence of this principle, and Holling's empha- the scale of description. Similarly, a claim that sis on resilience and flexibility must guide in- microbial communities are stable to perturba- formed practice (Holling 1986). Yet the rate of tions,suchastheintroductionofgeneticallyengi- change and variation will differ with the spatial neered organism>, results from the application of and temporal scales of interest and with the way a taxonomically broad filter, perhaps because in which components are aggregated in one's only a fraction of the microbial community can be view of the system. Unless we can identify a identified.Inecosystemsresearch,oneislikelyto single spatial, temporal, and organizational scale be concerned with a functional guild of microor- of interest, we must ask how a system's proper- ganisms that perform a particular service to the ties change with perceptual scale. And in any ecosystem and to use functional redundancy to case, to achieve sustainability, we must under- explain why ignoring changes within a guild is stand how the patterns described at one scale of acceptable. This is the key to scaling and interre- interest are influenced by events, both natural lating phenomena at different scales: knowing and anthropogenic, that are taking place on other what fine detail is relevant at the higher levels scales. and what is noise. The question of identifying which aspects of a There are several stages in the examination of system must be preserved is more problematic, the problem of pattern and scale. First, one must because it cannot be answered without reference have measures to describe pattern (Gardner and to how society values and wishes to use the others 1987; Milne 1988), so that criteria can be system. Ecosystems perform diverse services to established for relating that pattern to its causes humans byservingas sources of food and fiber, as and consequences. Cross-correlational analyses mediatorsof environmental quality,and asplaces can provide initial suggestions as to mechanisms to recreate. Decisions to dedicate segments of the but may miss emergent phenomena that arise landscape to particular functions are personal or from the collective behavior of smaller-scale pro- societal decisions, rather than scientific ones, cesses. Theoretical investigations of the various though decisions about land use fundamentally mechanisms through which patterncan arisepro- affect global sustainability. Explicitly recogniz- vide a catalog of possibilities and may suggest ing the role that external valuation plays in mea- relevant experiments to distinguish among hy- suring the health or capital of an ecosystem is a pothesized mechanisms. firststep towardachievingsustainability. Balanc- All ecological systems exhibit heterogeneity ing those decisions in ways that maintain the and patchiness on a broad range of scales, and global good is our greatest challenge. this patchiness is fundamental to population dy- Two fundamental and interconnected themes namics(Levin 1974; Roughgarden 1976), commu- in ecology are the development and maintenance nity organization and stability (Holling 1986; of spatial and temporal pattern and the conse- Kareiva 1987), and element cycling (Bormann quences of that pattern for the dynamics of popu- and Likens 1979). Patchiness is a concept that cuts lations and ecosystems. Central to these ques- across terrestrial and marine systems and pro- tions is the issue of how the scale of observation videsacommonground for populationbiologists influences the description of pattern; each indi- and ecosystem scientists. Patchiness, and the role vidual and each species experiences the environ- of humans in fragmenting habitats, is key to the 107 Defining and Measuring Sustainability: The Biogeophysical Foundations persistence of rare species and the spread of pest tionary responses of their components, rather species. The level of species diversity represents than some higher-level evolution at the ecosys- a balance between regional processes, such as tem level, Gaia notwithstanding (Lovelock 1972; dispersal and species formation, and local pro- see also Schneider and Boston 1991 for a wide cesses, such asbiotic interactions and stochasticity range of views). (Ricklefs 1987). That there is no single correct scale or leve. at Spatial pattern and patchiness have many con- which to describe a system does not mean that all sequences for the biota. Patchiness in the distri- scales serve equally well or that there are not bution of resources is fundamental to the way scaling laws. This is the major lesson of the theory organisms exploit their environment (Mangel and of fractals (Mandelbrot 1977; Milne 1988; Sugihara Clark 1986; Pulliam 1989; Schoener 1971; Wiens and May 1990). The power of methods of spatial 1976). Environmental heterogeneity provides a statistics-such as fractals, nested quadrat analy- diversity of resources that can lead to coexistence sis (Greig-Smith 1964; Oosting 1956), among competitors that could not coexist in ho- semivariogramsorcorrelograms(Burrough 1981, mogenous environments (Horn and MacArthur 1983a, 1983b; Sokal, Jacquez, and Wooten 1989; 1972;Levinl970,1974);buttheproblemofhowto Sokal and Oden 1978), or spectral analysis count the number of resources is vexing. Trivi- (Chatfield 1984)-or of allometry (Brown and ally, no environmentiscompletely homogeneous. Nicoletto 1991; Calder 1984; Harvey and Pagel But how different must resources be to support 1991; Platt 1985) is their capability to describe different species? This question, which has been how patterns change across scales. Thus, such central in community ecology (MacArthur 1970; methods have been used in ecology to quantify May and MacArthur 1972; Whittaker and Levin change in soilsand inecosystem propertiesat the 1977), goes to the heart of the problem of scale. level of subfields (Robertson and others 1988) or Species can subdivide the environment spatially, landscapes (Krummel and others 1987) and in concentratingondifferentpartsofthesameplant marine systems to quantify the distribution of (Broadhead and Wapshere 1966), different layers physical factors, primary producers, and con- of vegetation (MacArthur, Recher, and Cody sumers (Haury, McGowan, and Wiebe 1978; 1966), or different microenvironments; or they Levin, Morin, and Powell 1989;5teele 1978,1991; can subdivide it temporally, partitioning a suc- Weber, El-Sayed, and Hampton 1986). cessional gradient (Levin and Paine 1974) or a The simple statistical description of patterns is seasonal one. Thus, resource partitioning can re- a starting point, but correlations are no substitute sultin temporally constant, spatially nonuniform for mechanistic understanding (Lehman 1986). patterns, or spatially constant, temporally non- Modeling can play a powerful role in suggesting uniform ones, or spatiotemporal mosaics (Levin possible mechanisms and experiments, in explor- and Paine 1974; Paine and Levin 1981; Tilman ing the possible consequences of individual fac- 1988; Whittaker and Levin 1977). tors that cannot be easily separated experimen- All of this reinforces the recognition that there tally, and in relating fine-scale data to broad-scale is no single correct scale at which to view ecosys- patterns. tems; the individualistic nature of responses to Because there is no single scale at which eco- environment means that what we call a commu- systems should be described, there is no single nityor ecosystem is really just an arbitrary subdi- scale at which models should be constructed. vision of a continuous gradation of local species Methods from statistics and dynamical systems assemblages (Whittaker 1975). It also carries im- theory can play an important part in helping to portant implications for predicting how the biota determine the dimensionality of underlying will respond to global change and other stresses. mechanisms and of appropriate models (Schaffer Communities are not well-integrated units that 1981; Schaffer and Kot 1985; Sugihara, Grenfell, moveenmasse.Theyarecollectionsoforganisms and May 1990; Takens 1981). We need to have and species that respond individualistically to available a suite of models of different levels of temporal variation, as they do to spatial varia- complexity and to understand the consequences tion. This is also true, of course, of the evolution- of suppressing or incorporating detail. Models ary responsesof populations. Thus, if predictable that are insufficiently detailed may ignore critical patterns may be observed in what we define as internal heterogeneity, such as that which is communities and ecosystems, they have arisen responsible for maintaining species diversity through the individualistic ecological and evolu- (Holling 1986); it is clear, for example, that the 108 Scale and Sustainability: A Population and Community Perspective broadi brush of the general circulation models ability also can have major consequences for pat- ignores detail that is relevant for understanding terns of nutrient cycling (Bormann and Likens biotic influences on climatic systems, and vice 1979), persistence (Pimm and Gilpin 1989), and versa. On the other hand, overly detailed models patterns of the spread of introduced species provide littleunderstanding of what theessential (Durrett 1988; Mooney and Drake 1986). The key forces are, have more parameters and functional is to separate the components of variability into forms to estimate than the available data justify, those that inhibit persistence and coexistence, admit multiple basins of attraction, and are more those that promote them, and those that are noise prone to erratic dynamics that hamper prediction (Chesson 1986). and parameter estimation. Just as we would not To address such phenomena, we must find seek to build a model of human behavior by ways to quantify patterns of variability in space describing what every cell is doing, we cannot and time, to understand how patterns change expect to model the dynamics of ecosystems by with scale (for example, Steele 1978,1989; Dagan accounting for every individual, or for every spe- 1986), and to understand the causes and conse- cies (Ludwig, Jones, and Holling 1978). We must quences of pattern (Levin 1989; Wiens 1989). This determine what levels of aggregation and simpli- is a daunting task that must involve remote sens- fication are appropriate for the problem at hand. ing, spatial statistics, and other methods to quan- In an extremely instructivestudy, Ludwig and tify pattern at broad scales; theoretical work to Walters (1985) have shown clearly that in some suggest mechanisms and explore relationships; cases aggregated models can serve as better man- and experimental work, carried out both at fine agement tools than highly detailed models, even scales and through whole-system manipulations, when the data used to fit the parameters of the to test hypotheses. Together, these can provide model havebeen generatedby the detailed model; insightsas to how information is transferred across in retrospect, this should accord well with intu- scales and, hence, how to simplify and aggregate ition. The problem of aggregation and simplifica- models. tion is to determine the minimal level of detail The problem of relating phenomena across that is sufficient to the task (Levin l992b; Rastetter scales is the central problem in biology and in all and others 1992). of science. Cross-scale studies are critical to complement more traditional studies carried out on narrow single scales of space, time, and orga- Conclusions nizational complexity (Hollingl992a;Levin 1988, 1989; Meentenmeyer and Box 1987; Steele 1978, Classical ecological models (Scudo and Ziegler 1989),justasmeasuresoff-diversityareneeded to 1978) treated communities as closed, integrated, complement within-community measures of deterministic, and homogeneous. Such models A-diversity (Whittaker 1975). By addressing this are simplifications of real systems and provide a challenge, using the insights gained from similar place to begin analysis. However, each of these studies in other sciences and the unique ap- assumptions must be relaxed if we are to under- proaches that must be developed for ecological stand the factors governing the diversity and systems,wecanenhancegreatlyourunderstand- dynamics of ecosystems. Virtually every popula- ing of the dynamics of ecosystems and develop tion will exhibit patchiness and variability on a the theoretical basis necessary to manage them. range of spatial and temporal scales, so that the definition of commonness or rarity is a matter of scale (Schoener 1987). Virtually every ecosystem References exhibits patchiness and variability on a range of spatial, temporal, and organizational scales, sub- Agren, C. I., R. E. McMurtrie, W. J. Parton, J. stantial interaction with other systems, and sig- Pastor, and H. H. Shugart. 1991. "State-of-the- nificant influence of local stochastic events. These Art of Models of Production-Decomposition phenomena are critical for the maintenance of Linkages in Conifer and Grassland Ecosys- most species, which are locally ephemeral and tems." Ecological Applications 1, pp. 118-38. competitively inferior and which depend on the continual local renewal of resources and mecha- Allen, T. F. H., and T. B. Starr. 1982. Hierarchy: nisms such as dispersal to find those opportuni- Perspectives for Ecological Diversity. Chicago, ties. Fragmentation, local disturbance, and vari- Ill.: University of Chicago Press. 109 Defining and Measuring Sustainability: The Biogeophysical Foundations Armstrong, R. A. 1988. "The Effects of Distur- Colebrook, J. M. 1982. "Continuous Plankton bancePatchSizeonSpeciesCoexistence."jour- Records: Seasonal Variations in the Distribu- nal of Theoretical Biology 133, pp. 169-84. tion and Abundance of Plankton in the North Bormann, F. H., and G.E. Likens. 1979. Patternand Atlantic Ocean and the North Sea (Calanus Process in a Forested Ecosystem. New York: finmarchicus)." Journal of Plankton Research 4, Springer-Verlag. pp. 435-62. Broadhead, E., and A. J. Wapshere. 1966. Dagan, G. 1986. "Statistical TheoryofGroundwa- "Mesopsocus Populations on Larch in England: ter Flow and Transport: Pore to Laboratory, TheDistributionandDynamicsofTwoClosely Laboratory to Formation, and Formation to RelatedCoexistingSpeciesofPsocopteraShar- Regional Scale." Water Resources Research 22, ing the Same Food Resource." Ecological Mono- pp. 120S-134S. graphs 36, pp. 328-83. Denman, K. L., and T. M. Powell. 1984. "Effects of Brown, J. H., and P. F. Nicoletto. 1991. "Spatial Physical Processes on Planktonic Ecosystems Scaling of Species Composition: Body Masses in the Coastal Ocean." Oceanography and Ma- of North American Land Mammals." American rine Biology Annual Review 22, pp. 125-8. Naturalist 138, pp. 1478-512. Dickson, R. R., P. M. Kelly, J. M. Colebrook, W. S. Burkey, T. V. 1989. "Extinction in Nature Re- Wooster,andD.H.Cushing.1988."NorthWinds serves: The Effect of Fragmentation and the and Production in the Eastern North Atlantic." ImportanceofMigrationbetweenReserveFrag- lourmal of Plankton Research 10, pp. 151-69. ments." Oikos 55, pp. 75-81. Durrett, R. 1988. "Crabgrass, Measles, and Gypsy Burrough, P. A. 1981. "Fractal Dimensions of Moths:AnlntroductiontolnteractingPartideSys- Landscapes and Other Environmental Data." tems." Mathematical Intelligencer 10, pp. 37-47. Nature 294, pp. 240-42. Ehleringer, J., and C. Field, eds. 1993. Scaling . 1983a. "Multiscale Sources of Spatial Physiological Processes: Leafto Globe. San Diego, Variation in Soil. I: The Application of Fractal Calif.: Academic Press. Concepts to Nested Levels of Soil Variation." Elton, C. S. 1958. The Ecology of Invasions by Ani- Journal of Soil Science 34, pp. 577-97. mals and Plants. London: Methuen. . 1983b. "Multiscale Sources of Spatial Fahrig, L., and J. Paloheimo. 1988a. "Determinants Variation in Soil. 11: A Non-Brownian Fractal of Local Population Size in Patchy Habitat." Model and Its Applications in Soil Survey." Theoretical Population Biology 34, pp. 194-213. Journal of Soil Science 34, pp. 599-620. . 1988b. "Effect of Spatial Arrangement of Calder, W. A. III. 1984. Size, Function, and Life Habitat Patches on Local Population Size." History. Cambridge, Mass.: Harvard Univer- Ecology 69, pp. 468-75. sity Press. Gardner, R. H., B. T. Milne, M. G. Turner, and Chatfield, C. 1984. The Analysis of Time Series: An R. V. O'Neill. 1987. "Neutral Models for the Chroutfie C.3 18. TheoAnas a/Tm Seie:an Analysis of Broad Landscape Pattern." Land- antroduchon.3ded.London,England:Chapman scape Ecology 1, pp. 19-28. Chesson, P. 1986. "Environmental Variation and Gilpin, M. E., and 1. Hanski, eds. 1991. theCoexistence of Species." InJ.Diamond and Metapopulation Dynamics. London: Academic T. J. Case, eds., Community Ecology, pp.240-56. re New York: Harper and Row. Greig-Smith, P. 1964. Quantitative Plant Ecology. Clark, W. C. 1985. "Scales of Climate Impacts." 2d ed. London: Butterworths. Climatic Change 7, pp. 5-27. Hansen, J., I. Fung, A. Lacis, S. Lebedeff, D. Rind, Cohenl J.E.1977. pRatio ofPrytoPredatorsinCom- B. Ruedy, G. Russell, and P. Stone. 1987. Coe,t Food 1977."Rati re to pp. in5Coin '"Prediction of Near-term Climate Evolution: munityFood Webs." Nature 270,pp.165 77. What Can We Tell Decision-makers Now?" 1989. "Food Webs and Community In Preparing for Climate Change, pp. 35-47. Structure." In J. Roughgarden, R. M. May, Proceedings of the first North American con- and Simon Levin, eds., Perspectives in Theo- ference on preparing for climate changes, retical Ecology, pp. 181-202. Princeton, N.J.: October27-29.Washington,D.C.:Govemment Princeton University Press. Institutes, Inc. 110 Scale and Sustainability: A Population and Community Perspective Harvey, P. H., and M. D. Pagel. 1991. The Com- Levene,H.1953."GeneticEquilibriumWhenMore parativeMethod in Evolutionary Biology. Oxford, than One Ecological Niche Is Available." Ameri- England: Oxford University Press. can Naturalist 87, pp. 331-33. Haury, L. R., J. A. McGowan, and P. H. Wiebe. Levin, Simon A. 1970. "Community Equilibria 1978. "Pattems and Processes in the Time- and Stability, and an Extension of the Com- spaceScalesof PlanktonDistributions." InJ.H. petitive Exclusion Principle." American Natu- Steele,ed.,SpatialPatterninPlanktonCommuni- ralist 104, pp. 413-23. ties, pp. 277-327. New York: Plenum. . 1974. "Dispersion and Population Inter- Holling, C. S. 1986. "The Resilience of Terrestrial actions." American Naturalist 108, pp. 207-28. Ecosystems: Local Surprise and Global _ . 1978. "Pattern Formation in Ecological Change." In W. C. Clark and R. E. Munn, eds., Communities." In J. H. Steele, ed., Spatial Pat- Sustainable Development of the Biosphere, pp. tern in Plankton Communities, pp. 433-66. New 292-317. Cambridge, England: Cambridge York: Plenum. University Press. . 1981. "Mechanisms for the Generation - .1992a. "Cross-scale Morphology, Geom- and Maintenanceof Diversity." In R. W. Hiorns etry, and Dynamics of Ecosystems." Ecological and D. Cooke, eds., The Mathematical Theory of Monographs 62:4, pp. 447-502. the Dynamics of Biological Populations, pp. 173- .1992b. "TheRole of Forest Insectsin Struc- 94. London: Academic Press. turingtheBorealLandscape." InH.H.Shugart, . 1988. "Pattern, Scale, and Variability: An ed., A Systems Analysis of the Global Boreal For- EcologicalPerspective." InA.Hastings,ed.,Com- est, pp. 170-91. Cambridge, England: Cam- munity Ecology, pp. 1-12. Lecture Notes in Bio- bridge University Press. mathematics 77. Heidelberg: Springer-Verlag. Horn, H. S., and R. H. MacArthur. 1972. "Compe- . 1989. "Challenges in the Development of tition among Fugitive Species in a Harlequin a Theory of Community and Ecosystem Struc- Environment." Ecology 53, pp. 749-52. ture and Function." In J. Roughgarden, R. M. Hutchinson, G. E. 1953. "The Concept of Pattern May, and Simon A. Levin, eds., Perspectives in in Ecology." Proceedings oftheNationalAcademy Ecological Theory, pp. 242-55. Princeton, N.J.: of Sciences 105, pp. 1-12. Princeton University Press. Jarvis, P. G., and K. G. McNaughton. 1986. "Sto- . 1992a. "The Problem of Pattem and Scale matalControlofTranspiration:Scalingupfrom in Ecology." Ecology 73:6, pp. 1943-67. Leaf to Region." Advances in Ecological Research . 1992b. "The Problem of Relevant Detail." 15, pp. 1-49. In S. Busenberg and M. Martelli, eds., Differ- Kareiva, P. M. 1987. "Habitat Fragmentation and ential Equations Models in Biology, Epidemiol- the Stability of Predator-Prey Interactions." ogy, and Ecology, Proceedings, Claremont 1990, Nature 321, pp. 388-91. pp. 9-15. LectureNotes in Biomathematics92. Kareiva, P., and M. Anderson. 1988. "Spatial As- Berlin: Springer-Verlag. pects of Species Interactions: The Wedding of - . 1993. "Concepts of Scale at the Local ModelsandExperiments."InAlanHastings,ed., Level." In J. R. Ehleringer and C. B. Field, eds., Community Ecology, pp. 35-50. Lecture Notes in Scaling Physiological Processes: Leaf to Globe, pp. Biomathematics 77. Berlin: Springer-Verlag. 7-19. San Diego, Calif.: Academic Press. Kratz, T. K., T. M. Frost, and J. J. Magnuson. 1987. Levin, Simon A., J. E. Levin, and R. T. Paine. 1977. "Inferences from Spatial and Temporal Variabil- "Snowy Owl Predation on Short-eared Owls." ity in Ecosystems: Long-term Zooplankton Data The Condor 79, p. 395. fromLakes."AmericanNaturalistl29,pp.830-46. Levin, Simon A., A. Morin, and T. H. Powell. Krummel, J. R., R. H. Gardner, G. Sugihara, R. V. 1989. "Patterns and Processes in the Distribu- O'Neill, and P. R. Coleman. 1987. "Landscape tion and Dynamicsof Antarctic Krill." In Scien- Patterns in a Disturbed Environment." Oikos tific Committee for the Conservation of Antarctic 48, pp. 321-24. Marine Living Resources: Selected Scientific Pa- Lehman, J. T. 1986. "The Goal of Understanding pers, Part 1, pp. 281-99. SC-CAMLR-SSP/5. in Limnology." Limnology and Oceanography Hobart, Australia: Committee for the Conser- 31, pp. 1160-66. vation of Antarctic Marine Living Resources. 111 Defining and Measuring Sustainability: The Biogeophysical Foundations Levin, Simon A., and R. T. Paine. 1974. "Distur- Mooney, H. A., and J. A. Drake. 1986. Ecology of bance, Patch Formation,and CommunityStruc- Biological Invasions of North America and Hawaii. ture." Proceedings of the National Academy of New York: Springer-Verlag. Sciences 71, pp. 2744-47. Nee, S., and R. M. May. 1992. "Dynamics of Levin, Simon A., and D. Pimentel. 1981. "Selec- Metapopulations: Habitat Destruction and tion of Intermediate Rates of Increase in Para- Competitive Coexistence." Journal of Animal site-host Systems." American Naturalist 117, Ecology 61:1, pp. 37-40. pp. 308-15. Norman,J. M. 1980. "Interfacing Leaf and Canopy Lovelock, J. E. 1972. "Gaia as Seen through the Light Interception Models." In J. D. Hesketh Atmosphere." Atmospheric Environment 6, pp. and J. W. Jones, eds., Predicting Photosynthesis 579-80. for Ecosystem Models. Vol. 2, pp. 49-67. Boca Ludwig, D., D. D. Jones, and C. S. Holling. 1978. Raton, Fla.: CRC. "Qualitative Analysis of Insect Outbreak Sys- Nuernberger, B. D. 1991. "Population Structure tems: The Spruce Budworm and Forest." Jour- of Dineutus assimilis in a Patchy Environment: nal of Animal Ecology 44, pp. 315-32. Dispersal, Gene Flow, and Persistence." Ph.D. Ludwig, D., and C. J. Walters. 1985. "Are Age- diss., Cornell University, Ithaca, N.Y. structured Models Appropriate for Catch-Ef- Odum, H. 1983. Systems Ecology: An Introduction. fort Data?" Canadian Journal of Fisheries and New York: John Wiley. Aquatic Sciences 42, pp. 1066-72. Oosting, H. J. 1956. The Study of Plant Communi- MacArthur, R. H. 1970. "Species Packing and ties. 2d ed. San Francisco, CA: W. H. Freeman. Competitive Equilibrium among Many Spe- Paine, R. T. 1966. "Food Web Complexity and cies." Theoretical Population Biology 1,pp. 1-11. SpeciesDiversity."AmericanNaturalistl00,pp. . 1972. Geographical Ecology. New York: 65-75. Harper and Row. . 1980. "Food Webs: Linkage, Interaction MacArthur, R. H., H. Recher, and M. Cody. 1966. Strength and Community Infrastructure. The "On the Relation between Habitat Selection Third Tansley Lecture." Journal of Animal Ecol- and Species Diversity." American Naturalist 100, ogy 49, pp. 667-85. pp. 319-32. Paine, R. T., and Simon A. Levin. 1981. "Inter- Mandelbrot, B. B. 1977. Fractals:Form, Chance,and tidal Landscapes: Disturbance and the Dy- Dimension. San Francisco, Calif.: Freeman. namics of Pattern." Ecological Monographs Mangel, M., and C. W. Clark. 1986. "Towards a 51:2, pp. 145-78. Unified Foraging Theory." Ecology 67, pp. Parton, W. J., C. V. Cole, J. W. B. Stewart, D. S. 1127-38. Ojima, and D. S. Schimel. 1989. "Simulating May, R. M., ed. 1984. Exploitation of Marine Com- Regional Patternsof SoilC,N,andP Dynamics munities. Berlin: Springer-Verlag. in the U.S. Central Grassland Region." In L. May, R. M., and R. H. MacArthur. 1972. "Niche Bergstrom and M. Clarholm, eds., Ecology of Overlap as a Function of Environmental Vari- Arable Land, pp. 99-108. Dordrecht, the Neth- ability." Proceedings of the National Academy of erlands: Kluwer Academic. Sciences 69, pp. 1109-13. Pickett, S. T. A., and P. S. White, eds. 1985. The McGowan, J. A. 1990. "Climate and Change in Ecology of Natural Disturbance and Patch Dy- Oceanic Ecosystems: The Value of Time-Series namics. Orlando, Fla.: Academic Press. Data." Trends in Ecology and Evolution 5, pp. Pimm, S. L. 1982. Food Webs. Population and 293-99. Community Biology Series. New York: Meentenmeyer, V., and E. 0. Box. 1987. "Scale Chapman and Hall. Effectsin LandscapeStudies." In M.G.Turner, Pimm, S. L., and M. E. Gilpin. 1989. "Theoretical ed., Landscape Heterogeneity and Disturbance, Issues in Conservation Biology." In pp. 15-34. New York: Springer-Verlag. J. Roughgarden, R. M. May, and Simon A. Milne, B. T. 1988. "Measuring the Fractal Geom- Levin, eds., Perspectives in Theoretical Ecology, etry of Landscapes." Applied Mathematics and pp. 287-305. Princeton, N.J.: Princeton Univer- Computation 27, pp. 67-79. sity Press. 112 Scale and Sustainability: A Population and Community Perspective Platt, T. 1985. "Structure of the Marine Ecosys- Schneider, S. H., and P. J. Boston, eds. 1991. Scien- tem: Its Allometric Basis." In R. E. Ulanowicz tists on Gaia. Cambridge, Mass.: M.I.T. Press. and T. Platt, eds., "Ecosystem Theory for Schoener, T. W. 1971. "Theory of Feeding Strate- Biological Oceanography." Canadian Bulletin gies." Annual Review of Ecology and Systematics of Fisheries and Aquatic Sciences 213, pp. 55- 2, pp. 369-404. 75. _ _. 1987. "The Geographical Distribution of Powell, T. M. 1989. "Physical and Biological Scales Rarity." Oecologia (Berlin) 74, pp. 161-73. of Variability in Lakes, Estuaries, and the _ . 1989. "Food Webs from the Small to the Coastal Ocean." In J. Roughgarden, R. M. May, Large." Ecology 70, pp. 1559-89. and Simon A. Levin, eds., Perspectives in Theo- retical Ecology, pp. 157-80. Princeton, N.J.: Scudo, F. M., and,. R. Ziegler. 1978. The Golden Age Princeton University Press. of Theoretical Ecology: 1923-1940. Lecture Notes in Pulliam, H. Ronald. 1989. "Individual Behavior Biomathematics 22. Berlin: Springer-Verlag. and the Procurement of Essential Resources." Sokal, R. R., and N. L. Oden. 1978. "Spatial In J. Roughgarden, R. M. May, and Simon A. Autocorrelation in Biology 2: Some Biological Levin, eds., Perspectives in Theoretical Ecology, Implications and Four Examples of Evolution- pp. 25-38. Princeton, N.J.: Princeton Univer- ary and Ecological Interest." Biological Journal sity Press. of the Linnean Society 10, pp. 229-49. Radach, G. 1984. "Variations in the Plankton in Sokal, R. R., G.M. Jacquez, and M. C. Wooten. 1989. Relation to Climate." Rapports et Proces-Verbaux "Spatial Autocorrelation Analysis of Migration des Reunions (Conseil International pour and Selection." Genetics 121, pp. 845-56. l'Exploration de la Mer) 185, pp. 234-54. Steele, J. H. 1978. "Some Comments on Plankton Rastetter, E. B., A. W. King, B. J. Cosby, G. M. Patches." In J. H. Steele, ed., Spatial Pattern in Hornberger, R. V. O'Neill, and J. E. Hobbie. Plankton Communities, pp. 1-20. New York: 1992. "Aggregating Fine-scale Ecological Plenum. Knowledge to Model Coarser-scale Attributes . 1989. "Discussion: Scale and Coupling in of Ecosystems." Ecological Applications 2, pp. Ecological Systems." In J. Roughgarden, R. M. 55-70. May, and Simon A. Levin, eds., Perspectives in Ricklefs, R. E. 1987. "Community Diversity: Rela- Theoretical Ecology, pp. 177-80. Princeton, N.J.: tive Roles of Local and Regional Processes." Princeton University Press. Science 235, pp. 167-71. . 1991. "Can Ecological Theory Cross the Robertson, G. P., M. A. Huston, F. C. Evans, and Land-Sea Boundary?" Journal of Theoretical Bi- J. M. Tiedje. 1988. "Spatial Patterns in a Succes- ology 153, pp. 425-36. sional Plant Community: Patterns of Nitrogen Stommel, H. 1963. "Varieties of Oceanographic Availability." Ecology 69, pp. 1517-24. Experience." Science 139, pp. 572-76. Roughgarden,J. 1976. "Influence of Competition Sugihara, G. 1982. "Niche Hierarchy: Structure, on Patchinessin aRandom Environment." Theo- Organization, and Assembly in Natural Com- retical Population Biology 14, pp. 185-203. munities." Ph.D. diss., Princeton University, Schaffer, W. M. 1981. "Ecological Abstraction: Princeton, N.J. The Consequences of Reduced Dimensional- Sugihara, G., B. Grenfell, and R. M. May. 1990. ity in Ecological Models." Ecological Monographs "Distinguishing Error from Chaos in Ecological 51, pp. 383-401. TimeSeries."PhilosophicalTransactionsof theRoyal Schaffer, W. M., and M. Kot. 1985. "Nearly One- Society of London Bulletin 330, pp. 235-51. Dimensional Dynamics in an Epidemic." Jour- Sugihara, G., and R. M. May. 1990. "Applications nal of Theoretical Biology 112, pp. 403-27. of Fractals in Ecology." Trends in Ecology and Schimel, D. S., W. J. Parton, T. G. F. Kittel, D. S. Evolution 5, pp. 79-86. Ojima, and C. V. Cole. 1990. "Grassland Bio- Takens, F. 1981. "Detecting Strange Attractors in geochemistry: Links to Atmospheric Pro- Turbulence." In D. A. Rand and L. S. Young, cesses." Climate Change 17, pp. 13-25. eds., Dynamical Systeems and Turbulence: Warwick Schneider, S. H. 1989. "The Greenhouse Effect: 1980, pp. 366-81. Lecture Notes in Mathemat- Science and Policy." Science 243, pp. 771-81. ics 898. Berlin: Springer-Verlag. 113 Defining and Measuring Sustainability: The Biogeophysical Foundations Tilman, D. 1988. Plant Strategies and the Dynamics Whittaker, R. H., and S. A. Levin. 1977. "The and Structure of Plant Communities. Princeton, Role of Mosaic Phenomena in Natural Com- N.J.: Princeton University Press. munities." Theoretical Population Biology 12, Watt, A. S. 1947. "Pattern and Process in the Plant pp. 117-39. Community." Journal of Ecology 35, pp. 1-22. Wiens,J. A. 1976. "Population ResponsestoPatchy Weber, L. H., S. Z. El-Sayed, and 1. Hampton. Environments." Annual Review of Ecology and 1986. "TheVarianceSpectra of Phytoplankton, Systematics 7, pp. 81-120. Krill, and Water Temperature in the Antarctic . 1989. "Spatial Scaling in Ecology." Func- Ocean South of Africa." Deep-Sea Research 33, tional Ecology 3, pp. 385-97. pp. 1327-43. Yodzis, P. 1989. Introduction to Theoretical Ecology. Whittaker, R. H. 1975. Communities and Ecosys- New York: Harper and Row. tems. New York: Macmillan. 114 Scale and Sustainability: A Population and Community Perspective Commen ts ments drives the push for sustainable develop- ment today. Recent calculations suggest that humans now use and co-opt something like 40 Charles H. Peterson percent of the present net global terrestrial pro- ductivity (Vitousek and others 1986). Even in the absence of further human population I wish to acknowledge support from the National growth, the economiesof many populouscoun- ScienceFoundation ardtheStateofNorth Carolina tries are expanding, thus ensuring further ex- for support under projects of the Cooperative ploitation of global production. The human Institute of Fisheries Oceanography and to thank population of the globe is perilously close to its W. Ellington, M. E. Hay, H. Lenihan, and F. sustainable limit today. No false sense of secu- Micheli for their comments on the manuscript. rity over the power of ecology to design strate- gies for sustainable use can be allowed to de- flect attention from attempts to solve the hu- Simon Levin in his 1992 MacArthur award lec- man population crisis. ture printed in Ecology (Levin 1992) and again Although Levin's essays on the role of scale in his contribution to this volume provides in ecology and in achieving sustainability of irrefutable justification for the need to design, resource use follow the optimistic prescription conduct, and integrate analyses of ecological set forth by the Ecological Society of America's problems on multiple spatial and temporal SustainableBiospherelnitiative(Lubchencoand scales. This mandate to environmental scien- others 1991), it is important to temper this en- tists is as necessary and as compelling in the thusiasm withacorresponding analysis of how study of sustainability of resource use as it is in the spatial and temporal scales of human soci- the study of any basic problem in ecology. The eties constrain and otherwise affect the imple- fundamental nature, durability, and articulate mentation of advice from natural science on expression of the wisdom and advice provided sustainable resource use. The ecological prob- by Levin in these contributions are likely to lems of human use of resources do not differ ensure their longevity as conceptual roadmaps intrinsically from the ecological challenges guiding the strategies of ecological problem posed by the exploitation of other species in solving for some long time into the future. natural ecosystems. What is different is the I see a continuing, unfulfilled need for an need to recognize explicitly the implications of analogous contribution in the social sciences on human social scales in achieving sustainability. the implications of scale to our ability to achieve Some ecologists (for example, Ehrlich 1968; sustainable use of our planet's natural resources. Hardin 1993; Murdoch 1980) have devoted sub- There is a serious danger implicit in the recent stantial effort to addressing the important con- publicity over sustainable development and nection between natural and social sciences in the role of ecological science in achieving dealing with the problems of human popula- sustainability. The danger is that unrealistic tion and resource exploitation. Nevertheless, expectations of what ecology and natural sci- need still exists for an introspective and inte- ences might provide may deflect attention away grative overview of the role of scale in interac- from the need to address the social science tionsbetween natural and social science,analo- context and the core social problem that re- gous to what Levin has done here for natural quires us even to question sustainability: the science alone. The political, social, and eco- unchecked growth of the human population. nomic landscapes of the human societies on Reliance on some unspecified, future tech- Earth provide the practical, realistic framework nological innovation has long been used as an in which sustainability must be implemented. excuse for failing to use resources in a sustain- In the absence of a sophisticated understanding ablefashion. Although Malthus(1806) mayhave of the interactions between social scales of hu- failed to predict accurately the time scale on man populations and scales of impacts on the which resources would limit growth of the hu- natural resources on which they rely, the work man population, the principle of ultimate re- of ecologists on sustainability of resource use source limitation on which he built his argu- will be of little practical value. 115 Defining and Measuring Sustainability: The Biogeophysical Foundations Ref erences P. G. Risser. 1991. "The Sustainable Biosphere Initiative: An Ecological Research Agenda." Ehrlich, Paul R. 1968. The Population Bomb. New Ecology 72, pp. 371-442. York: Ballantine. Malthus, T. R. 1806. "An Essay on the Principle Hardin, G. 1993. Living within Limits: Ecology, of Population." 3rd ed. London: J. Johson Economics, and Population Taboos. New York: Publishers. Oxford University Press. Murdoch, W. W. 1980. The Poverty of Nations: Levin,Simon A. 1992. "ThePatternof Patternand Population, Hunger, and Development. Balti- Scale in Ecology." Ecology 73, pp. 1943-67. more, Md.: Johns Hopkins University Press. Lubchenco, J., A. M. Olson, L. B. Brubaker, S. R. Vitousek, P. M., P. R. Ehrlich, A. H. Ehrlich, and Carpenter, M. M. Holland, S. P. Hubbell, P. A. Matson. 1986. "Human Appropriation Simon A. Levin,J. A. McMahon, P. A. Matson, of the Products of Photosynthesis. Bioscience J. M. Melillo, H. A. Mooney, C. H. Peterson, 36, pp. 368-73. H. Ronald Pulliam, L. A. Real, P. J. Regal, and 116 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate Meinrat 0. Andreae and Robert E. Dickinson Theatmosphere maintainslifeon Earth through * To screen out harmful components of solar its physical weather and climate svstem and and cosmic radiation. through the effects of its chemical composition. * To redistribute, remove, and eliminate gases In a broad sense, there is no question but that and particles emitted by the biota that would these systems are natlurally sustainable; they be harmful if they were to accumulate in the will keep operating, no matter what. Life has atmosphere. been present on Earth for 4 billion years, and it seems highly likely that it will, in some form, A workable definition of atmospheric extend that far into the future. In this chapter, sustainability must include reference to the we use the term sustainability in a narrower present biota. Over longer time spans, living sense, adapting the definition of strong things can adapt to environmental change, ini- sustainability laid out by Daly (1991), which tially by changing the community's structure requires that both man-made and natural capi- and composition and eventually by evolving tal remain intact in the course of sustainable genetically. At issue here is the well-being of all development or in a sustainable global thepresentbiotaandtheirdescendants,reflect- economy. In this sense, a sustainable activity is ing the definition of sustainable development one that leaves intact the ability of the atmo- offered by the Brundtland Commission, as de- sphere to support the present biota. In other velopment that "meets the needs of the present words, for an activity to be sustainable, it should without compromising the ability of future gen- not diminish the abilitv of the atmosphere to erations to meet their own needs" (World Com- perform the following essential services for the mission on Environment and Development birta currently present: 1987). From the point of view of the biosphere, the * To regulate the global budget of energy and essential function of the atmosphere is to provide to redistribute energy between high and low a reasonably stable climate. Climate in a broad latitudes and between oceansand continents. sense encompasses the chemical and radiative * To establish surface temperature and mois- propertiesoftheatmosphereaswellastheclassi- ture regimes, growing seasons, and so forth cal meteorological parameters. "Reasonably suitable for the biota that have previously stable" requires that the rates of environmental adopted to such. change do not exceed those to which species and communities-or, in the human context, econo- * To provide visible solar radiation at levels mies and societies-are adapted or can respond needed for photosynthesis by plants. without serious negative consequences. Defining and Measuring Sustainability: The Biogeophysical Foundations The effect of human activities on global cli- perhaps more likely, because other changing mate in the narrower, meteorological sense is processes have up to now countered the green- being intensely discussed both within the sci- house warming. entific community and within society at large. There is a substantial reflection of solar ra- (The current state of the art on this topic is diation by tropospheric aerosols resulting from presented in the report of the Intergovernmen- human activities; this reflection could have up tal Panel on Climate Change-see Houghton, to now partially countered the greenhouse gas Callander, and Varney 1992; Houghton,Jenkins, warming (Charlson and others 1991). Due to and Ephraums 1990-and in a less technical, increased atmospheric aerosol loading, cloud but scientificallyaccurate formbyNilsson 1992.) properties could be changing in such a way as Human activities are altering the gaseous com- to cool the planet. Natural phenomena can also position of the atmosphere in previously un- act to slow greenhouse warming. For example, known ways and as a result are apparently we know that a major volcano, Pinatubo, warming global average temperatures to out- erupted last yearand injected massive amounts side the envelope of the past million years. This of sulfate aerosol into the stratosphere, which can be achieved relatively easily, because the should bring greenhouse warming to a halt for last 10,000 years have already been near the a year or two. Furthermore, some large-scale warm limits of this period. Furthermore, the adjustment may have occurred in the internal atmosphere has a relatively small mass, and the dynamics of the oceans. However, all these addition of trace gases that make up only a possibilities for natural or inadvertent mitiga- small fraction of the atmosphere's total mass tion of greenhouse warming would be unlikely can significantly affect radiative or chemical to continue to grow substantially with further balances. The old balances will not return within increases in greenhouse gases. This is because the next millennia. Can we achieve a new state carbon dioxide as well as the chlorofluorocar- that can still adequately maintain our natural bons that also threaten the ozone layer once systems, or will we continue to drift away from injected into the atmosphere remain there for a balance? century or more. By contrast, stratospheric aero- Changes in global climate over the last cen- sols remain only a year or two, and tropo- tury by themselves do not seem particularly spheric aerosols remain just a week or two. threatening; global temperatures have risen by The radiative climateon Earth is significantly fits and spurts, but by no more than 0.5 C or so. different from pre-industrial times. Besides the There is nodirect meansof establishing that the effects of the greenhouse gases on infrared ra- warming we see is that expected from green- diation, we are especially concerned with a house gases. What we see unequivocally is the very narrow band of wavelengths in the near incessant year-to-year rise of CO2 (carbon diox- ultraviolet. But other issues of harmful radia- ide) and other atmospheric greenhouse gases. tion might arise in the future. The inhabitants We can calculate with considerable confidence of the Earth are now flooded with electromag- what the role of these gases, and of their in- netic radiation over a very broad spectrum, crease, is in trapping thermal infrared radiation from the very low-frequency region of power and so in warming the planet. We also are fairly transmission lines, through radio frequencies confident that if the concentrations of CO2 were and microwaves, "light pollution" of the night ordersof magnitude smaller than they are, Earth sky, to ultraviolet and ionizing gamma radia- would be too cold to be habitable, perhaps tion.ltisusuallyassumedthatthechangeinthe completely covered with ice. However, present- radiative climate for photons lessenergetic than day science reaches its limits in trying to deter- the ultraviolet has relatively little effect or is, at mine precisely the change in climate expected most, annoying. However, this has not been from the present or future concentrations of demonstrated convincinglyeveninthecontext greenhouse gases. The increases in tempera- of acute and chronic impact on humans, and ture up to now, besides not being unequivo- studies on ecological effects are even less well cally a resultof greenhouse gases, are at the low investigated. Presently, the most threatening end of what might be expected. This may be perturbation of radiative climate is the weaken- eitherbecause the climate system is more stable ingof the stratosphericozonelayer, which shields to the greenhouse forcing than expected or, the biosphere from ultraviolet radiation. 118 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate Characteristic patterns of temperature, properties: (1) therelatively rapid mixingof the winds, and so forth in the atmosphere define atmosphere results in an intrinsic averaging the physical climate. Similarly, the patterns of and integrating effect, and (2) the low mass of chemical composition of the atmosphere, which the atmosphere makes it a sensitive indicator of are subject to characteristic temporal and spa- chemical and physical change. The study of the tial variations, represent the chemical climate. time record of atmospheric CO2providesa clear Trace gases, for example, are distributed example: the atmospheric concentration of this through the atmosphere in patterns of concen- gas reflects the balance of inputs and outputs of tration reflecting their sources and sinks, and CO2 from respiration, photosynthesis, weath- their concentrations are subject to diurnal, sea- ering, fossil fuel burning, deforestation, and so sonal, and secular variations. Since many trace forth. It would be impossible to measure all of gases and aerosols have effects on organisms these processes individually to account for the directly or through their influence on other overall balance with the accuracy required to atmospheric properties-for example, global deduce global trends. Yet the increase of atmo- temperatures-the perturbation of the chemi- spheric CO2 is readily measured, and its magni- cal climate by human activities is of serious tude provides a benchmark against which any concern. model of the carbon cycle must be matched. The atmosphere is closely connected to the Other compounds exist in the atmosphere that biosphere. In fact, both the major and the trace can serve as indicators of a changing chemical gases that constitute the atmosphere are ex- climate; these will be discussed in detail below. creted from the biota: nitrogen from microbial denitrification, oxygen from plant photosyn- thesis, methane from bacterial fermentation, Changng physical climate and so on. The atmosphere acts chemically to and carbon dioxide convert photochemically labile substances (such as hydrocarbons and reduced sulfur com- Over the lifetime of our planet, life has flour- pounds) into photochemically inert compounds ished over a very wide range of climatic and (such as CO2. sulfate, and nitrogen, N2), which . chemical conditions. However, that would not are then taken up again by the biosphere. In this have been possible if the planet had become too context, the atmosphere has two crucial func- cold or too hot to maintain liquid oceans. Geo- tions: first, to remove chemical emissions from physical and biological processes have pro- their source and prevent accumulation to toxic vided carbon dioxide and other greenhouse levels, and second, to combust trace gases pho- gases at concentrations suitable for maintain- tochemically so that they can be removed again ing a habitat for the unicellular life that existed from the atmosphere. Without this combus- over most of the history of the planet. Ad- tion, the biosphere would very soon choke on vanced life forms have only been present over its own gaseous emissions. a considerably narrower range of conditions, Human activity has placed an additional bur- after development of the ozone shield and tem- den on this system, releasing new kinds of peratures not dissimilar to those of today. Peri- emissions into the atmosphere and increasing ods with temperatures warmer by as much as many of the fluxes already emitted by natural 10 C over the present ones occurred in the processes. If the removal rates for these com- Cretaceous and earlier periods (1 to several pounds could speed up at the same time as hundred million years ago). emissions increase, this change could in prin- The extreme warmth of past geological peri- ciple be sustainable. However, the atmosphere ods is now generally explained as a conse- is apparently not able to accommodate this quence of CO2 concentrations up to ten times increasedburden.Insomerespects,theincreas- higher than present levels. Geophysical pro- ing input of trace gases into the atmosphere cesses are capable of providing such changes even reduces the rate at which they can be over tens of millions of years, and no other removed, providing a situation both unstable mechanism has been found capable of giving and with unpredictable consequences. such drastically different climates. The correla- In assessing sustainability, there are two sub- tion between temperatureand atmospheric CO2 stantial advantages to monitoring atmospheric is clearly seen in the analysis of Antarctic ice 119 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 8-1: Analysis of Air Trapped in Antarctic Ice Cores 2 0 -2 Temperature -4 J I -6 t -8 1990 CH 8 -l 0 _ tJ 4 700 level: 1t7bo ppbv _ _ %\ Methane 600 * ~~500 400 0 ,,, E 300 - 7SZ 228602 - 8 Carbon dioxide OO ~~260- 240 - 1990 co 220- 200 . 180 160 120 80 40 0 Age (thousand years before present) Note: Concentrations of methane and carbon dioxide were dosely correlated with the local temperature over the last 160,000 years. Source: Adapted from Houghton, Jenkins, and Ephraums 1990. cores (see figure 8-1). Human cultures evolved The Pleistocene climate has been one of ice over the last million years of Pleistocene cli- ages; large ice sheets have formed many times mate and have therefore experienced a much over the continents of the Northern Hemisphere narrower range of conditions. Civilization has as global temperatures dropped by several de- evolved within the last 10,000 years; much of grees from interglacial conditions. The extreme the technological development happened only cold of ice ages is now explained in part by the in the last century. lowconcentrationsof carbondioxideduringthese 120 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate periods, not much more than half of present con- Figure 8-2: Schematic Diagrams of Variations centrations. Current temperatures are near or at in Global Temperature since the Pleistocene the upper limit of that which has occurred over Period on Three Time Scales the last million years (see figures 8-2 and 8-3). Present interpretations of the history of our planet thus show that it has experienced and . survived large but slow fluctuations in carbon dioxide and temperature and that life has sur- vived these fluctuations. The fiery temperatures of Venus, which are the result of a massive atmo- sphere comprised almost entirely of carbon diox- EI ide, demonstrate that planetary habitability is not E Previou ice ages ast ice age inevitable. Furthermore, even if we were not to l l distinguish habitats of humans from those of 800,000 600,000 400,000 200,000 dinosaurs, it is remarkable that the current rates Years before present of change are hundreds to thousands of times faster than those of past geological changes. u Holocene maximum Humans began to affect atmospheric chemis- ice try when they learned to control fire, some 1.5 million to 2 million years ago, and began to use it on a large scale in Africa (James 1989). The use of fire by indigenous populations evidently changed the landscape quite dramatically, and on a nearly E continental scale, an activity that might not be sustainable by our definition. This led to a new equilibrium based on fire-managed landscapes 10,000 8,000 6,000 4,000 2,000 0 in the African savannas (Schule 1990). But once Years before present this transformation had occurred, the new bal- ance between ecology and human activity could u probably havebeen sustainable indefinitely. Since Little ice age the fires only short-circuited an otherwise bal- anced carbon cycle, and the emission of long- lived trace gases from the fires was small com- B pared to that from natural sources, chemical cli- . -e\e/ mate probablydid not undergo a dramatic change aE rmen de at this time. The industrial revolution, based on the avail- ability of abundant and cheap energy from the 1,000 AD 1,500 AD 1,900 AD combustion of fossil fuels, has resulted in a very Years before present different situation. Carbon is now being mobi- Source: Houghton, Jenkirs and Ephraurns 1990. lized into the atmosphere from a very large geo- logical reservoir (coal, petroleum, gas; see figure 84). There is about ten times as much carbon Carbon dioxide is a nearly inert gas. Its major readily accessible in fossil fuel deposits as there is path of removal is the ocean, mixing into deeper in atmospheric CO2, suggesting the potential for ocean layers and eventually being expelled as a majorchange inatmospheric composition when carbonate sediment. Also significant is the depo- a significant fraction of the fossil fuel deposits are sition of organic materials in sediments, the ini- combusted (see figure 8-5). tial source of our fossil fuels. About half of the From the beginning of the industrial revolu- incremental carbon dioxide supplied to the atmo- tion until now, concentrations of carbon dioxide sphere mixes into the upper layers of the oceans in the atmosphere have increased from 280 to 360 within a decade or two, and the remainder re- parts per million, due to the combined influence quiresa centuryor morebefore itsremoval. High- of the burning of fossil fuels and deforestation. precision measurements of the atmospheric con- 121 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 8-3: Global Mean Combined Land-Air and Sea-Surface Temperatures, 1861-89, Compared with the Average for 1951-80 0.4 1 -0.6 ! 1870 1890 1910 1930 1950 1970 1990 Year Source: Houghton, Jenkuns. and Ephraums 1990. Figure 8-4: Global Annual Emissions of CO2 from Fossil Fuel Combustion and Cement Manufacturing, 1860-1990 10 0~~~ 1 .0 _ ,z-: - 0 Po . N° ,.-" ~~~~~~~I, I , I , I , I I , I I 1 860 1880 1900 1920 1940 1960 1980 2000 Year Note: The average rate of increase in emissions between 1860 and 1910 and between 1950 and 1970 is about 4 percent a year. Source: Houghton, Jenkins, and Ephraums 1990. 122 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate Figure 8-5: Global Carbon Reservoirs and Fluxes DeforestationI Atmosphere 750l + 3/year 21 5 102 50 5G) 92 90 i Latd a l l ~~~~~~blota 55() 1 | + 50 X Rzers | ~~~~~~Surf ace oceain 10(0 + I/vea r Soil anid (letr tul -3\\ 15(0l Bi\ot.a3 Fossil tuiel1 j ln~~titermediate anid cdeep waters 38000 + 2, Xear 0.2 Note: Numbers apply to the pesent-day situation and represent tvpical values in the !iterature. Fluxes between atmosphere and surface ocean are gross annual exchanges. Numbers in italics indicate net annual accumulation of CO' due to the actions of hurnans. Units are gigatons of carbon (GtC; I gigaton = 109 metric tons = 10" kilograms) for reservoir sizes and GtC yr' for fluxes. Source: Houghton, Jenkins, and Ephraums 1990. centration of oxygen (0,) may help resolve some About 35 to 40 percent of the direct increase of of the remaining uncertainties in the atmospheric greenhouse warming over the last century comes balance of COV. Initial results show a decrease in from gases other than CO2: about 21 percent from atmospheric 02 consistent with the increase in methane(CH4),4 percent from nitrous oxide (N20), CO2 (R. F. Keeling, personal communication, 1992). and the rest from various chlorinated compounds. Because of this very long atmospheric lifetime Theozonelossfromchlorinatedcompoundsappar- of the remaining fraction of the fossil fuel C02 we ently largely compensates for its greenhouse warm- can only prevent further increase of carbon diox- ing (ozone being another important greenhouse ide in the long run by nearly stopping our use of gas), and the most important chlorinated contribu- fossil fuels. However, because some of the cur- tors to greenhouse warming are being phased out rent atmospheric excess will still move into the because of their proven destruction of the ozone oceans within a few decades, it is presently layer. Furthermore, methane has a relatively short possible to limit further growth by restricting decadal lifetime, so that its increasesmightberela- the use of fossil fuels to about one-third of its tivelvreversible.Hence,controllingfutureincreases present rate, and less ambitious limitations of CO2 is the primary question that needs to be would still be able to reduce significantly the addressed as a means of retarding future rates of rates of future increases. increase in greenhouse warming. 123 Defining and Measuring Sustainability: The Biogeophysical Foundations At present, decisions to curtail further growth cycle, future national economies, and possible re- of the use of fossil fuels cannot be made and are straints on the use of carbon because of concerns not made on the basis of accurate detailed projec- about the changes in climate (see figure 8-6). tions of future changes in climate. Rather, guid- Thus,onlysomewhatgeneralandsimplequan- ance is derived from the general magnitude and titative statements can be made about changes in probable nature of changes inferred by the re- climate that can be expected in the future. With search community on the basis of three-dimen- the present enormous use of fossil fuels, the incre- sional models of climate. These models provide mental greenhouse warming that has accumu- considerable detail as to possible climates in the lated over the last century will at least double in future, but these are not validated. Many aspects the next thirty years. Global temperatures should ofthemodelsarebasedonwell-establishedphysi- be warmer than the average of the twentieth cal laws, the large-scale hydrodynamics, thermo- century by I 'C to 3 C over land. Smaller increases dynamics, and radiative transfer of the atmo- are anticipated over the ocean because the large sphere. However, other aspects, such as those mass of the ocean's surface waters requires de- involvingtheformationofcloudsandtheirinterac- cades to warm. It is also because of the oceanic tion with radiation, processes of convection, and mass that temperature increases would be only vertical transport of water vapor, are based on very slowly reversible even if all the excess green- oversimplified and largely unverified assumptions. house gases were removed. Added to the uncertainties in the predictions Changes in temperature will be somewhat of the climate mnodels are the uncertainties in the smaller in the tropics and somewhat larger dur- furthergrowth of greenhouse gases and uncertain- ing winter in high-latitude areas than the global ties resulting from ignorance of details of the carbon average change. Models suggest significant mid- Figure 8-6: Global Emissions of Carbon Dioxide from Energy, Cement Production, and Deforestation CO2 emissions from energy, cement production, and deforestration 40 I 35- 30- IS92f 25 -SA90 20 -IS92a 1 25 0 IS92b O 10- = IS92d 5 ~ - -- t IS92c 0 1980 2000 2020 2040 2060 2080 2100 Year Nole: in a1992 supplement, the JPCCpresented six different scenarios of carbon dioxide emissions instead of the one business- as-usual scenario in the original report (called SA90 in the figure). The major difference from previous estimates is that higher population forecasts increase the emission estimates, while phase-out of halocarbons and more optimistic costs of renewable energy reduce them. The small difference between a and b is mainly accounted for by the commitment that many OECD countries have made to stabilize or reduce carbon dioxide emissions. Source: Adapted from Houghton, Callander, and Varney 1992; Nilsson 1992. 124 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate continental drought. Some largeshiftsin supplies Future global warming will be closely linked of water from the atmosphere are fairly likely. A to the world's vegetation. Over moist land re- further doubling of the incremental greenhouse gions, vegetation is a dominant control on the warming is likely by the end of the twenty-first flux of water between soil and atmosphere, and century unless society acts to restrain this in- its productivity in turn depends on the availabil- crease long before then (see figure 8-7). Without ity of water. Of special concern are the possible such restraints, the rate of increase of greenhouse effects of increased temperatures and accompa- warming will be considerably greater than it is nying shifts in rainfall patterns on the viability of today. natural forest ecosystems. The converse question, The increasing temperatures will be accompa- concerning the effects of forests on temperature nied by rises in sea level, most likely primarily and rainfall, also warrants attention. Climate from thermal expansion of the oceans and from modelers (for example,Nobre, Sellers,and Shukla melting of small glaciers. It is not known whether 1991) have been addressing the possible conse- the Greenland and Antarctic ice caps will grow or quences of converting the Amazon forest to grass- shrink over the next century, and this question land; these studies indicate that this conversion contributes considerable uncertainty to the over- would be accompanied by substantial decreases all projection of future sea levels given by the in rainfall that extend the duration of the dry report of the Intergovernmental Panel on Climate season, and by somewhat larger daytime tem- Change as between 0.3 and 1.1 meters by the year peratures. The implied changes could be irrevers- 2100 for the business-as-usual scenario ible on the southern margins of the forest. Addi- (Houghton, Jenkins, and Ephraums 1990). Sea tional effects would be expected from the large level rises of as much as a meter represent a volumesofsmokeintheatmosphereduringtropi- serious threat to many coastal areas of the world. cal dry seasons. Figure 8-7: Change in Temperature under Scenario IS92a 5 High climate 4 sensitivity 3 - Best climate -Q _ /sensitivity 2 - Low climate <,, 1 - // ~~ sensitivity 0 1 E .0 1990 2010 2030 2050 2070 2090 Year Note: Estimates of global mean temperature change IS92a using high (4.5 C), best-estimate (2.5 C), and low (1.5 C) values of climate sensitivity. The effects of sulphate aerosol and ozone depletion have not been taken into account. Source: Houghton, Callander, and Varney 1992. 125 Defining and Measuring Sustainability: The Biogeophysical Foundations A world 5C to 10C warmer than today by Changing chemical climate the middle of the twenty-second century is marvelous to contemplate for a climate scien- As already indicated, the chemistry of the at- tist as a natural experiment, but it is not a world mosphere has co-evolved with the biota. This he would likely want to live in or bequeath to includes the chemical composition of the atmo- his great-grandchildren. Long before then, sphere, the spatial and temporal distribution of many of the world's species would have been major and trace gases, and the chemical reac- stressed to extinction. We hope that reality will tions that influence the formation and destruc- be more benign than our best estimates, but a tion of gases and aerosol particles. Biological prudent world cannot afford to wait to see if we organisms were and still are major sources and are wrong; the only sensible path to a climate sinks of the constituents of the atmosphere. that will most probably support natural sys- However, over the last century, the use of tems is to proceed toward as much restraint in fossil fuel has grown to a magnitude where it our use of fossil fuels as the world can afford. not only influences the CO2 content of the atmo- Economic analyses show that initial reductions sphere but also supplies much of the trace gas in the use of fossil fuels through conservation emissions that are changing our chemical cli- would be beneficial (Lovins 1991), but that be- mate. At the same time, cheap and abundant yond some point, further reduction could be energy made possible the worldwide expan- prohibitively expensive. Taking the first easy sion of industrial and agricultural activity, steps today may allow us to avoid being forced which results in further emissions to the atmo- to make impossible decisions in the future. The sphere and increasing concentrations of trace possibility of a much-accelerated rate of warm- gases (see figure 8-8). The degree to which ing is more threatening than a slow continua- human society has taken over control of the tion of the present warming trends. biosphereisreflectedintheestimatebyVitousek Besides the uncertainty of present models of and others (1986, p. 368) that "nearly 40 percent climate, the difficulty of interpreting global tem- of potential terrestrial net primary productiv- perature records up to now in terms of green- ity is used directly, co-opted, or forgone be- house warming is used in arguments against cause of human activities." Human activities efforts to restrict further use of fossil fuels. In- now dominate the atmospheric cycle of many deed, the 0.5C rise in temperature experienced trace substances, resulting in widespread air over the last century is at the lower limit of that pollution and a change of chemical climate not expected. If it is furthermore imagined that this reversible on time scales of tens to hundreds of rise is partiallyor largely a result of other climatic years. processes, it might be inferred that climate is The relationships between the sources of at- much lessresponsive to changes in energy inputs mospheric trace gases and aerosols and the than has been inferred. However, such a conclu- environmental consequences of these emissions sion would fly in the face of past geological evi- can be represented in the form of a matrix (see dence that indicates a major role for greenhouse figure 8-9). In this figure, the drivers of atmo- warming in forcing changes larger than what we spheric change (agriculture, animal husbandry, now face. A more plausible interpretation of the fisheries, biomass burning, fossil fuel combus- up-to-now modest rise in temperature is that tion, and industrial processes) are represented some other aspect of atmospheric radiation has asonedimension,theeffectsontheatmospheric also been changing and in part canceling the environment (ultraviolet radiation, global greenhouse warming. Increases in sulfate and warming, photochemical smog, global oxida- smoke aerosols are one current suggestion for tion efficiency, acid rain, visibility, and corro- this factor. Most such interpretations would not sion) as the other dimension. In the following beexpected tobufferfuturegreenhousewarming sections, we discuss these threats to the health to the extent they may have in the past, so at best of the atmospheric environment and the ways in they might give us a little more breathing room. which human activity changes chemical climate. 126 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate Figure 8-8: Concentrations of Carbon Dioxide and Methane, 17501995 360 1800 Caifon Dioxide MIhIZl 1600 - 3420 - 8 1400- 320 -. *-1200- 30 1000 8280~~~~~~~~~~~ 280 ,,,,, . 0 ,o 8.0 260 600 1750 1600 1850 1900 1950 2000 1750 1800 1850 1900 1950 2000 Year Year 310 0.3 Nitrous Oxide CFC1 1 30 -j 0.2 8290 h 0.1 0 280~ 0 1750 1800 18&50 1900 1950 2000 .1750 1800 18590 1900 1950 2000 Year Year Note: Concentrations of carbon dlioxide and methane remained relatively constant up to the eighteenth century and have risen sharply since then, due to the activities of humans. Concentrations of nitrous oxide have increased since the midd-eighteenth century, especially in the last few decades. Chlorofluorocarbons were not present in the atmnosphere before the 1930s. Source: Houghton, Jenkins, and EphraUMnS 1990. Stratospheric ozone increasing inputs of chlorine and bromine spe- and ultraviolet radiation cies into the stratosphere. These species and their sources are as follows: chlorofluorocar- The solar radiation spectrum contains wave- bons are used as foaming agents, propellants, length regions energetic enough to cause chemi- and refrigerants; methyl chloroform and car- cal reactions in biological tissue. Much of this bon tetrachloride are used as industrial sol- damaging radiation is filtered out by the atmo- vents; halons, which contain both bromine and sphere. Of particular concern at present is the chlorine, are used as fire suppressants. Methyl observation that stratospheric ozone, which chloride and methyl bromide both have large serves to remove a part of the spectrum called natural sources, but anthropogenic emissions UV-B, is decreasing globally (see figure 8-10; may also be important: methyl chloride is emit- World Meteorological Organization 1992). The ted in great amounts from biomass burning; present rate of decrease is estimated to be on methyl bromide is used as a pesticide in house- the order of 5 percent a decade. In the polar holds and agriculture and may also have a sub- regions, ozone becomes greatly depleted dur- stantial source in biomass burning. Other trace ing the polar spring (the "ozone hole"), with gases that play a role in regulating stratospheric effects extending to the higher mid-latitudes ozone are nitrous oxide (N20), carbonyl sulfide (for example, Australia and northern Europe). (COS), methane (CH4), hydrogen (H), carbon Present consensus attributes this decrease to monoxide (CO), and the nonmethane hydrocar- 127 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 8-9: Atmospheric Life Support System Waste management* Srvice : ' ,tlr\ce> B :: i Xt t w 8 } 4i Source Land use change (deforestation) Anuture A=ra husbandrvsb _ ___ Biomass burnmng Fossil fuel bumnng Industrial acbstivitvws *ControlLing s Major Impact > Moderate C9 Some Note: Service hmctions provided by the atmosphere are shown in the columns, human activities that influence the viabiltv of these functions are shown in the rows. Symbols indicate the level of Impact that human activities have on the various functions. Source: Crutzen and Graedel 1986. bons (NMHC). All of these gases have natural and regional scales. Smog arises from photo- sources, but their anthropogenic emissions have chemical reactions between hydrocarbons, nitro- nowbecomesogreatthattheiratmosphericcycles gen oxides, and a number of oxygen-containing are significantly perturbed. This perturbation is molecules, most importantly the hydroxyl radi- clearly evident in the form of continuously in- cal, OH. These reactions are a normal part of the creasingatmosphericconcentrationsofN20,CH4' photochemical cycles in the atmosphere and act H2' and CO (see table 8-1). even in unpolluted environments to remove bio- genic hydrocarbons and nitrogen oxides emitted from soils by transforming them into less volatile Photochemical smog and acid rain and more water-soluble substances, which then are removed by precipitation or dry deposition. In contrast to the problems related to the deple- The environmental problems associated with tion of stratospheric ozone, photochemical smog photochemical smog and acid deposition are th- and acid rain are felt most seriously on the local result of an overloading of the self-cleaning ca- 128 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate Figure 8-10: Total Ozone Mapping Spectrometer pacity of the atmosphere. This overloading oc- Trends in Zonal Mean and Latitude, by Season, curs either as the result of fossil fuel buming and 1978-91 industrial activities in the industrial countries, or Trends Zn percentage per decade as a consequence of biomass burning in its many forms (savanna fires, deforestation bums, do- o -. _ mestic production of biomass energy, and agri- cultural buming of wastes) in the developing -2 - world. Recent studies suggest that biomass burn- ing worldwideproducesamountsof photochemi- -4 - . ,' / " \ cal smog comparable to those resulting from the ., J ,, . combustion of fossil fuels (Andreae 1991). Once -6 - __-8/ _, the capacity of the atmosphere to dissipate and remove the emissions is exceeded, products like -8- ozone and nitric acid are formed at concentra- /10 December - March tions that are toxic to biological organisms, corro- -10- /.- May -August sive to man-made and natural structures, and 12 --....... -.- September - November ecologically damaging through physicochemical 12 changes in the environment (for example, lake and soil acidification). Sulfuric acid is produced l l l l l from the sulfur contained in fossil fuels, espe- -60 -40 -20 0 20 40 60 cially coal, and also contributes to acid deposi- Latitude (negative numbers are south) tion. The environmental impact of photochemi- Note: Covers the period from November 1978 through cal smog and acid deposition has been discussed March 1991. in many publications (see, for example, Andreae 1991; Crutzen and Graedel 1986; NAPAP 1991). Source: World Meteorological Organization 1992. Table 8-1: Climatically and Chemically Active Atmospheric Trace Substances Tropospheric Annual concentration, growth Plant Biomass Fossil Function Substance 1989 (percent) Ocean andsod Agriculture burning fuel Industry Photochemical affected CFCs 5-500 ppt 3-10 - - - - - +++ - UV CH3CC3 135 ppt 3.7 - - - - - - UV cc, 107 ppt 1.2 - - - - - +++ - UV CH3C1 600 ppt ? + - - t+ - + - UV CH3Br 10-15 ppt ? ++ - + + - + - UV N O 308ppb 0.25 + ++ ++ + + + - UV,GE ch 1.69 ppm 0.8 + + + + + + - GE, UV, OX H 515 ppb 0.6 (+) (+) (+) ++ (+) (+) ++ UV cbs =500 ppt ? ++ + ? + + - ++ UV CO 50-150 ppb -0.3 + - - ++ + + ++ OX NO ppt-ppb t - + + + ++ + - OX, PL SO ppt-ppb t (++) (+) - + +++ + ++ GE, PL NNHC ppt-ppb t + + - + + + - OX, PL Aerosols ppt-ppb t ++ + + ++ ++ + ++ GE, PL CO2 352 ppm 0.5 - - ++ - +++ - + GE Tropospheric O3 ppbrange 1 - (+) (+) (++) (++) (+) +++ OX, PL, GE - None. Note: ppt, parts per trillion; ppb, parts per billion; ppm, parts per million; CFCs, chlorofluorocarbons; NMHC, non-methane hydrocarbons; UV ultraviolet radiation; GE, greenhouse effect; OX, oxidizing capacity of the atmosphere; PL, pollution. +, ++, and +++ indicate relative resource strength. Parentheses indicate that the source is indirect, that is, through release of a precursor photochemicaUy converted to the substance in question. 129 Defining and Measuring Sustainability: The Biogeophysical Foundations Global oxidation capacity Figure 8-11: Surface Volume Mixing Ratios of O3 for the Industrial and Preindustrial Periods Photochemnical reactions in the atmosphere are Ozone responsible for the removal of gaseous emissions from the biosphere and from human activities. As 60'N - an example of these processes, we will outline the O atmospheric oxidation of methane. In the first 30° - step, methane reacts with the hydroxyl radical to EQ - C> form a methyl radical, CH 3: CH4+OH => CH3+H2O. (8-1) 30° - 00 The subsequent reaction sequence depends on 600S- the ambient concentration of nitric oxide, NO. In an NO-rich environment, a reaction sequence occurs ..0 that has the net consequence of producing ozone: 180°W 1200 600 o0 600 120° 180 E CH-4 +402 =>CH2O + H2O + 2 03. (8-2) The OH consumed in the initial reaction is pro- Ozone duced again within this reaction sequence. In NO-poor environments, a different reaction chain O C is followed, which consumes the radicals OH and 0 HO2 according to: 30° - CH4 + OH + HO2 => CH20+2H 20. (8-3) EQ The further oxidation of formaldehyde (CH2O) to CO and finally to CO2 has similar consequences for 300 - ) O0 J C o the budget of ozone and the OH and HO2 radicals: 60°S - in NO-poor environments, the oxidants are used up; in NO-rich environments, more oxidants are 0 produced (Crutzen and Zimmermann 1991). 180 W 120 60° 0 60° 120° 180°E The effect of anthropogenic emissions of CO, hydrocarbons, and nitrogen oxides on the oxida- tion state of the atmosphere thus depends on the Note: The figures for the industrial period are for 1980. Ratios were calculated using our three-dimensional relative amounts of these pollutants. Here we tropospheric photochemical model. Surface is measured in need to look at effects at different scales as well. 1,000 hectoPascal; units are parts per villion by volume. Since nitrogen oxides have a much shorter life- time in the atmosphere than CO and CH4, the Source: Crutzen and Zimmerman 1991. relativeamountsof NO, CO, and CH4 will change with time (and therefore distance) followingemis- Figure 8-12: Percentage Change in the Calculated sion. Near the sources-that is, in polluted re- Daytime, Zonal Average, and Annual Average g,ons_th systemwillbeNO-ri, andConcentrations of Oxyl Radical since Preindustrial gionsthe system will be NO-rich, and ozone and Times other oxidants will be produced (see figure 8-11). Timec But in the remote atmosphere, the NO will have Oxyl Radical been removed, while much of the CO and CH4 is Pressure (hPa) still left, and therefore oxidants, particularly OH 100- - - radical, will be consumed (see figure 8-12). The 200- : . ./ 2 effect on the global atmosphere depends thus on 300- the details of the three-dimensional distribution 400 of the various trace gases and can only be simu- lated with fairly sophisticated models. :00 One of the most disquieting aspects of this 600- issue is that it constitutes a potentially unstable 700- ....... system due to a positive feedback. Since the oxi- 800- dation of CH4 and CO consumes OH, increasing 900 . / amounts of these trace gases in the atmosphere will consume even more OH. This would lead to 85 S 65 45 25 EQ 25 5 65 850N Latitude 130 Source: Crutzen and Zimmermann 1991. Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate a decrease in the concentration of OH, so that CO 8-2 shows, many other gases are also increasing in and CH4 would be oxidized less rapidly and the Earth's atmosphere. The change in their global would build up to a higher concentration. This mean concentration tells us about changes in the would further suppress OH, and so on. The final budgetof theirsourcesand sinks, while theirspatial result of this scenario would be an atmosphere distribution around the globe gives us information very different from the present one; in fact so on the location of their most important sources. different that no models at present could predict Seasonal variations in their concentrations can be its composition with any reliability. For further used further to identify sources and sinks, because details, see the paper by Crutzen and biogenic sources vary with the growth cycle and Zimmermann (1991). photochemical sinks change with the seasonal flux of solar radiation (Prinn and others 1992). As pointed out in the section on global oxidiz- Detecting change ing capacity, changing OH levels in the atmo- sphereareoneof themostfundamental perturba- The chemical composition of the atmosphere is tions of atmospheric chemistry. Yet, there are at one of the most sensitive and unambiguous indi- present no reliable methods to measure OH on cators of global change. Although there is an the time and space scales required to validate our active discussion about the details of the pro- current concepts and to detect a change in atmo- cesses defining the rate of increase of CO2 in the spheric abundance and distribution. At present, atmosphere, there can be no argument that the we have to rely on models (which use measure- atmospheric levels of CO2 are changing and that ments of the distribution of chemicals that react they are higher than at any time within the last with OH) to estimate the atmospheric concentra- 160,000 years (the time covered by the oldest tion of OH (see, for example, Crutzen and samplesof Antarctic ice obtained so far). As table Zimmermann 1991; Prinn and others 1992). The Table 8-2: Emissions of Carbon Dioxide and Other Trace Gases and Aerosols from Deforestation and Biomass Burning Emission ratio Biomass burning All sources Fraction Species (X 10,000) (tg per year) (tg per year) (percentage) Carbon dioxide Net from deforestation - 2,400 7,600 32 Gross from combustion 880 3,590 8,790 41 Carbon bumed - 4,080 - - Gases CO 80 290 1,100 26 CH, 8 29 380 8 NMHC 10 36 100 36 N2 4.7 39 150 26 N20 0.1 0.8 6 14 NO 2.3 9.6 40 24 NH, 1.3 5.5 44 12 SO 0.25 2.4 200 1.2 cOs 0.011 0.11 1.4 8 CH,CI 0.12 1.3 2.3 55 H2 27 16 36 45 0 30 430 1,100 39 Aerosols Organic carbon 25 90 1,500 6 Black carbon 17 59 180 33 Total 2.5 9 22 40 Note: Tg, teragram (1 trillion grams). 131 Defining and Measuring Sustainability: The Biogeophysical Foundations development of more direct measurement ap- Conclusions proaches is highly desirable but may still require substantial technological advances. In the preceding sections, we have outlined the major problems currently known to threaten the life support functions provided by the Earth's Monitoring change atmosphere: the greenhouse effect from gases and aerosols, the weakening ozone shield, the In order to apply measurements of atmospheric increase in local and regional pollution, and the properties to the assessment of sustainable devel- decreasing capability of the atmosphere to oxi- opment, we have to address two problems: (1) dize biogenic and anthropogenic emissions. The What level of atmospheric change is consistent symptoms that document the progressive, with sustainable development? (2) How must a nonsustainable deterioration of the atmospheric monitoring program be designed to detect this environment are readily detected by chemical critical level of change? analysis of the levels of trace gases in the atmo- It appears that, fundamentally, any secular sphere. The increase in temperature observed change of chemically or climatically important over the last century, and particularly in the last atmospheric properties is not sustainable. But a decade, is not as unambiguous as the chemical rate of change so slow that it would cause envi- record, but it points in the same direction. ronmental damage only in thousands of years is The presently observed changes in physical and clearly irrelevant, since the processes or human chemical climate appear still rather small: some activities causing this change cannot be forecast percentages to tens of percentages of the original over this time period. In order to be practicable, a level.This, however, may lull usinto a false sense of time frame over which this change is of concern security. One of the bewildering characteristics of and a rate of change that is acceptable to the climateandweather(bothchemicalandphysical)is present biota and human society will have to be its nonlinearity. For example, when air rises, it does selected. The selection of this time frame is as notslowlybecomemoreandmoreturbidandgradu- much a social and political issue as a scientific one. ally turn into a cloud; rather, the onset of cloud At present, several activities are under way to condensation is very sudden, and clear air turns assess and monitor global change. Under the intocloudoveradistanceofjusta fewmeters.A less auspices of the World Meteorological Organiza- commonplace, but much more dramatic, nonlinear tion, a network of monitoring stations has been phenomenon is the Antarctic ozone hole. When set up at remote oceanic sites to measure atmo- Molina and Rowland (1974) predicted a loss of spheric CO2, CH4, and a number of meteorologi- stratospheric ozone from the input of chlorofluoro- cal and radiation parameters. The GAGE (Global carbons, they projected a slow, incremental de- Atmospheric Gas Experiment) monitors the con- crease in ozone, on the order of a few percentages a centrations of N2O, CH4, and a series of haloge- decade. The appearance in the late 1970s of the hole nated compounds at a number of remote marine in the polarozone, where some50 percentof strato- sites. A large variety of regional networks moni- spheric ozone disappears within a few weeks, was tor local and regional pollution. In connection totally unpredicted in spite of the fact that the with theGlobal Change/ InternationalGeosphere- brightest minds in the community of atmospheric Biosphere Program, especially its International chemistshad been studying thelossof stratospheric Global Atmospheric Chemistry (IGAC) core ozone for more than a decade. The most important project,newnetworksfordetectingand monitoring lesson to be leamed here is that, in physical and the changing composition of the atmosphere are chemicalclimate,smallchangesinthedrivingforces being planned. Site selection is now moving away can sometimes lead to very large and unpredictable from the remote marine locations, which are highly effects. suitablefordetectinghighlyaveraged,global trends, Another reason why the linkagesbetween causes and toward more continental locations more suit- and effects are often so difficult to establish in the ableforidentifyingsourcesand processes. Attempts system of physical and chemical climate is the fact to monitor sustainable development will have to that the relevant phenomenacrossseveral scientific take into account these present and projected net- disciplines. Processes thatencompassmeteorology, works, and specific requirements will have to be photochemistry, oceanography, microbiology, and formnulated for future networks. other scientific disciplines are beyond the grasp of 132 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Clinate individual scientists and can only be investigated References by collaborative, cross-disciplinary efforts. There is only a short history of this type of collaborative Andreae, Meinrat 0. 1991. "Biomass Burning in approach in science; only a decade or two ago, fields the Tropics: Impact on Environmental Quality likebiogeochemistrywereconsideredirrelevantby and Global Climate." In K. Davis and M. S. most of the scientific community. And while the Bernstam, eds., Resources, Environment, and multidisciplinary approach is now accepted in the Population: Present Knowledge, Future Options, sciences dealing with the Earth system, actually pp. 268-91.NewYork:Oxford UniversityPress. implementing this new way of thinking will take Charlson, R. J., J. Langner, H. Rodhe, C. B. Leovy, considerable time. Progress is urgently needed in and S. G. Warren. 1991. "Perturbation of the the development of linked models that represent Northern Hemisphere Radiative Balance by the inherent linkages among the atmosphere, the Backscattering from Anthropogenic Sulfate oceans, and the biosphere. Aerosols." Tellus 43AB, pp. 152-63. In order to establish the scientific foundation for Crutzen, P. J., and T. E. Graedel. 1986. "The Role sustainable development, an even more daunting of Atmospheric Chemistry in Envirownent- task lies ahead: we will have to analyze the interac- Development Interactions." In W. C. Clark and tions between the natural sciences and the social R. E. Munn, eds., Sustainable Development of the and economic sciences. At present, these interac- Biosphere, pp. 213-50. Cambridge, England: tions are widely ignored by the practitioners on Cambridge University Press. either side. At best, they are given lip service (as is, Crutzen, P. J., and P. H. Zimmermann. 1991. "The admittedly, done here), but little actual research is ChangingPhotochemistryof the Troposphere." being undertaken. The ongoing efforts to put into Tellus 43AB, pp. 136-51. action the International Geosphere-Biosphere Pro- Daly, Herman E. 1991. "Sustainable Develop- gramandtheSustainableBiospherelnitiativeshould ment: From Concepts and Theory to Opera- include the investigation of these interactions as tional Principles." In K. Davis and M. S. central objectives. Forexample, research isurgently Bernstam, eds., Resources, Environment, and needed in the development of accounting tech- Population: Present Knowledge, Future Options, niques to represent the economic value of natural pp. 25-43. New York: Oxford University Press. resources. This will have to include evaluation of Houghton, J. T., G. J. Jenkins, and J. J. Ephraums. the economic benefits of "services" such as protec- 1990. Climate Change: TheIPCCAssessment. Cam- tion from ultraviolet radiation by stratospheric bridge, England: Cambridge University Press. ozone, fertilization of crops by atmospheric sulfate, Houghton, J. T., B. A. Callander, and S. K. Vamey. and removal of pollutantsby theoxidationpowerof 1992. Climate Change 1992. The Supplementary the troposphere. Conversely, the economic cost of Report to the IPCC Scientific Assessment. Cam- CO2 emissions and the resulting rise in temperature bridge, England: Cambridge University Press. will have to be qXuantified. , will haveto be quatified.lames, S. R. 1989. Hominid Use of Fire in the Although this chapter cannot provide detailed Lames S. M 1989. -Homine o Fie in the poliy reom2nendtios, sme undaentl diec- Lower and Middle Pleistocene: A Review of the policy recornunendations, sonme fundamental direc- Evidence.' Current Anthropology 30, pp. 1-26. hons are evident. We must curtail emissions of Evince." Curen Anhroplo , pp. 1-26. gases that lead toprogressiveand acceleratingdete- Lovis, A. B. 1991. "Energy, People, and Industri- rioration of the fundamental life support system of eds., Resources, Environment, and Population:a the Earth. This is clearly the case for the chlorofluo- Present Knowledge, Future Options, pp. 95-124. rocarbons, and the Montreal Protocols are an im- New York: Oxford University Press. portant step in the right direction. In the case of C02, current evidence is also strong enough that a pru- Molina, M. J., and F. S. Rowland. 1974. Strato- dent approach would clearly demand a reduction spheric Sink for Chlorofluoromethanes: Chlo- , - . . ~~~~~~rine Atom Catalyzed Destruction of Ozone." in emissions. Because of the multiple sources and Nature 249, pp. 810-14. sinks for many of the other gases and aerosols, an action to reduce one source will often lead to in- NAPAP (National Acid Precipitation Assessment creased emission of the same oranothergas from a Program). 1991. National Acid Precipitation different source. These complex interactions re- Assessment Program: 1990 Integrated Assessment quire that a detailed analysis of policy options and Report. Washington, D.C. consequencesbemadebeforemitigationeffortsare Nilsson, A. 1992. Greenhouse Earth. Chichester, undertaken. England: John Wiley. 133 Defining and Measuring Sustainability: The Biogeophysical Foundations Nobre, C. A., P. J. Sellers, and J. Shukla. 1991. impacts include changes in hydrology and the "Amazonian Deforestation and Regional Cli- transportofsedimentsandnutrients.Developing mate Change." Journal of Climate 4, pp. 957-88. the capability to predict such changes would Prinn, R., D. Cunnold, P. Simmonds, F. Alyea,R. providea robustbasisfordevelopingsustainable Boldi, A. Crawford, P. Fraser, D. Gutzler, D. alternatives. Hartley, R. Rosen, and R. Rasmussen. 1992. ThepresentchemicalcompositionoftheEarth's "Global Average Concentration and Trend for atmosphere would not be the same if living or- Hydroxyl RadicalsDeduced from ALE/GAGE ganisms did not exist, especially those that can Trichloroethane (Methyl Chloroform) Data for synthesize organic compounds through photo- 1978-1990." Journal of Geophysical Research 97, synthesis and those that mediate and maintain pp. 2445-61. the biogeochemical cycles. Therefore, the present Schiile, W. 1990. "Landscapes and Climate in chemical composition of the Earth's atmosphere Prehistory: Interactions of Wildlife, Man, and is a consequence of physicochemical activities Fire." In J. Goldammer, ed., Fire in the Tropical associated with living organisms throughout the Biota: Ecosystem Processes and Global Challenges, history of our planet. pp. 273-318. Berlin: Springer-Verlag. Although we can also say that humans have Vitousek, Peter M., Paul R. Ehrlich, A. H. Ehrlich, played a role since they first dominated fire, the and P. A. Matson. 1986. "Human Appropria- average chemical composition of the Earth's tion of the Products of Photosynthesis." atmosphere was largely controlled by natural BioScience 36, pp. 368-73. processes until the beginning of the nineteenth World Meteorological Organization. 1992. Scien- century. From there on, with the onset of the tificAssessmentof OzoneDepletion, 1991.Geneva, industrial revolution and the deforestation as- Switzerland. sociated with development, the concentration World Commission on Environment and Devel- of greenhouse gases in the atmosphere has in- opment (The Brundtland Commission). 1987. creased considerably. This is especially true for opment the last few decades. The emission rates of University Press. greenhouse gases has reached a point at which the natural regulatory systems can no longer prevent the increase of these gases in the atmo- sphere. Comments If we want to relate sustainability to climate, we should be aware of the global dimension, which was thoroughly reviewed in Andreae and Eneas Salati and Reynaldo Luiz Victoria Dickinson's work, and the regional dimension, which should be analyzed considering: The chapter by Andreae and Dickinson is an * Naturalecosystems,eitherundisturbedorsub- excellent review of state-of-the-art knowledge of jected to very little human intervention the role that the atmosphere's chemical composi- tion plays in the global climate. It clearly shows Agricultural ecosystems the importance of the increase in the concentra- * Human activity,includinglarge,medium,and tionof greenhousegases, which could jeopardize small cities. both the sustainability of the Earth's productive The natural ecosystems reached equilibrium system, and the interactions between the bio- throughcontinuousinteractionsbetween thebio- sphere and atmosphere. Our comments are in- sphere and the atmosphere. In some areas, the tended as contributions to, not critiques of, this Amazon forest for instance, the present dynamic excellent report. They are also relevant to discus- equilibrium of the atmosphere depends on the sions of the sustainability of the Earth's climate forest; in other words, the present climate de- system. Being Brazilians, and also because tropi- pends on the forest (Salati 1985). Therefore, any cal forestsareat thecenterof discussionsof global change in the parameters that define the climate, changes in climate, we offer examples mostly eitherproducedbydeforestationorglobalchanges from the Brazilian Amazon. Recent deforestation in climate, may alter the present ecological equi- inthatregionisnowalteringtheecosystemstruc- librium of the region and consequently its ture of certain subbasins, where potential regional sustainability. 134 Sustainability and the Changing Atmosphere: Assessing Changes in Chemical and Physical Climate The following climate parameters may bear species or other aquatic organisms may also be relation to sustainability and should therefore be used as indicators of loss of sustainability of measured: natural ecosystems. (a) Temperature, precipitation, solar radiation, (e) Change in the flux of water vapor. Another wind, cloudiness, and so forth. It is important to fingerprint for change in climate is the change in analyze not only their average values but also thefluxofwatervapormeasuredbyrawinsondes. extremes events. In Manaus, for instance, mea- For agroecosystems, the relationship between surements of temperature for the last three possible changes in climate and sustainability is decades do not show any trend for variation in more complex. The sustainability in this case theaverage, eithermaximumor minimum, tem- might be maintained at a certain level with the perature (Ribeiro 1991), but they do show an use of technology. For instance, in the cerrados of increasing trend for theabsolutemaximum tem- central Brazil, a change in climate leading to a perature (Villa Nova and Salati, unpublished decrease in precipitation may be overcome with data). the use of irrigation techniques, therefore main- (b) Parameters that integrate several factors, like taining agricultural productivity, as long as care river discharges and balance of water in watersheds. is taken to maintain the underground water res- Another important variable is the chemical char- ervoir of the region. It is evident that the use of acterization of surface waters, since it reflects the technology to overcome possible changes in cli- interrelations between the biogeochemical and mate has limits, which are generally dictated by hydrological cycles. its economic or social sustainability. Thus, sus- (c) Isotopic composition of rainfall. The 'IO and D tainable agricultural production that is devel- concentration of rainfall water is directly related oped in a region as a function of its previous to the history of the water vapor from which it climatemaybecomeunsustainablewithachange originates. It depends on previous occurrence of in climate. precipitation, temperature of evaporation and For urban ecosystems, it is also difficult to use condensation, type of cloud, and so forth. Based a simple concept of sustainability. In this case, on the isotopic composition of rainfall in the technological development and improvement of Amazon region, Salati and others (1979) esti- economic conditions may also permit different mated thatroughly50 percentof the precipitation levels of sustainability. Thus, the sustainability of is water recycled through evapotranspiration. a way of life is dependent on both climatic and Morerecently,GatandMatsui(1991)andVictoria economic conditions; human comfort may be and others (1991) showed that there is as yet no improved by artificially controlling the climate clear distinction between the relative contribu- (temperature, humidity, and light). In general, tion of evaporational and transpirational pro- the standard of living in cities of industrial coun- cesses to the total evapotranspirational flux. There tries is well above that of developing countries. If is, however, a strong indication that evaporation the later is to be improved, new projects of archi- may contribute up to 40 percent of total evapo- tecture and urban planning will be necessary, transpiration during the dry season. The loci of which will certainly require an increase in energy evaporation, whether from floodplain, canopy, usage and consequently have a positive feedback or open bodies of water (lakes and rivers), cannot on changes in climate. yet be defined with the available data. Whether It is, and will be, difficult to find parameters to the isotopic features found for the Amazon are measure the sustainability of urban planning; common to all other tropical areas is not yet several mightbe useful: percapita income, useof known, merely for lack of data, but we are certain potable water, energy consumption per capita, that, in the event of global or regional changes, soil erosion, crime level, education level, and so the isotopic concentration of rainwater will be forth. Change in climate may have a relation to one of the first to change. such indexes. (d) Biological indicators. In addition to the cli- The use of climate resources to increase the matological parameters, several biological indi- production of biomass in the tropics represents cators may also be used. For instance, some or- an excellent opportunity to transfer carbon from ganisms living on the canopy of trees in humid the atmosphere to the biosphere. In this region, regions may have their population altered by the net primary production rate is among the changes in the hydrological regime. Some fish highest in the planet for forests, aquatic plants, 135 Defining and Measuring Sustainability: The Biogeophysical Foundations and crop stands. The limiting factor for biomass - . 1993. "Water Supply Systems Utilizing production is generally associated with the fertil- the Edaphic-Phytodepuration Technique." In ity of both terrestrial and aquatic systems. Constructed Wetlandsfor WaterQualityImprove- One way to solve the problem, and which is ment:AnInternational Symposium. Proceedings already showing positive results (Salati 1987), is of a conference held in Pensacola, FL, October the recycling of urban and industrial sewage into 21-24. Lewis Publishers. production systems. The DHS system Ribeiro,A.1991."Analisesdasvariac6esclimaticas (Manfrinato, Salati Filho, and Salati 1990, 1993; observadas na regiao de Manaus (AM)." Salati Filho 1991) is being used to produce biom- Master's thesis, University of Sao Paulo, ass from rice and aquatic plants, decreasing the Piracicaba. level of pollution of waters contaminated by ur- Salati, Eneas. 1985. "The Climatology and Hy- ban sewage at the same time. To extend the con- drolo of the Amazonia." In G. Prance and cept to other systems, including forestry, is rela- . . tively simple. The technology to produce char- T. Lovejoy, eds., Amazonia: Key Environments, coal or biogas from biomass is readily available. pp. 18-48. Oxford, England: Pergamon Press. This is an important field for scientific and tech- .1987. "Edaphic-Phytodepuration: A New nological development that could bring a better Approach to Wastewater Treatment." In K. R. and longer sustainability to tropical regions, in- Reddy and W. H. Smith, eds., Aquatic Plantsfor creasing the standard of living (social WaterTreatmentandResourceRecovery.Orlando, sustainability) at the same time. FL: Magnolia Publishing. Salati, Eneas, A. Dal'Ollio,J. R. Gat, and E. Matsui. 1979. "Recycling of Water in the Amazon Ba- References sin: An Isotope Study." Water Resources Re- search 15, pp. 1250-58. Gat, J. R., and E. Matsui. 1991. "Atmospheric Salati Filho, Eneas. 1991. "HDS System: Utiliza- Water Balance in the Amazon Basin: An Isoto- tion in theTreatment of Secondary WaterFlow pic Evapo-Transpiration Model."lournalofGeo- in the Cellulose and Paper Industries." In Con- physical Research 96:D7, pp. 13, 179-88. structed Wetlands for Water Quality Improve- Manfrinato, E. S., E. Salati Filho, and E. Salati. ment:AnlnternationalSymposium. Proceedings 1990. "Water Supply Systems Utilizing the of a conference held in Pensacola, FL, October Edaphic-PhytodepurationTechnique." In P. F. 21-24. University of West Florida. Cooper and B. Findlater, eds., Constructed Wet- Victoria, R. L., L. A. Martinelli, J. E. Richey, and lands in Water Pollution Control. Oxford, En- J. Mortatti. 1991. "Mechanisms of Water Recy- gland: Pergamon Press. cling in the Amazon: Isotopic Insights." Ambio 20:8, pp. 384-87. 136 Sustainability at Landscape and Regional Scales R. V. O'Neill, C. T. Hunsaker, D. Jones, J. M. Klopatek, V. H. Dale, M. G. Turner, R. H. Gardner, and R. Graham Many of the important threats to sustainability of the biosphere involve changes in patterns of land use as a result ofan expandinghuman population. Loss of habitat, decreased biodiversity, tropicaldeforestation, and desertification all involve changes in the extent and patterning of natural systems on the landscape. Reversing current trends requires a new emphasis on the interaction between spatial patterning and ecologicalfunction. This chapter recommends a numberof measures of landscapepattern, basedprimarily on remote sensing, and suggests critical research areas needed to address sustainability of heterogeneous spatial systems. For decades, ecology has focused on paradigms cal systems is the isolation of natural systems on that assume spatial homogeneity. Heterogeneity the landscape, which affects the spread of and and pattern were invariably observed in field ability to recover from disturbances (Turner and studies, but results were nevertheless averaged others 1989,1991). In fact, mostof thephenomena across space. The dominant paradigms held that we label as disturbance-from the eruption of the critical regularities were contained in the MountSt.Helens,toinvasionsoftheGypsyMoth, means. Heterogeneity was the "nasty bit" left to damage from hurricanes in the Everglades, to over and relegated to unexplained variance. fires in Yellowstone-are spatial disruptions at Today,aswediscussthesustainablebiosphere, the landscape and regional scale. it is impossible to ignore spatial pattern: 5.5 bil- Tropical deforestation provides a strong moti- lion humans are carving the terrestrial biosphere vation for considering landscape and regional into a mosaic of land uses that threaten the subsi- scales. The Brazilian Amazon has the largest tract diesprovidedbyintactecosystemsonintactland- of moist tropical rain forest in the world scapes.' For purposes of this chapter, we define (Molofosky, Hall, and Myers 1986). Remotely the landscape scale as involving 10,000 to 10 sensed imageryrevealsthatbyl987,41,000square million hectares, with each spatial unit or pixel kilometershadbeencleared inRondoniafor coo- involving 1 to perhaps 100 hectares. nization (Booth 1989; Malingreau and Tucker Landscape and regional changes are impli- 1988; Stone, Brown, and Woodwell 1989). At cur- cated in all of the problems identified in the rent rates, the forest in Rondonia will be gone by Sustainable Biosphere Initiative (Lubchenco and 2000. The ecological effects of changing tropical others 1991). The primary terrestrial contributor forest landscapes include decreased biodiversity to global change in atmospheric carbon is the (Wilson 1988), disruption of hydrological re- disruption of tropical forest landscapes (Dale gimes (Shukla, Nobre, and Sellers 1990), degra- 1990a, forthcoming; Dale, Houghton, and Hall dation of soil (Hecht 1981), and changes in 1991). The primary risk to biodiversity comes greenhouse gases that could induce changes in from landscape fragmentation and loss (Wilson climate (Houghton and others 1983; Post and 1988). An important threat to sustainable ecologi- others 1990). Defining and Measuring Sustainability: The Biogeophysical Foundations Problems such as deforestation lead us to ask cover, for example, reflect loss of habitat and can how sustainability is related to the mosaic of be directly translated into increased risk to wild- ecological systems at landscape and regional life species. O'Neill, Gardner, and Turner (1992) scales. Basically, these large scales are relevant to present a list of rare and endangered species in sustainability insofar as the arrangement of sys- southeastern United States that are uniquely de- tems in space influences (1) the intrinsic stability pendent on a particular type of habitat. In specific of ecosystems and (2) the ability of ecosystems to regions, such as the sandy scrub of central Florida, provide a variety of human subsidies, including the list permits an immediate interpretation of food, fiber, water, and recreation. Factors influ- habitat loss. encing intrinsic stability include increase in the Attention can also be focused on specific tran- frequency or extent of disturbances, significant sitions from one land cover to another. These loss of species, and decreased ability to recover include loss of rare land covers (such as wet- fromdisturbances. Factors influencing sustained lands), loss of windbreaks, or formation of con- services to humans include erosion, water qual- tiguous agriculture adjacent to a stream or lake. ity, and loss of wildlife. We would also be concerned with connectivity Given the importance of changes in land use, it between landscapes as reflected in corridors be- is providential that three technological break- tween patches of natural habitat (Forman and throughs have made the study of large-scale spa- Godron 1986; Harris and Scheck 1991). tial changes practical. First, advances in satellite imagery (Dale 1990b; Tucker and others 1986) Ecotones provide large-scale data sets. Second, develop- The length of ecotones or edges can also be re- ments in geographic information systems facili- lated to suitability of the landscape for wildlife tate manipulation of data (Dearstone and others, (Gardner and others 1991). In general, loss of forthcoming).Third,rapiddevelopmentsinland- edge can be related to decreased biodiversity scape ecology are providing the theory (Graham since ecotones normally have higher species di- and others 1991; Hunsaker and others 1990) versity. Ecotones often form unique habitats that and models (Dale and others 1993; Southworth, are associated with rare and endangered species Dale, and O'Neill 1991a, 1991b) for regional (O'Neill, Gardner, Turner, and Romme 1992). applications. It is also relevant to examine the relationship of The purpose of this chapter is to suggest how edges to sizes of patches. For example, cowbirds landscape ecology can be applied to monitoring on forest edges are nest predators on warblers. large geographic areas for sustainability. We of- Forest patches have to be sufficiently large so that fera variety of elementary measures of landscape nest sites are far enough away from edges that pattern that are both practical (based on current cowbirds cannot find them. If patches get too technology) and interpretable (based on current small, warbler populations start to decline. ecological understanding). We also recommend areas of research that must be addressed if we are Patch configurations to assure the sustainability o' heterogeneous spa- Much of what we understand about the influence tial systems. of landscape pattern on ecological processes is based on the configuration of patches. For ex- ample, the frequency distribution of patches by Measuring pattern at landscape scales size can be important because some species re- quire a minimal size of patch. The largest patch Measurementsatthelandscapescalemustsatisfy may also serve as a reservoir that maintains a four criteria. They must (1) be practical, (2) cap- population on the landscape. Fragmentation of a ture significant aspectsof pattern, (3) be sensitive landscape into many isolated patches has been to changes through time or across systems, and shown to reduce biodiversity (Bierregard 1990; (4) be interpretable in terms of ecological pro- Lovejoy,Bierregard,andRylandsl986;Saunders, cesses (O'Neill, Hunsaker, and Levine 1992). Hobbs, and Margules 1991). It would also be possible to consider the fre- Habitat coverage quency distribution of distancesbetween patches. The simplest landscape measure is the number of These nearest-neighbor distances are related to pixels (the smallest spatial units on the map) that the difficulty experienced by wildlife moving change land use. Changes in natural vegetative across the landscape. Some spatial arrangements 138 Sustainability at Landscape and Regional Scales of patches are particularly vulnerable to frag- Percolation theory (Gardner and others 1987) mentation. Isolated habitat is often configured in providesa framework for relating pattern of land- a longitudinal pattern, like a string of pearls. scapeto theabilityofanorganismtomoveacross Examples include alpine tundra along ridge tops the landscape. Rates of diffusion can be calcu- of the Rockies, dune vegetation along beaches, lated and interpreted in terms of wildlife use or and granite outcrops. Removing a single patch spread of fires. The percolation backbone defines from this configuration may split the entire habi- the fewest steps needed to traverse the landscape. tat in two if the gap exceeds the ability of the Percolation theory also defines critical values populations to disperse. of habitat coverage (Gardner and O'Neill 1991). Attentiondoesnotneedtobelimitedtopatches On a random square lattice, the critical value is of natural vegetation. It is also possible to inter- 59.28 percent. If the percentage of cover for habi- pret changes in the extent and pattern of clearing tat is less than this value, the landscape becomes relative to risk of erosion. Erosion is significantly dissected into isolated patches. The resource uti- increased on long slopes of uninterrupted, lization scale measures the scale at which an unvegetated surfaces. On this basis, clearings organism must operate to use the resources on a might be weighted by slope, size, and proximity landscape (O'Neill and others 1988b). Similarly, to other clearings. epidemiology theory can be combined with per- colation theory to calculate the probability that a Weighting change in habitat disturbance or pest will spread or become endemic Interpretations of change in landscape can be (O'Neill, Gardner, Tumer, and Romme 1992). enhanced by weighting the transition of indi- Several other landscape indexes have been vidual pixels. Forexample, a transition to agricul- developed in recent studies. Dominance is an ture might be weighted by distance to water to information theoretic index that indicates the ex- evaluate the impact on water quality. Greater tent to which the landscape is dominated by a weight might be given to a transition that frag- single land use (O'Neill and others 1988a). Em- ments a large patch. Similarly, a transition could pirical studies confirm that the fractional dimen- be weighted by the probability of forming a bar- sion of patches indicates the extent to which hu- riertothemovementofanimalsorthebreakupof mans have reshaped the structure of the land- a corridor. It would be important to distinguish scape (Krummel and others 1987; O'Neill and between 100 pixels scattered randomly and 100 others 1988a). Humans prefer simple shapes; na- pixels in a line, forming a new barrier to the ture generates complex configurations. Conta- movement of animals. gion (Haban Li and Reynolds, forthcoming; Individual transitions can also be weighted by O'Neill and others 1988a) expresses the probabil- characteristics of the entire landscape. In an area ity that land uses are more clumped than the with very little wetland (or riparian or critical random expectation. habitat), loss of a pixel is more important than in One method to assess a land cover would be to a region where the habitat is abundant. ask, "How well is the land being used, compared Many possibilities emerge through the use of with its potential?" This suggests an index that ancillary data, such as agricultural censuses, hu- compares each pixel with an overlying map of man population numbers, or forest surveys. As potential vegetative cover and calculates the per- an example, loss of a pixel of forest with recre- centage difference. Another index might com- ational value would be weighted more heavily in pare present land use with suitability for various regions with large urban populations. human applications, such as agriculture. Similarly, it should be possible to devise a measure of landscape suitability for animals. Dealing with changes in all types of land cover Consider a square "window" of pixels, the size of can be a daunting task, and some consideration an organism's home range. Within the window, mustbegiventoreducingtheamountofdata. For we could consider a variety of habitat require- example,the"U"index(O'Neillandothersl988a) ments, such as mixture of vegetation and avail- is a simple measure of overall impact of humans. ability of water. We could then place the window The index is calculated as the number of pixels in over a corner of the map and calculate a correla- types of natural land cover divided by the pixels tionbetween the existinglandscapeand thehabi- in human land uses, such as agriculture and tat requirements. The window could then be urbanization. moved systematically over the map to obtain an 139 Defining and Measuring Sustainability: The Biogeophysical Foundations overall score that indicates the suitability of the Landscape pattern and water quality landscape for this organism. We could design a Itisalsopossibletoassesstheriskofwaterquality suite of windows for insects, birds, mammals, degradation due to changes in terrestrial land- and so forth. This approach provides a simple, scapes (Hunsaker and others, forthcoming). easily interpreted method to compare two land- Across a region, increase in agricultural and ur- scapes or to evaluate the impact on animals of a ban land uses or decreases in natural vegetation change in landscape pattern. indicate a risk of future problems with water Indicators of economic activity quality. The changes in basic cover could be We must be careful not to miss obvious measures weighted by distance to water, soil type, and of potential human impact. Oneof theseis simply slope calculated from digital elevation models. the number of miles of roads. Vehicles that use Essentially, the same data set can be used with the roads kill wildlife. Road surfacesalterhydrologic Universal Soil Loss Equation to estimate erosion. pathways and, when they intersect streams, af- A second type of indicator might focus on the fect the quality of water. Roads also provide risks of undesirable hydrologic events. For ex- fpattheays quality fc ater.he Roadispalsof provie ample, a flood indicator could include vegetative pathays hatccelratetheisprsa of pe's cover, wetlands that modify peak flows, and It is well established in economic theory that . . the number of miles of new roads (and their Surficial geology. quality) indicates future development and eco- Riparian zones are important buffers for nomic activity (Jones and O'Neill n.d.). In sim- maintaining the quality of water in streams. plest terms, products must be transported to Changes in the width of buffers, weighted by market. As thedistance,to market increases trans- slope, are an important indicator. The actual ,raincssan. t mdex might be average width or miles of ripar- proftreat cos eat toerofis. At somistance, ian zone that are narrower than desirable. The profits reach zero, and there is no motivation for CadinTmeMngentGde(tro farmin&. logn ormnn.Jne n 'el Canadian Timber Management Guide (Ontario fArmin, login, or miin (.nsad_el Ministry of Natural Resources 1988) sets stan- 1993a, 1993b, and forthcoming). Distance to mar- dardsfrt of burzes thatang ket can also determine the motivation of farmers dards for the wsdth of buffer zones that range to employ conservation measures in their farm- on slopes aT stadar ould be ing activities (Jones and O'Neill n.d.). on slopes above 45a. This standard could be It would also be of interest to evaluate land- pplied by counting pixels that encroach on the scapes by distance to the nearest urban pixel. recommended buffers. One could calculate the average distance for a scene or just draw an area of urban influence. Given global trends toward urban sprawl, this Research needs n landscape could be taken as a measure of risk for develop- and regional studies ment or overuse. The research agenda for landscape studies can be Hierarchical scales of landscape pattern stated simply: How do ecological processes inter- Empirical studies (O'Neill and others 1991 a, 1991b) act with the environment to create patterns, and have confirmed the prediciion from hierarchy how do the patterns influence ecological func- theory (O'Neill and others 1986; O'Neill 1988, tion? Basically, there is a need to develop mea- 1989; O'Neill, Johnson, and King 1989) that land- sures of spatial pattern and to correlate pattern scapes should show pattern at distinct scales. with ecological processes in field studies. With- This approach uses statistical analysis of transect out attempting to beall-inclusive, several areas of data to identify multiple scales of pattern (Turner research seem particularly relevant. andothersl991).Thenumberofscalesisaunique To what extent does spatial pattern affect the index that focuses on the ecological processes that ability of systems to recover from disturbance? We influence landscape pattern. know that northern hardwoods may take sixty to Disruptions of this scaled structure mean that eightyyears to replacebiomassand nutrientslost in ecological processesdeterminingaparticularscale harvesting (Likens and others 1978). How would have been disrupted. Although the detailed rela- thisrecoverytimechangeif distancestoseed sources tionship between processes and scaled patterns is were increased or if erosion set in? still a matter of research, we have hierarchical To what extent is spatial pattern critical to the neutral models that are suitable for testing hy- sustainability of plant and animal communities? notheutr iodel that areld (O'uitable forstesang h For example, pollinators and nectar feeders fly, Levine 1992). which permits them to cover large areas and explicitly integrate across spatial heterogeneity. 140 Sustainability at Landscape and Regional Scales They require something to be blooming some- References where throughout the season. If the spatial scale of the heterogeneity is increased to exceed their Bierregard, R. 0. 1990. "Avian Communities in the ability to integrate, the result could be important Understoryof Amazonian ForestFragments." In changes in the plant community. A. Keast, ed., Biogeography and Ecology of Forest We need to identify ecological systems that are Bird Communities, pp. 333-43. The Hague, the particularly sensitive to spatial disturbances. The Netherlands: SPB Academic Publishing. proverbial erosion effects of tire tracks in the Arctic tundra are well known, but arid lands may be Booth, W. 1989. "Monitoring the Fate of Forests equally sensitive. Even the casual observer can see from Space." Science 243, pp. 1428-29. howsmallalterationsinnatural land formsresultin Dale, V. H. 1990X. "Strategy for Monitoring the major changes in the vegetation of arid lands. Effects of Land Use Change on Atmospheric Thepotentialsensitivityof arid landsalsoalerts Carbon Dioxide Concentrations." In Proceed- us to the need to identify critical thresholds in ings of Global Natural Resources Monitoring and landscape pattern. Percolation theory indicates Assessments. Vol. 1, pp.422-31. Bethesda, Md.: that small changes in land cover can critically American Society for Photogrammetry and alter connectivity between landscapes. But we Remote Sensirg. also know from tragic experience in the American D plains and the African Sahel that other critical Dale, V. H., ed.m1990b. "Report of a Workshop on thresholds exist. Beyond these thresholds, cas- Using Remote Sensing to Estimate Land Use cading effects or positive feedbacks cause bifur- Change."ORNL/TM-11502.0akRidge,Tenn. cations that move the system into new modes of . Forthcoming. Effects of Land-use Changeon operation (Schlesinger and others 1990). Atmospheric Carbon Dioxide Concentrations: To address sustainability at large scales, we Southeast Asia as a Case Study. New York: need to integrate socioeconomic theory with ecol- Springer-Verlag. ogy. Landscapes change because of human deci- Dale, V. H., R. A. Houghton, and C. A. S. Hall. sions. To predict such change requires an under- 1991. "Estimating the Effects of Land Use standing of the economic forces that determine Chan on Global Atmospheric Carbon Diox- land tenure and land use (Jones and O'Neill 1993a, g. 1993b,and forthcoming).Toreversecurrent trends ide Concentrations." Canadian 8ournal of Forest requires that ecological consequences be trans- Research 21, pp. 87-90. lated into feedbacks, into the real cost of supply- Dale, V. H., F. Southworth, R. V. O'Neill, and ing environmental subsidies. The development A. Rosen. 1993. "Simulating Spatial Pattemsand of this interdisciplinary theory is critical, yet al- Socioeconomic and Ecologic Effects of Land-use most no support is available for this research. Change in Rondonia, Brazil." In R. H. Gardner, ed., SomeMathematical Questions in Biology. Provi- dence, R.I.: American Mathematical Society. Conclusions Dearstone, K. C., V. H. Dale, R. H. Frohn, F. Southworth, R. V. O'Neill. Forthcoming. "Link- The analysis presented in this chapter seems to inSptaDtaoaMdeofClizin justify three conclusions. First, an emphasis on ing Spatial Data to a Model of Colonization landscape ecology is required since most threats and Deforestation of Rondonia, Brazil." to a sustainable biosphere involve changes in land use at large scales. Second, the combination Forman, R. T. T., and M. Godron. 1986. Landscape of remote sensing, geographic information sys- Ecology. New York: Wiley and Sons. tems, and landscape theory permits us to define a Gardner, R. H., B. T. Milne, M. G. Turner, and R. V. significant number of new metrics that are both O'Neill. 1987. "Neutral Models for the Analysis practical and interpretable. Third, continued of Broad-scale Landscape Pattern." Landscape progress involves difficult new areas of research. Ecology 1, pp. 19-28. New research, particularly interdisciplinary so- Gardner, R. H., and R. V. O'Neill. 1991. "Pattern, cioeconomic-ecological studies, requires new Process, and Predictability: The Use of Neutral sources of funding that currently do not exist. ModelsforLndscapeAnalysis."InM.G.Tumer and R. H. Gardner, eds., Quantitative Methods in Note Landscape Ecology, pp. 289-307. New York: 1. A billion is 1,000 million. Springer-Verlag. 141 Defining and Measuring Sustainability: The Biogeophysical Foundations Gardner, R. H., M. G. Turner, R. V. O'Neill, and . 1993a. "Human-environmental Influences S. Lavorel. 1991. "Simulation of the and Interactions in Shifting Agriculture." In Scale-dependent Effects of Landscape Bound- T. R. Lakshmanan and P. Nijkamp, eds., Essays aries on Species Persistence and Dispersal." In on Space and Time, pp. 297-307. New York: M. M. Holland, P. G. Risser, and R. J. Naiman, Springer-Verlag. eds., Ecotones: The Role of Landscape Boundaries _ . 1993b. "Human-environmentalInfluences in the Managernent and Restoration of Changing and Interactions in Shifting Agriculture When Environments, pp. 76-89. New York: Chapman Farmers Form Expectations Rationally." Envi- and Hall. ronment and Planning 25, pp. 121-36. Graham, R. L., C. T. Hunsaker, R. V. O'Neill, and --. Forthcoming. "Land Use with Endog- B. L. Jackson. 1991. "Ecological Risk Assess- enous Environmental Degradation and Con- ment at the Regional Scale." Ecological Applica- servation." Resources and Energy. tions 1, pp. 196-206. _ . n.d. "Rural Land Use and Deforestation Haban Li, and J. F. Reynolds. Forthcoming. "A with Costly Investment in Transportation In- New Contagion Index to Quantify Spatial Pat- frastructure." M ss. terns of Landscapes." Landscape Ecology. Krummel, J. R., R. H. Gardner, G. Sugihara, R. V. Harris, L. D., and J. Scheck. 1991. "From Implica- O'Neill, and P. R. Coleman. 1987. "Landscape tion to Applications: The Dispersal Corridor Patterns in a Disturbed Environment." Oikos Principle Applied to the Conservation of Bio- 48, pp. 321-24. logical Diversity." In D. A. Saunders and R. J. Likens, G. E., F. H. Bormann, R. S. Pierce, and Hobbs, eds., Nature Conservation 2: The Role of W. A. Reiners. 1978. "Recovery of a Deforested Corridors, pp. 189-220. London: Surrey Beatty Ecosystem." Science 199, pp. 492-96. and Sons. Hecht S. B. 1981. "Deforestation in the Amazon Lovejoy, T. E., R. 0. Bierregard, and A. B. Rylands. H ts. B. 1981. Deforestan 1986. "Edge and Other Effects of Isolation on Basin: Magnitude, Dynamnics, and Soil Resource Amazon goetFamns" nM .Sue Eff-cs. SudesinThird World Societies 13, pp. AmznForest Fragments." In M. E. Soule, Effects." Studies in ed., Conservation Biology: The Science of Scarcity and Diversity, pp. 275-85. Sunderland, Mass.: Houghton, R. A., J. E. Hobbie, J. M. Melillo, B. Sinauer Associates. Moore, B. J. Peterson, G. R. Shavers, and G. M. Woodwell. 1983. "Changesin theCarbon Con- Lubchenco, JM, A. M. Olson, L. B. Brubaker, S. R. tent of Terrestrial Biota and Soilsbetween 1860 Levin, J. A. MacMahon, P. A. Matson, J. M. and 1980: Net Release of Carbon Dioxide to the Melillo, H. A. Mooney, C. H. Peterson, H. R. Atmosphere." Ecological Monographs 53, pp. Pulliam, L. A. Real, P. J. Regal, and P. G. Risser. 235-62. 1991. "The Sustainable Biosphere Initiative: Hunsaker, C. T., R. L. Graham, G. W. Suter, R. An Ecological Research Agenda." Ecology 72, V. O'Neill, L. W. Barnthouse, and R. H. pp. 371-412. Gardner. 1990. "Assessing Ecological Risk Malingreau, J. P., and C. J. Tucker. 1988. on a Regional Scale." Environmental Manage- "Large-scale Deforestation in the Southeast- ment 14, pp. 325-32. ern Amazon Basin of Brazil." Ambio 17:1, pp. Hunsaker, C. T., D. A. Levine, S. P. Timmins, B. 49_55 L. Jackson, and R. V. O'Neill. 1992. "Land- scJapCac teon,andRizatiO'Neill. for9Asseing - Molofosky, J., C. A. S. Hall, and N. Myers. 1986. scape Characterization for Assessing Re- "A Coprso of TrpclFrsuvy. gional Water Quality." In D. H. McKenzie, "A Comparison of Tropical Forest Surveys. D. E. Hyatt, and V. J. McDonald, eds., Eco- DnEtnB D. W logical Indicators, pp. 997-1006. New York: ington, .C. Elsevier. Ontario Ministry of Natural Resources. 1988. "Timber Management Guidelines for the Pro- Jones, D. W., and R. V. O'Neill. 1992. "Endog- tection of Fish Habitat." Toronto, Canada. enous Environmental Degradation and Land Conservation: Agricultural Land Use in a O'Neill, R. V. 1988. "Hierarchy Theory and Glo- Large Region." Ecological Economics 6, pp. bal Change." In T. Rosswall, R. G. 79-101. Woodmansee,aid P. G.Risser,eds.,Spatialand 142 Sustainability at Landscape and Regional Scales Temporal Variability in Biospheric and Geospheric Saunders, D. A., R. J. Hobbs, and C. R. Margules. Processes, pp. 29-45. New York: John Wiley 1991. "Biological Consequences of Ecosystem and Sons. Fragmentation: A Review." Conservation Biol- - . 1989. "Perspectives in Hierarchy and ogy 5:1, pp. 18-32. Scale." In J. Roughgarden, R. M. May, and Schlesinger, W. H., J. F. Reynolds, G. L. Simon A. Levin, eds., Perspectives in Ecological Cunningham, L. F. Huenneke, W. M. Jarrell, Theory, pp. 140-56. Princeton, N.J.: Princeton R. A. Virginia, and W. G. Whitford. 1990. "Bio- University Press. logical Feedbacks in Global Desertification." O'Neill, R. V., D. L. DeAngelis, J. B. Waide, Science 247, pp. 1043-48. and T. F. H. Allen. 1986. A Hierarchical Concept Shukla, J., C. Nobre, and P. Sellers. 1990. "Ama- of Ecosystems. Princeton, N.J.: Princeton Uni- zon Deforestation and Climate Change." Sci- versity Press. ence 247, pp. 1322-25. O'Neill, R. V., R. H. Gardner, B. T. Milne, M. G. Southworth, F., V. H. Dale, and R. V. O'Neill. Turner, and B. Jackson. 1991a. "Heterogeneity 1991a. "Contrasting Patterns of Land Use in and Spatial Hierarchies." In J. Kolasa and Rondonia, Brazil: Simulating the Effects on S. T. A. Pickett, eds., Ecological Heterogeneity, Carbon Relea.ie." International Social Science pp. 85-96. New York: Springer-Verlag. journal 43, pp. 681-98. O'Neill, R. V., S. J. Turner, V. I. Cullinen, D. P. . 1991b. "Modesopposesd'occupationdes Coffin, T. Cook, W. Conley, J. Brunt, J. M. sols au Rondonia, Bresil: Simulation de leurs Thomas, M. R. Conley, and J. Gosz. 1991b. effets sur les emissions de carbone." Revue "MultipleLandscapeScales: AnlntersiteCom- Internationaledes Sciences Sociales 130, pp. 729-46. parison." Landscape Ecology 5, pp. 13744. Stone, T. A., F. Brown, and G. M. Woodwell. 1989. O'Neill, R. V., R. H. Gardner, and M. G. Turner. "Estimates of Land Use Change in Central 1992. "A Hierarchical Neutral Model for Land- Rondonia, Brazil, by Remote Sensing." Forest scape Analysis." Landscape Ecology 7, pp. 55-61. Ecology and Management 38, pp. 291-304. O'Neill, R. V., R. H. Gardner, M. G. Turner, and Tucker, C. J., J. R. G. Townsend, T. E. Goff, and W. H. Romme. 1992. "Epidemiology Theory B. N. Holben. 1986. "Continental and Global and DisturbanceSpread onLandscapes." Land- Scale Remote Sensing of Land Cover." In J. scape Ecology 7, pp. 19-26. Trabalka and D. E. Reichle, eds., The Changing ' ~~~~~~~~~Carbon Cycle: A Global Analysis, pp. 221-41. O'Neill, R. V., C. Hunsaker, and D. Levine. 1992. New York: Springer-Verlag. "MonitoringChallengesand Innovative Ideas." Ner,M G,riHgardner,aV. In D. H. McKenzie, D. E. Hyatt, and V. J. O'Neill. 1989. "Predicting the Spread of Dis McDonald, eds., Ecological Indicators, pp. 1443- trNeslacross'Hedtergten Landc apes." 60. New York: Elsevier.' turbances across Heterogeneous Landscapes." 60. New York: Elsevier. Oikos 55, pp. 121-29. O'Neill, R. V., A. R. Johnson, and A. W. King. Turner, M. G., R. H. Gardner, and R. V. O'Neill. 1989. "A Hierarchical Framework for the Analy- 1991. "Potential Responsesof LandscapeiStruc- sis of Scale." Landscape Ecology 3, pp. 193-205. ture to Global Environmental Change." In M. O'Neill, R. V., J. R. Krummel, R. H. Gardner, G. M. Holland, P. G. Risser, and R. J. Naiman, Sugihara, B. Jackson, D. L. DeAngelis, B. T. eds., Ecotones: The Role of Landscape Boundaries Milne, M. G. Turner, B. Zygmunt, S. in the Management and Restoration of Changing Christensen, V. H. Dale, and R. L. Graham. Environments, pp. 52-75. New York: Chapman 1988a. "Indices of Landscape Pattern." Land- and Hall. scape Ecology 1, pp. 153-62. Turner, S. J., R. V. O'Neill, W. Conley, M. R. O'Neill, R. V., B. T. Milne, M. G. Turner, and R. H. Conley, and H. C. Humphries. 1991. "Pattem Gardner. 1988b. "Resource Utilization Scales and Scale: Statistics for Landscape Ecology." and Landscape Pattern. Landscape Ecology 2, InM.G.TurnerandR.H.Gardner,eds.,Quan- pp. 63-69. titative Methods in Landscape Ecology, pp. 17-41. New York: S rin er-Verlag. Post, W. M., T.-H. Peng, W. Emanuel, A. W. King, r rg and D.L.DeAngelis.1990."TheGlobalCarbon Wilson, E. O., ed. 1988. Biodiversity. Washington, Cycle." American Scientist 78, pp. 310-326. D.C.: National Academy Press. 143 Part -) Case Studies Indicators of Biophysical Sustainability: Case Study of the Chaco Savannas of South America Enrfque H. Bucher Even lacking a precise definition (Sarachchandra domesticated hooved mammals and began to 1991), sustainable development obviously im- control the movement and size of their herds. plies the management of at least three basic com- Many areas have been lost to production entirely, ponents: the biological machinery that provides whereas much larger areas have had their pro- resources and services, the human beings who ductivity seriously impaired and continue on a use those resources and services, and the socio- downward trend. Although not as diverse as the economic scenario in which the development pro- tropical rain forests, semi-arid savannas are still cess takes place. This chapter concentrates on very important in terms of their biodiversity. In possible indicators of sustainability related to the these regions, human populations are dependent first component (biophysical machinery) and on thecontinued productivity of natural resources applicable to savannas in general and to the Gran such as livestock, fuelwood, and wildlife, but the Chaco region of South America in particular. It natural environment also directly affects indi- also tests the proposed criteria by applying them vidual well-being via parasites, diseases, and the to different styles of land use common in the like. These linkages influence human social and region. This restriction in scope does not mean cultural systems, which in turn affect patterns of that I ignore the critical importance of the other land use (Ellis and Swift 1988; Schofield and components, nor that I am fully convinced of the Bucher 1986). advantages of discussing each one of them in isolation. Unless we start dealing with the specif- ics, we run the permanent risk of being lost in The Gran Chaco endless and fruitless debates on vaguely defined concepts. Totranslateabstract conceptsinto man- The vast plain known as the Gran Chaco is a agement, economists and ecologists clearly need natural region of about 1 million square kilome- a common language and conmmon criteria. This ters extending over parts of Argentina, Bolivia, need includes the development of indicators of and Paraguay. The climax vegetation is a sub- sustainability that can be collected easily and tropical savanna (Bucher 1982). The eastem part frequently and can be used to monitor the is characterized by the presence of abundant sustainability of large regions and even entire swamps, reed beds, and gallery forests, whereas continents. the predominant vegetation in the western Chaco This issueisparticularlyimportantin the semi- is a medium-tall, xerophilous subtropical forest arid savanna ecosystems of the world. Conver- with a ground layerof grassesand manycactiand sionof productivesavanna land intobarren waste- terrestrial bromeliads. Rainfall decreases west- land has been taking place since humans first ward from a maximum of around 1,200 millime- Defining and Measuring Sustainability: The Biogeophysical Foundations ters along the Paraguay-Parana rivers to a mini- Recent occupation of the Paraguayan mum of 450 millimeters in the southwest. This Chaco decrease in rainfall is accompanied by the length- ening of the winter dry season from two to Because of lack of access and the war between seven months. Bolivia and Paraguay during the 1930s, the Para- The Chaco supports considerable biological guayan Chaco remained almost empty until the diversity (Bucher 1982). Its floristic richness in- 1980s,whenconstructionofthetrans-Chaco road cludes eighty-three genera in the east, fifty-six in between Asunci6n and Santa Cruz de la Sierra the center, and sixty-five in the west (Sarmiento began. Occupation of the area has been based on 1972). Grass diversity is especially important: large-scale ranching in the west and mixed agri- more than fiftyspecies havebeen found inJoaquin culture and ranching in the center (Mennonite V. Gonzalez, Salta Province (C. Saravia Toledo, colonies), where annual rainfall is more than 700 unpublished data). The vertebrate fauna includes millimeters. In both cases, development is based about thirty-five species of amphibians, twenty on capital-intensive technology, usually financed of lizards, and at least twenty-five of snakes by loans from multinational agencies. It follows (Bucher 1982). The Chaco avifauna includes 408 the Brazilian model of occupation of the Amazon species (Capurro and Bucher 1988; Short 1975). forest, in which nearly all native vegetation is Invertebrate diversity is also great, although it is eliminated and replaced with introduced pas- incompletely known. Overall, ants are extremely tures or crops according to rainfall. As a result, a important in the Chaco, particularly leaf-cutting large portion of theoriginal biodiversity hasbeen ants, which are significant as herbivores, lost. Encroachment of the woody vegetation re- detritus-reducers, and soil-modifying agents sulting from land clearing is a serious problem. (Bucher 1982). Removal of all palatable native grass and shrubs may result in lack of availability of proteins and The impact of the European colonization green material during the dry season, affecting in Argentina the condition of cattle, particularly in years when frost damages the introduced pastures (Saravia The Chaco's primeval landscape was a parkland and Bucher, personal communication). withpatchesof hardwood intermingled withgrass- lands (Bucher 1982). This mosaic of vegetation was Sustainable management of the Chaco: kept stable by occasional flooding of low-lying ar- The Salta Project eas in the east, but more importantly by periodic firescausedbylightningorsetbylndians,ina kind In the Argentine province of Salta, a group of of pulse equilibrium common to many semi-arid researchers and landowners have proved the fea- savannas throughout the world (Huntley and sibility of managing the Chaco in a sustainable Walker 1982). After Europeans colonized the area, way by restoring and maintaining the native for- the frequency and intensity of fires decreased, par- est and grasslandsin a productivecycle. TheSalta ticularly in the dry westem Chaco, as the Indians Project, started more than twenty-five years ago, withdrewanddomesticcattlegrazedthefuelneeded is based on 300,000 hectares of public land and by fires. Consequently, woody vegetation rapidly 60,000 hectaresof privateland locatednearJoaquin invaded grassland patches throughout western V. Gonzalez, Salta (Bucher and Schofield 1981; Chaco, to the point at which grasslands disap- Saravia Toledo 1987). The scheme preserves the peared completely (Bucher 1982; Bucher and natural vegetation and dedicates a small portion Schofield 1981; Morello and Saravia Toledo 1959). of the exploited area (less than 10 percent) to A second stage in the alteration of landscape introduced pastures. The process includes two started in the 1880s when railways expanded into consecutivesteps:first,restorationofthedegraded the Chaco. This expansion not only allowed inten- vegetation, and second, institution of a multiple- sive forest cutting but also helped the Chaco species ranching system based on the sustainable campesinos (puesteros) expand into the intervening exploitation of cattle, forest products (charcoal, areas between rivers. Crucial to this expansion was fuelwood, and timber), and (possibly) wildlife. the introduction of railway transportation and new The scheme is organized around the manage- technologies, particularly the tools for digging the ment of large units of land under a system of rest artesian wells that allowed campesinos to maintain rotation of grazing by cattle, cutting of shrubs for cattlearound watering points throughout the year. charcoal and fuelwood, and felling of trees for 148 inuicarors of biG opyszcal bustainability: Case Study of the Chaco Savannas of South America fence posts and railroad ties. The smallest unit is scape fragmentation patterns, edge configura- around 5,000 hectares. Each unit is fenced to keep tions, and patch-boundary characteristics (Wiens, out cattle and goats. Then, in progressive sectors Crawford, and Gosz 1985). of about 100 hectares, all fallen wood is removed, Results after twenty-five years of management and almost all the trees and scrubs are harvested indicate that the productivity of both forage and to provide hardwood timber and fuel for the beef increase dramatically in managed areas (see charcoal ovens. The forest is then exploited se- table 10-1). After eight years, 75 percent of the quentially for timber and charcoal in cycles of original productivity can be reached (Saravia about twenty to forty years allowing adequate Toledo 1987). By this careful management, the timeforthehabitatandthegrasscovertorecover. productive cycle can be maintained, providing Seeds of wild grasses are added where regrowth profit and employment for people living in rural is poor, and the area is left undisturbed and areas. protected from cattle until the young hardwood During the last decade, it became evident that saplings are large enough to be immune to graz- wildlife could be incorporated in this production ing, although limited grazing may be allowed to system. For example, exports of Tupinambis liz- help disperse grass seeds. Once recovery is com- ards and parrots have increased steadily. At plete, cattle is reintroduced to carrying capacity present, more than 2.5 million lizard skins are density under a controlled grazing regime based exported annually from Argentina, and 141,000 on rest rotation (Saravia Toledo 1987). This care- parrots were exported in 1987, mostly from the ful level of stocking improvessubstantiallyon the Chaco (Traffic Uruguay, unpublished data). As a present capacity of the western Chaco, which in result, wildlife has become an important source its degraded form now rarely supports more than of income for the campesinos, and the possibility one cow per 15 to 30 hectares. of integrating its exploitation into the existing After about five years, when the young trees projectsunderasustainable,multispeciesscheme are about 2 meters tall, the saplings are thinned, is being explored (Beissinger and Bucher 1992; and all undesirable woody plants are removed to Bucher 1989). relieve competition with the young hardwoods The Salta approach to managing theChaco has and then used to produce charcoal. At intervals of the potential to enhance the social, medical, and about twenty to forty years, the mature hard- economic well-being of human populations in a woods can be harvested. sustainable fashion, while preserving a very high Thus, the whole area is composed of sectors proportion of the region's original biodiversity. each in a different stage of exploitation. While Moreover, the scheme may have very important some sectors are producing good-quality hard- effects on social welfare and public health. The wood timber, other areas are producing beef and cycle of overexploited land, poverty, disease, and charcoal. This management model has important urban migration has been interrupted, not just by consequences at the landscape level. Instead of treatingthediseaseorbyalleviatingpovertywith having a homogeneous cover of secondary charity but by using the land in an ecologically shrubland with isolated areas of highly degraded and economically sustainable manner (Bucher soil, a much greater degree of heterogeneity de- and Schofield 1981; Solbrig 1988). Moreover, in- velops, which in turn has the potential for sup- creased prosperity and the construction of new porting greater biodiversity, since each manage- wooden housing appear to be responsible for mentunitisinterspersedwithothersinanirregu- reducing Chagas disease. That is, rather than lar pattern that creates a complex mosaic of land- treating the problems of natural resource degra- Table 10-1: Productivity of Cattle Ranching under Different Management Systems in Salta, Argentina Parameter Traditional system Chaco del Norte Carrying capacity (number of cattle per 100 hectares) 4.35 20.00 Productivity (kilograms per hectare a year) 1.21 21.78 Profit (U.S. dollars per hectare a year) 0.36 8.00 Source: Saravia Toledo 1987. 149 Defining and Measuring Sustainability: The Biogeophysical Foundations dation, poverty, and disease as separate prob- SUSTAINABILITY OF FRESHWATER RESOURCES lems, it may be more effective to deal with them As in every region of the world, availability of simultaneously as interrelated problems (Bucher fresh water is limited. Campesinos obtain water and Schofield 1981). from shallow artesian wells that are replenished by rainfall. Large-scale ranching programs like those developed in Paraguay and Salta may re- Indicators of sustainability quire larger amounts of water, which has to be obtained from deeper groundwater sources. At For any model of development, biophysical indi- least some of these sources may not be sustain- cators of sustainability should deal with different able since they come from past geological times scales of environmental health and cover at least and are not replenished fast enough. Adequate the following broad aspects. evaluation of the sustainability of water resources should be an important prerequisite for any de- At the global scale, velopment program in the Chaco. * Contributions to the balance of atmospheric SOIL CONSERVATION gases and impact on climatic stability Pastoral activities may affect soil mostly through * Contributions to pollution of air, water, and soil. erosion and loss of nutrients in overgrazed areas. At the regional scale, This process is very serious in areas occupied by campesinos, whereasitshould be negligible in well- * Sustainability of freshwater resources managed natural pastures, like in the Salta model. * Preservation of soil integrity PRESERVATION OF BIODIVERSITY * Preservation of biodiversity. Loss of biodiversity is high (although generally slow) under the campesino model. However, at PRESERVATION OF THE ATMOSPHERIC BALANCE least some portion of the original biodiversity is Emissions of carbon dioxide (C02) and other preserved. Losses are still higher in the Brazilian greenhouse gases in the Chaco relate mostly to model, which also implies the introduction of the burning of vegetation. Fire is an important alien species of plants. On the contrary, the Salta ecological factor in all semi-arid savannas, and it model preserves a very high proportion of the cannot be easily eliminated. Fuelwood is a cheap originalbiodiversity(see table 102 fora description and sustainable source of energy that may help to of the impact of different management models). decrease the consumption of fossil fuels: during World War ll, the Argentine railroad system ran Synthetic indicators of sustainability its steam engines exclusively on fuelwood. Fi- nally, the restoration of vegetation may be impor- Measurement techniquesand recommended lim- tant in increasing CO2 fixation and in reducing its for these indicators are already available and albedo from bare soil. easy to use. However, a set of synthetic indicators of rangeland sustainability can provide a more POLLUTION OF AIR, WATER, AND SOIL reliable and comprehensive indication of Pastoral land use produces, on average, a limited sustainability. This set includes two well-known amount of pollutants. concepts among range managers: condition and Table 10-2: Impact of Different Management Systems on Biophysical Sustainability in the Chaco Subsystem Campesino Paraguay Salta Climate Low Medium Low Pollution None Low Low to none Water resources None Low Low Soil High Medium Low Biodiversity loss High Very high Very low 150 Indicators of Biophysical Sustainability: Case Study of the Chaco Savannas of South America trend. Both have been widely tested in many References semi-arid ecosystems of the world with an annual rainfall of around 300 millimeters and apply per- Beissinger, S. R., and Enrique H. Bucher. 1992. fectly well to the Chaco. "Can Parrots Be Conserved through Sustain- able Harvesting?" Bioscience 42, pp. 164-73. CONDITION Bucher, Enrique H. 1982. "Chaco and This term covers the assessment of the status of a Caatinga-South American Arid Savannas, particular area of savanna in relation to its poten- Woodlands, and Thickets." In B. Huntley and tial, classifying its condition as excellent, good, B. Walker, eds., Ecology of Tropical Savannas, fair, and poor (Dasmann 1951). In areas where pp. 47-79. Berlin: Springer-Verlag. savanna is considered the climax vegetation, range _ . 1989. "Conservaci6n y desarrollo en el condition is measured as the degree of departure neotr6 ico: En bnsqueda de alternativas." Vida from the stable, climax mixture of grass, forb, and neotrp opica 2 e pp. 3ida woody species, along a gradient of increasing Silvestre Neotropical 2, pp. 36. grazing pressure. Bucher, Enrique H., and C. J. Schofield. 1981. "Economnic Assault on Chagas Disease." New TREND Scientist 92, pp. 320-24. A companion measurement that must accom- Capurro, H. A., and Enrique H. Bucher. 1988. pany the assessment of condition concerns trend. "Lista comentada de las aves del bosque This states whether a particular area is deteriorat- chaquefio de Joaquin V. GonzAlez, Salta, Ar- ing, improving, or stable, under current manage- gentina." Hornero 13, pp. 39-46. ment, which requires successive measurements Dasmann, W. P. 1951. "Some Deer Range Survey of condition along time. For example, an area in Methods." California Fish and Game 37, pp.43- excellent condition showing a deteriorating trend 52. indicates inadequate management, whereas an- otherinfairconditionbutshowinganimproving Ellis, J. E., and D. M. Swift. 1988. "Stability of trend indicates good management. African Pastoral Ecosystems: Alternate Para- Measurements of condition are based on an digms and Implications for Development." assessment of change in vegetation and rely on Journal of Range Management 41, pp. 450-59. the presence or absence of indicator species that Huntley, B., and B. Walker, eds. 1982. Ecology of occur only in a well-defined range along the Tropical Savannas. Berlin: Springer-Verlag. condition axis (Dasmann 1951). Indicator species Morello, J., and C. SaraviaToledo. 1959. "Elbosque may include not only plant but also wildlife spe- chaquefio I. Paisaje primitivo, paisaje natural y cies if suitable. For example, medium-size ro- paisaje cultural en el oriente de Salta." Revista dents like the Chaco cavy (Pediolagus salinicola) Agron6mica del Noroeste Argentino 3, pp. 5-81. and the vizcacha (Lagostomus maximus) increase in degraded areas (Bucher 1982); whereas the Sarachchandra, M. L. 1991. "Sustainable Devel- blue-fronted Amazon parrot (Amazona aestiva) opment: A Critical Review." World Develop- breeds only in old-growth forests where snags 'pp and old trees provide holes suitable for nesting SaraviaToledo,C.1987."Restorationof Degraded (Beissinger and Bucher 1992). Pastures in the Semi-arid Chaco Region of The advantage of using synthetic indicators Argentina." In Proceedings of the UNESCO In- such as condition and trend is that they are simple ternational Symposium on Ecosystem Redevelop- and reliable. Even if the techniques for assessing ment: Ecological, Economic, and Social Aspects, them for a given region need to be developed by April 1987, pp. 25-37. Budapest: UNESCO. specialists, the resulting guidelines can be used by Sarmiento, G. 1972. "Ecological and Floristic nonspecialists. Of course, the only prerequisite Convergences between Seasonal Plant Forma- would be the ability to recognize the indicator spe- tions of Tropical and Subtropical South cies of the various successional stages. At the same America." Journal of Ecology 60, pp. 367-410. time, evaluation of condition and trend provides a Schofield, C., and Enrique H. Bucher. 1986. "In- much better and more reliable synthesis of the dustrial Contributions to Desertification in sustainabilityofanymanagementsystemthan many South America." Trends in Ecology and Evolu- specific measurements taken in isolation. tion 1, pp. 78-80. 151 Defining and Measuring Sustainability: The Biogeophysical Foundations Short, L. L. 1975. "A Zoogeographic Analysis of Wiens, J. A., C. S. Crawford, and J. R. Gosz. 1985. the South American Chaco Avifauna." Bulletin "Boundary Dynamics: A Conceptual Frame- of the American Museum of Natural History 154, work for Studying Landscape Ecosystems." pp. 163-352. Oikos 45, pp. 421-27. Solbrig, 0. 1988. "Destrucci6n o transformaci6n del paisaje tropical sudamericano?" Interciencia 13, pp. 79-82. 152 The Sustainability of Natural Renewable Resources as Viewed by an Ecologist and Exemplified by the Fishery of the Mollusc Concholepas concholepas in Chile Juan Carlos Castilla Iwish toacknowledgefinancialsupportfrom FONDECYT-CHILEandfrom thelnternational Development Research Center-Canada. I benefited enormously from presentations given during the fourth Cary Conference at the Institute for Ecosystem Studies held in Millbrook, N.Y., in 1991, and from the symposium on the resource known as loco, Comitj4e las Ciencias del Mar, held in Santiago, Chile, in May 1992. Discussions with Dr. Jane V. Hall, Department of Economics, California State University, Fullerton, and Dr. Michael Berg, Department of Zoology, University of Cape Town, South Africa, enriched my understanding of economic theory and strategiesfor managing natural resources. I sincerely appreciate their contributions. To sustain, to support, or to uphold are verbs different economic and social realities, questions commonly used in the catchphrases permeating such as the following should be addressed in the diverse documents dealing with develop- every case: (a) what is to be.sustained? (b) for ment issues, the ecological basis of human life, whom? (c) under which circumstances? (d) for environmental problems, or the future of the how long? and, perhaps more critically, (e) is Earth (see, for example, Tolba 1984a, 1984b; development synonymous with growth in the WCED 1987; Brundtland 1989; McNeill 1990; consumption of materials? (Lele 1991; see also GoodlandandElSerafyl991;IUCN1991;Mead- Goodland and El Serafy 1991). ows,Meadows,andRanders1992;seealsochap- These questions should not be addressed ex- ter I of this volume). Moreover, the concept of clusively by economists, because the answers sustainable development itself has become per- should consider theavailableknowledge derived vasive. Lele (1991) published a critical review on from disciplines such as biology, chemistry, soci- the concept of sustainable development, showing ology, physics, and so forth. In the same vein, the that this new paradigm harbors weaknesses, in- lack of research across disciplines leads to (a) adequacies, contradictions, and in some cases undervalorization of the current knowledge, (b) inconsistencies. Intellectual accuracy and rigor failure to identify the critical issues to be tackled are needed in its use. in future research, and (c) inadequate use of the Indeed, there is a need to attach real substance deeply rooted perceptions of scientists working in to the definition of sustainable development as differentdisciplinesandusingdifferentparadigms. the development that meets the needs of the Therefore, this chapter addresses, first, thebasic present without compromising the ability of the tenets that form partof the paradigm of the West- future generations to meet their own needs. Since ern mechanistic price system (as understood by the concept cannot be applied equally well under an ecologist), confronting them with those that Defining and Measuring Sustainability: The Biogeophysical Foundations we, as ecologists, use frequently when dealing renewable resource complexes. Both, the mecha- with natural renewable resources. Second, it fo- nisticpricesystemschemeandthenaturalrenew- cuses-perhaps simplistically, but as a way to able resource complexes can be extremely contro- illustrate some of the main conflicting issues-on versial issues. I take full responsibility for their two basic and contrasting models of growth: one interpretation. This is an important exercise be- relating to the mechanistic price system and an- cause it highlights the major assumptions, rules, other commonly used in fishery management and articles of faith involved in these paradigms. strategies. Third, it describes the history of Chil- According to Hall (1993), humansreact mecha- ean fishery of an economically extremely impor- nistically and in predictable ways to economic tant mollusc, Concholepas concholepas, which is signals, and therefore, understanding of the hu- foundexclusivelyinthesoutheasternPacificcoast. man factor is based on assumptions that indi- This example illustrates the biological knowl- vidual maximization optimizes the system, that edge and economic perceptions developed by a the past is the past (and does not count), that very marine ecologist who has followed the ups and simple signal (prices) encapsulating extraordi- downs of this mollusc's fishery. Indeed, this ex- nary complex information are captured in such ample properly illustrates the saga of many ma- signals (and are adequate measures of value), rine resources exploited in developing countries, that each economic actor knows what they are where economic pressure generated by the exter- doing and that growth is inevitable, unlimited, nal debt and social problems, use of the Western and desirable. mechanistic price system, and lack of integration The main tenets of the mechanistic price system of the peculiar characteristicsof natural resources scheme can be simplified in six major concepts: to such models have led to the overexploitation of many unique marine resources. The concept of Optimzat w t p o y equity, which is also at the center of sustainable personal level development paradigm, is not addressed. * Rationality, which takes place also at a per- sonal level a Values, which means that all values are mon- The mechanistic price system and the main etized tenets of natural renewable resources * Marginalism, which indicates that the optimi- zation process takesplaceat the margin (past is The Western economic mechanistic price system, past) the related models, and their main assumptions * Impersonality, which means that transactions and tenets have been discussed by numerous happen to be distant (few peoples buy fish authors.Thisdiscussion summarizes the viewsof hapen to b istant (fe popes Hall and Hall (1984), Hall (1993), and Hall, Btajer, directly from fishermen at the cove) andRowe(1991),whichenablesmetoextractthe * Continuing and unlimited growth, which is main principles of the present mechanistic price highly desirable (Costanza 1989), and system scheme operating in most Western societ- avoidance of the issues of equity and ies (Chile included) and, furthermore, to relate anthropocentric values, which are difficult to them to the main tenetscharacterizing the natural resolve (Hall 1993). Table 11-1: Economic Tenets of Natural Renewable Resource Complexes Compared with the Price System Scheme Paradigm Price system scheme Natural renewable resources complexes Optimization: personal Externalities: personal and collective (pollution, overexploitation) Rationality: personal Rationality: personal and public goods Values: monetized Value: monetized and unpriced or underpriced (ecosystems, aesthetics) Marginal: past is past Holistic: past, present, and future are important Impersonal transactions: distant Perceptions: direct and indirect; distant and close Growth: continuous, unlimited, Growth: limited, not continuous, oscillating (carrying capacity and desirable of environment, energy flow scarcity) 154 The Sustainability of Natural Renewable Resources Table 11-1 shows the major tenets of the mecha- Figure 11-1: Two Contrasting Models: Capital nistic price system scheme compared with those versus Interest and Stock Size versus Harvest operating in natural renewable resource com- plexes. The list is selective rather than exhaustive, Monetary grow but the differences are striking. Indeed it can be Resource in the bank said that the tenets referring to the natural renew- overexploited i able resource complexes (at the level of popula- Max/( num sustainable yield tion, community, or ecosystem) are basically the l opposite of those ruling mechanistic price system : Biornass and opposite ~~~~~~~~~~~~~~~~~~~~~renewable resource schemes. , (growth in thesea) Hence, externalities are important tenets in Resource natural renewable resource complexes, both at unexploited personal and at collective levels. For instance, (-) Capital (+) pollution, overexploitation, or aesthetic values (-) Stock size (+) are externalities of natural renewable resource Note: Capital versus interest represents monetary growth in complexes. The rational approach, being personal, a bank. Stock size (or biomass) versus harvest represents a also includes public goods. Value includes both renewable resource in the sea. Source: Taken from a conference given by M. Berg of monetized and unpriced or underpriced values the University of Cape Town, South Africa, in Santiago, of the system, for example, aesthetics or unpriced Chile, May 1992. values of the ecosystems or communities to which the monetized natural renewable resource com- the case of natural renewable resource complexes plexes belong. In natural populations, communi- paradigm the model features the relationship ties, or ecosystems, the history in evolutionary or between stock size or biomass and harvest of a ecological time matters. The resilience capabili- natural resource at a given level of population. In ties and stability of natural renewable resource this case-as for instance in the basic Gordon- complexes are key factors: the perception of the Schaefer model (Gordon 1954; Schaefer 1954; see observer (actor) is both direct or indirect and also Ricker 1975) and the maximum sustainable close or distant. Furthermore, interactions are - yield harvesting model-the argument of critical extremely important as are the feedback mecha- biomass harvesting around the greater excess nisms and the transitive or intransitive interactions. above maintenance requirements is used. This The most dramatic differences between mechanis- excess (surplus production) is the maximum sus- tic price system schemes and natural renewable tainable yield, and the exploitation rate that pro- resource complexes are related to growth. In natu- duces maximum sustainable yield is likely to be ral populations, growth is limited and never totally near a cliff edge. Slightly higher exploitation rates continuous.Thereare different growth ratesatpar- could drive the stock toward extinction. ticular stages, and population growth can follow a Thisillustratesthe deepconceptual differences sigmoidmodeltendingtoreachacarryingcapacity that can arise around the main tenets of both (K) or can follow a I model. Energy flow is a critical mechanistic price system schemes and natural factor, and both scarcity and abundance can be renewable resource complexes paradigms. This spotted, which could be related to resource oscilla- is not to say that al ternative economic approaches tions along time. Biological interactions, perturba- regarding the exploitation of natural resources tions, and environmental heterogeneity can modu- cannot be worked out. For instance, G6mez-Lobo late population growth. and Jiles (1991) have addressed the use of the Gordon-Schaefer growth model (Gordon 1954; Schaefer 1954), in conjunction with models of Two simple contrasting models economic return, to achieve a bioeconomic equi- librium for a given renewable population. Not- Figure 11-1 shows two contrasting models. In the withstanding, past experience (particularly in the case of the mechanistic price system scheme para- developing world) has shown specifically that digm, the model exemplifies the relationship be- the management of marine resources is not done tween capital and growth (monetary growth) in a on a sustainable basis, but according to the main bank account. There is a direct relationship re- philosophicaltenetsofthemechanisticpricesystem sulting in an ever-increasing growth of capital scheme paradigm, such as continuous and unlim- and better interest rates if money is not drawn. In ited growth (see Goodland and El Serafy 1991). 155 Defining and Measuring Sustainability: The Biogeophysical Foundations The Gordon-Schaefer (1954) model has been Chile between 1960 and 1992. As is true of most used to manage pelagic fisheries for more than artisanal fisheries, information related to effort is twenty-five years (for a review of economic mod- not available. Nevertheless, four periods can be eling see Clark 1985). Because it is simple and distinguished. static rather than complex, the model is attractive The first is a period of internal or local (Chil- to managersand, asdemonstrated by Clark (1985), ean) consumptionbetween 1960 and 1974 in which has a general theoretical validity. Nonetheless, landings fluctuated around 3,000 to 5,000 metric there are numerous instances around the world tons. The fishery operated on a free-entry basis in which pelagic fisheries have collapsed even and was unlicensed. No closed seasons occurred when this or other related dynamic bioeconomic during this period, and external markets were fisherymodelshavebeenused: thecollapseof the nonexistent. During this period, Chileans con- Peruvian anchovy fishery in the 1960s, the Nor- sumed locos as they have done since pre- wegian overexploitation of marine resources in Columbian times (Ramirez and others 1991; the 1960s, the capital overinversion and foreign Jerardino and others 1992), and nobody thought fleets affecting island fisheries, and so forth. In that by doing so, they were jeopardizing the most of these instances, negative externalities, abilityof futuregenerationstoconsumethisprized lack of adequate biological knowledge, environ- mollusc. The general perception was that the mental perturbations, and probably the use of resource could be sustained a long time. single-speciesmodeling, in conjunction with eco- The second is a period of local consumption nomic pressures, should be blamed for the col- plus heavy exportation of the resource between lapse. It must be remembered that in the mecha- 1975 and 1981. In a very short period, landings nistic price system scheme paradigm, such exter- increased to 25,000 metric tons a year (in 1980). nalities do not exist. The fishery still operated on a free-entry basis and Furthermore, it must be kept in mind that was unlicensed; no closed seasons occurred; ex- models such as the one described above and the ternal markets opened mainly in Southeast Asia lateral bioeconomic models attached to them fo- (Hong Kong, Japan, and Korea). Moreover, as a cus exclusively on the resource of interest. It is not matter of policy, the Chilean government encour- a community approach. In fact, the community aged exports of all sorts of natural renewable and and ecosystem approach, wherein the species of nonrenewable resources (Castilla 1990). Particu- interest islocated,isconsidered tobeanexternality. larly strong economic incentives (signals) oper- ated with regards to the exportation of nontradi- tional resources (see Bustamante and Castilla The example: The fishery of the mollusc 1987). Chileans continued to have local access to Concholepas concholepas in Chile the loco, until fishery authorities realized that the rate of exploitation was unsustainable: during Concholepas concholepas, known in Chile as loco, is 1981 landings dropped from about 25,000 to about a unique muricid mollusc found along the entire 18,000 metric tons. Right or wrong, it was esti- Chilean and Southern Peruvian coastline. It is mated that the future capability to export locos fished exclusively by divers operating from small- was being jeopardized. At that time, the value of scale artisanal boats. Thelife history, biology, and the loco in international markets reached around fishery of this valuable resource is described in $20 million. numerous papers (for example, see reviews by The third isa period of regulatory measuresas Castilla 1982; Bustamante and Castilla 1987; an answer to probable (though never verified) Castilla 1988; Geaghan and Castilla 1988). overexploitation. Hence, during 1982, 1983, and In 1987 the Chilean loco fishery operated only 1984 reproductive closed seasons-several forty-five days, and more than 21,000 metric tons months a year-were established. Landings were were landed. The total exportation value about 16,000 to 19,000 metric tons, and the exter- amounted to more than $42 million (all dollars nal market values were around $18 million to $25 are U.S. dollars). For the same year, the total million. From the beginning of May 1985 to the exportation value of the Chilean shellfish fish- end of May 1987, the loco fishery wasclosed, and, ery-based on several dozen species-amounted for the first time, a global quota of 4,000 metric to more than $120 million. Figure 11-2 shows the tons was established (and greatly exceeded dur- landings and exportation values of the loco in ing 1985 and 1986). 156 The Sustainability of Natural Renewable Resources Figure 11-2. Landings of Concholepas concholepas in Metric Tons and Exportation Values in llions of U.S. Dollars, 1960-92 25 -50 Concholepas concholepas5 *-O Landings / i 20 - 2 O 0 port values0 20~~~~~~~~~~~~~~~~~~~~~~~~~~~ 215 30 o 30\ - 0\I 10 20 KO El 5 V10 Exportabon pbask Overexploited I Total aiUnsustanab Repgulatrx | Closure Local conmsumpbon , . ˘ ' 1 O - Sustainable L 0 1960 1965 1970 1975 1980 1985 1990 In 1987, at the end of the closure, the loco lars). In fact, this fourth period can be identified fisherywasopenedforonlyforty-fivedays.More as the illegal or chaotic period. Indeed, it is not than 21,000 metric tons were fished during that known whether or not this period can be consid- short period of time, with a market value of $42 ered sustainable or unsustainable from the million. Moreover, during 1988 the fishery was fishery's point of view. Since no legal fishery is opened for only fifteen days with a maximum takingplace,itisimpossibletoevaluatetheactual total quota of 4,000 metric tons for the whole condition of the resource. Moreover, not enough country. The fishery effort was so elevated and research isbeing conducted toevaluate the stocks centered in the most productive fishing grounds of loco in the country. that more than 11,000 metric tons were fished in a few days, with a total export value of more than $35 million. G6mez-Lobo and Jiles (1991) have The tragedy of the locos used this example to illustrate economic ineffi- ciency. The global-quota strategv used in this "Each man is locked into a system (economic case was not appropriated, system) that compels him to increase his herd The fourth period-1990, 1991, and 1992- withoutlimit-inaworldthatislimited"(Harding represents the total closure of the loco fishery, at 1968, p. 1243). The example of the Chilean loco least legally and in theory. Actually, the amount fishery portrays extremely well the so-called trag- of illegal fishing of locos amounted to around edy of the commons, which in this case was 5,000 to 7,000 metric tons a year (see the arrows in clearly forced by an economic vector. The ques- figure 11-2). The loco resource has such an el- tlonsaddressedearnhernthischapterareparticu- evated value in the intemational market that ille- larly relevant here: gal actors have devised and used all sorts of tricks * What is tobe sustained? If onlythepopulation to continue exploitation and exportation. The of locos is to be sustained, then, sustainability losses to the country in taxes forgone are ex- hasnotbeenachievedinChileinthepastthirty tremely high (probably several millions of dol- years. 157 Defining and Measuring Sustainability: The Biogeophysical Foundations • For whom is it to be sustained? For the small- been applied. In each case, the mollusc extrac- scale fishermen? For the country? For the in- tions have been tightly regulated, and the policy dustry? For the whole Chilean population that of allowing free entry to the fishery of Concholepas has consumed locos since ancient times? has been abandoned in the country. * For how long is it to be sustained? For a de- cade? For the next generation? References * How do we know or assess the damage al- ready inflicted on the loco population? Or Brundtland, G. H. 1989. "Global Change and Our perhaps there is no such a damage? Common Future, Washington, D.C., Benjamin * What are the social, environmental, or FranklinLecture(May2)."Environment31,pp. ecological consequences? 16-20, 40-43. Indeed, so far this chapter has only focused on Bustamante, R., and Juan C. Castilla. 1987. "The this resource at the level of population. Neverthe- Shell Fishery in Chile: An Analysis of Twenty- less, the loco belongs to a rich and unique coastal six Years of Landings (1p960985)." Biologia community and ecosystem. In a series of papers, Pesquera 16, pp. 79-97. Castilla and co-authors (Castilla and Duran 1985; Castilla, Juan C. 1982. "Pesquerias de moluscos Castilla 1988; Duran and Castilla 1989) have high- gastropodos en Chile: Concholepas concholepas, lighted several singularities or ecological exter- un caso de estudio." In Juan C. Castilla, ed., nalities of this mollusc, which must be considered Segundo seminario-taller, bases biol6gicas para el a keystone species (Paine 1966) in coastal benthic uso y manejo de recursos naturales renovables: ecosystems of Chile. Its presence or absence is Recursos bioldgicos marinos, pp. 199-212. critical to the future structure and dynamics of Monografias Biol6gicas 2. Santiago de Chile: these ecosystems. This is an unpriced aspect that Pontificia Universidad Cat6lica de Chile, cannot be fed into biocconomic models. Last, but Vicerectoria Academica. not least, the species is unique and present only . 1988. "Una revisi6n bibliografica (1980- along the coasts of Chile and Peru. What is the 1988)sobreConcholepasconcholepas(Bruguiere, price of such externality? 1789) (Gastropoda, Muricidae): Problemas The case of the loco fishery in Chile is only pesqueros y experiencias en repoblaci6n." one among many examples of dramatically af- Biologia Pesquera 17, pp. 9-19. fected stocks of autochthonous species of in- _ . 1990. "Clase magistral: Importancia y vertebrates and marine algae in the country. proyecci6n de la investigaci6n en Ciencias del The concepts of growth and development have Mar en Chile." Revista de Biologia Marina, both been wrongly equated with economic growth. Undoubtedly, a strong policy focusing Valparaslso 25:2, pp. 1-18. on the economic exploitation of natural renew- Castilla, Juan C., and L. R. Duran.1985. "Human able resources-containing tenets of both eco- Exclusion from the Rocky Intertidal Zone of nomic and renewable resource models-is ur- Central Chilc: The Effects on Concholepas gently needed. This is particularly imperative concholepas (Gastropoda)." Oikos 45, pp.391-99. in the developing world. Clark, C. W. 1985. Bioeconornic Modeling and Fish- eries Modeling. New York: Wiley-lnterscience. Note Costanza, R. 1989. "What Is Ecological Econom- ics?" Ecological Economics 1, pp. 1-7. Since June 1992, the date of the conference on Duran, R. L., and Juan C. Castilla. 1989. "Varia- which this volume is based, the fishery of the tioilandPersistenceoftheMiddleRockylnter- molluscConcholepasconcholepasinChilehasexpe- tidal Community of Central Chile with and rience(. several changes. The most important re- without Human Harvesting." Marine Biology fers to three experimental openings of the fishery 103, pp. 55542. (following nearly four years of total closure) for Geaghan, J. P., and Juan C. Castilla. 1988. "As- short periods: January and July 1993 for approxi- sessment of the Present Capacity for Manage- mately seven days in each case and August 1994 ment of the 'Loco' Concholepas concholepas for thirty days. New regulations based on the (Bruguiere, 1789) (Gastropoda, Muricidae) in 1991 Chilean Fishery and Aquaculture Law have Chile." Biologfa Pesquera 17, pp. 57-72. 158 The Sustainability of Natural Renewable Resources G6mez-Lobo, A., and J. Jiles. 1991. "Regulaci6n McNeill, J. 1990. "On the Economics of Sustainable pesquera: Aspectos te6ricos y experiencia Development." Paper prepared for the USAID mundial." Notas Tecnicas 142. Corporaci6n de workshop, Washington, D.C., January 23-26. Investigaciones Econ6micas para Meadows, D. H., D. L. Meadows, and J. Randers, Latinoamerica (CIEPLAN), Santiago, Chile. eds. 1992. Beyond the Limits: Global Collapse or a August. Sustainable Future. London: Earthscan Publica- Goodland, R., and S. El Serafy, eds. 1991. "Envi- tions. ronmentally Sustainable Economic Develop- Paine, R. T. 1966. "Food Web Complexity and ment Building on Brundtland." Working Pa- SpeciesDiversity." AmericanNaturalist lO0,pp. per46. Environment Department, World Bank, 65-75. Washington, D.C. Ramirez, J. M., N. Hermosilla, A. Jerardino, and Gordon, H. S. 1954. "The Economic Theory of a Juan C. Castilla. 1991. "Analisis bio- Common Property Resource: The Fishery." arqueol6gico preliminar de un sitio de Journal of Political Economy 62:2, pp. 124-42. cazadores recolectores costeros: Punta Hall, J. V. 1993. "The Iceberg and the Titanic: Curaumilla-1, Valparaiso." In H. Niemeyer, Human Economics Behavior in Ecological ed.,ActasdelXlCongreso NacionaldeArqueologfa Models."lnM.J.McDonnellandS.T.A.Pickett, Chilena. Vol. 2, pp. 81-93. Santiago, Chile: eds., Humans as Components of Ecosystems: The MuseoNacional de Historia Natural ySociedad Ecology of Subtle Human Effects and Populated Chilena de Arqueologia. Areas, pp. 51-60. New York: Springer-Verlag. Ricker, W. E. 1975. "Computation and Interpreta- Hall, D., and J. V. Hall. 1984. "Concepts and tion of Biological Statistics of Fish Popula- Measures of Natural Resources Scarcity with a tions." Bulletin of the Fisheries Research Board of Summary of Recent Trends." Journal of Envi- Canada 191, pp. 382. ronmental Economics and Management 11, pp. Schaefer, M. B. 1954. "Some Aspects of the Dy- 363-79. namics of Populations Important to the Man- Hall, J., V. Btajer, and R. Rowe. 1991. "The Values agement of Commercial Marine Fisheries." of Cleaner Air: An Integrated Approach." Con- Bulletin of theInter-American Tropical Tuna Com- temporary Policy Issues 9, pp. 81-91. mission 1, pp. 25-56. Harding, G. 1968. "The Tragedy of the Com- Tolba, M. K. 1984a. "The Premises for Building a mons." Science 162, pp. 1243-48. Sustainable Society." Address to the World Commission on Environment and Develop- IUCN (International Union for the Conservation ment, October. United Nations Environment of Nature). 1991. Caringfor the Earth: A Strategy Program, Nairobi, Kenya. for the Future of Life. Gland, Switzerland. _ . 1984b. "Sustainable Development in a Jerardino, A., Juan C. Castilla, J. M. Ramirez, and Developing Economy." Address to the Inter- N. Hermosilla. 1992. "Early Coastal Subsis- national Institute, Lagos,Nigeria, May. United tence Patterns in Central Chile: A Systematic Nations Environment Program, Nairobi, Study of the Marine-Invertebrate Fauna from Kenya. the Site of Curaumilla-1." Latin American An- WCED(World Commission on Environment and tiquity 3:1, pp. 43-62. Development). The Brundtland Commission. Lele, S. 1991. "Sustainable Development: A Criti- 1987. Our Common Future. Oxford, England: cal Review." World Developtnent 19:6, pp. 607-21. Oxford University Press. 159 1> Sustaiinable Development and the Chesapeake Bay: A Case Study Christopher F. D'Elia lack Greer, director of the Coastal and Environmental Policy Program, made helpful comments on the manuscript. Charles Spooner of the U.S. Environmental Protection Agency, Chesapeake Bay Program, provided data on population growth in the Chesapeake watershed; Robert Summers and Darcy Austin of the Maryland Department of the Environment provided the three-dimensional figures of nitrate concentration from the Chesapeake Bay Monitoring Program; and Eric Itsweire of the National Science Foundation and Lawrence W. Harding, Jr. of UMCEES provided the Ocean Data Acquaisition System image showing levels of phytoplankton biomass in the Chesapeake Bay. The Chesapeake Bay stretches nearly 200 nauti- No other U.S. estuary serves as a better model cal miles and has a drainage basin of more than for assessing the prospects of sustainability in 160,000 square kilometers that includes large ecological, sociological, or economic terms. This parts of Maryland, New York, Pennsylvania, chapter has several goals. The first is to provide and Virginia, small parts of Delaware and West some background information about current en- Virginia, and all of the District of Columbia (see vironmental and demographic concerns relevant figure 12-1). Approximately one in twenty-five to the Chesapeake Bay, viewed from the context Americans lives in its watershed. Because it is of management activities and governance. The the largest estuary in the forty-eight contigu- second is to consider how sustainability has been ous U.S. states and is located near the nation's defined by others and to use the bay as a case capital, it has long garnered attention from the study for defining and measuring sustainability, public and the nation's leaders. Because the in ecological terms, foraland-marginecosystem. earliest European colonists settled on its shores, The third is to address the role of monitoring and human activities have had varying effects for researchindefiningandmeasuringsustainability. nearly four centuries. Over that period, agricul- The final goal is to raise crucial questions that will tural yields have been high, and the bounty of bear on our ultimate success in defining and the bay has been rich harvests of oysters, fin- achieving sustainability for the bay. Although I fish, and crabs. Unfortunately, this productive mainly focus here on the role of science in obtain- estuary no longer yields such rich harvests, and ing an understanding of how the Chesapeake Bay fewer and fewer people earn their keep as system functions or changes ecologically, I take watermen. Instead, this land of pleasant living some polemical liberty in commenting about issues now attracts more and more people and, with that must be addressed if we are to be truly able to them, rampant development. achieve sustainability in the Chesapeake system. Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 12-1: Map Showing the Chesapeake Bay and Its Watershed LAJKE ONTARIOS \ ~~NEW YORK_ SL_________ _G~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I__________ A/A' lf you ask an ecologist, a sociologist, an econo- get a single answer: no, although probably for dif- mnist, a public official, a watermnan, a recreational ferentrcasons.Ineffect,wea1lknowthattheCl.,sa-. fisherman, or for that matterasuffering wageearner peake is not a sustainable system, even though to and taxpayer to define sustainability, you are likely each of us sustainability means somnething slightly to get seven different definitions, probably sharing different. Before offering my own definition of some common aspects. If you ask any of them sustainability in the context of the Chesapeake, I whether we have achieved sustainable develop- first provide a brief history and description of the ment for the Chesapeake Bay, you will inevitably effects of human activity on that system. 162 Sustainable Development and the Chesapeake Bay: A Case Study Brief description and history Figure 12-2: Typical Effects of Nutrient Additions to of the Chesapeake Bay the Chesapeake Bay The Chesapeake Bay is considered on average to Runoff Atmospheric Sewage be a partially mixed estuary, owing to its basin epIto morphology and flow regimes (Pritchard 1955). I K Much of the bay is stratified, particularly in the Nutnent warmer months. A current topic of interest for the inputs Chesapeake is nutrient enrichment and its ulti- mate effect of promoting oxygen depletion in deep waters. The process by which enrichment Increased nutrient occurs is outlined in figure 12-2. In brief, nutrient concentrations in water inputs from point sources (sewage treatment fa- | cilities) and nonpointsources (runoff, groundwa- lncreased ter, atmospheric deposition) increase the nutrient growth and biomass concentrations in the estuary. These elevated con- I centrations, in effect, fertilize and promote the I growth of phytoplankton, which produces or- Increased turbidity, and biochemnical ganic matter and increases turbidity of the water oxvgen demand column. When freshwater flows are strong, the bay is most prone to stratification in which denser and more saline water moving from the sea is Accumulation below overlain by fresher water moving down the estu- pycnocLine ary. This, in turn, isolates deep waters from reaeration at the surface, and when organic mate- rial from primary producers moves to deep wa- and of organeics ters where it is consumed by microbes, oxygen is depleted. There is a direct relationship between I the amount of organic material produced in sur- Hypoxia or anoxia face waters and the expression of oxygen deple- tion in deep waters (Malone 1991,1992), and it is many areas since the turn of the last century, the clearthatnutrientinputsfrompoint andnonpoint application of nutrients to agricultural lands, in sources, in turn, control the production of organic the form of fertilizers, has increased nutrient load- material by phytoplankton. ings from agricultural runoff. Paleostratigraphic analysis of sediment cores The degradation of water quality appears to representing depositions from the first European have accelerated in the last half century, as in- colonization verifies the bay's natural suscepti- ferred from the elevated iron pyrite content in bility to the effects of nutrient enrichment and surficial sediments, an indicator of increasingly oxygen depletion (Cooper and Brush 1991). A reduced conditions attendant to severe oxygen relatively small colonial population caused pro- depletion (Cooper and Brush 1991). Beginning in nounced ecological changes, principally through the nineteenth century, this change was accom- its land use practices. When land clearance and panied by an accelerated shift in the species com- attendant erosion were greatest-during the late position of microalgaepreserved in the sediments eighteenth through the late nineteenth century- (Cooper and Brush 1991): centric,pelagicdiatoms sedimentation rates and the preservation of total are now more abundant than benthic, pennate organic carbon, nitrogen, and sulfur accelerated diatoms. The increased use of fertilizer has in- accordingly. During that period, more than 80 creased the production of algae in the water col- percent of the arable land in the Chesapeake Bay umn and decreased production of algal species region wascleared of forests and used foragricul- associated with the surface of benthic sediments. tural purposes (Mackieman 1990). Since then, Inaddition,vascularplantsrootedinthebenthos increases in nutrient inputs associated with have declined severely (Orth and Moore 1983). changes in land use have been further supple- Such changes are characteristic of enriched water mented by inputs from industrial and domestic bodies in whichnutrients stimulate the production sources. Although reforestation has occurred in of phytoplanktontothedetrimentofbenthicplants. 163 Defining and Measuring Sustainability: The Biogeophysical Foundations The Chesapeake as a management case study sistent to prove a bay-wide change in water qual- ity (although recent study of the sedimentary In many ways, the Chesapeake Bay is an ideal record by Cooper and Brush 1991 has led to more case study for the analysis of both sustainability clear-cut conclusions). However, in certain areas, and environmental management policy in the such as the meschaline region of the Patuxent United States. For the better part of two centuries, River, changes were dramatic enough to raise it has played an important role in policy develop- substantial concern (Heinle and others 1980). With ment, legislation, and litigation (Capper, Power, respect to the second goal, the disappearance of and Shrivers 1983; D'Elia 1987; Malone and oth- benthic vascular plants was generally attributed ers 1993). Of the present major environmental totheeffectsofnutrientenrichment.Finally,with concernsfor thebay(see, forexample,Mackiernan respect to the third goal, it was readily apparent 1990 for more details)-effects of nutrient enrich- that most phosphorus came from point sources, ment, lossof livingresources, destructionof habi- at a reasonably constant rate year-round, while tats, and toxification-the effects of nutrient en- nitrogen came from nonpoint sources, especially richment have received the most attention be- during the high-flow period of the spring. cause they loom the largest in terms of degrada- Increases in nutrient loadings derived from tion of environmental quality. effluent and runoff were typically considered to Hypoxia is the most significant ramification of have af fected the duration and extent of estuarine nutrient enrichment. Newcombeand Horne(1938) hypoxia, but specific mechanisms were not well first documented hypoxia in the deep channel of described or understood. Accordingly, consider- the Chesapeake Bay's main stem during the sum- able argument arose about appropriate strategies mer. In 1938, fisheries yields in the estuary were for reducing nutrients and, in particular, over still high, and the public was not concerned about whether nitrogen or phosphorus or both should water quality. However, by the early 1970s, bur- becontrolled(D'Elia,Sanders,andBoyntonl986). geoning population, urbanization in the water- Sincemost nitrogencame from nonpointsources, shed, and apparent declines in water quality and and most phosphorus came from sewage efflu- fishery yields raised concern over the bay's envi- ent, and considering EPA's policy to avoid ad- ronmental quality. Problems related to nutrient vanced wastewater treatment other than the re- enrichment in freshwater reaches of the bay's moval of chemical phosphorus, initial manage- tributaries, such as the Potomac River, had re- mentplansemphasizedcontrolofthepointsource sulted in an earlier focus on the construction of a and phosphorus. Most agencies did not recog- sewage treatment plant in the Washington, D.C., nize the importance of understanding nutrient metropolitan area, but the early 1 970s marked the cycles or estuarine trophic dynamics. Especially first serious public outcry for attention to issues unrecognized was the role of benthic sediments affecting estuarine portions of the Chesapeake as a site for denitrification and seasonably vari- system. Accordingly, the U.S. Congress directed able regeneration of phosphorus and nitrogen the U.S. Environmental Protection Agency (EPA) (see, for example, Boynton and Kemp 1985). to initiate a study focusing on the main stem of the Nonetheless, scientific progress wasrapid, and bay,and theChesapeake Bay Program wasestab- studies of nutrients and trophic dynamics soon lished with an executive council composed of the improved our understanding of the ecological governors of Maryland, Pennsylvania, and Vir- function of the bay and suggested that the inputs ginia, the mayor of the District of Columbia, the of both nitrogen and phosphorus should be re- administrator of the U.S. Environmental Protec- duced substantially. (Two recent summaries of tion Agency, and, beginning in 1987, the chair of key scientific findings relating to nutrient enrich- the Chesapeake Bay Commission. ment and hypoxia provide additional details rc- The EPA study had several early goals: (1) to garding the basic understanding of system func- document historical changes in water quality in tion achieved in the 1980s: D'Elia and others 1992; the bay, (2) to understand the cause of the disap- Harding, Leffler, and Mackiernan 1992.) In 1987, pearanceof submerged vascularplants,and(3) to the executive council of theChesapeake Bay Pro- estimate nutrient inputs from point and nonpoint gram established a goal of a 40 percent reduction sources. With respect to the first goal, unfortu- in the inputs of both nitrogen and phosphorus to nately, even for such a well-studied body of wa- thebay,adecisionbasedasmuchoninstinctason ter, the data recorded were too sparse and incon- early results of a complex mathematical model. 164 Sustainable Development and the Chesapeake Bay: A Case Study Few doubt the boldness of the executive meet theirown needs"(WCED 1987, p. 8). Mead- committee's goal, for it has huge economic rami- ows, Meadows, and Randers (1992, p. 209) have fications. However, recent EPA calculations on defined a sustainable society as "one that can nutrient inputs are sobering in their suggestion persist over generations [and is wise enough] not that more than 80 percent of the nutrient inputs to to undermine either its physical or social systems the bay may ultimately be considered anthropo- of support."ArecentbillbeforetheU.S.Congress genic (D'Elia and others 1992; Thomann and others referred to sustainable use simply as "using re- 1994; U.S. Environmental Protection Agency sources at rates within their capacity for renewal." 1992). For purposes of thischapter, which seeks to Regardless of whatever definitional criteria understand ecological sustainability from a bio- might be used, clearly neither the scientific com- physical perspective, however, it is an interesting munity nor the lay public regards the present exercise to consider how this decision relates to condition of the Chesapeake system as some- sustainability for the Chesapeake Bay. thing that onedesires to sustain. In fact, thepublic will is to remediate and restore the bay to a previous idyllic (and not well-defined) condition Sustainability and the Chesapeake that is beautiful to observe, bountiful in its sea- food harvests, and well oxygenated, while being In considering sustainability from a biophysical totally accessible to anyone who wants to use it perspective, this chapter focuses primarily on the for any purpose. This is what people really want ecologicalaspectsoftheChesapeake,whilestress- to sustain-in effect, an oxymoron! ing that virtually any vision of ecological A very worrisome aspect of sustainability is sustainability has human dimensions and is in- that, in general, many people appear to want evitably related to human uses of ecological sys- sustainability to pertain to a previous ecological tems. Accordingly, thinking of sustainability in condition that may or may not have existed and strictly scientific terms is merely an academic for which we have relatively little quantitative exercise, for clearly the issue of sustainability is information to describe. A goal of ecologists related primarily to managing local impacts of should be to help the public understand (1) what humans rather than global impacts or those of arereasonablepossibilitiesgivencompetinguses, natural events. Inevitably, decisions about what (2) what measures (and costs) will be required to system states should be sustained must be made achieve those states, and (3) whatlimitationsexist by public consensus, which one hopes occurs onexploitation.Thisislargelybeyond thepresent with the advice and counsel of scientists. One also state of the art. hopes that scientists are willing to play an active role in this process. The tendency even for scientists to view Sustainability and growth sustainability in terms of human uses is immedi- ately evident to anyone who reviews definitions The Chesapeake and its watershed have ben- that have been used. Fisheries scientists have for efited from having generally enlightened leaders yearsused sustained yield to indicatemaintaining who consider the consequences of population constantharvestswithoutdepletingthebreedingstocks growth and development. The executive council necessary to replenish what is removed. Within a of the Chesapeake Bay Program commissioned a larger context, numerous papers have offered study of population growth and development in definitions of sustainability (for example, the Chesapeake watershed for thirty years hence Shearman 1990). In formulating its Sustainable (Year 2020 Panel 1988). The report did not ad- Biosphere Initiative, the Ecological Society of dress sustainability per se or attempt to identify a America (Lubchenco and others 1991) has used humancarryingcapacityforthewatershed,butit sustainability "to imply management practices did reach conclusions that undoubtedly apply to that will notdegrade theexploited systemsorany these issues. The U.S. EPA Chesapeake Bay Pro- adjacent systems" (after Tumer 1988, p. 394). gram (C. Spooner, personal communication) esti- From an even more human-use perspective, mated that population will continue to grow at a the World Commission on Environment and De- substantial ratein the next threedecades (seefigure velopment has defined a sustainable society as 12-3).SimilarestimatesbytheNationalOceanicand one that "meets the needs of the present without Atmospheric Administration (Culliton and others compromising the ability of future generations to 1990) project comparable increases. 165 Figure 12-3: Population in Chesapeake Bay Watershed, 1950-90 and Projected until 2020 ' 15 cIx 10 CL. 1950 1960 1970 1980 1990 2000 2070 2020 Year Note: Population includes residents of the District of Columbia, Maryland, Pennsylvania, and Virginia. Sustainable Development and the Chesapeake Bay: A Case Study Two things especially caught my attention as I nize the implications of this in microcosm, for it reread the report of the Year 2020 Panel. First was applies more broadly to our global situation. In the acknowledgment that how the land is used is essence, human population growth and uncon- abasic factorin the ecological healthof the Chesa- trolled development have disrupted our natural peake Bay. This observation is important because systems to the point where increasingly draconian it provides an implicit rationale for applying the controls are necessary to mitigate what humans analytical tools and principles of landscape ecol- do to the environment. In essence, if we do not ogy to understanding the bay's response to its check our rate of population growth, and we watershed. Second was the recognition that with cannot individually undertake our social respon- 2.6 million new people by the year 2020, the states sibility to live in concert with nature, we will have and their local governments need to adopt a more to preempt our own freedom through increased highly integrated approach to planning and to government regulation. Even if we were able to directing and managing growth. This is impor- adapt our institutions and technology to obviate tant because it connects growth of the human the effects of population growth, this will have to population with environmental decline and be- be done at the expense of individual freedoms cause it implies that the result of such growth and increased government regulations. must inevitably lead to increased government While sympathetic with concerns about exces- regulation and control. sive government regulation, I am truly amazed FortheChesapeake,twophasesofhumanuses that certain political elements of society, who of the land apparently affect the water quality of ostensibly cherish individual freedom so greatly, the bay. In the early colonial era, the activities of havesuchadifficulttimecomingtogripswiththe relatively few individuals had large impacts, prima- effect that uncontrolled population growth is rily in the form of substantial deforestation and havingonindividualfreedombypromotingmore clear-cutting for agricultural purposes. While the stringent regulation. On the other side of the changes are observable in the sedimentary record, political spectrum, although I appreciate that the the impacts then were much less severe than one distribution and use of resources are both inequi- might expect today for the same changes in land table and unfair, I cannot imagine how, even in a use because fertilizer application, among other fairer world, we can sustain exponential increase things, was not then the factor it is now. in population at the same time that we reduce our Withsignificantreforestation in thebasin from use of resources and our impact on the environ- the late nineteenth century until recently, some ment. improvementsoccurred in thequalityof the bay's water. This reforestation came not because of concern for environmental quality or apprecia- Monitoring and understanding sustainability tion of the effect of land use on water quality, but instead because of industrialization and urban- From a scientist's point of view, what might be said ization. That the early effects of industrialization about the present status of the bay? What particular were manifested in improved environmental measurements are helpful in understanding the quality of the bay is somewhat ironic. status quo and thusbear on sustainability? How, in Recent trends toward degradation of water fact, can one consider the issue of sustainability quality are probably more insidious and difficult without addressing the status quo or understand- to reverse. Present concerns over the quality of ing how the system functions? water relate more to the cumulative impact of many In scientific terms, we must at least understand the individuals, which will be much more difficult to basics of the system, its processes, and its reactions to reverse because of the sociological complexity perturbations before we can say anything about involved with changing the behavior of a large sustainability.Iknowofnoquickindexestomeasure number of people. ecologicalsustainabilityinland-marginecosystems The 2020 report explicitly recognizes the con- no substitute exists for proper research and moni- sequences of human demographic change and toring. This section first explains briefly why such land use. It also recognizes that the only rational research and monitoring are necessary and then response for government is to plan land use more describes several aspects of the Chesapeake Bay wisely and to develop regulations and controls Program's monitoring effort that contribute to the that will cause development to affect the water- program's potential success in assessing shed less. It is important for everyone to recog- sustainability. 167 Defining and Measuring Sustainability: The Biogeophysical Foundations Scientifically designed monitoring IMPORTANCE OF LONG-TERM DATA Programs such as the Long-Term Ecological Re- Among the many ironies raised in this chapter, searograms suchR Togram the liahan Lo m E ia]Rep- none is more perplexing than the fact that so search (LTER) program (Callahan 1984) havepro- little of the huge amount of environmental data vided more than a decade of long-term data on collected is applicable to or useful for deter- selected ecosystems. The particular advantages mining long-term trends. Ecologists are accus- of such programs derive from their continuous, tomed to understanding the need for long-term long-term,and scientificallyoriented design. Data medeligunderstarch nusor long- are scrutinized as they are collected and consid- modeling and research (Magnuson 1990); envi- ered in perspective with other relevant informa- ronmental managers and policymakers often tion. These management features assure that the are not. Monitoring is rarely oriented toward qualtyo theatashigandthat the moni toring solving problems or asking questions, and this quality ofuthedata isphigh andthttemniori ng causes two problems: the persons or agencies done is coupled with process-oriented research who collect data feel no sense of responsibility and is driven by questions about ecological func- or ownership, and they do not assess the qual- tionand aboutchange. Althoughprogramssuch ity of the data produced. as LTER are generally hailed as high-quality Other problems with monitoring also pre- scientific programs, they are also roundly criti- vent the acquisition of reliable long-term data. cized because they appear to eschew issues of Standard procedures are not applied, or they practical importance to managers. In contrast, are misapplied (D'Elia, Sanders, and Capone monitoring programs conducted by govern- 1989). Littleattentionispaid to improvingmeth- mental agencies are often criticized because odologies or to understanding matrix effects. theylacka scientificapproach. TheChesapeake Quality assurance and quality control proce- Bay Program's monitoring effort has achieved dures are often disregarded or, in the other a successful balance between scientific quality extreme, are arbitrary and bureaucratic. The and practical problem orientation (although uses of data and the quality or grade of data are scientistsarealwaysconcerned thatnotenough rarely considered. We have research-grade process-oriented work is being done). chemicals and technical-grade chemicals-each useful for a particular application-yet para- SEnARATiNGNATURALANDANT-itROriOGENICEFFECTS doxically, we do not have grades for data. Far Scientifically driven monittoring and research too much of our data has been collected in such programs seek to differentiate different caus- mindless exercises as Section 302e (Clean Wa- ative agents of ecological effects (for example, ter Act) reporting. change in climate, anthropogenic effects, and local or regional perturbations). Few people PROCESS-ORIENTED MEASUREMENTS outside the scientific community understand Understanding system variability alone, al- the difficulties scientists face in differentiating though necessary, is not sufficient. Rates of natural, large-scale (usually climatic) effects ecological processes must also be understood, and anthropogenic, local e'fects. This is some- for these processes control state variables. For what ironic, since personal experience provides many ecosystems, studies of the regulation of enough evidence to the contrary: everyone ex- primary production, of the transfer of carbon periences the variability inherent in weather, through the trophic system, and of the cycling for example. We are all susceptible to being andassimilationofnutrientshaveprovenespe- surprised by heat waves, cold spells, droughts, cially valuable. In the case of the Chesapeake, floods, and storms, and we tend to view every in addition to research studies supported by event as unusual or as the harbinger of a global change in climate. This sort of inherent human such as the National Science Foundationl Na- myopia challenges the scientific community suchas thea atn Ascen cFudaion,sNa- continually to remind policymakers and the tional Oceanic and Atmospheric Administra- public (if not ourselves, too) not only of the tion (Sea Grant and the Coastal Ocean Pro- nature of variability but also of the inherent gram), Fish and Wildlife Service, and EPA, the difficulties in distinguishing natural from an- basic monitoring program includes process- thropogenic causes of change. It also requires oriented monitoring for sediment nutrient us to recognize the value of and promote the fluxes, primary productivity, and other key gathering and analysis of long-term data. functions of the system. 168 Sustainable Development and the Chesapeake Bay: A Case Study Major indexes in monitoring index is determining the rate and extent of change in habitat types. Until recently, this had Monitoring programs aveity fosedaon to be done by exhaustive in situ sampling and colctn inomto navait fadr observation. Recent advances In remote sens- topics. These programs have met with mixed ibsadVata mngent usin geographic success, often for institutional reasons. Manage- information systems offer substantial promise ment agencies have for years pursued the Holy in providing both real-time and long-term in- Grail of a singlegeneral index of ecological health formain on changsin abitat. or condition that can be used to assess human impacts. Although I do not arbitrarily exclude the TROPHIC STATE AND WATER QUALITY possibility that this can work, in a realistic sense, Probably the best overall indicator of I do think such a goal is unnecessary and imprac- sustainability for the Chesapeake is trophic state tical. However, some practical indicators of hu- (Harding, Leffler, and Mackiernan 1992; Smith man impact do exist that require some under- Leffler, and Mackiernan 1992; i a th, Leffler, and Mackiernan 1992; Taft and others standing of ecosystem function, and several of 1980). The expression of trophic state is deter- these are discussed next. mined by nutrient inputs, freshwater flow, tidal height, basin moi phology, and physical oceano- LIVLnoGRESOURcES(ESPECIALLYFIS-ERISANDI-ABITAT) graphic conditions (Sanford, Sellner, and We are notoriously poor at managing fisheries Breitburg 1990; Seliger, Bogg, and Biggley 1985; worldwide, and the Chesapeake is no exception. Webb and D'Elia 1980). Present evidence sug- Fisheries are extremely variable, and statistical gebb at higher nutien input hvenle thgh reporting of catch is often inconsistent or unreli- levels and growth rates of phytoplankton biom- able. Figure 12-4 provides examples of data from ass (Malone 1991, 1992; Officer and others 1984 catches of oysters, blue crabs, and striped bass Seliger, Bogg, and Biggley 1985), a shift to a that illustrate this. Landings of oysters are at microbiaggy d nd deco8poshifood w historical lows. There is considerable debate over (Jonas12 doss o e sor and hab whether yields are down because of degrading t(Jonas 1992), and loss of oving resource and habi- habiat nd ate qulit or verishng.Strped tat (Breitburg 1992). Certain components of es- has,toabitatad a er q salitysorcoierfishinog. Srp tuarine and coastal systems, such as the benthos, bass, too, have exhibited seriousdeclines, although provide help in integrating time-variable re- there is recent evidence that stringent manage- ponie hus cn provin imporTanteclue ment actions have been dramatically successful. tsponses and thus can provrde an M mportant clue Similar debate exists about the cause (in this case e yatica s aroaches torunderstndn thi overarvstig ad dstrctin o haita). lue ematical approaches to understanding trophic overharvesting and destruction of habitat). Blue interactions (Baird and Ulanowicz 1989) offer crabs have not shown serious declines as yet, but promise for analysis of trophic webs that may landings are extremely variable. y To complicate the issue, it is also proving par- prove to be useful for assessing sustainability. ticularly difficult to manage fisheries data. Juris- Chesapeake Bay program monitoring dictional boundaries, problems in obtaining ac- curate catch statistics from commercial and recre- Monitoring approaches in which oxygen, chloro- ational fishing interests, and political interests phyll, nutrients, and other traditional, related cause further complications. Accordingly, even variables for trophic state are determined within though public pressures to do so may be strong, a specified temporal and spatial matrix have managers should resist using harvestable species proven enormously useful for attaining a basic as key indicators of sustainability, because they understanding of how the Chesapeake system are not, as a practical matter, sustainable and functions. Four factors, in particular, account for because fishery recruitment, natural mortality, the success of this approach. First, the methods and so forth are intrinsically highly variable. used are by and large reliable, sensitive, and Living resources such as habitats are another appropriate for the saltwater sample matrix. Sec- issue, however, and indexes of quality and ex- ond, quality assurance and quality control are tent of habitat such as marsh, nontidal wet- appropriately applied to ensure that samplesare lands, sea grass cover, and so forth are poten- collected and analyzed properly and that data are tially important indicators of sustainability that recorded correctly and without error. Third, pro- integrate over at least several years of time. The fessional-level staff at state and federal agencies most difficult problem in using habitat as an review the data continually and analytically to 169 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 12-4: Landings of Oyster, Crab, and Striped Bass in the Chesapeake Bay, 1940-90 108II I I 1 eggS_; .. as*O * ysters S.~ ~ IE.I 10 1930 1940 1950 1960 1970 1980 1990 Year assess their meaning (see, for example, Magnien, for concentrations of nitrate in surface and deep Austin, and Michael 1991). Fourth, monitoring is waters in the bay over a seven-year period. It is complemented by process-oriented research that obvious even to the relatively uninitiated ob- helps to put monitoring data in perspective of server that patterns exist in the availability of functional aspects of the system. In many re- nitrate with time and location on the bay. These gards, this approach is embodied also in the Na- patternsvary considerably. When combined with tionalScienceFoundation-supportedLong-Term data for other key parameters, when viewed in Ecological Research Program (Callahan 1984). context with research on nutrient cycling and Because most monitoring programs do not com- input, and when used critically to verify math- bine these essential elements, and because they ematical models based on realistic conceptual are not sustained for long periods of time, their models, such data offer powerful opportunities utility is limited for understanding how a system to understand variability in the ecosystem's state functions and for differentiating natural from variables: clearly, if one is to understand anthropogenic effects. sustainability, one must also understand vari- With respect to the kinds of comprehensive ability. Practical problems exist, however, in sus- data now collected for the Chesapeake Bay, fig- taining the fundingand effort necessary toobtain ure 12-5 provides an example of a time-space plot this sort of information. 170 Sustainable Development and the Chesapeake Bay: A Case Study New approaches to monitoring interest. Geostationary satellites are constrained to high orbits, so in any case, orbiting satellitesare Mersonitl.Orng istnsiver andreuiescmany typically employed for high-resolution color-sens- peersonnel. One must consider not only the costs igcvrg.Acrigy o ealdtmoa of monitoringbutalso thehumanresourcesavail- coverage. Accordengly, for detailed temporal able to collect and analyze information. It is there- rote sesn us ing intdb sisathe foreextemey imortnt o dvelo moitoing remote sensing using instrumented buoys is the fore extremely important to develop monitoring alternative. The state of the art is developing well techniques that are cost-effective and offer rapid for this method of remote sensing, given our and facile assessments of wide geographical ar- present capabilities to develop detectors for nu- eas, in as close to real time as possible. merous parameters such as oxygen and chloro- Remote sensing is becoming an increasingly phyll fluorescence and our ability to store data important tool for monitoring. Although it oper- using microelectronic technology or, better yet, to atesprimarilyat the experimental level at present, telemeter data back to the laboratory. Since sam- it will inevitablybecome a major tool for environ- pling is seldom done over a twenty-four-hour mental monitoring in the next several decades. period, and we are often remiss in considering For example, remote color sensing of estuaries factors that occur at night, instrumented buoys and coastal regionsby satellite- and aircraft-borne do offer particular opportunities. instrumentation has already demonstrated its potential for understanding temporal and spatial variations in primary productivity. Although in Crucial uestions situ sampling from ships has the capability of q providing depth-integrated estimates of chloro- Iwouldliketoconcludeb presentingaseriesof phyllbiomassatselected sites, spatialand tempo- y ral coverage is poor and cost is high. Satellite- questions that will be of considerable importance borne sensors such as Landsat Thematic Mapper, if sustainability, by virtually any definition, is to Spot, and the proposed SeaWiFS and Modis in- be achieved. Some of these questions are rhetori- struments offeropportunities for obtaining excel- cal; some will require additional research and lent estimates of primary productivity with wide study; and some will depend on the ability of spatial and temporal coverage. In the absence of scientists, policymakers, and others to provide satellite-borne sensors, which is likely to occur education and leadership to the public. with continuing emphasis on manned missions Is sustainability realistic in the face of population to space, excellent sampling coverage can also be growth? This is a question that few dare to ask. obtained using aircraft-borne sensors such as The issue of population growth is a political hot ODAS (Ocean Data Acquisition System). Figure potato.Fewpoliticiansoneithersideof thepoliti- 12-6 shows an example of the estimation of phy- cal spectrum want to confront this issue head on. toplankton biomass obtained for the Chesapeake Meadows and others (1992) have reevaluated BayusingtheODASsensor.Thistypeofinforma- their original Limits to Growth scenario because tion, when obtained routinely and managed us- some of the dire predictions made did not come ing new computer techniques such as geographic true in the 1980s. Even though it is difficult to put information systems, offers enormous promise limits on results of models developed to estimate for determining the response of phytoplankton sustainability in the face of population growth, it both to nutrient inputs and physical factors and seems obvious that the ultimate test of accuracy forassessingchangesinhabitat(Harding,ltsweire, of the most dire predictions-an environmental and Esaias 1992). Armageddon-is the last thing the modelers want Satellite,aircraft,and shipboard measurements, to see. But exponential growth has inexorable although capable of offering wide spatial cover- consequences: the worst isthatacceleratingchange age, are necessarilylimited in temporal coverage. occurs and that negative effects can thus be ex- Clearly, for ships, only so many stations can be pected to occur at an accelerating pace. Must we occupied per cruise, and continuous operation at wait to verify the most dire predictions of models any single station is extremely costly when the by experience? daily costs of a ship start at a minimum of $1,500 Can technology always provide the solutions? The a day. For satellites using passive methods to public has an inordinate trust in technology and collect color data, coverage is limited to light human inventivenessasa countermeasure tobad periods, when satellites pass over the area of judgment or intemperate use of resources. 171 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 12-5: Time-Space Plots of Nitrate Concentration in Chesapeake Bay Program Monitoring Study, 1984-91 SURFACE LAYER NITRATE + NITRITE (mg/l) N 2 a JAJ&4 4V X J ** JJLao 7 JAJ7 44 JAzA AWN AIM JAb'M BOTTOM LAYER NITRATE + NITRITE (mg/1) N 2 AAM4 W jL8, AI U j .Le J te JAue Source: Data from the Maryland Department of the Environment, Chesapeake Bay Program, Water Quality Monitoring Program. 172 Sustainable Development and the Chesapeake Bay: A Case Study Figure 12-6: Concentration of Chlorophyll in the Chesapeake Bay, April 23, 1990, as an Indicator of Phytoplankton Abundance A_-__23-Apr-1 990 Chl [mg/m3] 37.5 j: __ _ _-6 37 0 *t.~l~ s-Aj1 60. 40. 39.0 7 30. 25. 1~ ~ ~ ~ ~~~~7 I ~~1 6. 12. 37.5 6. -v`~~~~~~- 37-0~~~~~~~~~~~~~~~r 77.0 76.5 76.0 75.5 Source: Data were collected by Harding, Jr., ltsweire, and Esias (1992) using aircraft remote sensing with the National Aeronautics and Space Administration's Ocean Data Acquisition System. 173 Defining and Measuring Sustainability: The Biogeophysical Foundations Is thereadequatepolitical will? (Or better,are there easy answers at present and am not optimistic sufficient economic resources?) If technology is the that this situation can improve in a more popu- answer to our sustainability problems, we must lous world. also have the economic wherewithal to afford to use it. In the next two decades, federal and state governments in the United States will be sorely Conclusions pressed by demographic changes that will put great demands on social spending, even if there is Considerable and long-term interest in the Chesa- a political will to deal with the effects of human peake Bay and its watershed have provided op- impactonsuchresourcesasthebayanditswater- portunities for in-depth scientific research and shed. Clearly, other coastal waters besides the scientifically oriented monitoring. From these Chesapeake Bay are subject to rapidly increasing activities, we are obtaining a comprehensive un- population and its effects (Culliton and others derstanding of the ecological functions of the bay 1990). Tertiary treatment of sewage, for example, and its natural variation and anthropogenic per- is expensive in terms of initial capital investment turbations. Such activities are essential for en- but also in terms of operation and management. lightened understanding and management of a If sustainability of environmental quality in the complicated system. Many would view face of growth requires technology, then will the sustainability for the bay in terms of a few eco- requisite funding be there to sustain technology? logical indicators: (1) reduced nutrient enrich- We may need to think of management solutions ment and hypoxia, (2) improved and bountiful in terms of sustainable (or appropriate) technol- fishery yields, and (3) improved habitat, prima- ogy (Schumacker 1973). We will have to be care- rily for rooted aquatic vegetation. ful not to buy what will be too expensive to For the Chesapeake, how we use the land sustain in the future. We must find ways of set- appears to hold the secret to sustaining a desir- ting priorities for environmental mitigation that able ecological state in the bay itself. Although we prove most cost-effective. Are we even close to may, with difficulty, have some success in sus- being able to provide the economic analysis nec- taining the trophic state and yields of the system essary to do this, much less to see wise decisions as a whole, it is hard to see how we can sustain implemented through the public process? historic yields on a per capita basis. Can the public be educated adequately? Many of Withouta public will tounderstand and recon- the pollution abatement issues we must deal with cile the issue of what is sustainable by almost any will require a cooperative and educated populace. criteria in the face of human population growth, Land use, refuse and sewage disposal, and other all speculation about sustainability (by any defi- issues will require enthusiastic public support. nition) amounts to little more than an academic Can we improve the links between science and man- exercise. Leaders in science and public policy agement?Scienceisconsiderecdbymanytobesuper- must redouble their efforts to promote public fluous to the public process of environmental man- understanding of this problem and to find ways agement.Manyscientistsarercluctanttoeniterwhat for controlling population that are effective and they see primarily as a political arena. Yet, as never acceptable to the largest percentage of people before, science must be part of the problem-solving possible. The challenge is awesome and the topic process. Ourability to sustain ourselves as a human controversial-it is also the core issue relating to race will ultimately require a better interaction be- sustainability on Earth. tween science and nmanagement. Will weaccept morecontroland tighter regulation? With more people and more impact, tighter con- References trol and regulation seem inevitable. At what point do fundamental constitutional issues relating to Baird, D., and R. E. Ulanowicz. 1989. "The Sea- individual freedoms (life, liberty, and the pursuit sonal Dynamics of the Chesapeake Bay Eco- of happiness) come in conflict with the regula- system." EcologicalMonographs 59,pp.329-64. tions necessary to sustain a healthy environment? This is ultimately the most difficult question to Boynton,W.R., andW.M.Kemp.1985."Nutrient answer and the one that argues most strongly for Regeneration and Oxygen Consumption by humanity's need to self-limit its population Sediments along an Estuarine Salinity Gradi- growth and its use of natural resources. I sec no ent." Marine Ecology Program Series 23, pp. 45-55. 174 Sustainable Development and the Chesapeake Bay: A Case Study Breitburg, D. L. 1992. "Episodic Hypoxia in the phasis on Effects of Enrichment." U.S. Envi- Chesapeake Bay: Interacting Effects of Recruit- ronmental Protection Agency, Chesapeake Bay ment, Behavior, and Physical Disturbance." ProgramFinal Report,GrantR806189010.Pub- Ecological Monographs 62, pp. 525-46. lication 84. Chesapeake Research Consortium, Callahan, J. T. 1984. "Long-term Ecological Re- Inc., Annapolis, Md. search." BioScience 34, pp. 363-67. Jonas, R. 1992. "Microbial Processes, Organic Capper, J., G. Power, and F. R. Shrivers, Jr. 1983. Matter, and Oxygen Demand in the Water Chesapeake Waters: Pollution, Public Health, and Column." In David E. Smith, Merrill Leffler, Opinion, 1602-1972. Centreville, Md.: Tidewa- and Gail Mackiernan,eds., Oxygen Dynamics in ter Publishers. the Chesapeake Bay, pp. 113-48. College Park, Cooper, S. R., and G. S. Brush. 1991. "Long-term Md.: Maryland Sea Grant College. History of Chesapeake Bay Anoxia." Science Lubchenco, J., A. M. Olson, L.. B. Brubaker, S. R. 254, pp. 992-96. Carpenter, M. M. Holland, S. P. Hubbell, S. A. Culliton, T. J., M. A. Warren, T. R. Goodspeed, D. Levini, J. A. MacMahon, P. A. Matson, J. M. G. Remer, C. M. Blackwell, and J.J. McDonough Melillo, H. A. Mooney, C. A. Peterson, H. R. Ill. 1990. "Fifty Years of Population Change Pulliam,L.A."eal,P.J.Regal,andBP.G.Risser. along the Nation's Coasts." National Ocean 1991. "The Sustainable Biosphere Initiative: Service, National Oceanic and Atmospheric An Ecological Research Agenda." Ecology 72, Administration, U.S. Department of Com- pp. 371-412. merce, Rockville, Md. Mackiernan, G. B. 199J. "State of the Chesapeake D'Elia, C. F. 1987. "Nutrient Enrichment of the Bay." Water Environment and Technology 9, pp. Chesapeake Bay: Too Much of a Good Thing." 60-67. Environment 29, pp. 6-11, 30-33. Magnien, R. E., D. K. Austin, and B. D. Michael. D'Elia, C. F., L. W. Harding, Jr., M. Leffler, and G. 1991. "Chemical/Physical Properties Compo- Mackiernan. 1992. "The Role and Control of nent: Level I Data Report." Maryland Depart- Nutrients in Chesapeake Bay." Water Science ment of the Environment, Special Projects Pro- and Technology 26, pp. 2635-44. gram Report. Annapolis, Md. D'Elia, C. F., J. G. Sanders, and W. R. Boynton. Magnuson, J. J. 1990. "Long-term Ecological Re- 1986. "NutrientEnrichmentStudiesinaCoastal search and the Invisible Present." BioScience Plain Estuary: Phytoplankton Growth in Large- 40, pp. 495-508. scale, Continuous Cultures." Canadian journal Malone, T. C. 1991. "River Flow, Phytoplankton of Fisheries and Aquatic Sciences 43, pp. 397-406. Production, and Oxygen Depletion in Chesa- D'Elia, C. F., J. G. Sanders, and D. G. Capone. peake Bay." In R. V. Tyson and T. H. Pearson, 1989. "Analytical Chemistry for Environmen eds., Modern and Ancient Continental Shelf An- tal Sciences: A Question of Confidence." Envi- oxia, pp. 83-93. Special Publication 58 of the ronmental Science Technology 23, pp. 768-74. Journal of the Geological Society (London). Harding, L. W., Jr., E. C. Itsweire, and W. E. . 1992. "Effects of Water Column Processes Esaias. 1992. "Determination of Phytoplank- on Dissolved Oxygen, Nutrients, Phytoplank- ton Chlorophyll Concentrations in the Chesa- ton, and Zooplankton." In David E. Smith, peake Bay with Aircraft Remote Sensing." Re- Merrill Leffler, and Gail Mackiernan, eds., Oxy- mote Sensing and the Environment 40, pp. 79- gen Dynamics in the Chesapeake Bay, pp. 61-112. 100. College Park, Md.: Maryland Sea Grant College. Harding, L. W., Jr., M. Leffler, and G. B. Malone, T. C., W. Boynton, T. Horton, and C. Mackiernan.1992.DissolvedOxygenintheChesa- Stevenson. 1993. "Nutrient Loadings to Sur- peake Bay: A Scientific Consensus. Technical Re- face Waters: Chesapeake Bay Case Study." In port. Maryland Sea GrantCollege, College Park, T. C. Malone, ed., Keeping Pace with Science and Md. Engineering, pp. 8-38. Washington, D.C.: Na- Heinle, D. R.,C. F. D'Elia,J. L.Taft,J. S. Wilson, M. tional Academy Press. Cole-Jones, A. B. Caplins, and L. E. Cronin. Meadows, D. H., D. L. Meadows, and J. Randers. 1980. "Historical Review of Water Quality and 1992. Beyond the Limits. Post Mills, Vt.: Chelsea Climatic Data from Chesapeake Bay with Em- Green Publishing Co. 175 Defining and Measuring S ustainability: The Biogeophysical Foundations Newcombe, C. L., and W. A. Horne. 1938. "Oxy- Taft, J. L., W. R. Taylor, E. 0. Hartwig, and R. gen-poor Waters of the Chesapeake Bay." Sci- Loftus. 1980. "Seasonal Oxygen Depletion in ence 88, pp. 80-81. Chesapeake Bay." Estuaries 3, pp. 242-47. Officer, C. G., R. B. Biggs, J. L. Taft, L. E. Cronin, Thomann, R. V., J. R. Collier, A. Butt, E. Casman, M. A. Tyler, and W. R. Boynton. 1984. "Chesa- and L. C. Linker. 1994. "Response of the peake Bay Anoxia: Origin, Development, and Chesapeake Bay Water Quality Model to Significance." Science 23, pp. 22-27. Loading Scenarios." Chesapeake Bay Pro- Orth, R. J., and K. A. Moore. 1983. "Chesapeake gram Office, U.S. Environmental Protection Bay: An Unprecedented Decline in Submerged Agency, Annapolis, Md. Aquatic Vegetation." Science 222, pp. 51-53. Turner, M. G., ed. 1988. Sustainable Environmen- Pritchard,D. W. 1955. "EstuarineCirculation Pat- tal Management: Principles and Practice. Boul- terns." Proceedings of the American Society of der, Colo.: Westview Press. Civil Engineers 81, pp. 1-11. U.S. Environmental Protection Agency. 1992. Sanford, L. P., K. G. Sellner, and D. L. Breitburg. "Progress Report of the Baywide Nutrient 1990. "Covariability of Dissolved Oxygen with Reduction Reevaluation." Chesapeake Bay Physical Processes in the Summertime Chesa- Program, Annapolis, Md. peake Bay." Journal of Marine Research 48, pp. Webb, K. L., and D'Elia, C. F. 1980. "Nutrient 567-90. and Oxygen Redistribution during a Spring/ Schumacker, E. F. 1973. Small Is Beautiful. Lon- neap Tidal Cycle in a Temperate Estuary." don: Blond and Riggs. Science 207, pp. 983-85. Seliger, H. H.,J. A. Bogg, and W. H. Biggley. 1985. WCED (World Commission on Environment "Catastrophic Anoxia in the Chesapeake Bay and Development). 1987. Our Common Fu- in 1984." Science 228, pp. 70-73. ture. Oxford, England: Oxford University Shearman, R. 1990. "The Meaning and Ethics of Press. Sustainability." Environmental Management 14, Year 2020 Panel. 1988. "Population Growth and pp. 1-8. Development in the Chesapeake Bay Water- Smith, D. E., M. Leffler, and G. Mackieman, eds shed to the Year 2020." Chesapeake Execu- 1992. Oxygen Dynamics in the Chesapeake Bay tive Council, Chesapeake Bay Commission, College Park, Md.: Maryland Sea Grant College. Annapolis, Md. 176 13 Restoration of Arid Lands G. Pickup and S. R. Morton People of European origins began settling in Australia 200 years ago, and they quickly realized that the vast grasslands and shrublands of the interior could be used for pasturing stock. Today, most of the arid zone is classed as rangeland, where the dominant land use is pastoralism. This chapter provides an overview of recent research designed to assist sustainable mannagement of the natural rescurces of and Australia. General characteristics of Australia's arid zone soil and nutrients by wind and water also creates areas of better soils. Even in a landscape that may As well as supporting an extensive pastoral in- appear flat and featureless, water and nutrients dustry, the Australian arid zone has high value as concentrate around bushes and trees or in gentle a tourist destination and is a uniquebiogeographi- depressions. These sites of accumulation are key cal region of substantial importance for nature to the productivity of an area and are therefore conservation. In addition, it is home to a large importantforthepersistenceofmanynativeplants proportion of the Australian aboriginal popula- and animals as well as for the pastoral industry. tion, and many of these people live on their tradi- Because these productive areas are easily over- tional lands and continue many of their original grazed and eroded, special emphasis must be land management practices. Of the total area of placed on their management. the rangelands (5.6 million square kilometers), 66 The Australian rangelands are of low produc- percent is pastoral land, 14 percent is aboriginal tivity, and pastoralism is conducted at low densi- land, 4 percent is in conservation reserves, and 16 ties of human population. Properties in Australia percent remains unallocated (see figure 13-1). (stations) typically range from 1 0,000 hectares up The principal feature of the arid Australian envi- to 30,000 square kilometers and carry free-ranging ronment is the highly unpredictable year-to-year sheep or cattle within five to forty fenced pad- fluctuation in rainfall. Since plant growth is largely docks. A single property frequently supportsonly deerindyvalblmis h p ctin ' one family and rarely more than ten. Land is determaned by availableymorethenpodcto leased from state or territory govemments; in of biomasscanvarybya factorofl0betweenyears. some cases, the leases are perpetual, but even Plant establishment occurs intermittently, perhaps where they are not, they apply for decades. Often once every five years or more, and drought is a the lessee is also the manager, although particu- natural part of the unpredictability. Above all else, larly in the northern subtropical rangelands com- sustainable management involves coming to terms panies often own leases and employ managers. In with this variability. all areas, land use is governed by covenants writ- In general, rangeland soils are highly weath- ten into the leases under legislation. ered and very infertile compared with soils of Despite the arrangements for use of land un- arid regions elsewhere on the globe (see Stafford der leasehold covenants in arid Australia, poor Smith and Morton 1990). Soil fertilitv i s predomi- understanding of management requirements in nantly linked to the underlying rock, but within the region has led to overgrazing by both domes- any particular type of rock, the redistribution of tic and wild animals. This overgrazing has been Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 13-1: Map of Australia Showing the Extent of the Rangelands and the Areas Devoted Principally to Grazing Sheep or Cattle Arid & semi- arid lands: pastoral land use _-_ Sheep Cattle the most extensive cUntributor to land degrada- Sustainable pastoral management tion. Intensive management and restoration of degraded land in the arid zone is frequently un- Although sustainability has entered the popular economic due to the vastness of the region and political jargon in Australia, a precise definition low land valuesperunit of area. Furtherdegrada- remains elusive (Commonwealth of Australia tion needs to be avoided; thus, sustainable land 1990; Hamblin 1991). Problems arise because management is more a matter of prevention than sustainability must be defined for a particular of cure. Our examination of sustainability begins type of operation and also for particular scales in with pastoralism because it is the most extensive time and space. For example, sustainable man- use of land; later in the chapter we discuss land agement for pastoral purposes may have differ- management approaches that might be necessary ent objectives from those of management forfauna when conservation of biodiversity is incorpo- and flora conservation. Further, goals for sustain- rated into the definition of sustainability. able management of a region will almost cer- 178 Restoration of Arid Lands tainly differ from those for individual properties equilibrium conditions, the system experiences or paddocks, and activities that are sustainable negative feedback that dampens the effects of over a few years may not result in century-long disturbance within a limited domain of attrac- sustainability. We believe that the term should be tion. In practical terms, this may mean that high used at the scales at which land is managed and grazing pressure reduces the production of for- should consider periods at least as long as a age, which, in turn, reduces attractiveness for decade (and usually longer). grazing. However, this situation occurs where Given specific spatial and temporal scales, sus- forage is largely produced continuously and is tainable management involves using a natural therefore more typical of areas with high rainfall system so that it can produce output at the end of than of arid rangelands. a specific time similar to what it produced at the The alternativ2 nonequilibrium approach as- beginning. This definition avoids issues such as sumes that ecosystem behavior is stochastic but whether the output is economic or biological, that an element of persistence is maintained. In a whether management inputs change over time, rangeland ecosystem, persistence implies a con- and whether the type of production can vary. If tinuing ability to grow plants even after the long- land is managed basically to produce livestock, estdroughtortheheaviestgrazing.Severalmecha- thena relatively narrow definitionof sustainability nisms have been lproposed to explain persistence may be used. For example, Pickup and Stafford in the face of the high probability of extinction Smith (1993) define sustainable pastoral use of that is inherent to stochastic domination land as thecommercial production of livestockon (DeAngelis and Waterhouse 1987). These include rangelands that, at a minimum, seeks to: the setting of upper aAd lower limits on process rates by factors such as limited water and nutri- *Maintain the long-term capacity of the ecosys- et rdsrprintl ag eutosi tem to produce forage from rainfall (although grazinireswreionlyarfe padataens the composition of that forage may change, as grazing inareas whereonlya few palatableplants may the short-term capacity to produce it). survive. They also include the fact that landscape variability, and the larger range of conditions * Produce acceptable financial as well as present at greater spatial scales, reduces the risk nonfinancial returns for the manager and of extinction in a whole system even though that dependents (the preferred standard of living risk is high at individual points within it. may include intangibles such as lifestyle). Recolonization from adjacent areas may also off- To make theseprinciples operational, we must set local extinctions. Table 13-1 summarizes the now determine how the natural system behaves main differences between the equilibrium and and how management can operate sympatheti- nonequilibrium views of how rangeland ecosys- cally with that behavior. It may then become tems function. possible to put current pastoral activity on a more Until recently, arid rangelands were treated as sustainablebasis and, through conservativeman- equilibrium systems. More recently, non-equilib- agement, to restore many of our grazed ecosys- rium concepts that allow for highly stochastic tems to a more productive condition (for ex- behavior have gained favor (Behnke, Scoones, amples, see Bastin 1991; Friedel, Foran, and and Kerven 1993; Ellis and Swift 1988; Westoby, Stafford Smith 1990; Lange, Nicolson, and Walker, and Noy-Meir 1989). The stochastic ele- Nicolson 1984; Purvis 1986). ment results from the fact that these ecosystems are driven by rainfall and therefore dominated by highly discontinuous abiotic processes. This ap- Functioning of pastoral ecosystems proach is something of an abstraction, because rangeland ecosystems can display elements of Two theoretical approaches to the functioning of both biotic and abiotically controlled behavior, grazed rangeland ecosystems have been devel- depending on the time scale (Stafford Smith and oped, but a rangeof intermediate possibilities lies Pickup 1993). At the short-term scale of the indi- between these two extremes (DeAngelis and vidual rainfall event or sequence of events, sys- Waterhouse 1987; Ellis and Swift 1988). The equi- tem behavior is dominated by abiotic processes librium approach assumes that the system exists and is highly stochastic. In the longer term, biotic in some fopr of homeostatic state that is main- factors may change the response of the system to tained by interactions among its componentsm In short-term abiotic forces such as rainfall. For ex- tamed b tample, erosion induced by grazing may reduce 179 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 13-1: Characteristics of Equilibrium and Nonequilibrium-based Grazing Systems Characteristics Equdibrium-based systems Nonequilibrium-based systems Abiotic patterns Abiotic conditions relatively constant Stochastic/variable conditions Plant-growing conditions relatively invariant Variable plant-growing conditions Plant-herbivore interactions Tight coupling of interactions Weak coupling of interactions Feedback control Abiotic control Herbivore control of plant biomass Plant biomass abiotically controlled Population patterns Density dependence Density independence Populations track carrying capacity Carrying capacity too dynamic for close population tracking Limit cycles Abiotically driven cycles Community 'ecosystem Competitive structuring of communities Competition not expressed Characteristics Self-controlled systems External forces critical to system dynamics Source: Ellis and Swift 1988. the soil's capacity to store moisture and thereby growth period, moisture becomes limiting and limit the amount of vegetation that grows as a plant activity is reduced. Much of the biomass result of subsequent rainfalls. The short-term re- produced during the growth period is then lost as sponse of vegetation is then increasingly domi- plants die, are consumed by grazers, are con- nated by antecedent biotic conditions, and the verted to litter, and decay. During the period stochasticelementassociated withabioticcontrol when moisture is available, plants grow faster is reduced. In resilient systems, these biotic ef- than they are consumed by grazers, so growth fects may be short-lived, and the system will and herbivory do not occur at the same rate or at recover. However, where resilience is biotically the same time. reduced,systemsmaylosetheircapacitytorecover. Vegetation pulse and decay behavior can be Traditional approaches to assessing described by a time-series model with an input sustainability of use employ the mix of plant series (the rainfalls), a response function, a decay species present as an indicator (Lauenroth and function (biomass consumption and decay), and Laycock 1989). However, species composition an output series (the amount of biomass present can be relatively insensitive to a range of ecologi- at a particular time). The response function de- cal processes such as soil erosion and can vary scribes the magnitude of the plant growth re- extensively with climatic factors (Friedel, Pickup, sponse to a given rainfall and, eventually, the bio- and Nelson 1993; Westoby 1980). Recent work logicalproductivityofthesystem.ltcanvarythrough also shows that the performance of grazing ani- timeinresponsetoabioticfactorssuchasasequence mals may be largely unrelated to the pasture's of rainfalls,eachcomponentofwhichoffersparticu- composition of species and that animal produc- laropportunities forgrowthand changes themixof tion in nonequilibrium systems is overwhelm- species present (Westoby 1980). It can also be sub- ingly influenced by total production of grass (Ellis ject to biotically induced change, for example when and Swift 1988; Hodgkinson 1992; Mentis and heavygrazingafteronevegetationpulsecausessoil others 1989; Wilson and MacLeod 1991). Under erosion and dampens the next pulse. these circumstances, the continuing ability of a Some examples of the behavior of landscape to produce a large quantity of forage rainfall-driven systems are shown in figure 13-2. from rainfall may be the best measure of sustain- If the rainfall in the time-series model maintains able pastoral use. its statistical characteristics over time and the In moisture-limited arid ecosystems, vegeta- response and decay functions do not change, the tion grows in a series of rainfall-generated pulses output series will maintain a given mean, vari- as production and reproduction occur (Friedel ance, and autocorrelation structure. Thus, the 1984; Noy-Meir 1973; Westoby 1980). After the stateofthesystemvariesintheshortterm,butits 180 Restoration of Arid Lands behavior in response to rainfall is persistent. Be- in erosion or damaged the soil in other ways, it havior is no longer persistent when the response may take a very long time for the response func- functionchanges,producingagradualor sudden tion to shift back to previous conditions, even at change in the output series. If the output series high rainfalls. In some types of soil, it may even continues to drift over time, the response func- require the deposition of new soil transported tion has become unstable. If the drift ceases, the from upslope or the addition of aeolian material response function is stable once more, but the for recovery to occur at all (Pickup 1985). Less productivity of the system has changed. serious changes in species composition may be Changes in the characteristics of the rainfall reversed by a suitable sequence of rainfall response function through time are normal be- (Westoby 1980). The type of change in response havior in rainfall-dominated systems. If they re- function that is most easily reversed occurs when sult from the sequence of antecedent rainfalls, the grazing reduces cover but leaves the seed banks, associated change in the production of forage store of moisture in the soil, and pool of soil cannot be equated with a deterioration of or im- nutrients relatively intact. The loss of soil mois- provement in the condition of the rangeland. If ture due to runoff associated with reduced veg- the production of forage is reduced as a result of etative cover will reduce the response of plants to grazing, then there is a change in condition, but rainfall, but good conditions for plant growth in its seriousness depends on how easily the system wet years or a reduction in grazing will allow can recover. For example, if grazing has resulted relatively quick recovery. Figure 13-2: Changes in Biomass through Time in a Rainfall-driven System That is Initially Stable 4 - .Stable - Drifting Stable Tirne Note: The response function begins to drift and subsequently restabilizes at a lower level of producvitNy. 181 Defining and Measuring Sustainability: The Biogeophysical Foundations Measuring sustainability Grazing gradients develop because animals in the biological system graze away from waterpoints and return to them at regular intervals to drink. The distance trav- The short-term variations in biomass in a eled varies with factors such as water salinity, rainfall-driven system make it difficult to sepa- availability of forage, meteorological conditions, rat naura chngefrom that induced by grazing. and the species, breed, and condition of the ani- rate natural change ifro to induedty grazing nmals themselves (Pickup and Chewings 1988a; They also make it difficult to identify shifts in the Stafford Smith 1984). There is, however,a general vegetation response function, whether they be decline in animal activity and impact on plants as natural or induced by grazing. It is, however, the distance from water grows. Some grazing possible to identify spatial patterns in the rate of gradients may disappear with the next major vegetative growth and total biomass that are ex- rainfall. Others remain for longer periods and clusively the result of grazing and thereby to indicatethatgrazinghashadamorelastingimpact. separate change induced by grazing from other Grazing gradients resulting from the activities types of variability. Furthermore, when these of either sheep or cattle can be detected by mea- patterns are examnined in association with behav- suring differences in ground cover on satellite ior of the system through time, they make it images acquired at different times (Pickup and possible to separate short-term change in the Chewings 1988b; Cridland and Stafford Smith productivity of vegetation from longer-term dam- 1992). Simple examples of grazing gradients in a age. These patterns are known as grazing gradi- paddock with a uniform type of vegetation are ents, and a technology for identifying and using shown in figure 13-3. When average cover is them to assess range condition has been devel- plotted against distance from water without any oped recently (Pickup 1989, 1992; Pickup and grazing, a straight line function should result, Chewings 1992). This technology uses data from indicating no changeacross the landscape. Where satellites, which can readily be converted into grazing occurs, cover is reduced, but this effect is measures of the amount of vegetative cover progressively smaller as the distance from water present (Graetz, Pech, and Davis 1988; McDaniel grows, until it is no longer discernible; this trend and Haas 1982). The use of satellite data also is shown by the line marked dry period. After solves many of the logistical problems encoun- rain, vegetation recovers, and the grazing gradi- tered in measuring grazing gradients with con- ent disappears. An increased level of cover then ventional ground-based techniques in highly di- exists over the whole paddock, as the upper line verse paddocks from tens to hundreds of square in figure 13-3 shows. Thus, the previous level of kilometers in size (Friedel 1990; Pickup 1989). grazing may be regarded as sustainable. Figure 13-3: Schematic Diagram Showing Temporary and Permanent Normal Grazing Gradients Full recoverv Partial recovery No degradation Some degradatibon Max_murn Maximum Wetpenod x index op i Wet period 0~~~~~~~~~~~~~~~~~~ Miuunmum Mu llmum 0 2 4 6 8 10 n 2 4 6 8 10 DLstance from water (kilometers) Distance from water (kilometers) 182 Restoration of Arid Lands If a grazing gradient does not disappear, even spatial variability in the landscape such as an after very large rainfalls, then the land has been erosion cell mosaic (see Stafford Smith and Pickup damaged. Many landscapes are not damaged 1993). This requires a set of special procedures in by grazing and recover fully after moderate which the variance of plant cover is compared at rainfall. Others recover only partially and re- different distances from water (Pickup and quire a sequence of large rainfalls for the graz- Chewings 1994). ing gradient to disappear fully. They still retain Not all grazing gradients involve a simple a limited capacity for recovery, therefore, even increase in vegetative cover with distance from though it may be many years before a suitable water. It is also possible to find inverse grazing sequence of large rainfalls occurs. The distinc- gradients in which cover decreases with distance tion between a temporary grazing gradient and from water except in the immediate vicinity of the a permanent one is consequently somewhat waterpoint (figure 134). These features indicate arbitrary. Pickup and Chewings (1992) classi- grazing-induced increases in the proportion of fied a permanent gradient as one that persisted unpalatable herbage species or shrubs close to during the best rainfalls of the past ten years, a water. The inverse gradient develops because period chosen because it represents the time for animals avoid the unpalatable cover and graze which archived Landsat MSS data are readily farther from water, where progressively more available in Australia. palatable forage is available. If this is the case, the Real grazing gradients can be very complex. inverse gradient should intensify with time since Where several types of vegetation occur in a the last rain because grazing progressively re- paddock, some may be particularly affected be- duces vegetative cover. The gradient may disap- causeanimalspreferthemoverothers(seePickup pear partly or fully after the next rainfall. Where and Chewings 1988b). The landscape must there- it does disappear, the capability of the landscape fore be subdivided into types. Perhaps the big- to produce vegetative cover in response to rain- gest problem in detecting a grazing gradient oc- fall has not been reduced, but the ability to pro- curs when it is superimposed on other types of duce usable forage has. Figure 13-4: Inverse Grazing Gradients Derived from a Landsat MSS Vegetation Cover Index for Two Land Systems in Central Australia (A) Inverse gradient (B) Composite gradient 200 200 190 150 190 1 190 180~ 140 180 ~~~~~~~~~180 C.. g 170 1305 170 -g_170 a3 a a_ 170 170 8 to Wetperiod EL A E E 160 - X 120 160 -s 160 $$ \ _ 150I | 150 150D 110 140 140 140 100 0 1 2 3 4 5 6 7 0 2 4 6 8 10 Distance from water (kilometers) Distance from water (kilometers) 183 Defining and Measuring Sustainability: The Biogeophysical Foundations Normal and inverse grazing gradients are rates but evades climatic downturns or droughts sometimes combined in the same type of land- byearlydecisionsonagistmentorsaleofanimals. scape to produce a composite gradient. In this The second strategy is essentially one of gam- situation, cover decreases with distance from bling against short-term climatic variability and, water up to a particular distance and then begins if errors are made, can result in a property man- to increase (figure 13-4). Composite gradients ager having too many sheep or cattle, which he that persist even in dry times usually develop cannot sell except at a loss. The temptation is then where there is an increase in shrubs or a buildup to retain them through a drought even though the of unpalatable species close to water. Composite pasture resource may not be adequate to support gradients that persist for only a short time after them. Land degradation can result. rain indicate a flush of ephemeral species close to If we accept that there is always some risk of water that is quickly destroyed by trampling and, land degradation, the best strategy is to manage possibly, grazing. These effects displace the nor- landscapesand pastoral operations in such a way mal grazing gradient and push it farther from the as to minimize it. This involves using monitoring waterpoint. data from satellites to identify the most vulner- In short, the technology of grazing gradient able parts of the landscape and then designing provides a measure of the extent to which the paddocks with fencelinesand waterpointslaid out capacity of an ecosystem to produce forage is to avoid heavy concentrations of aniimals in these being reduced by grazing. Variations on the pro- areas. Two techniques have been developed re- cedure make it possible to determine whether centlv in Australia toachievethisoperation:erosion quality as well as quantity of forage are affected forecastinlgand grazingdistributionmodeling. Used (Pickup 1994), and large areas can be surveyed together, they can provide low-risk farm layouts quickly and cheaply. Thus, changes in the indica- that will, in future, be coupled with herd economic tor variables make it possible to determine whether and stocking strategy models to allow more sus- pastoral management is sustainable or not. tainable use of land for pasture. Erosion forecasting is based on changes in the statistical and spatial characteristicsof land- Restoring and maintaining sustainability scapesthatoccurastheyprogressivelydegrade in grazing lands (Pickup and Chewings 1988b). The first step in makinig a forecast is to map erosion and depo- The effectsof grazing always involve some risk of sition in the area of interest from satellite imag- land degradation, however small. Removal of cry. For central Australian landscapes, this may vegetative cover can produce adverse changes in be done using a soil stability index whose value pasture species, increased runoff, and crosion of is low for eroded areas, high for deposition exposed soil. Trampling breaks down soil aggre- zones, anid internmediate for stable or inactive gates, making it easier for wind and water to partsof tlelanidscape(Pickupand Nelson 1984). transport them, encourages gullying by concen- The frequency distribution and spatial trating flow along paths and tracks, andi reduces autocorrelation function of this index change in infiltrationand capacity to hold moisturebyconm- a consistent mann1ier as land degradationi pro- pacting the soil. The risk of degradation is espe- ceeds, reflecting changes in thc behavior of cially high where rainfall is unreliable because erosion cells. Forecasting involves fitting a spa- vegetation may not be restored to bare areas for tial interaction modiel whose parameters de- long periodsduringdroughts, therebyexacerbat- scribe the variance and spatial autocorrelation ingerosionand furtherreducinig theability of the functioni of the soil stability index in an area. An landscape to recover when it does rain. inverse filtering operation is then carried out The grazing management strategies adopted with the parameters of the fitted model to ob- in Australia'srangelandidsiayexaggeratetherisk tain a noise or underlying pattern series. This of degradation. Stafford Smith and Foran (1988), series is filtered with the parameters of a simi- for example, have described two contrasting ap- lar spatial interaction modiel derived for a more proaches for cattle properties: one buffers itself eroded prototype area to produce a forecast. In against land degradation and income fluctuationi effect, this procedure tranisfers the varianiceandi by low use of pasture and high rates of produc- spatial autocorrelationi tunction properties of tion per animal; the other involves high-i stocking prototype areas to areas for which a forecast is 184 Restoration of Arid Lands required but does not change the location of Figure13-5:PredictedPatternsofGrazingaround erosion cells. This locational information is held Four Waterpoints where Vegetation is Uniform in the underlying pattern series. The method appears to be reasonably accurate, and tests for a range of landscape types are presented by Pickup and Chewings (1988b). Like erosion, grazing does not occur uni- formly across a landscape. Instead, animals graze out from waterpoints and use gradually decreases as the distance from water grows in a -- broadly sigmoidal shape. This produces a con- 3 centric pattern around each waterpoint in pad- docks with uniform vegetation (see figure 13- 5). Where more than one type of vegetation is present, which is usually the case in the large paddocks used in Australia, the concentric pat- tern becomes distorted into a star-shaped one with distinct corridors of heavier grazing in- tensity where certain types of vegetation are more favored. Distortions can also result from the availability or otherwise of shade and, in the case of sheep, the presence of overnight camping sites and the prevailing direction of the wind. The corridors are areas of intense _- trampling in which the network of tracks lead- ing to water can result in gullying. Complex models of animal physiology have been used to derive models of the distribution of grazing and trampling under Australian con- ditions (Stafford Smith 1984). These models have substantial data requirements and are fre- quently not suitable for operational use. Sim- pler procedures based on linear regression models for sheep (Stafford Smith 1988) and Note: Lighter shading indicates more-intense grazing. convection-diffusion analogs for cattle (Pickup and Chewings 1988a) are now becoming avail- able. These models have been successfully cali- brated from remotely sensed data using spatial Management for conservation patterns of change over time in Landsat MSS Band 5 as a surrogate for the removal of vegeta- Some brief background is required to show why tive cover by grazing. conservation management needs to be improved. The ability to calibrate grazing distribution Arid Australia exhibits a poor record of extinc- models quickly for a range of different types of tions and contractions of range among its native vegetation means that the effect of alternative mammals since European settlement: eleven spe- waterpoint and fence line locations can be mod- cies have become extinct, five have disappeared eled. More even distribution of grazing may from the mainland and are now confined to then be established to avoid the creation of off-shore islands, and fifteen more have declined grazing and trampling degradation hot spots. dramatically and persist only in the semi-arid The modeled patterns may also be overlaid on fringe (Morton 1990). This sorry chronicle repre- erosion forecast maps to minimize use of the sents the worst record of any continental region most erosion-prone areas and to reduce the over the past 500 years (see Diamond 1984). The movement of animals along the principal d irec- basic reason for these problems is damage caused tions of water flow. by introduced animals. Domestic stock and feral 185 Defining and Measuring Sustainability: The Biogeophysical Foundations herbivores (particularly rabbits) ate out the best resource-rich areas because of their metabolic country and prevented native species from using requirements. Because these herbivores have not their drought refuges, and introduced foxes and been controlled by ecological forces in the way cats put further pressure on relict populations that native herbivores originally were, they have (Burbidge and McKenzie 1989; Morton 1990). had a markedly degrading effect on the most Populations of remaining mammals and other productivecountry(Foran,Low,andStrongl985; animals are not universally declining (Curry and Griffin and Friedel 1985; Low, Muller, and Hacker 1990), but there is still considerable con- Dudzinski 1980; Pickup and Chewings 1992; cern that the ecological changes brought about by Purvis 1986). In areas of poorer soils, introduced European settlement will continue to cause losses herbivores focus almost entirely on resource-rich of biodiversity. patches. In richer soils, use is more widespread, Our approach to sustaining and restoring the but patterning still occurs in relation to watering biodiversity of arid lands is less sophisticated points (cattle and sheep) or suitability of sub- than if we were only interested in maintaining strate for burrowing (rabbits). This patterning pastoral values because much remains unknown has immense practical significance for conserva- about the functioning of natural ecosystems. We tion management. also lack the tools for measuring and forecasting Introduced grazing animals may affect native system state under current or potential manage- species in resource-rich areas in two ways. The ment regimes. This lack of technology cannot be first is that grazing frequently exaggerates the used as an excuse for inaction given the rate of variabilityof plantproduction(Foranl986;Friedel extinctions described above. We must therefore 1990; Pickup and Chewings 1992). The second is proceed on the basis of existing knowledge and that the absolute rate of plant production is al- accept that errors may be made. tered, usually downwards. These effects led to the extinction of many native mammals and con- tinue to threaten the maintenance of biodiversity Functioning of the natural ecosystem (see figure 13-7). Our approach to the problem of developing a system for conservation management in arid Measuring sustainability Australia is based on our understanding of the dominantecological forcesat work.StaffordSmith Earlier we showed that it is possible to monitor and Morton (1990) argue that two primary forces the status of soil and vegetation on pastoral lands underlie ecological functioning in the arid zone: using remotely sensed grazing gradients and ero- the supply of moisture and the availability of sion forecasting, together with targeted nutrients. At the most basic conceptual level, they ground-based techniques. Such monitoring tech- suggest that two different types of landscape niquesarevitaltothedevelopmentofanindustry exist. One consists of relatively rich soils, such as that is sustainable in economic terms. But how those of the Mitchell grasslands and the cheno- does one monitor for sustainability of land uses pod shrublands. Theother (and more common) is from the perspective of native plantsand animals dominated by poor soils, such as the spinifex that live in these pastoral lands? grasslands, but scattered throughout this vast All the problems associated with creation of a expanse are run-on areas that tend, through the monitoring system for vegetation and soil exist accumulation of water and nutrients, to be more when native biota are considered, in particular productive. These patterns of resource richness the difficulties of identifying long-term trends in may arise from nutrients, from water run-on, or biotic populations in an environment character- from the accumulation of both, but from the per- ized by capricious rainfall. These problems are spective of ecological functioning, it is not espe- compounded by lack of information about the cially important which is most influential at any determinants of distribution and abundance of one place. The contrasts drawn here are concep- organisms in arid Australia. Detailed understand- tual only; in real environments, these types of ing of all but a few groups of angiosperms is landscapeareonlypointsalongacontinuum(see lacking. Distribution maps of higher vertebrates figure 13-6). are available, but those of reptiles are unreliable Stafford Smith and \lorton (1990) suggest that and newspeciesarediscovered everyyear. Habi- introduced herbivores preferentially select tat requirements of only a few vertebrates have 186 Restoration of Arid Lands Figure 13-6: An Area of Central Australia Showing Patterns of Moisture Run-on fH~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~' Note: Pale areas shed water after rainfall, but darker areas receive run-on and thus are more productive and are described in the text as resource rich. been analyzed quantitatively. A vast array of mass of largely undescribed species apart from invertebrate species remains unstudied and usu- those isolated examples just mentioned, doing so ally undescribed. How can monitoring be con- would clearlybeimpossibleonfinancial grounds. ducted effectively under such condifions? Every arid area around the globe would face The short answer to this question is that most idenfical problems. Thus, some form of surrogate native species cannot be monitored effectively. must be found. Monitoring populations of just three kangaroo We recommend a twofold approach to this species in the southern, western, and eastern problem. The first principle of monitoring is to parts of the arid zone costs the various state and focus on those areas of land that have experi- federal wildlife services in the vicinity of $A1 enced regular use, most often pastoralism and to million a year (G. M. Maynes, personal communi- give lower priority to monitoring poorer and less cation). A few monitoring operations are under frequentlyusedland.Thisprinciplemaybepecu- way to follow population trends in rare or endan- liar to the Australian arid zone, where substantial gered mammal and bird species, but these efforts areas have been only mildly altered by human arebytheirnaturelimitedgeographically.Evenif activity. Virtually all parts of the country that the technical capacity existed to monitor the vast have been degraded are relatively rich in mois- 187 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 13-7: Suggested Chain of Events Leading biological diversity. This recommendation is to the Extinction of Native Australian Mammals purely pragmatic, because focusing on perennial Following Arrival in the Arid Zone of European vegetation appears to .e the only way to define Grazing Stock and Feral Antimals eeaonapastbehenlwytodfe adequate reserve networks and to search for mea- Localized populatior, ot medium-size mammals sures of ecological sustainability in rangelands I under human use. The monitoring that begins Graztng stock Rabbit under this set of assumptions could subsequently l n\ imsion ot the be refined as further research improves our un- patches (f habitat derstanding of the relationship between spatial Linevitable patterning of perennial vegetation and diversity drought of other organisms.This final point automatically Competition or simplx highlights an area of research that should be of habitat change high priority. Alteredi fire / Introduced Neither of the principles presented here has patterns i ../. predators been fully instituted in arid Australia. Our re- Increamed probabilities search group is developing the remote-sensing of local (iLsappearance and geographic information system tools neces- More droughts sarv to test the suggestions, and plans are under way to test the techniques in our Central Austra- ZnI Ii hlian studv area. Soircem Morton 144U. Implementing a sustainable conservation management scheme ture and nutrients. Thus, to institute such moni- toring, it is necessary first to identify patterning The management of sustainable conservation in the landscape that gives rise to the differential depends foremost on development of a represen- distribution of intensity in use. With techniques tative network of conservation reserves; the tech- similar to those used to analyze grazing gradient, niques for doing so are relatively well under- remotely sensed maps of the landscape can be stood in Australia (Margules, Pressey, and produced showing the distribution of productive Nicholls 1991). The Australian government rec- orvulnerable habitats. When thisremotely sensed ognizes that the present reserve system does not information is combined with digital data for the adequately sample Australia's biological diver- region from a geographic information system sity (Commonwealth of Australia 1992). How- (such as data on geology and topography), the ever, there are difficulties in achieving complete networks of such habitats can be examined at the representation in the inland environmentbecause regional scale, and decisions can be made about unpredictable fluctuations from heavy rainfall to the most effective distribution of monitoring ef- drought mean that no one place can be consid- fort. We are currently building such a system for ered permanently secure for many plants and the Central Australian mountain ranges, an area (especially) animals. It would beexceedingly hard of 120,000 square kilometers in the center of the to develop a suite of reserves that could take into arid zone. account all possible climatic sequences and The second requirement is to choose a surro- thereby guarantee the conservation of all species, gate to monitor. The recommended principle is to across the vast landscape of outback Australia. use a combination of remote sensing and ground The solution to this dilemma is twofold. The monitoring to assess the status of perennial veg- first step is to develop a hierarchy of reserves, etation in the areas deemed of importance. The rather than to rely purely on national parks (or inherent assumption is that maintaining biologi- whatever term is used in a particular location). cal diversity is dependent on maintaining the full National parks would remain the most signifi- range of perennial vegetation; without this, bio- cant component of the system but would be logical diversity will inevitably decline. We do complemented by a network of smaller reserves not assume that maintaining perennial vegeta- incorporating habitats that appear to be vital to tion is sufficient to guarantee the maintenance of the regional persistence of the biota but that can- 188 Restoration of Arid Lands not be declared as national parks because the including introduction of new diseases and de- areas concerned are too small or because the velopmentofimmuno-sterilizationvectorstolimit relevant authority has insufficient resources to reproduction (Cooke 1991; CSIRO Division of manage them as parks. In this case, the current Wildlife and Ecology 1990). land manager, most often a pastoralist, could be Techniques for instituting sustainable grazing paid an allowance to carry out the necessary in the broader sense (including recognition of the conservationactivities.Inmanyinstancesitwould necessity to consider the impact on biological not be necessary to remove the parcel of land diversity) are clearly much less advanced than permanently from other uses; only at key times those concerned with sustainability of an enter- (say during or at the break of drought) might prise only on the basis of soil stability, palatable conservation concerns take priority. Such an ex- vegetation, and animal production. This dispar- panded but flexible reserve system would go a ity is not surprising, given the wider issues at long way toward providing long-term conserva- stake when biological diversity is considered, tion of biological diversity in the environment of and is undoubtedly characteristic of most arid arid Australia where rainfall is capricious and environments around the world. plant production temporally and spatially unpre- dictable (see Morton and others 1992 for a fuller development of these proposals). Conclusions A second element remains essential to the implementation of sustainable management: the Our work on sustainability of land use in arid development of sustainable grazing strategies in Australia, which is summarized in figure 13-8, the country that lies between the various compo- has led us to four principal conclusions. nents of the reserve system. The requirements for maintaining sustainability in terms of continued * Adequate monltoring systems are essential mi production of plants and animals in the pastoral the uncertain climate of all arid lands, particu- landsnow need tobebroadened to include main- marly inust where Rainfall is uncom- tenane ofbiolgica divesityunde graing. monly difficult to predict. Remote sensing can Aparte ofriomoge al diverstigtiuns (uryazind now be used to monitor the effects of pastoral Apart from general investigations (Curry and acivt onsi n ln rdcin idw Hacker 1990), no data havebeen published on the activity on soil and plant production, and we impact of different levels of grazing on any native anticipate it becoming broadly accepted as a anml exep lag agrorCuhe,Shp major tool in the Australian rangelands over animals except large kangaroos (Caughley,Shnep- th etfwyas herd, and Short 1987). Preliminary studies dem- the next few years. onstrate that grazing indeed does affect a variety * Control of weeds and mammalian pests re- of animal taxa (C. D. James, personal communica- mains a serious problem in arid Australia. tion), and analysis of these relationships through Major research efforts are under way to im- further research is of paramount importance. prove control of the rabbit. A critical problem has so far been set aside: . A broader network of conservation reserves is weeds and vertebrate pests. The principal plant necessary because the vegetation associations invaders in arid Australia are Acacia nilotica of arid Australia are inadequately sampled in (Mimosaceae), Parkinsonia aculeata (Caesal- the current system. The techniques for choos- piniaceac), Cenchrus ciliaris (Poaceae), Prosopis spp. ing representative areas of land are well devel- (Fabaceae), and Tamarixaphylla (Tamaricaceae); a oped, but research is necessary to determine variety of strategies is required to control these the degree to which vegetation associationsact species (Humphries, Groves, and Mitchell 1991). as surrogates for the diversity of other groups Of greater importance than all other weeds and of organisms. pests, though, is the European rabbit. It is wide- * Wealth-generating uses of land are not yet spread, often phenomenally abundant, and a pri- meshed with ecological sustainability. A major mary cause of land degradation and extinction of research problem is to determine grazinglevels native species (Lange and Graham 1983; Morton that are consistent with maintaining regional 1990; Myers 1971). Several strategies for improv- biological diversity. ing the control of rabbits are under investigation, 189 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 13-8: Some Major Requirements for Sustainable Land Management Flowing from the Biophysical Characteristics of Arid Australia Reserve~~~~~~~~~~~~~~~ Coneserv ation| g of I Ecological Paee ng of Pe,s t b isustainability misuetand _control foprage grazn t pastoralism Erosion forecasting G,raz ng gracienit Monitoring References Cooke, B. 1991. "Rabbits: Indefensible on Any Grounds." Search 22, pp. 193-94. Bastin, G. 1991. "Rangeland Reclamation on Cridland,S.,andM.StaffordSmith.1992."Devel- Atartinga Station, Central Australia." Australian opmentand Disseminationof Design Methods lournalof Soiland WaterConservation4,pp. 18-25. for Rangeland Paddocks Which Maximise Behnke, R. H., I. Scoones, and C. Kerven, eds. Animal Production and Minimise Land Deg- 1993. Range Ecology at Disequilibrium. London: radation." Technical Memo. Western Austra- OverseasDevelopment Institute, International lian Department of Agriculture. Institute for Environment and Development, CSIRO Division of Wildlife and Ecology. 1990. and Commonwealth Secretariat. Biennial Report 1988-90. Melbourne. Burbidge, A. A., and N. L. McKenzie. 1989. "Pat- Curry, P. J., and R. B. Hacker. 1990. "Can Pastoral terns in the Modern Decline of Western Grazing Management Satisfy Endorsed Con- Australia'sVertebrateFauna: Causesand Con- servabon Objectives in Arid Western Australia?" servation Implications." Biological Conservation lournal of Emnironment Management 30, pp. 295-320. 50, pp. 143-98. DeAngelis, D. L., and J. C. Waterhouse. 1987. Caughley, G., N. Shepherd, and J. Short, eds. "Equilibrium and Non-equilibrium Concepts 1987. Kangaroos: Their Ecology and Management in Ecological Models." Ecological Monographs in theSheep Rangelands of Australia. Cambridge, 57, pp. 1-21. England: Cambridge University Press.' Diamond, J. M. 1984. "Historic Extinctions: A Commonwealth of Australia. 1990. "Ecologically Rosetta Stone for Understanding Prehistoric Sustainable Development." Commonwealth Extinctions." In P. S. Martin and R. G. Klein, Discussion Paper. AGPS, Canberra. eds., Qua ternary Extinctions, pp. 824-62. Tuc- . 1992. "A National Strategy for the Con- son: University of Arizona Press. servation of Australia's Biological Diversity Ellis, J. E., and D. M. Swift. 1988. "Stability of (Draft for Public Comment)." Department of Afncan Pastoral Ecosystems: Alternate Para- the Arts, Sport, the Environment, and Territo- Afia Patoa Ecsses-lent aa theAs, Spobrt nd digms and Implications for Development." nies, Canberra. Journal of Range Management 41, pp. 450-59. 190 Restoration of Arid Lands Foran,B.D.1986."ThelmpactofRabbitsandCattle Lange, R. T., and C. R. Graham. 1983. "Rabbits on an Arid Calcareous Shrubby Grassland in and the Failure of Regeneration in Australian Central Australia." Vegetatio 66, pp. 49-59. Arid Zone Acacia." Australian Journal of Ecol- Foran, B. D., W. A. Low, and B. W. Strong. 1985. ogY 8, pp. 377-81. "The Response of Rabbit Populations and Veg- Lange, R. T., A. D. Nicolson, and D. A. Nicolson. etation to Rabbit Control on a Calcareous 1984. "Vegetation Management of Chenopod Shrubby Grassland in Central Australia." Aus- Rangelands in South Australia." Australian tralian Wildlife Research 12, pp. 237-47. Rangeland Journal 6, pp. 46-54. Friedel, M. H. 1984. "Biomass and Nutrient Lauenroth, W. K., and W. A. Laycock, eds. 1989. Changes in the Herbaceous Layer of Two Cen- Secondary Succession and the Evaluation of Range- tral Australian Mulga Shrublands after Un- landCondition.Boulder,Colo.:WestviewPress. usually High Rainfall." Australian Journal of Low, W. A., W. J. Muller, and M. L. Dudzinski. Ecology 9, pp. 27-38. 1980. "Grazing Intensity of Cattle on a Com- . 1990. "Some Key Concepts for Monitor- plex of Rangeland Communities in Central ing Australia's Arid and Semi-arid Range- Australia." Australian Rangeland Journal 2, pp. lands." Australian Rangeland Journal 12, pp.21- 76-82. 24. Margules, C. R., R. L. Pressey, and A. 0. Nicholls. Friedel, M. H., B. D. Foran, and D. M. Stafford 1991. "Selecting Nature Reserves." In C. R. Smith. 1990. "Where the Creeks Run Dry or Margules and M. P. Austin, eds., Nature Con- Ten Feet High: Pastoral Management in Arid servation: Cost Effective Biological Surveys and Australia." Proceedings of the Ecological Society Data Analysis, pp. 90-97. Melbourne: CSIRO. of Australia 16, pp. 185-94. McDaniel, K. C., and R. H. Haas. 1982. "Assessing Friedel, M. H., G. Pickup, and D. J. Nelson. 1993. Mesquite-grass Vegetation Condition from "The Interpretation of Vegetation Change in a Landsat." Photogrammetric Engineering and Re- Spatially and Temporally Diverse Arid Aus- mote Sensing 48, pp. 441-50. tralian Landscape." Journal of Arid Environ- Mentis, M. T., D. Grossman, M. B. Hardy, T. G. ments 24, pp. 241-60. O'Connor, and P. J. O'Reagain. 1989. "Para- Graetz, R. D., R. P. Pech, and A. W. Davis. 1988. digm Shifts in South African Range Science, "The Assessment and Monitoring of Sparsely Management, and Administration." South Af- Vegetated Rangelands using Calibrated rican Journal of Science 85, pp. 684-87. Landsat Data." International Journal of Remote Morton, S. R. 1990. "The Impact of European Sensing 9, pp. 1201-22. Settlement on the Vertebrate Animals of Arid Griffin, G. F., and M. H. Friedel. 1985. "Discon- Australia: A Conceptual Model." Proceedings of tinuous Change in Central Australia: Some the Ecological Society of Australia 16, pp. 201-13.' Implications of Major Ecological Events for Morton, S. R., D. M. Stafford Smith, M. H. Friedel, Land Management." Journal of Arid Environ- G. F. Griffin, G. Pickup, and A. S. Sparrow. ments 9, pp. 63-80. 1992. "Sand-dune, Saltbush, and Stock: A Vi- Hamblin, A. 1991. "Sustainability: Physical and sion for the Stewardship of Arid Australia." Biological Considerations for Australian Envi- Processed. ronments." Working Paper WP/19/89. Bu- Myers, K. 1971. "The Rabbit in Australia." In P. J. reau of Rural Resources, Canberra. den Boer and G. R. Gradwell, eds., Dynamics of Hodgkinson, K. C. 1992. "Elements of Grazing Populations, pp. 478-503. Wageningen, the Strategies for Perennial Grass Management in Netherlands: Centre for Agricultural Publish- Rangelands." In G. Chapman, ed., Desertified ing and Documentation. Grasslands: Their Biology and Managernent, pp. Noy-Meir, I. 1973. "Desert Ecosystems: Environ- 77-94. London: Linnean Society of London. ment and Producers." Annual Reviewof Ecology Humphries, S. E., R. H. Groves, and D. S. Mitchell, and Systernatics 4, pp. 25-51. eds.1991.Plantlnvasions:TheIncidenceofEnvi- Pickup, G. 1985. "The Erosion Cell: A Geomor- ronmental Weeds in Australia, vol. 2. Kowari: phic Approach to Landscape Classification in Australian National Parks and Wildlife Ser- Range Assessment." AustralianRangelandJour- vice. nal 7, pp. 114-21. 191 Defining and Measuring Sustainability: The Biogeophysical Foundations . 1989. "New Land Degradation Survey Stafford Smith, D. M. 1984. "Behavioural Ecology Techniques for Arid Australia: Problems and of Sheep in the Australian Arid Zone." Ph.D. Prospects." Australian Rangeland Journal 11, pp. diss., AustralianNational University,Canberra. 74-82. . 1988. "Modeling: Three Approaches to .1992. "Modelling Patterns of Defoliation Predicting How Herbivore Impact Is Distrib- by Grazing Animals in Rangelands." Journal of uted in Rangelands." Research Report 628. Applied Ecology 31, pp. 231-46. New Mexico State University, Agricultural Ex- Pickup, G., and V. H. Chewings. 1988a. "Estimat- periment Station, Las Cruces, N.Mx. ing the Distribution and Patterns of Cattle Stafford Smith, D. M., and B. D. Foran. 1988. Movement in a Large Arid Zone Paddock: An "Strategic Decisions in Pastoral Management." Approach Using Animal Distribution Patterns Australian Rangeland Journal 10, pp. 82-95. and Landsat Imagery." International Journal of Stafford Smith, D. M., and S. R. Morton. 1990. "A Remote Sensing 9, pp. 1469-90. Framework for the Ecology of Arid Australia." .1988b. "Forecasting Patterns of Erosion in Journal of Arid Environments 18, pp. 255-78. Arid Lands from Landsat MSS Data." Interna- Stafford Smith, D. M., and G. Pickup. 1993. "Out tional Journal of Remote Sensing 9, pp. 69-84. of Africa Looking in: Understanding Vegeta- .1992. "A Grazing Gradient Approach to tion Change." In R. H. Behnke, l. Scoones, and Land Degradation Assessment in Arid Areas C. Kerven,eds.,RangeEcologyat Disequilibrium, fromRemotelySensed Data." lnternationallour- pp. 196-226. London: Overseas Development nal of Remote Sensing 15, pp. 597-617. Institute, Intemational Institute for Environ- Pickup,G., and D.J.Nelson. 1984. "Useof Landsat ment and Development, and Commonwealth Radiance Parameters to Distinguish Soil Ero- Secretariat. sion, Stability, and Deposition in Arid Central Westoby, M. 1980. "Elements of a Theory of Veg- Australia." Remote Sensing of Environment 16, etation Dynamics in Arid Rangelands." Israel pp. 195-209. Journal of Botany 28, pp. 169-94. Pickup,G.,and D.M.StaffordSmith.1993."Prob- Westoby, M., B. Walker, and 1. Noy-Meir. 1989. lems, Prospects, and Procedures for Assessing "Opportunistic Management for Rangelands the Sustainability of Pastoral Land Manage- Not at Equilibrium." Journal of Range Manage- ment in Arid Australia." Journal of Biogeogra- ment 42, pp. 266-74. phy 20, pp. 471-87. Wilson, A. D., and N. D. MacLeod. 1991. "Over- Purvis, J. R. 1986. "Nurture the Land: My Philoso- grazing: Present or Absent?" Journal of Range phiesof Pastoral ManagementinCentral Austra- Management 44, pp. 475-82. lia." Australian Rangeland Journal 8, pp. 110-17. 192 5*14 Currencies for Measuring Sustainability: Case Studies from Asian Highlands P. S. Ramakrishnan The concept of sustainability as generally understood implies the use of ecological systems in a manner that satisfies current needs without compromising the needs or options of future generations. Achieving such an objective involves a variety of choices. These choices stemfrom clearly defined ecological concepts and less obvious and highly variable social concerns. In developing countries, population pressures are more intense than in the industrial world, and human societies are often strongly rooted to their traditional beliefs and influenced by socioeconomic, sociocultural, and sociopolitical considerations. In such a context, understanding the tradeoffs between meeting current needs and maintaining a variety of options for thefuture becomes increasingly complex. Ecological concepts are no doubt important, but social dimensions that are more elusive need to be given equal weight. Our understanding of the ecological processes that could form the basis for the sustainable use of a given ecosystem is farfrom adequate. Add to this the social dimensions, and we are left groping in the dark. It is in this context that someof thecase studies considered here, and the northeast Indian case study in particular, become significant. One of the important objectives of this chapter is to indicate the variety of routes that are available for developing the northeastern highlands in a sustainable way. While doing so, I evaluate a variety of parameters that may be considered for defining and measuring sustainability. The case study from northeastern India is a comprehensive multidisciplinary study, and the emphasis here is placed on work from the Asian highlands. However, a few scattered studies from elsewhere in the region are also considered to illustrate specific points under consideration. Shifting agriculture, upland forests, and a complex and large base of natural resources and sustainable development: Case study of a hilly terrain, provide the ecological diversity. northeastern India The people of the region, comprising many tribes, provide the sociocultural diversity. To build on It is well recognized now that conservation and such a profile on a sustainable basis is a challeng- sustainable development are two sides of the ingopportunity.Thefollowingsummarizesmore same coin: one cannot be achieved at the expense than 200 publications arising from this case study. of the other. From a human angle, such an inte- grated approach demands satisfying basic hu- The setting man needs in an equitable manner and maintain- Shifting agriculture (locally called jhum) is the ing and indeed promoting social, cultural, and chief land use system of the northeastern high- biologicaldiversityandecologicalintegrityof the lands in India. The forest farmer here and else- system.Thenortheastemhighlandsoflndia,with where in the humid tropics has managed his Defining and Measuring Sustainability: The Biogeophysical Foundations traditional shifting agriculture for centuries, kilograms of nitrogen, 30 kilograms of phospho- achieving optimum yield on a long-term basis, rous, and 30 kilograms of potassium per hectare rather than maximizing production for short-term a year; this input obviously is a major deterrent to considerations (Ramakrishnan 1984, 1992a; its acceptance. The system often also conflicts Ruthenberg 1971; Watters 1960). The farmer tries with the patterns of land tenure and with the to capitalize on the soil fertility built up through social structure of tribal communities, adversely natural processes during the forest fallow phase affecting the independence of the family unit. It and the nutrients released in a single flush during cannot be sustained by a single family. In view of the slash-and-bum operation. However, in the this and in view of the fact that it ignores tradi- northeastern region and elsewhere, the jhum cycle tional technologies of the tribal communities (the length of the fallow phase during two succes- (Ramakrishnan 1984,1992a), the ICAR model has sivecroppingsfromthesamesite)hasdeclined in not had a significant impact so far. Meanwhile, recent years to become a short cycle of four to five jhum has become even more untenable because of years, reaching less than three years in places. large-scaletimberextraction, increasing population This contrasts with the cycle of twenty years or pressure, and simultaneousdecline in landi area for more that was common in the not-too-distant jhum, all of which contributed to the drastic decline past and that ensured that the system was in inthecycle(Ramakrishnan]992a).1tisobviousthat harmony with the environment. Shortening the the jhumsystemasitisoperating now in the north- cycle has resulted in poor recovery of soil fertility east cannot be sustained. This prompted this effort and the consequent problems of sustainability of to define sustainability and detemiiine biophysical the agricultural system, both in economic and parameters for evaluating the sustainability of a ecological terms (Ramakrishnan 1985, 1992a). forest-agroecosystem complex. Further, shortened cycles do not permit the forest ecosystem to recover adequately, leading to loss Traditional patternts of resource use of forest cover, its replacement by weed cover, and in extreme cases total desertification of the The northeastern hill region is inhabited by more site (which means the decline in biological pro- ithan a hundred diffrent tribes that are highly dciiyof the ecosystem; see, for example, insulated, with their own language andi cultural ductivity she 1992b. identities. The region is also characterized by It is in this context that government agencies extreme variations in ecology. Altitude varies dealing with agriculture, forest, and soil conser- widely from sea level to about 3,000 meters in vation~~~~~ hav atepe.orpac hhmss Arunachal Pradesh. With an annual average rain- vation have attempted to replace the jhum sys- falo20cetmerfrthrgii s wl, tem. Formorethanfortyyears, theattempt wasto fall of 200centimeters for the region as as whole, convince farmers to practice sedentary agricul- rainfallomayereach 12 metersnintsome areas or in 1a ture largely through the development of terrace exceptional year even 24 meters, as it ,id in 1974 farming. lncentiveswereprovided in the formof in Cherrapunji, one of the wsettest sbots in the monetary subsidies for terracing and establish- world. The highly leached soils may be weakly to ing and maintaining terraces, for using high- strongly acidic, supporting subtropical forests at yielding varieties of seed, and for applying fertil- lower elevations and subtemperate broad leaved izers to sustain soil fertility. These subsidies were or conifer forests at higher al titudes. A variety of withdrawn after the initial introduction of alter- degraded typcs of forests such as bamboo forests nate technology, however, and the farmer re- and grasslands also develop, depending on biotic verted back to his traditional jhum system. In disturbances (Ramakrishnan 1992M). more recent times, the Indian Council of Agricul- Tribal commu nities in the region engage in a tural Research (ICAR) complex located in the variety of economic activities based on the avail- northeastern region developed a model to re- ability of resources. In the more remote areas ot place jhum (Borthakur and others 1978). This Arunachal Pradesh, tribes such as the Sulu ngsare largely hunter gatherers, although they may also model proposes a three-tier system for farming the hill slope. The upper third of the slopc is work in traditional agriculturh an d ainial hus- designated for forestry, the middle portion for and Ramakrishnan 1987). The more advaGnced horticulture and plantation crops, and the lower tribes such as the Apatanis of Arunacial laradesh third for terraced agriculture. This system de- engage in sedentary agriculture oni valley land mands inorganic fertilizer on the order of 60 and in traditionial animal husbandry (IKumlarand 194 Currencies for Measuring Sustainability: Case Studies from Asian Highlands Ramakrishnan 1990). All the tribal communities 1990). Rice is supplemented by Eleusine coracana in the region depend on the forest for collecting cultivated on elevated partition bunds between wild food and extracting fuelwood and fodder the rice plots. Pisciculture is done only with the for animal husbandry and for slash-and-burn late-maturing variety of rice, which improves the agriculture, which is the traditional system of use of resources and the consequent efficiencyof land use in the region. the system. THlE JIHUM SYSTEM HOME GARDENS The jhum system is highly heterogeneous. This An important agroforestry system of tribal com- mixed cropping system follows a slash-and-burn munities is the home garden, variously termed operation and varies considerably in its cropping the kitchen garden or forest garden. These gar- pattern, depending on the ecological conditions dens have highly diverse and stratified plant and socioeconomic and cultural background of species, dominated by woody perennials. With a each tribe. In the same region, cropping patterns mixture of annuals and perennials forming a also vary with the length of the jhum cycle. The multiple-storeyedstructure,theyresembleanatu- number of species in the crop mixture may vary ral forest. From a plot of 0.5 to 2 hectares located from eight to more than thirty-five, with much close to the habitation, the farmer obtains food, variation noted in the proportion of different fire wood, medicinal plants, spices, and oma- species in the mixture. This makes the jhum sys- mentals all year round. Apart from meeting the tem highly complex. needs of the farmer, they also perform social or aesthetic functions, serving as an indicator of VALLEY LAND AGROECOSYSTEM social status of the owner or improving environ- Agriculture is also practiced in valley land mental quality. throughout the region. It is a sedentary form of The cropping pattern in home gardens varies wet rice (Oryza sativa) cultivation and is comple- considerably. The Mikirs at lower elevations of mentary to jhum. It is done wherever the terrain Meghalaya, for example, emphasize the arecanut permits, on flat lands between hill slopes. Obvi- (betel nut,orArecacatechu) withbetel (Piper betel), ously, this system is restricted by topography. black pepper (Piper nigrum), and banana as cash The soil in valley lands is fertile due to nutrient crops (Maikhuri and Ramakrishnan 1990). The wash-out from the hilly slopes and therefore does Khasis in the Cherrapunji region also grow cash not need added fertilizers. The main advantage is crops such as the bay leaf (Cinnamomum that the land gives sustained yield yearafter year. obtusifoliurm), orange (Citrus sinensis), and jack This land use also varies widely in the number of fruit(Artocarpusheterophylls).Avarietyof ground- croppings a year, the mixture of crops used, the level specics may be grown. Indeed, this land use recycling of organic resources done to maintain system too varies from place to place, based on soil fertility, and the inclusion or otherwise of ecological and social considerations. pisciculture as an integral part of the cropping system. Thus, the Garos at lower elevations of CASE] CROP ECOSYSTEMS Meghalaya may raise two crops, whereas the A shift toward plantation and cash crops some- Mikirs in the same area or the Khasis at higher times accompanies the rapid shortening of the elevations may raise only one (Maikhuri and jhum cycle (Ramakrishnan 1992a). Broom grass Ramakrishnan 1990; Patnaik and Ramakrishnan (Thysanolalna maxima) is collected from the wild 1989). Apatanis of Arunachal Pradesh, who have and cultivated in many places, as for example, one of the most efficient and advanced forms of by the Khasis of Meghalaya (Gangwar and valley cultivation, use late or early ripening vari- Ramakrishnan 1989). The grass used for mak- eties of rice depending on whether waste is re- ing brooms may be part of a plantation crop cycled efficientlyor not. The late varieties go with such as Cinnamom lum obtusifolium. Bamboo more fertile soils, and the early varieties go with (Dendrocalamnus hamiltonii) and thatch grass less fertile soils. Since the organic waste is largely (imnperata cylindrica) are also grown or harvested generated within the village, the late ripening from the wild. Ginger, banana, and pineapple variety is grown closer to the village. The early are extensively cultivated in pure or mixed variety is sown farther away, where disturbance systems. Recently, government agencies have by wild animals and poor irrigation facilities can introduced rubber, tea, coffee, and cashew nut be major constraints (Kumar and Ramakrishnan into the region. 195 Defining and Measuring Sustainability: The Biogeophysical Foundations FALLOW AND SEDENTARY SYSTEMS Currencies for measuring sustainability As a result of increasing pressure placed on limited land by the growing population, more A variety of currencies is used here to evaluate intensive systems of land use have developed the land use systems as they operate now. These recently in the northeastern region. Largely currencies also form the basis for devising sus- around urban growth centers, where the land is tamable development strategies. This section at- devoid of forest cover and often in a state of tempts to reconcile the different currencies for desertification, a semi-permanent bush-fallow measuring sustainability. system of agriculture is being practiced (FAO/ SIDA 1974); this ultimately ends up in seden- CROP BIOMASS IN AGRICULTURE tary systems of agriculture, with appropriate A high diversity of species is characteristic of all crop rotation. At least two dozen cropping pat- traditional systems of agriculture-the home gar- terns are recognized in a 20-kilometer radius den and the phum system (with up to thirty-five around the Shillong township in Meghalaya. species in a plot of 2 to 3 hectares) and of many Often lesser-known crop species such as tropical agroecosystemns (Ramakrishnan 1992a). Flemingia vestita, a legume, either in tercropped High diversity contributes to stability of the ecosys- or rotated every three or four years, improves tern. With a multilayered canopy, a high leaf area soil fertility, apart from its value as a tuber indexiforicapturingolight efficiently, andpaflayered during the lean season, when traditional food distribution of root mass in the soil profile for using sources are scarce. nuret pial,crop productivity is high. i Indeed, the jhum farmer shifts his emphasis in ANIMAL HUSBANDRY SUBSYSTEM the crop mixture depending on the jhum cycle. Animal husbandry is a low-cost subsystem be- Under a short cycle of five years, for example, the cause natural resources are relatively available emphasisisontuberandrhizomatouscropssince and population pressure is minimal. Of all prac- they use nutrients more efficiently and can maxi- tices, swine husbandry is one of the cheapest to mizeoutputunderlow-fertilitysoils.Underlonger maintain, as it is based on efficient recycling of jhum cycles of twenty years, for example, the resources. The waste biomass from agriculture emphasisison cereal cropssuch asriceandmaize and domestic subsystems, including food unfit that use nutrients less efficiently. Such a shift in for human consumption, is recycled through emphasis can also be seen under a given jhum swine husbandry. This detritus-based system is cycle, with emphasis on cereals at the base of the closely interlinked with shifting agriculture slope (sites rich in nutrients) and emphasis in throughout the world (Ramakrishnan 1992aa tuber and rhizomatous crops at the top of the Rappaport 1971). slope (sites poor in nutrients). Therefore, the val- Poultry isanotherimportantactivityoftribal ues for primary productivity under jhum com- populations; goats and cattle are recent intro- pare with those of the secondary successional ductionsintotribalsocietiesandhavecnotfound fallows up to twenty years (Toky and acceptance, although theyare sometimesraised Ramakrishnan 1981, 1983a). for meat. However, the immigrant Nepalis and A charactenstic feature of the jhum system is Biharis maintain cattle for milk. Many tribes thehighrateofbiomassaccumulationinrelation maintain mithun (Bos frontalis), which are to economic output. The high rates of organic slaughtered during festivities. These semi-do- matter produced under jhum (16 to 22 tons per mesticated traditional animals are valuable for hectare) are close to values obtained for natural religion, for status, and forbarter. Because the plant communities (14.8 tons per hectare) of animals require large forested grazing lands, twenty-year-old forest fallow in the region. With mithun husbandry has declined as deforesta- higher crop diversity, it has been possible to tion has accelerated. combine the need for increasing the production of harvestable food with the need formaintaining DOMESTIC SUBSYSTEM a high content of organic biomass. Without this Tribal communities depend on forests forobtain- high production of organic matter, it would be- ing fuelwood for cooking food and for heating come necessary to import costly inorganic fertil- their huts during winter months. They also col- izers, which are hard to come by and whose lect a variety of plant and animal foods from the effectivencss in the faceof high temperatures and forests. heavy rainfall is questionable. 196 Currencies for Measuring Sustainability: Case Studies from Asian Highlands Sequential harvesting of crops is an effective duction, minimizes losses, provides a wide base way to recycle organic residues into the system of food resources for tribal society by providing over a period of time. Under the partial weeding cereals, legumes, vegetables, and even fiber, and done during jhum, even the weed biomass gets at the same time ensures leisure by effectively recycled into the plot in a phased manner so that spreading out labor all the year round. a layer of humus is always present on the soil surface. About 20 percent of the weed biomass BIOMASS DURING THE FALLOW PHASE left in situ during cropping also contributes to In a series of studies, we have shown that the biodiversity and helF s conserve water and nutri- weed potential of a site is aggravated under jhum entson the steep slope (Swamy and Ramakrishnan cycles shorter than ten years. The increased weed 1988). Sequential harvestingisan effective way to potential under cycles of four to six years is obvi- manage up to thirty-five or forty crop species over ously due to the presence of the same weed spe- both space and time. Thus, after early maturing cies in the four- to six-year.old plots that are species such as maize and Setaria italica are har- slashed, which helps to build up the soil seed vested, more space is devoted to rice at the peak of bank. Long cycles of ten years or more, in con- its growth period. Successive harvests of cereals trast, may be sustainable since weeds decline create additional space for the remaining perennial naturally during long fallow periods. crops, which also receive humus and nutrients. Continuous cropping under terrace cultiva- Mixed cropping also helps control biological tion results in even higher weed potential than pests (Litsinger and Moody 1976). The use of under short-cycle jhum. On old terraces, crop native varieties would probably ensure that a yield could be adversely affected because of in- high degree of natural chemical defenses is main- tense competition with weeds. Indeed, this is one tained (janzen 1973). Further, under mixed crop- of the major difficulties of sustaining the terrace ping, with 20 percent residual weed biomass, it is agroecosystem. unlikely that any one of the populations of in- Large-scaledeforestation for timbe-extraction sects, bacteria, or nematodes would reach epi- and shortened jhum cycles of about five years demic levels due to high genetic diversity. have resulted in large-scale invasion of weeds, The crop more than doubles soil cover be- many of them exotic (Ramakrishnan 1991; tween the ten- and thirty-year jhum cycle and Ramakrishnan and Vitousek 1989). This is a stage increases it more than fivefold between the five- in site degradation. In the final stage, the land- and thirty-year cycle (see table 14-1). In the ulti- scape is totally bald and desertified mate analysis, mixed cropping maximizes pro- (Ramakrishnan 1992a). Table 14-1: Characteristics under Five-, Ten-, and Thirty-Year Cycles for the Garos at Bumrihat in Meghalaya Crop Five year Ten year Thirly year Economic yields (tons per hectare a year) Seeds 0.107 1.153 2.180 Leaf fruit 0.129 0.074 0.024 Tubers 0.320 0.613 0.192 Total 0.556 1.840 2.396 ANP 14.060 11.576 15.213 Growth rate (grams per square meter a day) 3.8 3.2 4.2 NPP 18.461 14.709 17.746 Growth rate (grams per square meter a day) 5.1 4.0 4.9 Number of cultivars 8 12 14 LAI 0.59 1.49 3.20 H.l. 0.030 0.125 0.135 H.I. (grain + seed) 0.184 0.230 0.182 Labor (days per hectare a year) 149 305 436 Note. H.I. = Crop yield / NPP; H.I. (grain + seed ) = above-ground crop yield / ANP. Source, Toky and Ramakrishnan 1981. 197 Defining and Measuring S ustainability: The Biogeophysical Foundations BIODrVERSUIY The jhum farmer tries to capitalize on the lim- Perturbation of an ecosystem mnay promote or ited soil fertility that is highly transient, both in adversely affect biodiversity, depending on the space and time. As already noted, he emphasizes intensity and frequency of the events crops that use nutrients efficiently under short (Chandrasekhara 1991; Ramakrishnan 1992a). In jhum cycles of five years. He also places crops so thenortheastern hill region, a ten-yearcycle seems that the more nutrient-efficient crops are largely to be the cutoff point for biodiversity in the rain on top of the slope and the less-efficient ones are forests. However, higher diversity values were at thebase. By thisheisable toachieveahighleaf obtained beyond a thirty-year cycle. Further, tra- area index for optimizing photosynthesis from a ditional agroecosystems are significant for con- highly heterogeneous soil environment. Sequen- serving the biodiversity of crops (Ramakrishnan tial harvesting and the consequent addition of 1989b).The objective was to optimize biodiversity humus through recycled crop and weed biomass by conserving the keystone species that may have ensures optimal use of nutrients for the succeed- multiple uses or that may have possible value for ing crop species. A high level of synchrony be- the future. Biodiversity is a critical measure for tween nutrients released by the organic residues sustainable development. and the pattern of use by the crops ensures opti- mal yield under the given situation. NUTRIENT CYCLING DURING CROPPING The process by which nutrients are depleted AND FALLOW PHASES from the soil through the cropping phase also During slash and burn involving low- or high- continues through the fallow phase up to about intensity burn, a variety of physicochemical tenyears.Therapid transferofnutrieritsfromthe changes occur in the soil (Ramakrishnan and soil to the living biomass during the early succes- Toky 1981). Carbon and nitrogen are volatilized. sional phase is reversed when litterfall occurs A rapid increase in soil pH occurs, with its impli- either through leaf drop or complete turnover of cation for biological activities of the soil. Phos- the early herbaceous vegetation, whizh happens phorus and cations are released in a flush. Nitro- only after about ten years of fallow regrowth (see gen buildup is soon initiated through microbial -figure 14-1). This implies that a ten-year jhum fixation. However, during the cropping phase cycle is critical from the point of view of sustain- nutrientsare lost, partly through runoff and infil- able cropping under this system. Shorter cycles tration and partly through the removal of weeds not only do not permit nutrient recovery in the and crops. One of the chief conclusions arising soilbutalsoaccelerateavarietyoflos.sesfromthe from a detailed study of soil fertility under the system under frequent perturbations. More fre- cropping phase is the generally poor level of quent losses under low levels of soil fertility nutrients under which the system has to operate under cycles of four to five years eventually (Ramakrishnan 1992a). lead to desertification. Indeed, traditional sys- Figure 14-1: Economic Yield under Different Jhum Cycles of Thirty, Ten, and Five Years in Burnihat at Lower Elevations of Meghalaya 55003 5 00 15 20 25 30 35 40 45 50 55 60 65 YEARS Sourcc Ramaknshnan 1985 198 Currcncies for Measuring Sustainability: Case Studies from Asian Highlands tems with cycles of ten years or more are closer a five-year cycle (Ramakrishnan 1992a). There- to natural ecosystems where nutrient cycling fore, it is not surprising to find many of the jhum and maintenance of soil fertility are based on plots in Nagaland being integrated with the efficient internal controls, thus contributing to Nepalese alder (Gokhale and others 1985). In- their stability. deed this species also provides cash income Detailed nutrient budget analysis throws fur- through wood biomass harvested every five or ther light on the value of using soil fertility as a six years and regenerated through coppices. currency for evaluating the jhum system. To take Keystone species such as this areimportant for one example of suc'i an analysis, during one redeveloping jhum (Ramakrishnan 1989a). Thus, cropping seasoni it was shown that the system for example, the different species of bamboo would lose up to about 600 kilograms of nitrogen (Dendroclamus ha.niltonii, Bamboosa tulda, and B. per hectare (see table 14-2). It would take about ten khasiana) coming up in jhum fallows between ten years of fallow regrowth to recover this entire loss to thirty years of fallow regrowth tend to con- through a natural process of plant succession. How- serve nitrogen, phosphorus, and potassium in the ever, under a short cycle of five years, the system is system (Rao and Ramakrishnan 1989; Toky and able to recover only half of what it had lost, that is, Ramakrishnan 1983b). Indeed, in younger fal- about 300 kilograms of nitrogen per hectare. In lows of less than five years, even the exotic weed other words,undereach croppingundera five-year Mikania micrantha conserves potassium under cycle in a given plot, the system would lose about shorter cycles of five to six years (Swamy and 300 kilograms of nitrogen per hectare, which is Ramakrishnan 1987). never put back into the system. This and similar losses of otherelements from thesystem would lead ECONOMIC EFFICIENCY to desertification (Ramakrishnan 1992a). A series of studies done on jhum under cycles T he link between the soil nutrient budget and ranging from sixty-year cycles on one extreme to cycling processes during the cropping and fallow five-year cycles on the other suggests that from phases has implications for sustainability of the the point of view of economic yield and monetary jhum1 system under varied cycles. An obvious analysis, a ten-year cycle should be the cutoff conclusion is that if the jhum could be done with point (Ramakrishnan 1992a). The net economic a minimum cycle of at least ten years, it could be returns tothe farmer,aftermakingallowancesfor sustainable in the region. Under shorter cycles, a variety of labor inputs for slash and burn, is the system obviously needs to be redeveloped optimum under a ten-year cycle. The monetary through additional agroforestry inputs. An obvi- output tends to decline under successive five- ouschoicethathasfoundreadyacceptanceamong year cycles but remains stable under a ten-year the farmers of the northeast is Alnus nepalensis cycle or longer (see figure 14-2). For the (the Nepalese alder). Growing at an altitudinal sustainability of the system as currently prac- range of 500 to 1,900 meters in the northeast, this ticed, a minimal cycle of ten years is required. species could fix up to about 117 kilograms per Indeed, the farmer is able to obtain a higher hectarea year when young (Sharma and Ambasht economic returnundera ten-year jhumcycle than 1988),andthisspeciescouldrecoverallthenitro- under terrace cropping on the same site gen that the system loses during cropping under (Ramakrishnan 1984). Table 14-2: Net Change of Nitrogen in the Soil under Jhum at Shillong in Meghalaya at Five-, Ten-, and Fifteen-year Cycles (thousands of kilograms per hectare a year) Five year Soil pool Fifteen year Ten year First-year crop Second-year crop Before burning 7.68 7.74 6.40 5.98 At the end of cropping 7.04 7.15 5.98 5.60 Net difference 0.64 0.59 0.42 0.38 Source: Mishra and Ramakrishnan 1984. 199 Defining and Measuring Sustainability: The Biogeophysical Foundations The wide variety of jhum systems available in systems as models for development in an the northeast offers opportunity for manipula- energy-limited world are obvious. Therefore, tion so that the farmer can increase his returns. many traditional mixed cropping systems are Thus the jhum system where potato is empha- held up as models of ecological efficiency. Under sized gives the farmer up to five times higher jhum, for every unit of energy input, which itself returns than another where the emphasis is on ischieflyintheformofhumanlabor,fiftyormore rice. Indeed wide variations in economic yield units of energy are harvested (Toky and exist depending on the cropping pattern even Ramakrishnan 1982). under the same jhum cycle. Mere transfer of tech- The jhum system is more efficient than seden- nology from one area to another could improve tary terrace cultivation, which requires the subsi- the returns. dizing of fossil fuel energy in the form of fertiliz- ers. The energy cost of establishing and maintain- ENERGY EFFICaENCY ing terraces is high. Over a period of time, due to The increasing agricultural yields of the last half site degradation, the efficiency of fertilizer use century were made possible through the indus- under terraces declines drastically. trialization of agriculture involving large energy A comparison of jhum under different cycles subsidies and high-yielding varieties of crops suggests that if the cycle is long enough and the grown in pure stands. The drawbacks of such land is not a limiting factor, the input of solar Figure 14-2: Changes in Cumulative Quantity of Available Phosphorus (A), Potassium (B), Calcium (C), and Magnesium (D) within aSoil Column of 40 Centimeters inDepthunderJhum Fallows of Various Ages Available P (gm/m2) K(g eq/m2 ) 0 0.5 1.0 1.5 2.0 2.5 20.5 40.5 1 2 3 4 5 ~14- b - ° - P - ... I. ... 7, ...... V 28 _ .C b p o LfI 40 5 b0 5 1 o i o 15 so lo s 1 o Is so PHOSPHORUS POTASSIUM Ca (g eq/m2) Mg (g eq/m2) 0 1 2 3 4 5 6 7 1 3 5 7 9 11 13 7-- 14 -- w\ I C, 28 o q , 0 LI) 4 0 o o b 50 5 1015 01 so l s I o CALCIUM MAGNESIUM Source: Ramakrishnan and Toky 1981. 200 Currencies for Measuring Sustainability: Case Studies from Asian Highlands energy to a larger area of the jhum system could What is sustainability in the context offset the need to import fossil fuel energy, which of northeastern India? would ensure harmony of the system with the environment. Even when one uses a correction The northeastern case study shows that factorof 1/30, 1/10,orl/5forthirty-, ten-,and five- sustainability can have a short- and long-term year cycles, one finds that a ten-year jhum cycle is aspect to it. In the short-term context, two possi- thecutoffpointforenergyefficiency(output/input bilities exist. The first is to sustain jhum in the ratio) and land use (Ramakrishnan 1992a). present form at a minimal cycle of ten years since Keeping energy efficiency high, possibilities this is the cutoff point for efficiency. This could be exist for increasing crop production by strength- done by strengthening the valley land and home ening agroforestry, without departing too much garden ecosyster.ms. Here, transfer of technology from the traditional jhum system. In a wider from one area to another is a possibility, since all context of Indian agriculture, it should be pos- these systems are highly heterogeneous and offer sible to replace imported fertilizers with local a wide variety of socioeconomic returns to the resources based on biofertilizers and small-scale farmer. A valley land agroecosystem, such as that watershed management projects and to have a of the Apatanis whose complex agricultural sys- stable system of production. With a large rural tem is integrated with pisciculture, is attractive population of small and marginal farmers en- for transfer to other areas. Land uses other than gaged in agriculture, emphasizing agricultural jhum and even the animal husbandry component technologybasedonefficientrecyclingofnatural such as swine husbandry and poultry could be resources seems to be more appropriate. further strengthened through appropriate technol- ogy inputs. This could take the pressure off the land SOCIAL AND CULTURAL VALUE SYSTEM devoted to jhum so that a minimal ten-year cycle Jhum has been a way of life for tribal communities could be ensured, at least in some areas. along with other land uses such as valley and Alternatively, as a short-term strategy, a jhum home garden systems. Therefore, basing system with a short cycle of five years, for ex- sustainability on traditional technology and val- ample, could be redeveloped based on alder tech- ues and orienting sustainable development to- nology. Nepalese alder is now extensively used ward approaches with which these communities to strengthen the agroforestry component of the can relate become important. A variety of reli- jhum in places such as Nagaland (Gokhale and gious and cultural ceremonies are linked to the others 1985) so that the short-cycle system could jhum calendar, starting with slash and burn and besustained and slash-and-burnoperationselimi- continuing through sowing and harvesting the nated or at least minimized. crops (Ramakrishnan 1985). Indeed, the cultural On a long-term basis, a cooperative planta- link between traditional societies and their forest tion economy involving coffee, tea, rubber, fruit heritage-seen in the sacred grove forests main- trees, oreven timbercould bedeveloped, based tained extensively in the past-is still found dur- on the concept of the home garden. Such an ing the present (Khiewtam and Ramakrishnan economic initiative could be organized in small 1989). Some plant species such as bamboos and plots run by farmer cooperatives, which would the Nepalese alder are traditionally valued, and ensure participation. An economy based on farmers have often been unable to relate to other trees and forestry could be effectively done by tree species that are fast-growing and have the an appropriate mix of species that are based on appropriate architectural form from a biological efficient recycling of nutrients. Appropriate point of view (Ramakrishnan 1986). Nepalese rural technology could be introduced into the alders and bamboos species are also ecologically domestic sector. An integrated and holistic ap- significant keystone species for conserving nu- proach would ensure sustainability since it trients, asdiscussed earlier. Indeed, a parallel is would involve people in the developmental discernible between ecological and sociologi- process (see box 14-1). cal keystone species. A technology, however Identifying the key social issue and building effective, may not be relevant to a given society on it could be a sure way to ensure that farmers unless it is placed in the total social context. The participate and that development is sustainable human dimension of sustainability is critical to amongruralcommunities.Thecasestudyof north- acceptance. eastern Indian illustrates this very effectively. 201 Defining and Measuring Sustainability: The Biogeophysical Foundations Development of the Philippine highlands Box 14-1: Shifting Agriculture and A country of shifting agriculture, the Philippines Sustainable Development in North- experiences problems similar to those of the north- SustarnIndiab Deveopmnt i Noth-eastern highlands of India (Fujisaka, Sajise, and del eastern India Castillo 1986). Methods of controlling soil erosion and conserving soil through tillage are less labor For improving the system of land use and resource management in northeast- intensive and cost-effective than terracing. Rede- ern India, the following strategies are veloping the agroforestry system using traditional based on a multidisciplinary analysis. knowledge is the starting point. Contour intercrop- Many of these proposals have already ping of nitrogen-fixing tree hedgerows and food been put into practice. and cash crops is one possibility for recyding nutri- ents efficiently. Leucena leucocephala and Glircidia * Employ a wide variation in the pat- sepium tree hedgerows were found to be most ap- terns of cropping and yield under propriateandacceptable.Becausethedevelopment jhum and transfer technology among of technology for appropriate land use is often tribes, areas, or ecosystems (empha- based exogenously, acceptance has been limited. A sis on potato at higher elevations and rice at lower elevations has led to a major factor working against sustainable land use manifold increase in economic yield has been problems related to land tenure, which is despite low fertility of the more acidic oftendiscriminatory. Severalinteractivefactorsneed soils at higher elevations). to be considered for the sustainable development of * Maintain a jhum cycle of a minimum these highlands. These interactive integrative link- of ten years (which is critical for ages are shown in Figure 14-3. achieving sustainability) by empha- sizing other systems of land use such Figure 14-3. Integrative Diagram Linking the as the traditional valley cultivation or Three Major Issues and Many Sub-Issues Cru- home gardens. cial for Rural Rehabilitation * Speed up fallow regeneration after jhum by introducing fast-growing Landscape as a unit native shrubs and trees. Site specific * Condense the time span of forest suc- Timeframe (short/long-term strategy) cession and accelerate restoration of Strengthen internal controls and reduce subsidies degraded lands, based on an under- Soil and water conservation/management standing of tree growth strategies and Traditional/appropriate technology . . . . ~~~~~~~~Enhance biodiversity architecture, by adjusting the mix of Resource optimization species in time and space. * Improve animal husbandry through improved breeds of swine and poultry. co]ogi * Redevelop village ecosystems through the introduction of appropriate technology to relieve ura drudgery and improve energy Rehabilitation efficiency (such as cooking stoves, agricultural implements, biogas Socio- Institutional generation, small hydroelectric economic projects); promote crafts such as smithying and products based on leather, bamboo, and other woods. History and causes of degradation Village-level organization * Strengthen conservation measures Cost/benefit sharing Flexibility based on traditional knowledge and ComTnunity participation Monitoring value system. Role of women Credit/marketing Empowerment Links with GOs/NGOs/ Tenurial rights scientists Source: Ramakrishnan 1992a. Value system Incentives . _______________________________________________ _ .Source: Ramakrishnan and others 1994. 202 Currencies for Measuring Sustainability: Case Studies from Asian Highlands Case studies of other shifting The Indonesian transmigration program agricultural systems In many case studies, demographicpressure ham- pers sustainable development. Ill-conceived ag- Indonesia, Malaysia, and Thailand areamong the ricultural activities often are responsible for site countries in the region with highland shifting degradation and the consequent social disrup- agriculture (UNESCO 1983). Social problems re- tions.ndethe uman earemovedtoafrestarea lated to the transfer and implementation of tions. Before humans are moved to forest areas techolog hav bee a ajorimpeimen to us- that are clear-cut, land use needs to be carefully technology have been a major impediment to sus- planned. Such planning should take into consid- tainable development. In the Nabawan Project in eration the social dimensions. The Indonesian Sabah, for instance, farmers were settled to wet rice transmigration i a telling example that has at- cultivation, for which was not suitable for the eco- tracted considerable criticism from both within logicalconditionsofthearea.Theshortageofwater and outside the country (see box 14-2). was severe, and farmers were used to dry land rice cultivation. The project soon had to be abandoned. Case studies from the Himalayan region Problems often arise when alien crop species are introduced to replace upland rice because rice is In many hilly areas of the Himalaya, water is the culturally linked to the people's very existence. key social issue. Land use development and res- Similar social and cultural issues stand in the torationof forestecosystemsareoftenconstrained way of sustainable land use development in the by lack of adequate water outside the monsoon Thai highlands, too. With large-scale migration season. In a hilly terrain, uneven distribution of into the hills from the adjoining lowlands, social rainfall interacting with excessive land degrada- disruptions occur and are aggravated when tech- tion aggravates the shortage of water. nologies more suited to lowland social ethos are In the village Sukhomajri located in the foot- blindly imposed on the highlanders. Patterns of hills of the Himalaya, called the Shiwalik ranges land ownership, for example, are quitedifferent; the near Chand igarh in northwestern India, land was highlander has community-owned lands, whiereas highly degraded with frequent crop failures. This the low-lander has individually owned lands. compelled farmers to emphasize goat and sheep husbandry. Having identified water as the limit- The Nepalese forestry project ing factor, the Hill Resource Management Society (HRMS) of the local community, with scientific The Nepal-Australia forestry program for the support, initiated a watershed management plan Nepalese Himalaya emphasized the people and (Grewal, Mittal, and Singh 1990). Through a se- theecosystemasoneresilient whole(Griffin 1988). ries of earthern dams and water management, in Involving people in the early stages of the pro- just about five years, annual household income gram, rather than at the end, was the key to the increased an estimated Rs2,000 to Rs3,000, due to success of this forestry project. increased food production, better grass cover for Box 14-2: The Indonesian Transmigration Program This is the world's largest program for voluntary assisted migration, involving at least 2.5 million people since 1985. On the face of it, it is sensible to move people from overcrowded and degraded lands of Java, Madura Bali, and Lombok to sparsely populated outer islands. Although many settlers seem to be satisfied with their new environment, manysettlements havebeen established on infertile soilsand endangered ecosystems. Examplesare the endangered heath forests (Kerangas) on infertile white sands of Bangka Island, south Sumatra, and the ultra basic soil in Southeast Sulawesi. Migration also causes social disruption: sponsored migrants are followed by perhaps twice as many unassisted migrants, who have caused considerable environmental damage and social problems. In particular, theinteraction between migrantsand local inhabitants hascaused social disruption in the past, and this needs to be carefully analyzed to avoid repeating mistakes. On the basis of a detailed analysis, Whitten and others (1987) conclude that transmigration is not the answer to Javas's demographic problems. They rightly point out that, though the program has slowed down, it will continue. What is important is to seek the means of eradicating the root causeof the population explosion inJava and to minimize social disruption by supporting sustainable development in the settlements. 203 Defining and Measuring Sustainability: The Biogeophysical Foundations cattle, and better forest cover for fodder and often resolved through local institutions created fuelwood. People's participation was the key to for ecodevelopment; these institutions often in- success, and other communities in the region teract with nongovernmental voluntary groups initiated restoration programs based on the and government agencies in diverse ecological Sukhmajri model. situations (Agarwal and Narain 1989). The mes- In the restoration study initiated in the central sage that comes across loud and clear is that old Himalayan Kumaon region, water was a limiting social structures could form the basis for trans- factor and bamboo was an economically impor- forming the structureof societies where develop- tant but declining natural resource. Many species ment is initiated. Participatory action- oriented of bamboo,Thamnocalamusspathiflorus, T.falcuneri, research then could become meaningful. Some T. jaunsarensis, and Chimnobambusefalcata, were formoforganizationis,however,fundamental to emphasized as agroforestry and social forestry achieving full participation. species in private and common lands in a cluster of about fifty villages. Additional water was pro- vided through rainwater and subsurface seepage Conclusions harvesting tanks constructed cheaply using lo- cally available resources (Kothyari and others The Asian highlands are characterized by a vari- 1991). This participatory research cum develop- etyof traditional societiesoccupyingdistincteco- ment program involved a few hundred people logical niches. In such a context, sustainable de- from the local community during a short period velopment of local resources has to be based on a of six months! value system that people can understand and Now a series of rainwater harvesting tanks are appreciate and therefore choose to participate in operating in the Himalayan belt, and water is the the process of development. Traditional knowl- key factor catalyzing a variety of land use devel- edge and skills should be integrated. The unit for opment and restoration activities such as the re- development may be a village, a cluster of vil- development of hill agroecosystems, manage- lages, or a watershed. Location-specificity is im- ment of forests for fodder and fuelwood, and portant. Village-level institutions with participa- development of integrated watersheds. Sustain- tion from government and nongovernmental abledevelopment isbased on the participation of agencies and with special representation for people and communities. women (who often play a key role in traditional Building village-level organizations is criti- societies) could form the basis for a bottom-up cal in any effort to achieve sustainable develop- approach to institution building that scales up to ment. Impressive social forestry programs have the district level. been carried out in China over some decades In the Asian highlands, land degradation due through collective units such as people's com- toimproperuseofforestresourcesisapriorityfor munes and production brigades on their own people-oriented sustainable development. Initia- land; similarly, massive fuelwood plantations tivestopromotethesustainableuseofland should havebeenestablished and degraded forest lands considerecological processesand social concerns. rehabilitated in the Republic of Korea through A variety of currencies could be used to evaluate village forestry associations (FAO 1979, 1982). and monitor the system over a period of time and In Bankura in West Bengal, India, tribal women to allow possible and necessary reconciliations to organized themselves to restore the forest eco- be made, as illustrated by the northeast Indian system using a three-tier societal framework casestudy.Itisnolongerpossibletoviewecology starting at the village level; this effort has been merely as a natural science, although natural an income-generating and ecologically valu- science forms the foundation for studying eco- able activity. logical processes. One needs not only to draw Forest protection committees created by the parallels between studies of ecological and social Bengal Forest Development in India involve vil- processesbut also to seek out cross connections lagers living along the periphery of forest domi- (Ramakrishnan 1992a). We have very few com- nated by Shorea robusta (Sal), with effective and prehensive case studies on which to base our profitable results (Malhotra and Poffenberger conclusions. The International Sustainable Bio- 1989). Conflicts arising at the village level be- spherelnitiative(HuntleyandothersI991)should cause of social inequality and stratification are help to catalyze studies in this direction. 204 Currencies for Measuring Sustainability: Case Studies from Asian Highlands References Huntley, B. J., E. Ezucurra, E. R. Fuentes, K. Fugii, P. J. Grubb, W. Haber, J. R. E. Harger, M. M. Holland, S. A. Levin, J. Lubchenco, H. Agarwal,A.,and S.Narain. 1989. "TowardsGreen A. Mooney, V. Neronov, 1. Noble, H. Villages." CentreforScienceandEnvironment, Ronald Pulliam, P. S. Ramakrishnan, P. G. New Delhi. Risser, 0. Sala, J. Sarukhan, and W. G. Borthakur, D. N., A. Singh, R. P. Awasthi, and R. Sombrock. 1991. "A Sustainable Biosphere: N.Rai. 1978. "ShiftingCultivationin theNorth- The Global Imperative." Ecology Interna- eastern Region." in Proceedings of the National tional 20, pp. 5-14. Seminar Reso>urces, Development, and Environ- Janzen, D. H. 1973. "Tropical Agro-ecosystems." ment in the Himalayan Region, pp. 330-42. New Science 182 pp. 1212-19. Delhi: Government of India, Department of Siencam, p. 1212-9.S Science and Technology. Khiewtam, R. S., and P. S. Ramakrishnan. 1989. "Socio-cultural Studiesof theSacred Grovesat Chandrasekhara,U.M. 1991. "Studieson the Gap Cherrapunji and Adjoining Areas in North- Phase Dynamics of a Humid Tropical Forest." eastern India." Man in India 69, pp. 64-71. Ph.D.diss.,JawaharlalNehru University,New Kothyari, B. P., K. S. Rao, K. G. Saxena, T. Kumar, and P. S. Ramakrishnan. 1991. "Institutional FAO (Food and Agriculture Organization of the Approaches in Development and Transfer of United Nations). 1979. China: Mass Mobiliza- Water Harvest Technology in the Himalaya." tionofRural CommunitiesforReforestation. Rome. In G. Tsakiris, ed., Advances in Water Resources . 1982. Village Forestry Development in the Technology,pp.673-78. Rotterdam: ECOWARM Republic of Korea: A Case Study. FAO/SIDA, A. A. Balkema. Forestry for Local Community Development Kumar, A., and P. S. Ramakrishnan. 1990. "En- Programme, Rome. ergy Flow through an Apatani Village Ecosys- FAO/SIDA (Food and Agriculture Organization tem of Aru nachal Pradesh in Northeast India." of the United Nations/Swedish International Human Ecology 18, pp. 315-36. Development Agency). 1974. Reporton Regional Litsinger, J. A., and K. Moody. 1976. "Integrated Seminar on Shifting Cultivation and Soil Conser- Pest Management in Multiple Cropping Sys- vation in Africa. Rome. tems." In M. Stelly, ed., Multiple Cropping, pp. Fujisaka, S., P. Sajise, and R. del Castillo, eds. 293-316. Madison, Wisc.: American Society of 1986. Man, Agriculture, and the Tropical Forest: Agronomy. Change and Development in the Philippine Up- Maikhuri, R. K., and P. S. Ramakrishnan. 1990. lands. Bangkok: Winrock International. "Ecological Analysis of a Cluster of Villages Gangwar, A. K., and P. S. Ramakrishnan. 1987. Emphasizing Land Use of Different Tribes in "Agriculture and Animal Husbandry among Meghalaya in Northeast India." Agriculture theSulungsand Nishisof ArunachalPradesh." Ecosystem and Environment 31, pp. 17-37. Social Action 37, pp. 345-72. Malhotra, K. C., and M. Poffenberger. 1989. "For- - . 1989. "Ecosystem Function in a Khasi est Regeneration through Community Protec- Village of the Desertified Cherrapunji Area in tion." West Bengal Forest Department, Northeast India." Proceedings of the Indian Acad- Calcutta. emy of Science (Plant Science) 99, pp. 199-210. Mishra, B. K., and P. S. Ramakrishnan. 1984. Gokhale, A. M., D. K. Zeliang, R. Kevichusa, and "Nitrogen Budget under Rotational Bush Fal- T. Angami. 1985. "Nagaland: The Use of Alder low Agriculture (Jhum) at Higher Elevations Trees." Education Department, Kohima, of Meghalaya in Northeastern India." Plant Nagaland. and Soil 81, pp. 37-46. Grewal, S. S., S. P. Mittal, and G. Singh. 1990. Patnaik, S., and P.S. Ramakrishnan. 1989. "Com- "Rehabilitation of Degraded Lands in the Hi- parative Study of Energy Flow through Vil- malayan Foothills: Peoples Participation." lage Ecosystems of Two Co-existing Commu- Ambio 19, pp. 45-48. nities(theKhasisand theNepalis)ofMeghalaya Griffin, D. M. 1988. Innocents Abroad in the Forests in Northeast India." Agricultural Systems 30, of Nepal. Canberra: Anutech Pty. pp. 245-67. 205 Defining and Measuring Sustainability: The Biogeophysical Foundations Ramakrishnan, P. S. 1984. "The Science behind Rao, K. S., and P. S. Ramakrishnan. 1989. "Role of Rotational Bush Fallow Agriculture System BamboosinNutrientConservationduringSec- (Jhum)." Proceedings of the Indian Academy of ondary Succession following Slash and Burn Science (Plant Science) 93, pp. 397-400. Agriculture (Jhum) in Northeast India." Jour- . 1985. "Tribal Man in the Humid Tropics nal of Applied Ecology 26, pp. 625-33. of the Northeast." Man in India 65, pp. 1-32. Rappaport, R. A. 1971. "The Flow of Energy in an .1986. "Morphometric Analysis of Growth Agricultural Society." Scientific American 225, and Architecture of Tropical Trees and Their pp. 117-32. Ecological Significance," pp. 209-22. Compte- Ruthenberg, H. 1971. Farming Systems in the Trop- rendu du Colloque International l'Arbre, Sep- ics. Oxford, England: Clarendon Press. tember 9-14, 1985, Montpellier. Naturalia Sharma, E., and R. S. Ambasht. 1988. "Nitrogen Monspeliensia, numero hors serie. Montpellier. Accretion and Its Energetics in the Himalayan . 1989a. "Conservation Strategies: An Alder." Functional Ecology 2, pp. 229-35. Agroecologist's Viewpoint." In M. L. Trivedi, Swamy,P.S.,and P.S. Ramakrishnan. 1987."Con- B. S. Gill, and S. S. Saini, eds., Plant Science tribution of Mikania micrantha H.B.K. during Research in India, pp. 25-38. New Delhi: Today Secondary Succession Following Slash and and Tomorrow's Publications. Burn Agriculture (Jhum) in Northeast India. I: . 1989b. "Nutrient Cycling in Forest Fal- Biomass, Litterfall, and Productivity." Forest lows in Northeastern India." In J. Proctor, ed., Ecology and Management 22, pp. 229-37. Mineral Nutrients in 'Tropical Forest and Savanna _ . 1988. "Nutrient Budget under Slash and Ecosystems, pp. 337-52. Oxford, England: Burn Agriculture (Jhum) with Different Weed- Blackwell Scientific Publications. ing Regimes in Northeastern India." Acta .1991. "Biological Invasion in the Tropics: Oecologica-Oecologia Applicata 9, pp. 85-102. An Overview." In P. S. Ramakrishnan, ed., Ecol- Toky, 0. P., and P. S. Ramakrishnan. 1981. "Crop- ogy of Biological Inzvasion in the Tropics, pp. 1-19. ping and Yields in Agricultural Systems of the New Delhi: National Institute of Ecology. Northeastern Hill Region of India." .1992a. Shifting Agricultureand Sustainable Agro-Ecosystems 7, pp. 11-25. Development of Northeastern India. UNESCO- _ . 1982. "A Comparative Study of the En- MAB Series. Paris: Parthenon Publications. ergy Budget of Hill Agro-ecosystems with .1992b. TropicalForest: Exploitation,Conser- Emphasison theSlashand Burn System (Jhum) vation,andManagementIrnpact.Paris:UNESCO. at Lower Elevations of Northeastern India." Ramakrishnan, P. S., and 0. P. Toky. 1981. "Soil Agricultural Systems 9, pp. 143-54. Nutrient Status of Hill Agro-ecosystems and - . 1983a. "Secondary Succession Following RecoveryPattern afterSlashand Burn Agricul- Slash and Bum Agriculture in Northeastern ture (Jhum) in Northeastern India." Plant and India. I. Biomass, Litterfall, and Productivity." Soil 60, pp. 41-64. Journal of Ecology 71, pp. 735-45. Ramakrishnan, P. S., and P. M. Vitousek. 1989. . 1983b. "Secondary Succession Following "Ecosystem Level Processes and the Conse- Slash and Bum Agriculture in Northeastern quencesof Biological Invasions." InJ. A. Drake, India. II. Nutrient cycling." Journal of Ecology H. A. Mooney, F. di Castri, R. H. Groves, F. G. 71, pp. 747-57. Kruger, M. Rejmanck, and M. Williamson,eds., UNESCO (United Nations Educational, Scien- Biological Invasions: A Global Perspective, pp. tific,and Cultural Organization). 1983. Swidden 281-300. SCOPE 37. Chichester, England: John Cultivation in Asia. Vol. 2: Country Profiles. Wiley and Sons. Bangkok: UNESCO Regional Office. Rarakrishnan,P.S.,l.Campbell,L.Demierre,A.Gyi, Watters, R. F. 1960. "Some Forms of Shifting K. C. Malhotra, S. Mehndiratta, S. N. Rai, and Cultivation in the Southwest Pacific." Journal Sashidharan, E.M. 1904. Ecosystem Rehabilita- of Tropical Geography 14, pp. 35-50. tion of the Rural Landscape in South and Central Whitten, A. J., H. Haernman, H. S. Alikodra, and Asia: An Analysis of Issues. Special Publication, M.Thohari. 1987. Transmigration and the Environ- UNESCO, Regional Officeof Scienceand Tech- tnent in Indonesia. Gland, Switzerland: Intema- nology forSouth and Central Asia, New Delhi. tional Union for the Conservation of Nature. 206 Large Marine Ecosystems and Fisheries Kenneth Sherman "Environmental degradation is not inevitable; it is simply cheaper and easierfor some in the short term. Environmental health also is not inconsistent with economic imperatives and political realities. In fact, a healthy environment is the basis fora healthy economy. Ecosystem ecology provides an importan! and useful approach both for assessing and for helping to restore the "health" of the biosphere. " G. E. Likens 1992 Human intervention and changes in climate are 1990). Recent studies implicate changes in cli- sources of increasing variability in the natural mate and the natural environment as a prime productivity of the world's ocean. Overfishing driving force of variability in the level of fish hascaused multimillion-metric-ton flipsinbiom- populations (Alheit and Bernal 1992; Bakun 1992; ass among the dominant pelagic components of Kawasaki and others 1991). The growing aware- the fish community off the northeastern United ness that biomass yields are being influenced by States (Fogarty and others 1991; Sherman 1991; multiple but different driving forces in marine Sissenwine 1986). The biomass flip, wherein a ecosystems around the globe has accelerated ef- dominant species rapidly drops to a low level to forts to broaden research strategies to encompass be succeeded by another species, can generate the effects of food chain dynamics, environmen- cascading effects among other important compo- tal perturbations, and pollution on living marine nents of the ecosystem, including marine birds resources from an ecosystem perspective. (Powers and Brown 1987), marine mammals, and Mitigating actions to reduce stress on living zooplankton (Overholtz and Nicolas 1979; Payne resources within marine ecosystems are required and others 1990). Other sources of perturbations to ensure the long-term sustainability of biomass tomarinepopulationscausedbytheintervention yields. The principles adopted by coastal states of humans include incidental catches of marine under the terms of the United Nations Conven- mammals in fishing nets and the growing im- tion for the Law of the Sea have been interpreted pacts of pollution. Efforts to reduce stress and as supporting the management of living marine mortality amongmarinemammalscaught in fish- resources from an ecosystem perspective (Belsky ingnetsarebeingpursued(Bonner1982;Loughlin 1986,1989). However, at present no single inter- and Nelson 1986; Waringand others 1990). Pollu- national institutional regime has been empow- tion at the continental margins of marine ecosys- ered to monitor the changing ecological states of tems that affects cycles of natural productivity, large marine ecosystems and to reconcile the including eutrophication caused by high nitro- needs of individual nations with those of the gen and phosphorus effluent from estuaries, the community of nations (Myers 1990). In this re- presence of toxins in poorly treated sewage dis- gard, the need for a regional approach to imple- charge, and lossof wetland nurseryareasto coastal ment research, monitoring, and the mitigation of development, is also being addressed (GESAMP stress in support of the development and Defining and Measuring Sustainability: The Biogeophysical Foundations sustainability of resources at less than the global extensive areas of ocean space of approximately level has been recognized from a strategic per- 200,000 square kilometers or greater, character- spective (Malone 1991; Taylor and Groom 1989). ized by distinct bathymetry, hydrography, pro- From the ecological perspective, the concept that ductivity, and trophically dependent populations critical processes controlling the structure and (Sherman and Alexander 1989; Sherman and oth- function of biological communities can best be ers 1990). The concept of large marine ecosystems addressed on a regional basis (Ricklefs 1987) has defines the unit of resource interest on the order been applied to ocean space in the use of marine of thousands of kilometers in scale with regard to ecosystems as distinct global units for marine fish and fisheries yieldsand representsan energy research, monitoring, and management. The con- flow approach to factors determining variability cept of monitoring and managing renewable re- in theecosystem's productivity. In thisapproach, sources from the perspective of a regional ecosys- large-scale trophic, environmental, and climatic tem was the topic of a series of symposia and changes are examined in relation to the effects of workshops initiated in 1984 and continuing fishery removals on the long-term sustainability through 1992, wherein the geographic extent of of marine ecosystems. each region wasdefined on thebasisof ecological Temporal and spatial scales influencing bio- criteria. Under this approach, the regional units logical production in marine ecosystems have under consideration are referred to as large ma- becen the topic of a number of theoretical and rine ecosystems (see table 15-1). These units are empirical studies. The selection of scale in any Table 15-1: Countries and Large Marine Ecosystems Accounting for 95 Percent of the Annual Global Catch, by Share of the Total, 1987 Percentage of Country Lar,ge marine ecosystem global catch Japan Oyashio Current, Kuroshio Current, Sea of Okhotsk, Sea of Japan, Yellow Sea, East China Sea, West Bering Sea, East Bering Sea, and Scotia Sea 14.43 Former Soviet Union Sea of Okhotsk, Barcnts Sea, Norwegian Shelf, West Bering Sea, East Bering Sea, and Scotia Sea 12.63 United States Northeast United States Shelf, Southcast United States Shelf, Gulf of Mexico, California Currcnt, Gulf of Alaska, and East Bering Sea 7.03 China West Bering Sea, Yellow Sea, East China Sea, and South China Sea 6.72 Chile Humboldt Currcnt 5.98 Peru Humboldt Current 5.65 Subtotal 52.44 Korea, Federal Republic Yellow Sea, Sea of Japan, East China Sea, and Kuroshio Current 3.50 Thailand South China Sea and Indonesian Seas 2.48 Indonesia Indonesian Seas 2.45 Norway Norwegian Shelf and Barents Sea 2.40 India Bay of Bengal and Arabian Sea 2.09 Denmark Baltic Sea and North Sea 2.07 Iceland Icelandic Shelf 2.02 Korea, Dem. People's Rep. Sea of Japan and Yellow Sea 1.99 Philippines South China Sea and Sulu-Celebes Sca 1.78 Canada Scotian Shelf, Northeast United States Shelf, and Newfoundland Shelf 1.75 Subtotal 22.53 Cumulative total 74.97 (table continues on next page) 208 Large Marine Ecosystems and Fisheries Spain Iberian Coastal Current and Canary Current 1.69 Mexico Gulf of California, Gulf of Mexico, and California Current 1.55 South Africa Benguela Current and Agulhas Current 1.12 France North Sea, Biscay-Celtic Shelf, and Mediterranean Sea 1.00 Subtotal 5.36 Cumulative total 80.33 Ecuador Humboldt Current 0.84 United Kingdom & Scotland North Sea 0.82 Poland Baltic Sea 0.80 Viet Nam South China Sea 0.77 Malaysia Gulf of Thailand, Andaman Sea, Indonesian Seas, and South China Sea 0.74 Brazil Patagonian Shelf and Brazil Current 0.72 Turkey Black Sea and Mediterranean Sea 0.72 Argentina Patagonian Shelf 0.69 Namibia Benguela Current 0.64 Italy Mediterranean Sea O.f2 Morocco Canary Current 0.61 New Zealand New Zealand Shelf Ecosystem 0.54 Netherlands North Sea 0.53 Portugal Iberian Shelf and Canary Current 0.49 Faeroe Islands Faeroe Plateau 0.44 Subtotal 9.97 Cumulative total 90.30 Pakistan Bay of Bengal 0.42 Ghana Gulf of Guinea 0.40 Senegal Gulf of Guinea and Canary Current 0.35 Venezuela Caribbean Sea 0.34 Ireland Biscay-Celtic Shelf 0.31 U.K., England, Wales North Sea 0.30 Bangladesh Bay of Bengal 0.29 Hong Kong South China Sea 0.28 Sweden Baltic Sea 0.26 Australia North Australian Shelf and Great Barrier Reef 0.25 Cuba Caribbean Sea 0.25 Romania Black Sea 0.25 German Democratic Rep. Baltic Sea and Scotia Sea 0.22 Panama California Current and Caribbean Sea 0.21 Sri Lanka Bay of Bengal 0.19 Nigeria Gulf of Guinea 0.18 Uruguay Patagonian Shelf 0.17 Finland Baltic Sea 0.16 Subtotal 4.83 Cumulative total 95.13 Source: Based on fish catch statistics from FAO 1989. 209 Defining and Measuring Sustainability: The Biogeophysical Foundations study is related to the processes under investiga- ranean Sea large marine ecosystem. In others, tion. An excellent treatment of this topic can be geographic limits are defined by the scope of found in Steele (1988), which indicates that in continental margins. Among these are the U.S. relation to general ecology of the sea, the best- Northeast Continental Shelf, the East Greenland known work in the dynamics of fish populations Sea, and the Northwestern Australian Shelf. The are studies by Schaefer (1954) and Beverton and seawardlimitof largemarineecosystemsextends Holt (1957), following the earlier pioneering ap- beyond the physical outer limits of the shelves proach of Lindemann (1942). However, as Steele themselves to include all or a portion of the con- (1988)notes,thisarrayofmodelsisunsuitablefor tinental slopes as well. Care has been taken to considering temporal or spatial variability in the limit the seaward boundaries to the areas affected ocean.Thelargemarineecosystemapproachover- by ocean currents, rather than relying simply on comesthisdifficultybydefiningaspatialdomain the limits of the 200-mile exclusive economic based on ecological principles and, thereby, pro- zone or fisheries zone. Among the ocean current's viding a basis for focused scientific research and large marine ecosystems are the Humboldt Cur- monitoring in support of the long-term produc- rent, Canary Current, and Kuroshio Current. The tivity and sustainability of marine resources. Fish large marine ecosystems that together produce componentshave adapted reproductive, growth, approximately 95 percent of the annual yield of and feeding strategies to the distinct environ- global fisheries biomass are listed in table 15-1. mental conditions within the ecosystem. Changes Although the Food and Agriculture to the components of the system through the Organization's world fishery statistics show an removaloffishcantriggeracascadeeffectinvolv- upward trend in annual biomass yields for the ing higher trophic levels including birds and past three decades, it is largely the clupeids that marine mammals and lower trophic levels in- are increasing in abundance (FAO 1989). Large cluding zooplankton and the economies depen- numbers of stocks have been and continue to be denton theresources of theecosystem. Thetheory fished at levels above long-term sustainability. and modeling relevant to measuring the chang- The variations in levels of abundance among ing states of large marine ecosystems are imbed- species constituting the annual yield of global ded in contemporary studies of multistable eco- biomass are indicative of changing regional eco- systems (Beddington 1986; Holling 1973, 1986; system states caused by natural environmental Pimm 1984) and pattern formation and spatial perturbations, overexploitation, and pollution. diffusion in ecosystems (Levin 1978,1990). Although the spatial dimensions preclude a strictly controlled experimental approach to their study, large marine ecosystems are perfectly Large marine ecosystems as global amenable to the comparative method of science management units as described by Mayr (1982). Since 1984, thirty case studies investigating the major causes of Nearly95percentoftheusableannualyieldofthe large-scale perturbations in biomass yields of global biomass of fish and other living marine large marine ecosystems have been completed resources is produced in large marineecosystems (see table 15-2). within, and adjacent to, the boundaries of the exclusive economic zones of coastal nations lo- cated around the margins of theocean basins. The Historical perspective major biomass is caught within the geographic limits of forty-nine large marine ecosystems For nearly seventy-five years, beginning around (whose boundaries are depicted in figure 15-1). the turn of the century, fishery scientists were Criteria used for defining the geographic limits of preoccupied with assessing single-species stock, these ecosystems include distinct bathymetry, although biological oceanographers did not hydrography, productivity, and trophically de- achieve any great success in predicting fish yield pendent populations. Several occupy semi-en- based on food chain studies. As a result, through closed seas, such as the Black Sea, the Mediterra- the mid-1970s, predictions of the levels of biom- nean Sea, and the Caribbean Sea. Some can be ass yields for different regions of the world's divided into domains, or subsystems, such as the ocean were open to disagreement (Alverson, Adriatic Sea, which isa subsystem of the Mediter- Longhurst, and Gulland 1970; Lasker 1988; Ryther 210 Large Marine Ecosystems and Fisheries Figure 15-1: World Map of Large Marine Ecosystems Arcetic Ocan 17 Commonvmatth of 20 _z ! s 6Ft slKy ~~~~~~~~~~Commomvjealth of IndependentStaWs 48 United! iFT. N~orth 24 NW orth Pacific eovCn 1k F|sS {rnti, 21> 4 ~~~~~~~~27 Eauctor ~~~~~~~~~~~~16 \~ ~~~ ~ ~~~~~~~~ ~~ ~~~~~~~~~~~~~ n a n/ - 2S~~~~~~~~ \ndian Occan Ocgan AntarcUra 49 1. Eastern Bering Sea 25. Mediterranean Sea 2. Gulf of Alaska 26. Black Sea 3. California Current 27. Canary Current 4. Gulf of California 28. Guinea Current 5. Gulf of Mexico 29. Benguela Current 6. Southeast U.S. Continental Shelf 30. Agulhas Current 7. Northeast U.S. Continental Shelf 31. Somali Coastal Current 8. Scotian Shelf 32. Arabian Sea 9. Newfoundland Shelf 33. Red Sea 10. West Greenland Shelf 34. Bay of Bengal 11. Insular Pacific--Hawaiian 35. South China Sea 12. Caribbean Sea 36. Sulu-Celebes Seas 13. Humnboldt Current 37. Indonesian Seas 14. Patagonian Shelf 38. Northern Australian Shelf 15. Brazil Current 39. Great Barrier Reef 16. Northeast Brazil Shelf 40. New Zealand Shelf 17. East Greenland Shelf 41. East China Sea 18. Iceland Shelf 42. Yellow Sea 19. Barents Sea 43. Kuroshio Current 20. Norwegian Shelf 44. Sea of Japan 21. North Sca 45. Oyashio Current 22. Baltic Sea 46. Sea of Okhotsk 23. Celtic-Biscay Shelf 47. West Bering Sea 24. Iberian Coastal 48. Faroe Plateau 49. Antarctic 211 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 15-2: Twenty-Nine Large Marine Ecosystems and Subsystems for which Principal, Secondary, or Tertiary Driving Forces Controlling Variability in Biomass Yields Had Been Synthesized as of February 1991 Large marine ecosystem Author and reference U.S. Northeast Continental Shelf Sissenwine (Sherman and Alexander 1986) Falkowski (Sherman, Alexander, and Gold 1991) U.S. Southeast Continental Shelf Yoder (Sherman, Alexander, and Gold 1991) Gulf of Mexico Richards and McGowan (Sherman and Alexander 1989) Brown and others (Sherman, Alexander, and Gold 1991) California Current MacCall (Sherman and Alexander 1986) Mullin (Sherman, Alexander, and Gold 1991) Bottom (Sherman, Alexander, and Gold 1993) Eastern Bering Shelf Incze and Schumacher (Sherman and Alexander 1986) West Greenland Shelf Hovgaard and Buch (Sherman, Alexander, and Gold 1990) Norwegian Sea Ellertsen and others (Sherman, Alexander, and Gold 1990) Barcnts Sea Skjoldal and Rey (Sherman and Alexander 1989) Borisov (Sherman, Alexander, and Gold 1991) North Sea Daan (Sherman and Alexander 1986) Baltic Sea Kullenberg (Sherman and Alexander 1986; Sherman, Alexander, and Gold 1993) Iberian Coastal Wyatt and Perez-Gandaras (Sherman and Alexander 1989) Mediterranean-Adriatic Sea Bombace (Sherman, Alexander, and Gold 1993) Canary Current Bas (Sherman, Alexander, and Gold 1993) Gulf of Guinea Binet and Marchal (Sherman, Alexander, and Gold 1993) Benguela Current Crawford and others (Sherman and Alexander 1989) Patagonian Shelf Bakun (Sherman, Alexander, and Gold 1993) Caribbean Sea Richards and Bohnsack (Sherman, Alexander, and Gold 1990) South China Sea-Gulf of Thailand Piyakarnchana (Sherman and Alexander 1989) Yellow Sea Tang (Sherman and Alexander 1989) Sea of Okhotsk Kusnetsov (Sherman, Alexander, and Gold 1993) Humboldt Current Alheit and Bernal (Sherman, Alexander, and Gold 1993) Indonesia Seas-Banda Sea Zijlstra and Baars (Sherman, Alexander, and Gold 1990) Bay of Bengal Dwivedi (Sherman, Alexander, and Gold 1993) Antarctic Marine Scully and others (Sherman and Alexander 1986; Sherman, Alexander, and Gold 1993) Weddell Sea Hempel (Sherman, Alexander, and Gold 1990) Kuroshio Current Terazaki (Sherman and Alexander 1989) Oyashio Current Minoda (Sherman and Alexander 1989) Great Barrier Reef Bradbury and Mundy (Sherman and Alexander 1989) Kelleher (Sherman, Alexander, and Gold 1993) South China Sea l'aulv and Christensen (Sherman, Alexander, and Gold 1993) 1969). A milestone in fishery science wasachieved ersal finfish community before 1960 to become a in 1975 when the International Council for the dominant demersal community from the mid- Exploration of the Sea convened a symposium 1960s through the mid-1970s. Although no con- that focused on changes in the fish stocks of the sensus on cause and effect was reached, the con- NorthSea and theircauses. Thesymposium, which vener suggested that previous studies may have dealt with the North Sea as an ecosystem, follow- been too narrowly focused and that future stud- ing the lead of Steele (1974), Cushing (1975), ies should take into consideration, from an eco- Andersen and Ursin (1977), and others, was systems perspective, fish stocks, their competi- prompted by a rather dramatic slhift in the domi- tors, predators, and prey, and their interactions nance of the finfish species of the North Sea, with the environment, fisheries, and pollution which changed from a balanced pelagic and dem- (Hempel 1978). 212 Large Marine Ecosystems and Fisheries Perturbations and driving forces in large tion of the coastal zone of large marine ecosys- marine ecosystems tems is reduced and does not become a princi- pal driving force. Concerns remain regarding Marine scientists, geographers, economists, gov- the socioeconomic and political difficulties in ernment representatives, and lawyers are becom- management across national boundaries, as in ing more and more aware of the utility of taking the case of the Sea of Japan ecosystem, where a more holistic ecosystem approach to resource five countries share fishery resources (Morgan management (Alexander 1989; Belsky 1989;Byrne 1988), or the North Sea ecosystem, or the Carib- 1986; Christy 1986; Crawford, Shannon, and bean Sea ecosystem, where thirty-eight nations Shelton 1989; Morgan 1989; Prescott 1989). The share resources. principal driving forces for changes in biomass Somemarinescientistsconsiderchanges in the vary among ecosystems (Sherman and others ocean climate of the northern North Atlantic dur- 1990). On a global scale, the loss of sustained ingthelatel960sandearlyl970sasthedominant biomass yields from large marine ecosystems as a cause of change in the structure of the food chain result of mismanagement and overexploitation and biomass yields of at least three northern has not been fully investigated but is likely very North Atlantic large marine ecosystems. The large (Gulland 1984). It is clear that "experts" population of important fish stocks (such as cape- have been off the mark in earlier estimates of hn and cod) has declined on a large scale within globalyieldoffisheriesbiomass.Projectionsgiven the Norwegian Sea, Barents Sea, and West in The Global 2000 Report (U.S. Council on Envi- Greenland Seaecosystems. In theWestGreenland ronmental Quality 1980) expected the world's Sea ecosystem, cod stocks have been displaced annual yield to rise little, if at all, by the year 2000 southward since 1980, attended by a decrease in from the 60 million metric tons reached in the their averagesize and abundance. Biomassyields 1970s. In contrast, estimates given in The Re- declined from about 300,000 metric tons a year in sourceful Earth (Wise 1984) argue for an annual the mid-1960s to less than 15,000 metric tons in yield of 100 million to 120 million metric tons by 1985. Both changes appear to have been due to the year 2000. The trend is upward; the 1988 short-term cooling that influenced the stability of yields of marine global fishery reached 86.8 mil- water masses and the dynamics of the plankton lion metric tons (FAO 1990). The lack of a clear community, adversely affecting the growth and definition of actual and potential global yield is survival of early developmental stagesof cod and not unexpected, given the limited efforts pres- reducing recruitment. Since the 1920s, the annual ently under way to improve the base of global biomass yield of cod has been related to tempera- information on yields of living marine resources. ture, with catches increasing during warm peri- More and more attention has been focused ods and declining during cool periods. The ef- overthepast fewyearsonsynthesizingbiological fects of fishing mortality on the decline of the cod and environmental information on the natural are secondary to the major influence of climatic productivity of the fishery biomass within large conditions over the North Atlantic (Hovgaard marine ecosystems in an effort to identify the and Buch 1990). principal, secondary, and where important, the To the east, changes in the temperature struc- tertiary forces causing major shifts in the species ture of the Norwegian Sea ecosystem appear to be composition of biomass yields. Effective manage- the major force controlling the recruitment of ment from an ecosystems perspective will be importantcod stocks. Strongormediumproduc- contingent on identifying these forces. Man- tion of cod biomass is related to warmer tempera- agement of species responding to strong envi- tures. The conditions for growth and survival of ronmental signals will be enhanced by improv- early developmental stages of cod are enhanced ing our understanding of the physical factors during warmer years, when the larval cod are forcing biological changes, whereas in other maintained forlongerperiods withincoastal nurs- large marine ecosystems when the prime driv- ery grounds, where their most important prey ing force is predation-either by natural preda- organism-the copepod, Calanus finmarchicus- tors or by humans expressed as excessive fish- swarms in high densities under conditions of ing mortalities-options can be explored for well-defined thermocline structure and conse- implementing adaptive management strategies. quently under optimal conditions for feeding on Mitigation is required to ensure that the pollu- the abundant phytoplankton. 213 Defining and Measuring Sustainability: The Biogeophysical Foundations The changes in biomass yields of the Barents the southwest coast of Africa, the long-term fluc- Sea ecosystem have been attributed primarily to tuations in the abundance of pilchard, horse- changes in hydrographic conditions and second- mackerel, and hakes are attributed to changes in arily to excessive fishing mortality. The average the oceanographic regime (see figure 15-2). The annual biomass yield of the ecosystem in the Benguela large marine ecosystem is bounded by 1970s was about 2 million metric tons (fish, crus- warm water at both extremes: that toward the taceans, molluscs, and algae). However, by the equator and that toward the South Pole. Cold, 1980s annual yields declined to approximately nutrient-rich water is upwelled with moderate 350,000 metric tons. The decline of warm Atlantic intensity in thecentral section and more intensely water flowing into the Barents Sea ecosystem, in the northern and southern areas. Environmen- coupled withexcessivelevelsof fishingeffort, led tal conditions favor either the epipelagic or the to (1) collapse of the major fisheries of the region demersal species, never both simultaneously, and (cod, capelin, haddock, herring, redfish, and have been the principal driving force for large- shrimp), (2) subsequent disruption in the struc- scale shifts in abundance among the fish species. ture of the food chain, and (3) increase in the The effects of the fisheries on changes in species abundance of the shrimp-like euphausiids repre- abundanceare secondary. Changesinabundance senting a significant amount of biomass that is of pilchard stocks have led to detectable effects in underused in relation to the potential sustained the abundance of dependent predator species, yield of this ecosystem. Given the depressed state particularly marine birds (Crawford, Shannon, of the fish stocks, any restoration management and Shelton 1989). would need to consider significantly reducing The greatest increases in biomass yields in the the fishing effort of the fishermen of Norway and Pacific havebeenreported at theareaof confluence the former Soviet Union, the coastal nations that between the Oyashio and Kuroshio Current eco- share the resources of the Barents Sea ecosystem systems off Japan (Minoda 1989; Terazaki 1989) (Borisov 1991; Skjoldal and Rey 1989). and in theHumboldtCurrentecosystemoffChile. In the North Sea ecos.ystem, important species In the Oyashio and Kuroshio Current ecosys- have flipped from a position of dominance to one of subordination. This biomass flip in the North Sea occurred over the decade of the 1960s. The Figure 15-2: Estimated Biomass of Pilchard finfish stocks of the North Sea ecosvstem have (Sardinops ocellatus), Cape Horse-mackerel been subjected to intensive fishing mortality. The (Tracasurus capensis), and Cape Hakes (Merluccius yields of pelagic herring and mackerel decreased capcnsis and paradoxus) of the Benguela Current from 5 to 1.7 million metric tons, whereas small E fast-growing and cormmercially less desirable sand Biomiass of Pichard !iiotmass of i orse-mriackerel lance, Norway pout, and sprat increased by 1.5 (Allzo1os oJ totns) (Alli,ons of tons) million metric tons along with an approximate 36 percent increase in gadoid yields. The causes for Cape hake( [torso- the biomass flips are poorly understood. Several (1.3,1 4, 1i) rn.atkrul arguments correlate the flip with changing Cape hlakes - oceanographic conditions. Others support v (1.3,1.4,) overexploitation as the major cause. However, none of the arguments can be considered more than speculative at this time, pending rigorous Pitchard F analysisof morerecentinformation (Hempel 1978; Postma and Zijlstra 1988). 5 Farther to the south, the Iberian Shelf ecosys- tem has been examined recently in relation to variability of biomass yield. Alternation in the - abundance of horse mackerel and sardine within the Iberian Coastal ecosystem is attributed to 0 v o changes in natural environmental perturbation 55 60 65 70 ,- 80 of its thermal structure rather than to any density- 6 5ar dependent interaction between the two species. Similarly, in the Benguela Current ecosystem of Source. Crawford, Shannon, and Shelton 1((85( 214 Large Marine Ecosystems and Fisheries tems, the yield of Japanese sardines increased the northern and southemnhemispheresand (2) the from less than one-half million metric tons in 1975 dramatic decline-from about 12 million metric to just over 4 million metric tons in 1984 (see tons in 1970 to less than 2 million metric tons by figure 15-3). The yield of the Chilean sardine in 1976-in the biomass yields of anchovy in the the Humboldt Current ecosystem also increased northemareasof theHumboldtCurrentecosysterm from about 500,000 metric tons in 1974 to 4.3 in the early 1970s (Canon 1986; see figure 15-4). million metric tons in 1986. The increased yields Although less dramatic, the long-term shifts in havebeen attributed todensity-independentpro- the abundance of both sardines and anchovies cesses involvingan increase in productivity of the within the California Current ecosystem are con- lower food chain, made possible by coastward sidered the result primarily of natural environ- shifts in the boundary areas of the Oyashio and mental change and secondarily of intensive fish- Kuroshio systems and shifts of the water mass in ing, rather than of any density-dependent com- the Humboldt Current ecosystem. The effects of petition between the two species (see figure 15-5; fishing on the sardines in both areas are of sec- MacCall 1986). ondary importance compared with the enhanced Changes in biomass yields of two other Pacific productivity of the phytoplankton and zooplank- Rim large marine ecosystems have been the result ton components of the ecosystems, which im- of overexploitation. The introduction of highly proved the environment for growth and recruit- efficient modern trawlers to the Gulf of Thailand ment. Studies are under way to determine the ecosystem led to excessive fishing mortality and extent of the teleconnection between the Pacific- a marked reductioninannualyieldsof biomassof wide El Nifio events of the past decade and both fish for human consumption between 1977 and (I) the multimillion-metric-tonincreasesin yields 1982 (Piyakarnchana 1989; figure 15-6). Intensive of sardines occurring nearly simultaneously in fishery effort resulted in the depletion of the demersal fish stocks and dramatic reductions in the biomass yields of the Yellow Sea ecosystem. Figure 15-3: Catches of Japanese Sardines from the Between 1958 and 1968, fisheries yields declined Area of Confluence between the Oyashio Current Ecosystem and the Kuroshio Current Ecosystem off the Coast of Japan, 1975-84 Figure 15-4: Catches of Anchovies and Sardines 5 ffi from the Waters of the Humboldt Current Ecosystem off the Coasts of Chile and Peru, 1964-83 14 Total catch of Japan El Nifio El Nino El Nifno El Ninio o4 L --E 12 1965 1972-73 1976 1982-83 r 10 4,,Ź\/\ 4, 4, 4,~~~~~~~1 , , " Anchoveta ,5xs 2 0 ,. 4 0 . -- 0i 3 Chile Sardina Espanola Catch off Kush'rx Peru (ardnps s_gac) u 0) 1t 1 : | , 2 , 1 1964 1970 1975 1980 1976 1978 1980 1982 1984 Year Year Souzrce: Nnoda 1989. Source: Canon 1986. 215 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 15-5: Time Series of Sardine (Age 2+) and from 180,000 to less than 10,000 metric tons. The Anchovy Spawning Biomass (Log Scale) of the fishery then shifted to harvesting pelagic stocks, California Current Ecosystem, 1935-85 reaching a level of 200,000 metric tons in 1972, followed by a reduction to less than 20,000 metric ,10, tons in 1981. The fisheries of the Yellow Sea in 1982 X1 t Sardine lage 2+) shifted principally to anchovy and sardine, with a bsc ~ \ A .. total annual yield of all species 40 percent lower 10 a , than the 1958 level. The demersal fishery remains in v,,. a depleted state (Tang 1989; see figure 15-7). E \, ,^ . The importance of a natural predator driving , 101r A Anchovy (spa%%ming) anecosystemisevidentinthelarge-scalechanges A: \ in the community structure of the Great Barrier Reef ecosystem that extends over 230,000 square ____________________________________ kilometers of the Queensland continental shelf. o19- 3 1945 1955 1965 1975 1985 The predation by the crown-of-thorns starfish in the 1960s and 1970s resulted in a shift in the Year biomass of corals, community structure of the benthos, and a decoupling of energy transfer to Note: The area denoted A indLicates the approximnate anchovv several fish stocks. spawning biomass in 194041. To the north and west of Australia lies the Source: MacCall 1986 relatively pristine Banda Sea ecosystem, where no large-scale fisheries are presently conducted. The ecosystem is under the influence of mon- Figure 15-6: Total Catch of Caamivorous Feeding soon-induced seasonal periods of large-scale up- Species of Fish from the Gulf of Thailand Ecosystem, welling and downwelling. Biological feedback to these environmental signals is reflected in the 3000 changes in phytoplankton, mesozooplankton, ------ Nernipterus peroru micronekton, and fish. During upwelling events, , ---Saunda undosq productivity of the ecosystem is enhanced by a 2,600 uanus factor of 2 to 3. The biomass of pelagic fish re- sources is also higher during the upwelling pe- 2,200 * Lutianushneolatus riod. The fish biomass of the ecosystem is esti- \ ''-r-- Scomberomorus mated atbetween600,000and900,000metric tons -- \ \cormnterson in the peak upwelling season (August) and be- 1800 r \ k \ tween 150,000 and 250,000 metric tons in the downwelling period (February). The estimated e 1,400 \ sustained annual biomass yield of the ecosystem is ,... ' i approximately 30,000 metric tons of pelagic fish. 1,a0o \"0 \ / \ Management considerations Several ecologists have reviewed the empirical 2001 ' ' \.-. -u and theoretical aspects of yield models for large c .---. marine ecosystems. According to Beddington 1977 1978 1979 1980 1981 1982 19@ (1986), Daan (1986), Levin (1990), and Mangel 1977 1978 1979 1980 1981 1982 1983 (1991), published dynamic modelsof marineeco- Year systems offer little guidance on the detailed be- havior of communities. However, these authors Source: Plvakamchana 1989 concur on the need for covering the common groundbetweenobservationand theorybyimple- menting monitoring efforts on the large spatial and long temporal scales (decadal) of key compo- nents of the systems. Levin (1990) describes the 216 Large Marne Ecosystems and Fisheries Figure 15-7: Annual Catch of Dominant Species dated component models to provide predictive of the Yellow Sea Ecosystem, 1953-84 models for population dynamics and redistribu- 240 tion. This approach is consistent with the recent F F observation by Mangel (1991) that empirical sup- 200 r - port for the currently used modelsof largemarine *E A, , B^ S \ ecosystems is relatively weak and that a new E 160 F , D\ I \generation of models is needed to enhance the 120 ' ' ' ' | | link between theory and empirical results. rN 120 - , - \ Effective management strategies for large nu- 31 n8 i \ Arine ecosystems will be contingent on identifying 6 8n - , } V - the major forces causing large-scale changes in biomass yields. Management of species respond- ing to strong environmental signals will be en- _ _ _ _ hanced by improving our understanding of the 1960 197( 1980 physical factors forcing biological changes; where )e-ar the prime driving force is predation, options can be explored for implementing adaptive manage- ment strategies. Remedial actions are required to Not> A, small vellow croaker and hairtail. B, Pacific herring ensure that the pollution of the coastal zone is and Japanese mnackerel; C. Schpipoo tonh. anchovv, andi scaled sardine. reduced and does not become a principal driving Sozirc: lang 1989. force in any large marine ecosystem. For at least one system, the Antarctic, a management regime has evolved that is based on an ecosystem per- sequenceforimprovingourunderstandingof the spective in the adoption and implementation of possible mechanisms underlying observed pat- the Convention for the Conservation of Antarctic terns in large marine ecosystems as examination MarineLivingResources(Shermanand Ryanl988). of (1) statistical analyses of observed distribu- Efforts are also under way to manage the large tional patterns of physical and biological vari- marine ecosystemns of the United States' exdusive ables, (2) construction of competing models of economic zone, induding the northern California variability and patchiness based on statistical Current ecosystem (Bottom and others 1989). analyses and natural scales of variability of criti- A systems approach to the management of cal processes, (3) evaluation of competing mod- large marine ecosystems is depicted in table 15-3. els through experimental and theoretical studies These ecosystems represent the link between lo- of component systems, and (4) integrationofvali- cal events (such as fishing, pollution, environ- Table 15-3: Key Spatial and Temporal Scales and Principal Elements of a Systems Approach to the Research and Management of Large Marine Ecosystems Scale or element Description Scale Spatial Global (world's ocean), regional (exdusive economic zones), local Temporal Millennia-decadal, decadal-seasonal, seasonal-daily Unit Pelagic biogeographic, large marine ecosystems, subsystems Element Research element Spawning strategies; feeding strategies; productivity, trophodynamics; stock fluctuations, recruitment, and morutaity; natural variability (hydrography, currents, water masses, weather); human perturbations (fishing, waste dispoWl, petrogenic hydrocarbon impacts, aerosol contaminants, eutrophication effects) Management elements-options and Bioenvironmental and socoeconomic models; management to optimize advice-international, national, local fisheries yields Feedback loop Evaluation of ecosystem status, fisheries status, and management practices 217 Defining and Measuring Sustainability: The Biogeophysical Foundations ment) occurring on the daily-to-seasonal tempo- ment of marine communities, but all re- ral scale and their effects on living marine re- volvearoundthequestionofsustainability. sources and the more ubiquitous global effects of What levels of mortality imposed by a fish- climatic changes on the multidecadal time scale. ery will permit a sustainable yield? Are The regional and temporal focus of season to there levels below which a fish population decade is consistent with the evolved spawning will not recover? Can judicious manipula- and feeding migrations of fish, the keystone spe- tion of the catch composition of the fishery cies of most large marine ecosystems. These sea- alter the potential of the community to pro- sonal migrations occur over hundreds to thou- duce yields of a particular type, e.g., high sands of kilometers within the unique physical value species? Can a community be de- and biological characteristicsof the regional large pleted to a level where its potential for pro- marine ecosystem to which the species have ducing a harvestable resource is reduced? adapted. Because the fisheries represent most of With the exception of the first question the usable biomass yield of the large marine eco- qusn t, a systems and fish populations consist of several rarely explicitly addressed in the scientific age classes, it follows that measures ofvariability bodies of the various fisheries' organiza- in growth, recruitment, and mortality should be tions. Instead, such bodies concentrate on conducted over multivear time scales. The natu- the estimation of stock abundance and the rally occurring environmental events and the calculation of allowable catch levels, al- human-induced perturbations that affect the though often implicit in the advice demography of populations within the ecosys- thesebodiesto managementarea gienb theseodleso manzemenareasetof beliefs tem should be considered. Management options about the answers to such questions. from an ecosystems perspective should be based on scientific inferences of the principal causes of Given the increasing number of responsibilities variability in abundance and should give due consideration to socioeconomic needs. The final of mgovenmenta ieor() managinisheres, element in the system, with regard to the concept (2) mtgatgpolluton,(3) reducgenvronmen- of resource maintenance and sustained yield, is tal stress, and (4) restoring lost habitat, it is not the feedback loop that allows the effects of man- surprising that interest in pursuing resource agement actions to be evaluated at the fisheries mianagement problems from an ecosystem per- level (single species, multiple species) and the spective is growing. ecosystem level. The topic of change and persistence in marine It will be necessary to conduct supportive re- communities and the need for multispecies and ecosystem cers ectives in fishery management search on the processes controlling sustained pro- ys p pey g dutiitoflagytm . W n s relate to reports of changing states of marine ductivityof large marine ecosystems. Within sev- ecstm(Sghradhr]94.olps eral of these ecosystems, including the Northeast ecosystems (Sugihara and others 1984). Collapses Shelf, Gulf of Mexico, California Current, and of the Pacific sardine in the California Current Eastern Bering Sea, important hypotheses con- ecosystem, the pilchard in the Benguela Current ecosystem, and the anchovy in the Humboldt cerned wlth the growing impacts of pollution, Current ecosystem are but a few examples of overexploitation,mand environmental changeston cascading effects on other components of the sustained biomassyipldsare under investigation ecosystem, including marine birds (Burger 1988; (see table15-4). Comparing the results of research Crawford, Shannon, and Shelton 1989; Croxali among systems should allow us to learn how the 1987, MacCall 1986). systems respond and recover from stress and to narrow the context of unresolved problems and capitalize on current research efforts. The list of Ecosystem assessment and monitoring reports describing the effects of biological and physical perturbations on the fisheries biomass The National Marine Fisheries Service of the Na- yields of thirty large marine ecosystems given in tional Oceanic and Atmospheric Administration table 15-3 addresses questions similar to those (NMFS; NOAA) has focused greater emphasis posed a few years ago by Beddington (1984, p. 209): over the past decade on approaching fisheries There are a niumber of scientific questions research from a regional ecosvstem perspective Therc~~~~~~~~~i large marnembecosystemsnwithinquand adjacent which are central to the rational manage- m large marme ecosystems within and adjacent 218 Large Marine Ecosystems and Fisheries Table 154: Selected Hypotheses Concerning Variability in Biomass Yields of Large Marine Ecosystems Ecosystem Predominant variables Hypothesis Oyashio Current, Kuroshio Density-independent Increase in Clupeoid population: Current, California Current, natural environmental Predominant variables influencing changes in biomass of Humboldt Current, Benpuela perturbations clupeoids are increases in water-co[umn productivity result- Current, Iberian Coastal ing from shifts in the direction and flow velocities of the currents and changes in upwelling within the ecosystem Yellow Sea, U.S. Northeast Density-dependent Declines in fish stocks: Predpitous decline in biomass of fish Continental Shelf, Gulf of predation stocks is the result of excessive fishing mortality reducing Tlhailand the probability of reproductive success; losses in biomass are attnbuted to excesses of human predation expressed as overfishing Great Barrier Reef Density-dependent Change in ecosystem structure: Extreme predation of crown- predation of-thoms starfish has disrupted normal links in the food chain between benthic primary production and the fish component of the reef ecosystem East Greenland Sea, Barents Density-independent Shifts in the abundance of fish stock biomass: Major shifts in the Sea, Norwegian Sea natural environmental levels of fist stock biomass within the ecosystems are attrib- perturbations uted to large-scale environmental changes in water move- ments and temperature structure Baltic Sea Density-independent Changes in level of ecosystem productivity: Apparent increases pollution in level of productivity are attributed to the effects of nitrate enrichment resulting from elevated levels of agricultural contaminant inputs from the bordering land masses Antarcfic Marine Density-dependent Status of krill stocks: Annual natural production cycle of krill perturbations is in balance with food requirements of dependent predator populations; surplus production is available to support eco- nomically significant yields, but the sustainable level of fishing effort is not known Density-independent Shifts in abundance in krill biomass: Major shifts in abundance natural environmental levels of krill biomass within the ecosystem are attributed to perturbations large-scale changes in water movements and productivity to the exclusive economic zone of the United to the exclusive economic zone of the United States: the Northeast Continental Shelf, the South- States are designed to (1) provide detailed statis- east Continental Shelf, the Gulf of Mexico, the tical analyses of fish and invertebrate popula- California Current, the Gulf of Alaska, the East- tions constituting the principal yield species of ern Bering Sea, and the Insular Pacific, including biomass, (2) estimate future trends in biomass the Hawaiian Islands. These ecosystems, in 1989, yields, and (3) monitor changes in the principal yielded 9.0 billion pounds of fisheries biomass populations. The information obtained by these valued at approximately $17 billion to the programs helps managers understand the dy- economy of the United States (a billion is 1,000 namics of marine ecosystems and how these dy- million). namics affect harvestable stocks. Additionally, A description of the sampling programs pro- by tracking components of the ecosystems, these viding the biomass assessments within the U. S. programs can detect changes, natural or induced exclusive economic zone has been described in by humans, and warn of events with possible Folio Map 7 produced by the Office of Oceanog- economic repercussions. Although sampling raphy and Marine Assessment of NOAA's Na- schernes and efforts vary among programs (de- tional Ocean Service. The map depicts the seven pendingon habitats, species present, and specific ecosystems under investigation (see figure 15-8). regional concerns), they generally involve sys- Samplingprogramsnsupportingbiomassestimates tematic collection and analysis of catch statistics; in large marine ecosystems within and adjacent the use of NOAA vessels for fisheries-indepen- 219 Figure 15-8: Large Marine Ecosystems of the United States t1 0 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - UEA4 ECOSTS YEr Fishery Resource Assessment Programs S3mpNng Programs rrtMARMAAP Northeast Cl emcmist 8nd peiogiC 5ih.s. ichihyoplwioMhei. ... U. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~shrimp. z0oploni.ion. primary pautmy MAAnMAP Southea I \- I1__- 1TIIEAS T Demets.iernd pelagic fthes. khihyur&iWin , 1CO?Till?AENTAL laecricbral,1 ECOSYSTEFA ~ \ ( $ ) # f / / E LF SY EMr AP sout Atri,e5 rc acntiC ECO1 nse.^ a t SOUYtfiEASTEc a 3nd pelagic nl,hcs. lnWlCbr.nC% --- ~ ~~~~~~~SEAMAP,G'A of Mc.ko uc DI .. . 5 , ,+ ~~~~~16 Skates - - Other. 15 Source. Shenian, Cohen, and Langton 1990. dance. Catch-per-unit-of-effortand fishery-inde- 1960s level of abundance on Georges Bank pendent bottom trawling survey data were criti- (Murawski 1991; Smith and Morse 1990). cal sources of information in this case. However, the lower end of the food chain in the offshore waters of the ecosystem remained unchanged, as Changing ecosystem states described earlier by Bigelow (1926) and Riley, and health indexes Stommel, and Bumpus (1949), suggesting that the ecosystem remnained highly productive during a More and more effort has been directed over the period in which human intervention-fishing- past few years to synthesizing available informa- caused the species dominance of fish to shift tion on factors influencing the natural productiv- (Sherman and others 1983). The natural resilience i ty of fishery biomass and the changing states and of the ecosystem-its ability to recover from health of large marine ecosystems. The goal has stress-can be documented in the recovery of been to identify principal, secondary, and tertiary mackerel to former (pre-1960) levels of abun- driving forces causing major changes in ecosys- dance and the apparent recovery of herring to tem states and biomass yields. Ecosystem health 223 Defining and Measuring Sustainability: The Biogeophysical Foundations is a concept of wide interest for which a single states and health status among ecosystems. Ihe precise scientific definition is problematic. Eco- interrelations between the data sets and the se- system health is used here to describe the resil- lected parameters are indicated by the arrows ience,stability,andproductivityoftheecosystem leading from column 1 to column 2 in figure 15- in relation to its changing states. In present prac- 12. The measured components are depicted in tice, assessing the health of large marine ecosys- relation to the structure in a diagrammatic tems relies on a series of indicators and indexes conceptualizationof patternsand activities within (Costanza 1992; Karr 1992; Norton and Ulanowicz the large marine ecosystem at different levels of 1992; Rapport 1992). The overriding objective is complexity (see figure 15-13). The broad-spec- to monitorchanges in health froman ecosystems trum approach to research and monitoring of perspective as a measure of the overall perfor- large marine ecosystems provides a conceptual mance of a complex system (Costanza 1992). The framework for collaboration in process-oriented health paradigm is an evolving concept based on studies conducted by the National Science Foun- the multiple-state comparisons of resilience and dation and NOAA on the Northeast Continental stabilityoftheecosystem(Costanza 1992;Holling Shelf (GLOBEC 1991) and proposed for other 1986; Pimm 1984). Severai variables important to large marine ecosystems (California Current, the changing states and health of marine ecosys- Antarctic marine ecosystems) and in the pro- tems are defined in table 15-6. Following the posed study of the Indian Ocean-Somalia Cur- definition of Costanza (1992), to be healthy and rentecosystemplannedaspartoftheJointGlobal sustainable, an ecosystem must maintain its level Ocean Flux Studies (JGOFS). of metabolic activity and its internal structure Initial efforts to examine changes in the state of and organization and it must be resistant to exter- ecosystems and relative health within a single nal stress over time and space (see table 15-7). ecosystem are under way for four areas of the Among the indexes being considered as experi- NortheastShelf ecosystem: Gulfof Maine, Georges mental measures of changing ecosystem states Bank, Southern New England, and Mid-Atlantic and healthare(1) diversity, (2) stability,(3) yields, Bight. Initial studies of the structure, function, (4) production, and (5) resilience. and productivity of the system (Sherman and The data from which to derive the experimen- others 1988) report that the principal driving tal indexes are obtained from time-series moni- force affecting sustainable yield is fishing mortal- toring of key ecosystem parameters. A prototype ity expressed as predation on the fish stocksof the effort to validate their utility is being developed ecosystem and that long-term sustainability of by NOAA at the Northeast Fisheries Science Cen- high-economic-yield species depends on the ap- ter. The ecosystem sampling strategy is focused plication of adaptive management strategies on parameters relating to the resources at risk (Murawski 1991; Sissenwine and Cohen 1991). from overexploitation, species protected by legis- Several strategies for managing fish stocks of lative authority (marine mammals), and other the Northeast Shelf ecosystem are under consid- key biological and physical components at the cration by the New England Fishery Manage- lower end of the food chain (plankton, nutrients, ment Council and the Mid-Atlantic Fishery Man- hydrography; Sherman and Laughlin 1992). The agement Council. In addition to issues related to parameters of interest depicted in figure 15-12 the management of fisheries and significant bio- include the composition and biomass of zoop- mass flips among dominant species, the North- lankton, marine mammals, and finfish; structure east Shelf ecosystem is also under stress from the of the water column; photosynthetically active increasing frequencyof unusual planktonblooms radiation; transparency; chlorophyll a; nitrite; and eutrophication within the near-shore coastal nitrate; primary production; pollution; runoff; zone, which are the result of high levels of phos- wind stress; community structure and counts of phate and nitrate discharges into drainage ba- sea birds; domoic acid; saxitoxin; and paralytic sins. Whether increases in the frequency and ex- shellfish poisoning. These experimental param- tent of near-shore plankton blooms are respon- eters incorporate the behavior of individuals, the sible for the rising incidence of biotoxin-related resultant responses of populations and commu- shellfish closuresand deaths of marine mammals nities, as well as their interactions with the physi- remains an open question that is of considerable cal and chemical environment. The selected pa- concern to state and federal management agen- rameters, if measured in all large marine ecosys- cies (Sherman, jaworski, and Smayda 1992; tems, would permit us to compare the changing Smayda 1991). 224 Large Marine Ecosystems and Fisheries Table 15-6: Definitions of important variables Variable Definition Units Stability Homeostasis Maintenance of a steady state in living organisms by the use of feedback control processes Stability If, and only if, the variables all return to the initial Binary equilibrium following their being perturbed from it; a system is locally stable if this return applies to small perturbations and is globally stable if it applies to all possible perturbations Sustainable A system that can maintain its structure and function Binary indefinitely; all nonsuccessional (climax) ecosystems are sustainable, but they may not be stable (see resilience); sustainability is a policy goal for economic systems Resilience Speed with which variables return toward their equilibrium Time following a perturbation; not defined for unstable systems (Pimm 1984) or ability of a system to maintain its structure and patterns of behavior in the face of disturbance (Holling 1986) Resistance The degree to which a variable is changed, following a Nondimensional perturbation and continuous Variability Variance of population densities over time, or allied measures such as the standard deviation or coefficient of variation (standard deviation/mean) Complexity Species richness Number of species in a system Integer Connectedness Number of actual interspecific interactions divided Dimensionless by the possible interspecific interactions Interaction strength Mean magnitude of interspecific interaction: size of the effect of one species' density on the growth rate of another species Evenness Variance of the distribution of species abundance Diversity indexes Measures that combine evenness and richness with Bits a particular weighting for each; one important member of this family is the information theoretic index, H Ascendancy An information theoretic measure that combines the average mutual information (a measure of connectedness) and the total throughput of the system as a scaling factor (see Ulanowicz 1992) Other variables Perturbation Change to a system's inputs or environment beyond the Varies normal range of variation Stress A perturbation with a negative effect on a system Subsidy A perturbation with a positive effect on a system Source: Adapted and expanded from Costanza 1992. Present and future efforts to allow for useful comparisons among different processes influencing large-scale changes in the The topics of change and persistence in marine biomass yields of large marine ecosystems (Bax communities and the need for multispecies and and Laevastu 1990). ecosystem perspectives in fisheries management Among the ecosystems being managed from a were reviewed at the Dahlem Conference on Ex- more holisticperspectiveare theYellowSea ecosys- ploitation of Marine Communities in 1984 (May tem, where the principal effort is being made by the 1984). The designation and management of large Peoples Republic of China (Tang 1989); the marine ecosystems are, at present, evolving sci- multispecies fisheries of the Benguela Current eco- entificandgeopoliticalprocesses (Alexanderl989; system under management of the government of Morgan 1988). Sufficient progress has been made South Africa (Crawford,Shannon,and Shelton 1989); 225 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 15-7: Indexes of Vigor, Organization, and Resilience in Various Fields Component Related Existing related Field Probable method of health concepts measures of origin of solution Vigor Function GPP, NPP, GEP > Ecology Measurement Productivity GNP > Economics System Metabolism > Biology Throughput Organization Structure Diversity index Network analysis Biodivcrsity Average mutual information predictability > Ecology Resilience Scope for growth > Ecology Simulation modeling Combinations Ascendancy > Ecology Note. GPP, gross primary production; NPP, net primary production; CEP, gross ecosystem product; GNP, gross national product. Source: Costanza 1992. the Great Barrier Reef ecosystem (Bradbury and than single, large models that generally have Mundy 1989) and the Northwest Australian Conti- limited prediction capability. nental Shelf ecosystem (Sainsbury 1988), under 3. The development and evaluation of models us- management of the state and federal govemments ing health indicators that are directly applicable of Australia; and the Antarctic marine ecosystem, tomanagementdecisions; theyshouldbesimple under the Commission for the Conservation of in construction, allow for interaction with re- Antarctic Marine Living Resources and its twenty- source managers, and provide sufficient flexibil- one-nation membership (Scully, Brown, and ityfortestinghypothescsforarangeofscenarios. Manheim 1986; Sherman and Ryan 1988). Within TheNMFS-NOAAregionalprogramsconcerned the exclusive economic zone of the United States, withforecastingtrendsinbiomassyieldsarebeing the state governments of Washington and Oregon conducted in relation to the dynamics of large ma- have developed a comprehensive plan for manag- rine ecosystems. Results from several large-scale ing marine resources within the Northern Califor- studies, including the GLOBEC and JGOFS efforts nia Current ecosystem (Bottom and others 1989). oftheNationalScienceFoundation, shouldprovide To improve the definition of ecosystem health im. rtant new insights on the structure and func- so that it considers both natural environmental tion of large marine ecosystems. Ongoing studies perturbation and the effects of human interven- to flremrn csses non tde tion ton the agnsts of e yms, future are now yielding information on changes in the tion on the changing states of ecosystems, future productivity,habitat,and pollution stressof ecosys- effort will focus on: tems. Future efforts will be needed to provide an 1. The development of ecosystem change and improved,better-integratedbaseofinformationfor health indexes and ind icators for large marine assessing,understanding,and managingthenation's ecosystems. living marine resources within the boundary of 2. The development of component models of large large marine ecosystems, including wetlands and marineecosystemsincorporatingmeasurements estuaries, as critical habitats of fishes and inverte- of changing states and health indicators rather brate (shrimp, shellfish) resources. 226 Large Marine Ecosystems and Fisheries Figure 15-12: Schematic Representation of the Databases and Experimental Parameters for Indexing the Changing States of Large Marine Ecosystems Index Data set Parameter (based on 1960 value) Soop - Zooplankton composition Zooplankton Zoo°plankton biomass Diversity Ichthyoplankton Water columnn structure \ Ichthyoplankton / \ \\ ight Phytoplankton Transparency \>K- \ \ | F Stability Climatology A Chlorophylla \ \ Ocean physics '1 \ , Nitrate, nitrite 1- Nutrients Primary production Nutrients = X \\X! \ \ l Habitat eld Marine marmna biomass \ Bent.hos -/ , 7Ź a)VX Benthos Marine anianual composition / Marine mammals Runoff Marine birds Wmd stress, Productivity Bird cornmunity / Fish trawls structure Bird biomass - / Weather Fish composition Biotoxins > Fish biomass / Resilience Domoic acid '\ w Saxitoxin Note: The database represents time-series measurements of kev ecosvstem components from the U.S. Northeast Continental Shelf ecosystem. Indexes are based on changes compared wvith the state of the ecosystem in 1960. 227 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 15-13: Diagrammatic Conceptualization of Patterns and Activities at Different Levels of Complexity Ecosystem Level System Biornass System Productiviry ........ Energy Flux Nutinent Flux and Cycling Rrestence / Stabilitv Development Individual Level E Population Level Communitv Level C Growill Intraspecific Compcutiion * Interspecific Compeutior Repro(l;action* Population Age I Diversity MorijliiN Size Struciure Spatial Structure * Beh.ivior Populatiot Growth Rate Zonauon * Movement Population Cycles Succession Spatial Dtsinbutioii Invasion I Extinctioi Indirect Compctition ! Mutual Ism Note: Each sphererepresents an individual abioticor bioticentry. Abiotic isdefined as nonliving matter. Broad, double-headed arrows indicate feedback between entities and the energy matrix for the system. Thin arrows represent direct interactions between individual entities. Much of ecology is devoted to studying interactions between biotic and abiotic entities with a focus on how such interactions affect individuals (1), populations (P), or communities (C) of organisms. Ecosystem ecologists studv these interactions from the viewpoint of their effect on both biotic and abiotic entities and within the context of the system. The boundaries of the system must be established to conduct quantitative studies of flux. Source: Likens 1992. References Sherman, L. M. Alexander, and B. D. Gold, eds., Large Marine Ecosystems: Stress, Mitiga- Aiken, J. 1981. "The Undulating Oceanographic tion, and Sustainability, pp. 53-68. Washington, Recorder Mark 2." Journal of Plankton Research D.C.: American Association for the Advance- 3, pp. 551460. ment of Science Press. Alexander, L. M. 1989. "Large Marine Ecosys- Alverson, D. L., A. R. Longhurst,andj. A.Gulland. tems as Global Management Units." In K. 1970. "How Much Food from the Sea?" Science Sherman and L. M. Alexander, eds., Biomass 168, pp. 503-05. Yields and Geography of Large Marine Ecosys- Anderson, K. P., and E. Ursin. 1977. "A tems, pp. 339-44. AAAS Selected Symposium Multispecies Extension to the Beverton and II11. Boulder, Cola.: Westview Press. Holt Theory of Fishing with Accounts of Phos- Alheit, J., and P. Bemal. 1992. "Effects of Physical phorus Circulation and Primary Production." and Biological Changes on the Biomass Yield Meddelelser fra Danmarks fiskeri-og of the Humboldt Current Ecosystem." In K. havundersegelser (Ny serie) 7, pp. 319-435. 228 Large Marine Ecosystems and Fisheries Azarovitz,T.R.,and M. D. Grosslein. 1987. "Fishes Borisov, V. 1991. "The State of the Main Commer- and Squids." In R. H. Backus, ed., Georges Bank, cial Species of Fish in the Changeable Barents pp. 315-46. Cambridge, Mass.: M.I.T. Press. Sea Ecosystem." In K. Sherman, L. M. Bakun,A.1992. "TheCalifornia Current, Benguela Alexander, and B. D. Gold, eds., Food Chains, Current, and Southwestern Atlantic Shelf Eco- Yields, Models, and Management of Large Marine systems: A Comparative Approach to Identi- Ecosystems, pp. 193-203. Boulder, Colo.: fying Factors Regulating Biomass Yields." In Westview Press. K. Sherman, L. M. Alexander, and B. D. Gold, Bottom, D. L., K. K. Jones, J. D. Rodgers, and R. F. eds., Large Marine Ecosystems: Stress, Mitiga- Brown. 1989. "Management of Living Re- tion, and Sustainability, pp. 199-221. Washing- sources: A Research Plan for the Washington ton, D.C.: American Association for the Ad- and Oregon Continental Margin." NCRI-T-89- vancement of Science Press. 004. National Coastal Resources Research De- Bax, N. J., and T. Laevastu. 1990. "Biomass Poten- velopment Institute, Newport, Oreg. tial of Large Marine Ecosystems: A Systems Bradbury, R. H., and C. N. Mundy. 1989. "Large- Approach." In K. Sherman, L. M. Alexander, scaleShiftsinBiomassoftheGreatBarrierReef and B. D. Gold, eds., Large Marine Ecosystems: Ecosystem." In K. Sherman and L. M. Patterns, Processes, and Yields, pp. 188-205. Alexander, eds., Biomass Yields and Geography Washington, D.C.: American Association for of Large Marine Ecosystems, pp. 143-67. AAAS the Advancement of Science Press. Selected Symposium 111. Boulder, Colo.: Beddington, J. R. 1984. "The Response of Westview Press. Multispecies Systems to Perturbations." In R. Burger,J. 1988. "Interactionsof MarineBirdswith M. May, ed., Exploitation of Marine Communi- Other Marine Vertebrates in Marine Environ- ties, pp. 209-25. Berlin: Springer-Verlag. ments." In J. Burger, ed., Seabirds and Other . 1986. "Shifts in Resource Populations Marine Vertebrates, pp. 2-28. New York: Co- in Large Marine Ecosystems." In K. Sherman lumbia University Press. and L. M. Alexander, eds., Variability and Man- Byrne, J. 1986. "Large Marine Ecosystems and the agement of Large Marine Ecosystems, pp. 9-18. Future of Ocean Studies." In K. Sherman and AAASSelectedSymposium99.Boulder,Colo.: L. M. Alexander, eds., Variability and Manage- Westview Press. ment of Large Marine Ecosystems, pp. 299-308. Belsky, M. H. 1986. "Legal Constraints and Op- AAASSelectedSymposium99.Boulder,Colo.: tions forTotal Ecosystem Management of Large Westview Press. Marine Ecosystems." In K. Sherman and L. M. Canon, J. R. 1986. "Variabilidad ambiental en Alexander, eds., Variability and Management of relaci6n con la pesqueria neritica pelagica de la Large Marine Ecosystems, pp. 241-61. AAAS zona Norte de Chile." In P. Arana, ed., La pesca Selected Symposium 99. Boulder, Colo.: en Chile, pp. 195-205. Valparaiso: Escuela de Westview Press. Ciencias del Mar, Facultad de Recursos .1989. "The Ecosystem Model Mandate Naturales, Universidad Cat6lica de Chile. for a Comprehensive United States Ocean Christy, F. T., Jr. 1986. "Can Large Marine Eco- Policy and Law of the Sea." San Diego Law systems Be Managed for Optimum Yields?" Review 26:3, pp. 417-95. In K. Sherman and L. M. Alexander, eds., Beverton, R. J. H., and S. J. Holt. 1957. "On the Variability and Management of Large Marine Dynamicsof Exploited FishPopulations." Fish- Ecosystems, pp. 263-67. AAAS Selected Sym- ery Investigations (U.K. Ministry of Agricul- posium 99. Boulder, Colo.: Westview Press. ture, Fisheries, and Food) 2:19, pp. 1-533. Colebrook,J. M. 1986. "Environmental Influences Bigelow, H. B. 1926. "Plankton of the Offshore on Long-term Variabilityin MarinePlankton." Waters of the Gulf of Maine." Bulletin of the Hydrobiologia 142, pp. 309-25. Bureau of Fish and Wildlife 40:2. Government Costanza, R. 1992. "Toward an Operational Defi- Printing Office, Washington, D.C. nition of Ecosystem Health." In R. Costanza, Bonner, W. N. 1982. Seals and Man, A Study of B. G. Norton, and B. D. Haskell, eds., Ecosystem Interactions. Seattle: University of Washington Health: New Goals for Environmental Manage- Press. ment. Washington, D.C.: Island Press. 229 Defining and Measuring Sustainability: The Biogeophysical Foundations Crawford, R. J. M., L. V. Shannon, and P. A. Hempel, G., ed. 1978. "Symposium on North Sea Shelton. 1989. "Characteristics and Manage- Fish Stocks: Recent Changesand TheirCauses." ment of the Benguela as a Large Marine Eco- Rapport et Proces-Verbaux des Reunions (Conseil system." In K. Sherman and L. M. Alexander, International pour l'Exploration de la Mer) eds., Biomass Yields and Geography of Large Ma- 172, p. 449. rine Ecosystems, pp. 169-219. AAAS Selected Holling, C. S. 1973. "Resilience and Stability of Symposium 111. Boulder, Colo.: Westview Ecological Systems." InstituteofResourceEcol- Press. ogy, University of British Columbia, Croxall, J. P., ed. 1987. Seabirds: Feeding Ecology Vancouver, Canada. and Role in Marine Ecosystems. London: Cam- .1986. "TheResilienceofTerrestrial Eco- bridge University Press. systems Local Surprise and Global Change." Cushing, D. H. 1975. Marine Ecology and Fisheries. In W. C. Clark and R. E. Munn, eds., Sustainable London: Cambridge University Press. Development of the Biosphere, pp. 292-317. Lon- Daan, N. 1986. "Results of Recent Time-series don: Cambridge University Press. Observations for Monitoring Trends in Large Hovgaard, H., and E. Buch. 1990. "Fluctuation in Marine Ecosystems with a Focus on the North the Cod Biomass of the West Greenland Sea Sea."InK.ShermanandL.M.Alexander,eds., Ecosystem in Relation to Climate." In K. VariabilityandManagementofLargeMarineEco- Sherman, L. M. Alexander, and B. D. Gold, systems, pp. 145-74. AAAS Selected Sympo- eds., Large Marine Ecosystems: Patterns, Pro- sium 99. Boulder, Colo.: Westview Press. cesses, and Yields, pp. 36-43. Washington, D.C.: Dickson, R. R., P. M. Kelly, J. M. Colebrook, W. S. American Association for the Advancement of Wooster, and D. H. Cushing. 1988. "North Science Press. Winds and Production in the Eastern North ICES (International Council for the Exploration Atlantic." Journal of Plankton Research 10, pp. of the Sea). 1991. "Report of the Multispecies 151-69. Working Group." ICES C.M. 1991/Assess:7. FAO (Food and Agriculture Organization of the Copenhagen, Denmark. United Nations). 1989. FAO Yearbook of Fishery IOC (International Oceanographic Commission Statistics: Catches and Landings. Vol. 64 (for of UNESCO). 1992. "The Use of Large Ma- 1987). Rome. rine Ecosystem Concept in the Global Ocean .1990. FAO Yearbook of Fishery Statistics. Observing System (GOOS)." Prepared by K. Vol. 66 (for 1988) Rome. Sherman for the twenty-fifth session of the Fogarty, M., E. B. Cohen, W. L. Michaels, and IOC executive council, Paris, March 10-18. W. W. Morse. 1991. "Predation and the Regu- IOC/EC-XXV/Inf.7. Washington, D.C. lation of Sand Lance Populations: An Explor- Jossi, J. W., and D. E. Smith. 1990. "Continuous atory Analysis." ICES Marine Science Sympo- Plankton Records: Massachusetts to Cape sium 193, pp. 120-24. Sable, N.S., and NewYork to the Gulf Stream, . . ~1989." NAFO Series Document 90/66:1-li. GESAMP (Group of Experts on the Scientific As- Northwest Seric Fishent 90ganzatio. pects of Marine Pollution). 1990. "Thc State of Northwest Atlantic Fisheries Organization. the Marine Environment." Regional Seas Re- Karr, J. 1992. "Ecological Integrity: Strategies for ports and Studies 115. United Nations Envi- Protecting Earth's Life Support System." In ronment Program, Nairobi. R. Costanza, B. G. Norton, and B. D. Haskell, GLOBEC (Global Ocean Ecosystems Dynamics). eds., Ecosystem Health: New Goals for Environ- 1991. "Initial Science Plan. February 1991." mental Managment, pp. 223-38. Washington, Report 1. Joint Oceanographic Institutions, Washington, D.C. Kawasaki,T.,S.Tanaka,Y.Toba,and A.Taniguchi, Glover, R. S. 1967. "The Continuous Plankton Re- eds. 1991. Long-term Variability of Pelagic Fish corder Survey of the North Atlantic." Syimposium PopulationsandTheirEnvironment.Proceedings of the Zoological Society of London 19, pp. 189-210. of the Interational Symposium, Sendai, Japan, Gulland,). A.1984. "Epilogue." In R. M. May, ed., November 14-18,1989. Tokyo: Pergamon Press. Exploitan tio A. 19 ri4. "Epilogmue."ni p. May, 'd. Levin, S. A. 1978. "Pattern Formation in Ecologi- BerloitatSpriongofMriner Communities,pp. 335-37. cal Communities." In J. A. Steele, ed., Spatial Berlin: Springer-Verlag. 230 Large Marine Ecosystems and Fisheries Pattern in Plankton Communities, pp. 433-70. .1989. "Large Marine Ecosystems in the New York: Plenum Press. Pacific Ocean." In K. Sherman and L. M. . 1990. "Physical and Biological Scales, Alexander, eds., Biomass Yields and Geography and Modelling of Predator-prey Interactions of Large Marine Ecosystems, pp. 377-94. AAAS in Large Marine Ecosystems." In K. Sherman, Selected Symposium 111. Boulder, Colo.: L. M. Alexander, and B. D. Gold, eds., Large Westview Press. MarineEcosystems:Patterns,Processes,andYields, Myers, N. 1990. "Working towards One World. pp. 179-87. Washington, D.C.: American As- Book Review." Nature 344:6266, pp. 499-500. sociationfortheAdvancementofSciencePress. Murawski, S. A. 1991. "Can We Manage Our Likens, G. E. 1992. "The Ecosystem Approach: Its Multispecies F'sheries?" Fisheries 16:5, pp. 5-13. Use and Abuse." In 0. Kinne, ed., Excellence in Norton, B. G., and R. E. Ulanowicz. 1992. "Scale Ecology3.W-2124.Oldendorf/Luhe,Germany: and Biodiversity Policy: A Hierarchical Ap- Ecology Institute. proach." Ambio 21:3, pp. 244-49. Lindemann, R. L. 1942. "The Trophic Dynamic Overholtz, W. J., and J. R. Nicolas. 1979. "Appar- Aspect of Ecology." Ecology 23, pp. 399-418. ent Feeding by the Fin Whale, Balaenoptera Loughlin, T. R., and R. Nelson, Jr. 1986. "Inciden- physalus, and Humpback Whale, Megoptera tal Mortality of Northern Sea Lions in the novaeangliae, on the American Sand Lance, Shelikof Strait, Alaska." Marine Mammal Sci- Ammodytes americanus, in the Northwest At- ence 1, pp. 14-33. lantic." Fishery Bulletin, U.S. 77, pp. 285-87. MacCall, A. D. 1986. "Changes in the Biomass of Payne, P. M., D. N. Wiley, S. B. Young, S. Pittman, theCalifornia CurrentSystem." In K. Sherman P. J. Clapham, and J. W. Jossi. 1990. "Recent and L.M.Alexander,eds., Variability and Man- Fluctuations in the Abundance of Baleen agement of Large Marine Ecosystems, pp. 33-54. Whales in the Southern Gulf of Maine in Rela- AAASSelected Symposium99. Boulder, Colo.: tion to Changes in Selected Prey." Fishery Bul- Westview Press. letin, U.S. 88, pp. 687-96. Malone, T. C. 1991. "River Flow, Phytoplankton Pimm, S. L. 1984. "The Complexity and Stability Production, and Oxygen Depletion in Chesa- of Ecosystems." Nature 307, pp. 321-26. peake Bay." In R. V. Tyson and T. H. Pearson, Piyakamchana, T. 1989. "Yield Dynamics as an eds., Modern and Ancient Continental Shelf An- Index of Biomass Shifts in the Gulf of Thailand oxia, pp. 83-93. Special Publication 58. Boul- Ecosystems." InK.ShermanandL.M.Alexander, der, Colo.: Geological Society. eds., Biomass Yields and Geographyof LargeMarine Mangel, M. 1991. "Empirical and Theoretical As- Ecosystems, pp. 95-142. AAAS Selected Sympo- pects of Fisheries Yield Models for Large Ma- sium 111. Boulder, Colo.: Westview Press. rine Ecosystems." In K. Sherman, L. M. Postma,H.,andJ. J.Zijlstra,eds. 1988. Ecosystems Alexander, and B. D. Gold, eds., Food Chains, of the World 27: Continental Shelves. Amsterdam, Yields, Models, and Management of Large Marine Netherlands: Elsevier. Ecosystems, pp. 243-61. Boulder, Colo.: Powers, K. D., and R. G. B. Brown. 1987. "Sea- Westview Press. birds." In R. H. Backus, ed., Georges Bank, pp. May, R. M., ed. 1984. Exploitation of Marine Com- 359-71. Cambridge, Mass.: M.I.T. Press. munities. Berlin: Springer-Verlag. Prescott, J. R. V. 1989. "The Political Division of Mayr, E. 1982. The Growth of Biological Thought. LargeMarineEcosystemsintheAtlanticOcean Cambridge, Mass.: Harvard University Press. and Some Associated Seas." In K. Shermanand Minoda, T. 1989. "Oceanographic and Biomass L. M. Alexander, eds., Biomass Yields and Geog- Changes in the Oyashio Current Ecosystem." raphy of Large Marine Ecosystems, pp. 395-442. In K. Sherman and L. M. Alexander, eds., Bio- AAASSelectedSymposiuml11.Boulder,Colo.: mass Yields and Geography of Large Marine Eco- Westview Press. systems, pp. 67-93. AAAS Selected Sympo- Rapport, D. J. 1992. "What Is Clinical Ecology?" sium 111. Boulder, Colo.: Westview Press. In R. Costanza, B. G. Norton, and B. D. Haskell, Morgan, J. R. 1988. "Large Marine Ecosystems: eds., Ecosystem Health: New Goals for Environ- An Emerging Concept of Regional Manage- mental Management, pp. 144-56. Washington, ment." Environment 29:10, pp. 4-9, 26-34. D.C.: Island Press. 231 Defining and Measuring Sustainability: The Biogeophysical Foundations Ricklefs, R. E. 1987. "Community Diversity: Rela- Shenman, K., E. B. Cohen, and R. W. Langton. tive Roles of Local and Regional Processes." 1990. "The Northeast Continental Shelf: An Science 235:4785, pp. 167-71. Ecosystem at Risk." In V. Konrad, S. Ballard, Riley, G. A., H. Stommel, and D. F. Bumpus. 1949. R. Erb, and A. Morin, eds., Gulf of Maine: Sus- "Quantitative Ecology of the Plankton of the taiing Our Common Heritage, pp. 120-67. Pro- Western North Atlantic." Bulletin of Bingham ceedings of an international conference held at Oceanography College 12:3, p. 169. Portland, Maine, December 10-12,1989. Maine Ryther, J. H. 1969. "Relationship of Photosynthe- StatePlanningOfficeand theCanadian-Ameri- sis to Fish Production in the Sea." Science 166, can Center of the University of Maine, Augusta. pp. 72-76. Sherman, K., J. R. Green, J. R. Goulet, and L. Sainsbury, K. J. 1988. "The Ecological Basis of Ejsymont. 1983. "Coherence in Zooplankton of Multispecies Fisheries and Management of a a Large Northwest Atlantic Ecosystem." Fish- Demersal Fishery in Tropical Australia." In ery Buletn, U.S. 81, pp. 855-62. J. A. Gulland, ed., Fish Population Dynamics, 2d Sherman, K., M. Grosslein, D. Mountain, D. Busch, ed., pp. 349-82. New York: John Wiley and Sons. J. O'Reilly, and R. Theroux. 1988. "The Conti- Schaefer, M. B.154"SmeAspetsofheDy- nental Shelf Ecosystem off the Northeast Coast Schaefer, M. B. 1954. "Some Aspects of the Dy- ofteUidSaes"nI-.Ptm adJ.. namics of Populations Important to the Man- agement of the Commercial Marine Fisheries." Zilstra, eds., Ecosystems of the World 27: Conti- Bulletin of theInter-American Tropical Tuna Com- nental Shelves, pp. 279-337. Arnsterdam, Nether- mission 1:1, pp. 27-56. lands: Elsevier. Scully, R. T., W. Y. Brown, and B. S. Manheim. Sherman, K., N. Jaworski, and T. Smayda. 1992. "The Northeast Shelf Ecosystem: Stress, Mitiga- 1986. "The Convention for the Conservation of tion, and Sustainability, 12-15 August 1991 Sym- Antarctic Marine Living Resources: A Model ' S for arg Maine cosste Mangemnt. In posium Summary." Technical Memo NMFS-F/ for Large Marine Ecosystem Management." In NE94U..DprmnofCmeceNa K. Shermnan and L. M. Alexander, eds., Van- tnE9 UeS. deAtmof Commera- ability and Management of Large Marine Ecosys- tional Oceanic and Atmospheric Administra- tems, pp. 281-86. AAAS Selected Symposium tion, Washington, D.C. 99. Boulder, Colo.: Westview Press. Shermnan, K., and T. Laughlin, eds. 1992. "Large Sherman, K. 1991. "The Large Marine Ecosystem Marine Ecosystems Monitoring Workshop Re- Concept: A Research and Management Strat- port." Technical Memo NMFS-F/NEC-93. U.S. egyfoLiinMainR so e E Departmentof Commerce, NationalOceanicand egy for Living Marine Resources." Ecological Atopecdisrto,ahntnDC Applications 1:4, pp. 349-60. Atmospheric Adnministration, Washington, D.C. Applications 1:4, pp. 349-60. Technical Memo NMFS-F/NEC-93. Sherman, K., and L. M. Alexander, eds. 1986. Shern, K., and A. F. Ryan. 1988. "Antarctic Ma- Variability and Management of Large Marine Eco- systems. AAAS Selected Symposium 99. Boul- rine Living Resources." Oceanus 31:2, pp. 59-63. der, Colo.: Westview Press. Sissenwine, M. P. 1986. "Perturbation of a Preda- eds198dGeographyof Large tor-controlled Continental Shelf Ecosystem." Marine Ecosystems. Boulder, Colo.: Westview In K. Sherman and L. M. Alexander, eds., Vari- PressE ability and Management of Large Marine Ecosys- Press. tems, pp. 55-85. AAAS Selected Symposium 99. Sherman, K., L. M. Alexander, and B. D. Gold, Boulder, Colo.: Westview Press. eds. 1990. Large Marine Ecosystems: Patterns, Sissenwine, M. P., and E. B. Cohen. 1991. "Re- Processes,and Yields. Washington, D.C.: Amer- source Productivity and Fisheries Management can Association for thc Advancement of Sci- y g ence Press. of the Northeast Shelf Ecosystem." In K. Sherman, L. M. Alexander, and B. D. Gold, . eds. 1991. Food Chains, Yields, Models, eds., Food Chains, Yields, Models, and Manage- and Management of Large Marine Ecosystems. ment of Large Marine Ecosystems, pp. 107-23. Boulder, Colo.: Westview Press. Boulder, Colo.: Westview Press. . eds. 1993. Large Marine Ecosystems: Skjoldal, H. R., and F. Rey. 1989. "Pelagic Produc- Stress, Mitigation, and Sustainability. Washing- tion and Variability of the Barents Sea Ecosys- ton, D.C.: American Association for the Ad- tem." In K. Sherman and L. M. Alexander, eds., vancement of Science Press. Biomass Yields and Geography of Large Marine Eco- 232 Large Marine Ecosystems and Fisheries systems, pp. 241-86. AAAS Selected Symposium Taylor, P., and A. J. R. Groom, eds. 1989. GlobalIssues 111. Boulder, Colo.: Westview Press. in the United Nation's Framework. London: Smayda, T. 1991. "Global Epidemic of Noxious Macmillan. Phytoplankton Blooms and Food Chain Con- Terazaki, M. 1989. "Recent Large-scale Changes in sequencesinLargeEcosystems." InK.Sherman, the Biomass of the Kuroshio Current Ecosys- L. M. Alexander, and B. D. Gold, eds., Food tem." In K. Sherman and L. M. Alexander, eds., Chains, Yields, Models, and Management of Large Biomass Yields and Geography of Large Marine Eco- Marine Ecosystems. Boulder, Colo.: Westview systems, pp. 37-65. AAAS Selected Symposium Press. 111. Boulder, Colo.: Westview Press. Smith, W. G., and W. W. Morse. 1990. "Larval Ulanowicz, R. E. 1992. "Ecosystem Health in Terms DistributionPatterns: Evidence for theCollapse/ of Trophic Flow Networks." In R. Costanza, B. G. Recolonization of Atlantic Herring on Georges Norton, and B. D. Haskell, eds., Ecosystem Htalth: Bank."ICESC.M.1990/H:17.Copenhagen,Den- New Goals for Environmental Management, pp. mark: International Council for the Exploration 189-296. Washington, D.C.: Island Press. of the Sea. U.S. Council on Environmental Quality. 1980. The Steele,J. H. 1974. The Structuire ofMarine Ecosystems. Global 2000 Report to the President: Entering the Cambridge, Mass.: Harvard University Press. Tuwenty-first Century,3vols.WiththeU.S.Depart- . 1988. "Scale Selection for Biodynamic ment of State. Washington, D.C.: U.S. Govern- Theories." In B. J. Rothschild, ed., Towxarda Theory ment Printing Office. on Biological-physicallnteractionsintheWorldOcean, U.S. Department of Commerce. 1988. "A National pp. 513-26. NATO ASI Series C: Mathematical Atlas: Health and Use of Coastal Waters, United and Physical Sciences, vol. 239. Boston, Mass.: States of America." U.S. Department of Com- Kluwer Academic Publishers. merce, National Oceanic and Atmospheric Ad- Sugihara, G., S. Garcia, J. A. Gulland, J. H. Lawton, ministration, National Ocean Service, Office of H. Maske, R. T. Paine, T. Platt, E. Rachor, B. J. Oceanography and Marine Assessment, Wash- Rothschild, E. A. Ursin, and B. F. K. Zeitzschel. ington, D.C. 1984. "Ecosystem Dynamics: Group Report." In Waring, G. T., P. M. Payne, B. L. Parry, and J. R. R. M. May, ed., Exploitation of Marine Communi- Nicolas. 1990. "Incidental Take of Marine Mam- ties, pp. 13G-53. Berlin: Springer-Verlag. mals in Foreign Fishery Activities off the North- Tang, Q. 1989. "Changes in the Biomass of the east United States, 1977-88." Fishery Bulletin Yellow Sea Ecosystems." In K. Sherman and (United States) 88, pp 347-60. L. M. Alexander, eds., Biomass Yields and Geogra- Wise, J. P. 1984. "The Future of Food from the Sea." phy of Large Marine Ecosystems, pp. 7-35. AAAS In J. L. Simon and H. Kahn, eds., The Resourceful Selected Symposium 111. Boulder, Colo.: Earth, pp.113-27. New York: Basil Blackwell. Westview Press. 233 Paarg Ico Ml/anag,(ed Ecosysetems Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement Nigel J. H. Smith and Donald L. Plucknett Variousdefinitionsofsustainableagriculturehave lizes sometime in the middle of the next cen- surfaced, often heavily flavored by the disciplin- tury, and most of that growth will occur in the ary backgrounds of the proponents. Neverthe- warmer climates. Second, intensifying produc- less, a consensus has emerged that sustainable tion on existing cleared areas creates more op- agricultural systems incorporate four main dimen- tions for development or preservation of re- sions:thebiophysicalenvironment,theinstitutional maining spaces. and policy environment, social and cultural Third, sustainable agriculture is not synony- concerns, and economic viability. Unless all four mous with low-input agriculture. Some low-in- componentsareaddressedadequately,agricultural put systems are not sustainable, from either the development goals can be thwarted, often with ecological or social perspectives. Conversely, some severe environmental and social repercussions. high-input systems can damage the environment. Our intent here is not to add interpretations or A broad spectrum of input intensity provides the nuances to the definition of sustainable agricul- flexibility needed for achieving sound agricul- ture. Considerable thought has already gone tural development of the ecologically and cultur- into analyzing the dimensions to sustainability ally diverse tropics. in farming systems (Altieri, Letourneau, and This survey of sustainability issues in tropical Davis 1983; Brklacich, Bryant, and Smit 1991; agriculture includes a discussion of the special CGIAR 1988; FAO 1991; Smith 1990; York 1988). challenge of marginal lands, the management of Rather, our purpose is to focus on some of the pests and diseases, the positive and negative con- major conceptual issues confronting efforts to tributionsof weeds, the ecological and social role improve the sustainability of agriculture as a of tree crops, the interconnections among a mosaic framework for identifying ways to measure or of land uses, the intensification imperative, and the ascertain sustainable systems. need for a blend of traditional knowledge and mod- Before exploring various dimensions to sus- em science. Finally, we explore some issues in a tainable agriculture and their measurement, three research agenda for sustainable agriculture. overridingpointsareinorder.First,sustainability In this discussion of salient issues in tropical does not mean keeping yields at their current agriculture, we examine the biophysical mea- levels. Yields must continue to rise for the major surements required to monitor sustainability. crops, and more work is needed to upgrade the Where appropriate, we assess existing measure- productivity of the so-called minor crops that can ments and methods for measuring and predict- be important in certain regions. If yields do not ing sustainability on experiment stations and increase, it will be difficult to feed and improve under actual conditions of production. A case is incomes for the existing population in the trop- made for conductingbothon-farm and on-station ics, let alone future generations. The world's experiments and for protecting biodiversity in population is expected to doublebefore it stabi- natural and moderately disturbed ecosystems. Defining and Measuring Sustainability: The Biogeophysical Foundations The special challenge of marginal lands that require heavy fertilization to achieve their potential yield. As an overall strategy, it makes sense to concen- Soil erosion can be particularly serious in mar- trate food production on the optimal lands ginal environments, where steep slopes or sparse (Leonard 1989; Plucknett and Smith 1982). But a vegetation facilitate rapid runoff and wind ero- host of factors, ranging from inequitable land sion. Many societies have confronted this prob- ownership to population pressure, often forces lem, often with great success. Spectacular ter- farmers to settle on marginal lands. Areas with races adorn high mountains in many parts of the erraticorexcessive rainfall,poor soils, steep slopes, tropics and subtropics, including Java, Nepal, or inadequatedrainage pose daunting challenges northern Luzon in the Philippines, the central for agriculture. Andes, Kenya (Davidson 1969, p.35), and partsof It can be argued that rather than promote re- Guatemala (Cook 1909). Most of these complex search on ways to improve agriculture on mar- terrace systems were developed thousands of ginal lands, emphasis should be placed on the years ago for the dual purpose of irrigating the more productive environments. Although some land and conserving the topsoil. Such long-term scope may exist for encouraging people to leave land improvements are expensive to build and marginal environments-thereby allowing de- are difficult to maintain. graded habitats to recuperate-political, eco- In the case of the Great Lakes Region of Africa, nomic, and demographic forces are likely to keep classic bench terraces have been abandoned be- manypeopleonrmarginal landscapes. Byattempt- cause they are associated with colonialism. Soil ing to improve production on already disturbed erosion accelerated markedly in Rwanda, environments, pressure could be reduced on even Burundi, and eastern Zaire as such terraces were more fragile ecosystems. destroyed after independence. Soil erosion was Research strategies for achieving sustainable exacerbated as the population continued to ex- agriculture in marginal areas are similar in many pand rapidly in the already densely settled re- regards to those suitable for the better-endowed gion. Fortunately, some other soil conservation lands. Increased emphasis can be placed, how- techniques havebeen promoted recently,includ- ever,on developingcrop varieties that withstand ing grass or tree strips to trap soil (Egger and moisture stress, are adapted to poor soils, resist Martens 1987). Other methods to check soil ero- diseases and pests, and, in highland areas, toler- sion on hillsides in the humid tropics include ate cold. Marginal environments call for deploy- leaving bands of natural vegetation along con- ment of a range of practices that serve as insur- tours,pilingrocksandfielddebrisinparallelstrips, ance against late rains, lower than normal rain- and building earthen bunds to create a sloping fall, or epidemics of pests and diseases. A "fruit terrace (Plucknett 1976). Such relatively inexpen- basket" approach, for example, involves deploy- sive methods of checking soil erosion can be intro- ing a palate of different crops as well as several duced quickly to an area, particularly if the plant varieties of each crop that require different nutri- provides some uses, such as thatch or fodder. ents and have different abilities to tolerate envi- Although ways to encourage building terraces ronmental stresses. Farmers on marginal lands in marginal environments should be pursued, generally have fewer resources to combat such abandoned terraces built in ancient times could be challenges, such as access to irrigation and pesti- renovated. Between half and three-quarters of the cides. Genetic resistance, or tolerance, to environ- Inca terraces are no longer used (Bray 1990). Peru mental stresses is thus especially important in and Bolivia import food for their rapidly growing marginal areas. populations and could meet at least some of their Farmers in many parts of the lowland, humid food needs by repairing old terraces. tropics need varieties that tolerate acid condi- An important indicator of sustainability in tions and aluminum toxicity. The repertoire of mountainous areas, therefore, is the degree to traditional varieties maintained by farmers in which the farmers strive to slow soil erosion. Soil marginal areas often contains hardy land races: conservation techniques should beoneof the first breeders need to tap this material when develop- priorities for any farming system (Adiningsih, ing more productive varieties. Crops and variet- Sudjadi, and Setyorini 1988). The notion that fer- ies that produce reasonably well with low fertil- tilizerscan compensate forsoil fertilitylostdueto ity, such as cassava and mango, should be pro- erosion is an illusion, even in comparatively rich moted for marginal areas rather than genotypes industrial countries. Subsoils often havedifferent 238 Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement physical and chemical characteristics than more and pathogens are not checked by cold winters friable and fertile topsoils, and these properties and thus pose a ceaseless challenge for farmers can interfere with proper root development. and plantation owners (Janzen 1970, 1973). In Some areas of the tropics can be regarded as addition to environmental concerns, pesticides marginal for agricultural development because are often too costly for most crops in the humid of their low rates of photosynthesis. Year-round tropics. Pesticides may be economically viable for warmconditionsdonotnecessarilytranslateinto certain cash crops on a limited basis, but few higher plant productivity. Many parts of the trop- farmers-large or small-can afford significant ics that have high rainfall are cloudy most of the quantities of pesticides for basic staples. Further- time, thereby reducing photosynthesis. In such more, thenumberof arthropod speciesexhibiting cases, lands should never be cleared from forest resistance to pesticides doubled from 1970 to for any other purpose, including crop production 1980, and some pests have even developed resis- and grazing. Food production is more efficient in tance to biopesticides, such as i3acillus thuringiensis temperate or tropical areas that have abundant (Anderson 1992; Brattsten and others 1986; Dover sunshine (Chang 1968). and Croft 1986). In Amazonia, the largest remaining tract of Successful control of insects and pathogens in tropical forest in the world, the lowest photosyn- the tropics hinge.; on a better understanding of thetic potential is found in the western region of traditional methods of control and modern sci- thebasin,precisely where thegreatestbiodiversity ence. Traditional farmers employ a variety of is generally found. The crumpled ranges of the techniques to control pests and pathogens, rang- towering Andes force westward-moving air to ing from polyculture to the deployment of sev- rise, thereby triggering increased precipitation eral cultivars for each crop. Breeders can help by and more overcastdays than thoseexperienced in screening traditional varieties, wild populations, other portions of the basin, except for an enclave and near relatives of crops for genetic resistance, of high rainfall around Bel6m. Areas of lower then transferring such desirable traits to higher photosynthetic potential often coincide with zones yielding material. In certain situations, particu- of increased biodiversity and endemism. Plant larly on islands and in some perennial crops, diversity increases in western Amazonia, and the biocontrol methods can be effective in checking deeply cut valleys and rain-soaked ridges of the crop pests (Smith 1990). eastern Andes and associated lower, premontane Resistance to insects and disease has emerged slopes contain an unusual assortment of fruit asa major priority amongbreeders working with trees and other plants (Clement 1991; Clement, tropical crops. AttheMexico-based International Muller, and Chavez 1982; Myers 1990; Schultes Maize and Wheat Improvement Center 1991). The progenitor of the widely appreciated (CIMMYT), an estimated 40 percent of the breed- peach palm (Bactris gasipaes) arose in western ing effort is devoted to disease resistance Amazonia (Clement 1988), as well as a variety of (Winkelmann 1987). Near relatives of crops often other fruit crops, such as papaya (Carica papaya) contain the necessary genes for disease resis- and potential new domesticates (Smith and oth- tance, as in the case of rubber (Schultes 1977), oil ers 1992). Western Amazonia is also richer in palm (Gascon, Noiret, and Meunier 1989), coffee mammalian species than other parts of the basin (Eskes 1989), sugarcane (Brandes and Sartoris (Mares 1992; Pimm and Gittleman 1992). Areas 1936), cassava (Beck 1980; Hahn 1978; Hahn and with increased cloud cover, which are potentially others 1979; Hahn, Terry, and Leuschner 1980), less propitious for agricultural development, par- and potato (Creech 1970). ticularly for food crops, thus overlap with zones Biotechnologies are being harnessed to de- particularly important for conservation. velop and disseminate disease-resistant crop va- rieties. In Viet Nam, for example, "barefoot biotechnologists" havedevised simple procedures The pest and disease treadmill so that farmers can readily multiply pest-resis- tant potatoes. European potato varieties fare Year-round warm temperatures are a double- poorly in Viet Nam, but a Viet Namese edged sword for tropical farmers. Swidden fields, biotechnologist obtained more than 100 potato agroforestry plots, and dooryard gardens can samples from the gene bank maintained by the provideacontinualharvestofbasicstaples,fruits, International Potato Center (CIP) in Peru, se- andnuts,amongothergoods,yetarthropodpests lected those that fared well in Viet Nam, then 239 Defining and Measuring Sustainability: The Biogeophysical Foundations developed an inexpensivemethod that farmerscan countries to screen rice germ plasm; entries are use to multiply planting material using test tubes. supplied by IRRI and partners in the international CIP germ plasm coupled with innovative tissue ricenurseries.IRRI-coordinatednurserieshavebeen culture techniques have allowed VietNamese farm- responsible for developing various high-yielding ers to increase potato yields from about 10 tons per varieties of rice that are resistant to evolving strains hectareto around l8tonsperhectare,withoutusing of brown plant hopper, a serious rice pest in Asia pesticides (Economist, April 18, 1992, p. 87). Al- (IRRI 1983, p. 21). thoughgenebankshavebeen criticized as "morgues and museums" (Buttel 1992), the better conceived and managed ones have made tangible and signifi- Weeds as friend and foe cant contributions to upgrading agricultural pro- ductivity around the world (Huaman and Year-round warmth in the tropics allows weeds to Schmiediche 1991; Plucknett and others 1987; Rick proliferate (Kamarck 1973). Dry seasons slow but 1991). do not halt the growth of volunteer plants in fields, Single-gene resistance against a disease rarely pastures, and plantations. Weeds are often the ma- worksforverylongbecauseoftheevolutionofnew jor reason why farmers abandon their swidden pathotypes. Many pathogenic fungi and bacteria fields and move on to clear a fresh patch of forest have several races or strains that attack crops, even (Holm and others 1977; Joachim and Kandiah 1948; in the same area. Many disease-resistant genes are Plucknett 1976). In some cases, annual burning to specific to certain races of a pathogen (Bennetzen control weedsleadstoabuildupofperennialgrasses, and others 1988). A potpourri of genes is thus suchaslmperatacylindricainSoutheastAsia(Kellman necessary to counteract most diseases; the appro- 1969). When invading perennial grasses attain al- priate gene for resistance needs to be inserted in the most pure stands, farmers have little hope of con- variety of crop to be released where that particular trolling them. Weed invasion isa majorreason why patthotype prevails. Carefullyassembled and evalu- some pastures degrade in Amazonia, thereby low- ated collections of germ plasm are thus essential to ering the carrying capacity of ranches (Nepstad, backstop breeding efforts, particularly in the trop- Uhl, and Serrao 1991). ics. Every production system has it own problem Given the increased number of insect genera- with weeds, and various environmentally benign tions possible in the tropics, access to, and deploy- strategies can be employed to check them. Some of ment of, genetic resistance iscri tical to sustainabili ty. the more intensive agricultural systems may neces- Genetic resistance is cheaper in the long run for sitate rigorous control of weeds. Herbicides can be farmers, consumers, and the environment. Conser- appropriate in such cases, because chemicals used vation of wilderness is essential for this dimension to control weeds appear to have minimal environ- to sustainability, since resistance genes are some- mental impact and can be cheaper than manual times only found in wild populations of crops or labor or cutting with machinery. theirnearrelatives. Forexample,resistancetonema- One trick in weed control is to find intercrops tode pests has been located ir. close relatives of the that out-compete the weeds, while providing potato (Chavez and others 1988; Creech 1970; valuable fodder, mulch, or nitrogen fixation. Re- Hawkes 1977; Kahn 1984, p. 105). search is under way to find suitable nitrogen- An important measure of sustainability of agri- fixing plants, such as species of Centrosemna, that cultural systemsin thetropicsisthestatusof genetic can withstand trampling and heavy grazing. Im- resources of the crops involved. Comprehensive proving the productivity of pasture can relieve and well-managed genebankssupplemented by in pressure on the remaining forests. Pasture im- situ conservationarekeyindicatorsof sustainability. provement is an issue for both small- and large- Proper evaluation is essential if genetic resources scale operators in the Amazon. Cattle are an are to contribute to sustainable agriculture. For the integral part of many small farms in Amazonia majorcropplants,internationalnurserieshavebeen and other parts of the humid tropics, since they set up to screen material for resistance to pests and provide cash in emergencies, milk for children, diseases,amongothertraits. lnternational nurseries and manure for crops (Lambourne 1937). A re- serve as testing grounds to identify promising ma- lated tactic is to shade out weeds; for this reason, terial for further crossing. The International Rice various perennial cropping systems, particularly Research Institute (IRRI), for example, has helped agroforestry, can improve the sustainability of organize a network of collaborators in dozens of agricultural lands in the tropics. 240 Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement Yet weeds are not all bad. Weed is a cultural the tropics, from jungles to desert fringes and term. So-called weeds help protect the soil from from hot lowlands to the frost-prone slopes of the erosive power of tropical downpours and high mountains. Tree crops are virtually syn- may rapidly sequester nutrients that crop plants onymous with sustainable agriculture in many areunabletocapture.Inswiddensystems,weeds parts of the humid tropics, where multistory are vanguard species that start the process of home gardens and polycultural swidden fields forest regeneration. Even the notorious alang- abound (Juo 1989). alang grass (Imperata cylindrica) in South and Except for true deserts, any tropical landscape Southeast Asia has several uses. In southern devoid of trees should be a warning signal that Kalimantan, alang-alang forms part of the rice the landscapeisdegraded. Many tropical areas, fallow system; in northeastern India and on Bali, now denuded of perennial cover, once sup- the tough grass is employed for thatch (see chap- ported woodlands, even in semi-arid regions. ter 14 of this volume); and on Sumbawa Island, Several centuries ago, the Rajasthan School of game animals browse on succulent alang-alang artists in northwestern India depicted lush for- shoots promoted by burns (Conway, Manwan, ests full of game; barren scrubland has now and McCauley 1983). Some pasture weeds in the replaced these formerly rich woodlands (Roy Amazon fix nitrogen, such as species of Cassia. 1987). Vast areas have been stripped of trees Some of our crops started out as weeds, and and shrubs by fuelwood gatherers, goats, tim- uninvited plants in fields sometimes cross with ber traders, miners, farmers, and ranchers. Even crops, thereby enriching the domestic gene pool areas lush with annial crops or pasture may (Harlan 1965). In Mexico, for example, bees trans- mask ecological land mines created by the loss ferpollenfromwild fieldbeans(Phaseolusvulgaris) of tree cover. Soil erosion accelerated by wind to nearby planted beans. Farmers even recognize and torrential downpours will eventually take that the presence of "weedy" beans is a help to their toll, and in dry areas, trees that formerly their fields, since it fortifies their crop. In parts of trapped precious moisture from fog or low Mexico and Guatemala, maize (Zea mays) ex- clouds are often gone. changes genes spontaneously with annual te- One measure of sustainability in tropical agri- osinte (Zea mays mexicana), a weedy relative culture is the degree to which trees are incorpo- (Doebleyl990;Galinat,Mangelsdorf,andPiersen rated in development plans. Ideally, some natu- 1956). Introgression of weeds with crops has also ral woodlands should beallowed to remain stand- been recorded or suspected with pearl millet ing to secure supplies of a variety of goods, such (Harlan 1971), finger millet (Hussaini, Goodman, as water, wood, game, medicinal plants, and fruits. and Timothy 1977), groundnut (Williams 1989), For example, fruit pulp from the buriti (Mauritia potato (Hawkes 1977), various squash species, flexuosa) palm and the acai (Euterpe oleracea) palm and quinoa, a protein-rich crop of the high Andes is readily sold in Brazilian cities. Buriti and acai now being sold by some health food stores in the grow in forest along streams and rivers and pro- United States (Wilson 1990). vide widely appreciated fruits and, in the case of Although weedscertainly need tobecontrolled acai, heart-of-palm as well. as much as feasible within cultivated plots and Simply setting aside forest reserves is not plantations, a preoccupation with totally "clean" enough; these areas must provide tangible ben- fields and borders is not only too costly in many efits for the local population, and surrounding cases, but also undesirable. Ruderal plant com- settlements need to be supported by viable agri- munities may provide havens for biocontrol culture. As part of this strategy, tree planting agents thathelp suppresscroppests and mayalso should be encouraged, but in practice this noble provide food for pollinators of cultivated plants. idea is often difficult to implement. Two main considerations are necessary when promoting increased tree and bush cover in the Tree crops as pillars of sustainability tropics. First, the species involved must be suited to local ecological, cultural, and market condi- In addition to weed control, tree crops can pro- tions. Second, no one formula or blueprint will vide an array of other environmental services. necessarily work over a broad area. A top-down With proper spacing and management, planted approach to promoting perennial crops will not trees can help restore soil fertility and check ero- work; prescriptions must match the rich ecological sion. Tree crops are well suited for most areas of and cultural heterogeneity of the target area. 241 Defining and Measuring Sustainability: The Biogeophysical Foundations Local people must be canvassed about their A clearer appreciation of the complementary needs and aspirations, and market conditions interactions and antagonisms between land uses need to be analyzed, for tree-planting schemes to is a difficult task. Forces of change need to be be successful. Space is at a premium on most elucidated and payoffsbetween land use systems farms, ranches, and plantations, so selecting ap- pinpointed. Some land uses can be detrimental to propriate candidates for reforestation and for sustainableagriculture,yetinsomecircumstances planting living fences in backyards, in orchards, also beneficial. Al though often controversial, rais- and on ranches is critical. Also, varieties multi- ingcattleinrainforestareasoftenmakessensefor plied for distribution and sale should be able to smallholders,becausemanureisrecycledtocrops, withstand prevailing diseases and pests and be biogas digesters provide lighting and cooking backed up by breeding and selection programs in fuel, and cows provide milk for children. Larger case the situation of pathogens and insect pests ranches may degrade the environment but, if changes. Mistakes made with the adoption of properly managed, can be sustainable and pro- perennial crops are much more costly to rectify vide much-needed off-farm employment; such than those made with the adoption of annuals. outside income often makes the difference be- Trees and bushes can be planted in a wide tween the survival or failure of a farm. Improved array of cropping patterns, from polycultural pasture management is essential, however, par- plots to small, monocultural orchards and large- ticularly on grazing lands cleared from forest. scale plantations. Agroforestry is already widely A mosaic of land use systems has evolved in practiced, and further insights can be gained by most areas of the tropics, particularly in wetter studyingindigenoussystems. Butagroforestryis areas, in response to the rich cultural fabric and no panacea: some speciesdo notgrow well under patchwork of biomes and ecosystems of the re- mixed-croppingconditions.Single-speciesstands gion. This rich quilt of farms, fields, and planta- can be sustainable. Diversity does not always tions is a heritage of in situ tradition supple- mean greater stability or resiliency. mented by the diffusion of new technologies and ideas. Diversity in time and space, as exemplified by different uses of the land, can help provide a The land use mosaic buffer against catastrophic disease and pest epi- demics, among other environmental threats. Thus Just as a range of planting arrangements are ap- diversity of cropping and livestock systems on propriate for perennials, so a diverse array of individual farms, as well as between different land use systems can improve sustainability of properties,isfundamentaltothesustainabilityof agriculture. A farm, ranch, or plantation is em- tropical agriculture. bedded in a large socioeconomic and environ- Developmentplansforatropicalregionwould mental arena, and a better understanding of the do well to take into account the numerous ongo- system's interactions and synergisms is critical to ing experiments devised by operators of rural devising more productive and environmentally enterprises, from small to large scale. Successful benign rural enterprises. analogs can often be found among some ranchers, A single land use strategy for sustainable agri- farmers, and plantation owners, rather than on culture will not work in most tropical areas. Dis- experiment stations. Although still needed on cussions on sustainability typically reflect the experiment stations, carefully controlled trials particular interests of their proponents, such as often investigate componentsof farming systems agroforestry or extractive reserves. Although and are more likely to be successful if local expe- agroforestry in its various forms and forest rience is taken into account. preserves where people harvest products cer- In addition to tapping local knowledge on tainly have their place, they are only pieces of a successful cropping and livestock enterprises, broader pattern of land use. A variety of inter- scientists can enhance the impact of their workby locking annual and perennial cropping systems, conducting as much research on-farm as feasible. orchards, ranches, and plantations is often found In this manner, extension work is built into the in the tropics. The relative area occupied by each experiment,sincefarmerswill morereadilyadopt and the crops involved are usually highly dy- technologies they see benefiting their neighbors. namic. Identifying and understanding such shift- Progress toward sustainability will be quickened ing relationships in land use are fundamental to if both traditional knowledge and the frontiers of sustainable development. science are harnessed (Swaminathan 1990). 242 Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement Researcheis also need to be sensitive to the Another critical dimension to intensification is driving forces behind changes in land use. The monitoring the health of soil. Soil erosion is only waxing and waning of particular cropping pat- one dimension to the productive capacity of the terns, the turnover of varieties, the encroachment land. Intensification will entail a periodic report of cattle ranching all need to be understood and card on facets of soil such as organic matter, pH, this knowledge used to help set research priori- level of phosphorus, salinity, and compaction. ties. A clearer understanding of the dynamics of Inorganic fertilizerscannot correct fordiminished changes in land use can also help fine-tune fiscal levels of organic matter and compaction. Regular and socioeconomic policies. monitoring of soil conditions is expensive, but a necessary price to pay for more intensively man- aged systems. Several temperate-zone countries Intensification: A pantropical imperative have recently established environmental moni- toring systems to assess stresses created by vari- At first glance, intensification, environmental ous agricultural activities (see chapter 17 of this health, and sustainability might seem incompat- volume); determining how to adapt such sys- ible. The notion of intensification often conjures tems, and bring down their costs, represents a up images of tractors causing plow pans, fertiliz- major challenge for the tropics. ers causing eutrophication, herbicides destroy- ing nontarget plant communities, and pesticides entering food chains. The case for intensification A blend of traditional knowledge has been alluded to in our discussion of marginal and modern science lands and genetic resources. It also applies to the mosaic of land use systems: each system will have Answers to sustainability in tropical agriculture to be managed more efficiently and intensively to will be found among traditional farmers, field coax more goods and services for a given area. workers from such disciplines as anthropology, An enormous research agenda lays ahead in geography, and agricultural scientists, including this global imperative to intensify agriculture, biotechnologists. The appropriate mix of tradi- including agroforestry, plantation systems, and tional knowledge and modern science will vary livestock raising. How can this be done without widely, depending on ecological constraints and destroying the resource base? This complex issue market opportunities. In some highly intensive will require careful consideration by a range of agricultural systems on the optimal farm lands, disciplines, but a few indexes can be laid out here. technologies to raiseand sustain yieldsmaycome Oneconsiderationin intensification is thedegree more from test tubes than from thecrucible of folk to which unanticipated outputs, which economists experience. In other situations, such as marginal sometimes call externalities, might destroy the pro- environments, traditional systems of resourceman- ductive capacity of the land or damage the environ- agement are likely to have much more to offer. ments-andpeoplewholiveinthem-downstream The imperative of tapping the frontiers of sci- from farming activities. Specifically, metabolic by- ence and plumbing the depths of indigenous products from farms and ranches, such as fertilizer knowledge underscores the value of an interdis- not captured by plants and pesticides, can ruin ciplinary approach to sustainability. The ethno- water supplies for rural and urban folk, poison fish, botanist, anthropologist, and biotechnologist all or destroy aquatic environments so that game fish have contributions to make in the search for and turtle species are unable to survive. sustainability. More-intensive agricultural systems will ne- cessitategreateruseof energy. In the short term at least, the consumption of fossil fuel will thus An odyssey without end increase in rural areas, and the implications of this for greenhouse gases warrant assessment. Sustainability is a ceaseless journey, rather than a Some intensive agricultural systems, such as the destination. Successful research for sustainable cultivation of paddy rice, emit appreciable agriculture hinges on a multidisciplinary and amountsof methane,oneof thegreenhousegases. highly flexible agenda, rather than a method- Intensification of agriculture thus entails careful ological straightjacket. It also hinges on contin- monitoring for pollutants or gaseous emissions ued investments in research, an important mes- that may affect the local or even global climate. sage to convey to governments. 243 Defining and Measuring Sustainability: The Biogeophysical Foundations As frustratingas this somewhat imprecisedefi- conserved, both in situ and in collections of germ nition of sustainability may appear, dealing with plasm, to uphold breeding efforts. peoples' cultures and economies, as well as the In regard to genetic resources, several points intricacies of ecosystems modified by humans, warrant emphasis. First, it is important to con- raises the task of identifying and measuring all servewildemess,whichsometimescontainswild the factors that impinge on raising sustainability populations of existing crops and domesticated to an even higher orbit. Resiliency is an important animals, as well as theirnear relatives. Biodiversity property in this quest for sustainability, both in is essential for the long-term success of agricul- humnan institutions as well as in the managed ture (Pimentel and others 1992). Second, modi- ecosystems that societies have created. fied habitats can also be important for preserving To find the paths to sustainable agriculture in genetic resources, as long as they are not de- the tropics, inputs from many disciplines need to stroyed. Third, existing crop gene banks need be tapped, ranging from ecology to plant pathol- further support to improve their coverage, stor- ogy and the social sciences. Sustainability is re- age conditions, and evaluation of their acces- ally an approach or a perspective, rather than a sions.Morecollectionsofgermplasmareneeded, specific set of practices that can be easily mea- especially for the so-called minor crops, and col- sured and thus quantified. lections should be duplicated to avoid irrevers- Although more work is needed to improve our ible losses. Development withoutconcern forcon- understanding of thedimensions to sustainability servation will not work, nor will conservation and how they can be measured, flexibility should without development that improves living con- always be paramount. Rigid definitions and com- ditions. Both are inescapably linked. partmentalization, while making the task of mod- eling and prediction apparently easier, may lose some of the dynamism and complexity of tropical Ref erences agriculture. Satellite imagery and geographic information Adiningsih, G. S. A., M. Sudjadi, and D. Setyorini. systems can help researchers obtain a better over- 1988. "Overcoming Soil Fertility Constraints view of land use systems in the tropics as well as in Acid Upland Soils for Food Crop Based identify early warning signals of environmental Farming Systems in Indonesia." Indonesian Ag- distress. Remote sensingcan detect the early stages ricultural Research and Development Journal 10:2, of a disease or pest epidemic and can alert pp. 49-58. policymakerstothespreadingimpactofadrought. Altieri, M. A., D. K. Letoumeau, and J. R. Davis. Microcomputer-based systems have been devel- 1983."DevelopingSustainableAgroecosystems." oped for receiving satellite data that are small and BioScience 33, pp. 45-49. inexpensive enough for use in the field. Such systems are already being used in Botswana to Anderson, C. 1992. "Researchers Ask for Help to assist in land and water management and in Save Key Biopesticide." Nature 355, p. 661. Ethiopiatomonitordroughts(Disney1991).Geo- Beck, B. D. A. 1980. "Historical Perspectives of graphic information systems can map spatial in- Cassava Breeding in Africa." In Root Crops of formation on farm production, population clus- Eastern Africa: Proceedings of a Workshop, 23 ters,and othersocioeconomicand ecological data. November 1980, Kigali, Rwanda, pp. 13-18. Ot- Now a better integration of these two powerful tawa: IDRC. tools is needed to further ourefforts tobuild more Bennetzen, J. F., M. M. Qin, S. Ingels, and A. H. sustainable agricultural systems in the tropics. Ellingoe. 1988. "Allele-specific and Mutator- In spite of the difficulty of even gaining a firm associated Instability at the Rpl Disease-resis- grasp on all the facets of sustainability, certain tance Locus of Maize." Nature 332, pp. 369-70. conceptual features emerge from our discussion. Brandes, E. W., and G. G. Sartoris. 1936. "Sugar- Several key threads link the various facets of cane: Its Origins and Improvement." In Year- sustainable agriculture in the tropics. Soils need book of Agriculture, pp. 561-623. Washington, to be conserved, and methods of recycling nutri- D.C.: U.S. Department of Agriculture. ents need to be improved. Farmers, ranchers, and Brattsten, L. B., C. W. Holyoke, J. R. Leeper, and plantation owners need to be canvassed for their K. F. Raffa. 1986. "Insecticide Resistance: Chal- local knowledge as a prelude to introducing new lenge to Pest Managementand Basic Research." technologies. Genetic resources need to be better Science 231, pp. 1255-60. 244 Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement Bray, W. 1990. "Agricultural Renaissance in the Doebley, J. 1990. "Molecular Evidence for Gene High Andes." Nature 345, pp. 385. Flow among 2ea Species." BioScience 40, pp. Brklacich, M., C. R. Bryant, and B. Smit. 1991. 443-48. "Review and Appraisal of Concept of Sustain- Dover, M. J., and B. A. Croft. 1986. 'Testicide Resis- able Food Production Systems." Environmental tance and Public Policy: Resistance Management Management 15:1, pp. 1-14. Could Become the Key to Continuing Effective Buttel, F. H. 1992. "The Environmentalization of Pest Control." BioScience 36, pp. 78-85. Plant Genetic Resources: Possible Benefits, Egger, K., and B. Martens. 1987. "Theory and Possible Risks." Diversity 8:1, pp. 36-39. Methods of Ecofarming and Their Realization CGIAR, Technical Advisory Committee (Consul- in Rwanda, East Africa." In B. Glaeser, ed., The tative Group on International Agricultural Green Revolution Revisited, pp. 150-75. London: Research). 1988. Sustainable Agricultural Pro- Allen and Unwin. duction: Implications for International Agricul- Eskes, A. B. 1989. "Resistance." In A. C. tural Research. Rome. Kushalappa and A. B. Eskes, eds., Coffee Rust, Chang, J. 1968. "The Agricultural Potential of the pp. 171-291. Boca Raton, Fla.: CRC Press. Humid Tropics." Geographical Review 58:3, pp. FAO (Food and Agriculture Organization of the 333-61. United Nations). 1991. Criteria, Instruments, Chavez, R., M. T. Jackson, P. E. Schiemidiche, and and Tools for Sustainable Agriculture and Rural J. Franco. 1988. "The Importance of Wild Po- Development. Proceedings of the Food and Ag- tato Species Resistant to the Potato Cyst Nema- riculture Organization/Netherlands confer- tode, Globodera pallida, Pathotypes P4A and ence Agriculture and the Environment, 'S- P5A in Potato Breeding. I. Resistance." Hertogenbosch, Netherlands, April 15-19. Euphytica 37, pp. 9-14. Rome. Clement, C. 1988. "Domestication of the Pejibaye Galinat, W. C., P. C. Mangelsdorf, and L. Piersen. Palm (Bactris gasipaes): Past and Present." Ad- 1956. "Estimates of Teosinte Introgression in vances in Economic Botany 6, pp. 155-74. Archaeological Maize." Leaflets of the Botanical . 1991. "Amazonian Fruits: Neglected, Museum of Harvard University 17, pp. 101-24. Threatened, and Potentially Rich Resources Gascon, J. P., J. M. Noiret, and J. Meunier. 1989. Require Urgent Attention." Diversity7:1-2, pp. "Oil Palm." In G. Robbelen, R. K. Downey, and 56-59. A. Ashri, eds., Oil Crops of the World: Their Clement, C., C. H. Muller, and W. B. Chavez. Breedingand Utilization, pp. 475-93. NewYork: 1982. "Recursos geneticos de especies frutiferas McGra w-Hill. nativas da Amazonia Brasileira." Acta Hahn,S. K. 1978. "Breeding Cassava for Resistance Amazonica 12:4, pp. 677-95. to Bacterial Blight." PANS 24:4, pp. 480-85. Conway, G. R., I. Manwan, and D. S. McCauley. Hahn, S. K., E. R. Terry, and K. Leuschner. 1980. 1983. "The Development of Marginal Lands in "Breeding Cassava for Resistance to Cassava the Tropics." Nature 304, pp. 392. Mosaic Disease." Euphytica 29, pp. 673-83. Cook, 0. F. 1909. "Vegetation Affected by Agri- Hahn, S. K., E. R. Terry, K. Leuschner, 1. 0. culture in Central America." Bureau of Plant Akobundu, C. Okali, and R. Lal. 1979. "Cas- Industry Bulletin 145. U.S. Department of Ag- sava Improvement in Africa." Field Crops Re- riculture, Washington, D.C. search 2, pp. 193-226. Creech, J. L. 1970. "Tactics of Exploration and Harlan, J. R. 1965. "The Possible Role of Weed Collection." In 0. H. Frankel and E. Bennett, Races in the Evolution of Cultivated Plants." eds., Genetic Resources in Plants: Their Explora- Euphytica 14, pp. 173-76. tionand Conservation, pp. 221-29. Philadelphia, . 1971. "Agricultural Origins: Centres Penn.: F. A. Davis. and Non-centres." Science 174, pp. 468-74. Davidson, B. 1969. A History of East and Central Hawkes, J. G. 1977. "The Importance of Wild Africa to the Late Nineteenth Century. Garden Germ Plasm in Plant Breeding." Euphytica 26, City, N.Y.: Anchor Books. pp. 615-21. Disney,J 1991. "Resource Assessmentand Farm- Holm, L. G., D. L. Plucknett, J. V. Pancho, and J. ing Systems." Resource 4, p. 4. P. Herberger. 1977. The World's Worst Weeds: 245 Defining and Measuring Sustainability: The Biogeophysical Foundations Distribution and Biology. Honolulu: University Mares, M. A. 1992. "Neotropical Mammals and Press of Hawaii. the Myth of Amazonian Biodiversity." Science Huaman, Z., and P. Schmiediche. 1991. "The 255, pp. 976-79. Importance of Ex Situ Conservation of Germ Myers, N. 1990. "The Biodiversity Challenge: Plasm: A Case Study." Diversity 7:1-2, pp. Expanded Hot-spots Analysis." The Environ- 68-69. mentalist 10:4, pp. 243-56. Hussaini, S. H., M. M. Goodman, and D. H. Timo- Nepstad, D. C., C. Uhl, and E. A. S. Serr3o. 1991. thy. 1977. "Multivariate Analysis and the Geo- "Recuperation of a Degraded Amazonian graphical Distribution of the World Collection Landscape: Forest Recovery and Agricultural of Finger Millet." Crop Science 17, pp. 257-63. Restoration." Ambio 20:6, pp. 248-55. IRRI (International Rice Research Institute). 1983. Pimentel, D., U. Stachow, D. A. Takacs, H. W. Research Highlights for 1982. Los Bafios, Philip- Brubaker, A. R. Dumas, J. J. Meaney, J. A. S. pines. C'Neil, D. E. Onsi, and D. B. Corzilius. 1992. Janzen, D. H. 1970. "Herbivores and the Number "ConservingBiologicalDiversityinAgricultural/ of Tree Species in Tropical Forests." American Forestry Systems." BioScience 42, pp. 354-62. Naturalist 104, pp. 501-28. Pimm, S. L., and M. E. Gittleman. 1992. "Biologi- .1973. "Tropical Agroecosystems: These cal Diversity: Where Is It?" Science 255, p. 940. Habitats Are Misunderstood by the Temper- Plucknett, D. L. 1976. "Hill Land Agriculture in ate Zones, Mismanaged by the Tropics." Sci- the Humid Tropics." In Hill Lands, Proceedings ence 182, pp. 1212-19. of International Symposium, 3-9 October, West Joachim, A. W. R., and S. Kandiah. 1948. "The Virginia University, Morgantown, pp. 29-38. Effect of Shifting (Chena) Cultivation and Sub- Journal Series 2074. Honolulu: Hawaii Agri- sequent Regeneration of Vegetation on Soil cultural Experiment Station. Composition and Structure." Tropical Agricul- Plucknett, Donald L., and Nigel J. H. Smith. 1982. ture 54:1, pp. 3-11. "Agricultural Research and Third World Food Juo, A. S. R. 1989. "New Farming Systems Devel- Production." Science 217, pp. 215-20. opment in the Wetter Tropics." Experimental Plucknett, Donald L., Nigel J. H. Smith, J. T. Agriculture 25, pp. 14543. Williams, and N. M. Anishetty. 1987. Gene Kahn, E. J. 1984. The Staffs of Life. Boston: Little, Banks and the World's Food. Princeton, N.J.: Brown, and Company. Princeton University Press. Kamarck, A. M. 1973. "Climate and Economic Rick, C. M. 1991. "Tomato Resources of South Development." Finance and Development 10:2, America Reveal Many Genetic Treasures." Di- pp. 2-8. versity 7:1-2, pp. 54-56. Kellman, M. 1969. "Some Environmental Com- Roy, R.1987. "Trees: AppropriateToolsforWater ponents of Shifting Cultivation in Upland and Soil Management." In B. Glaeser, ed., The Mindanao." Journal of Tropical Agriculture 28, Green Revolution Revisited, pp. 111-25. London: pp. 40-56. Allen and Unwin. Lambourne, J. 1937. "Experiments on the Eco- Schultes, R. E. 1977. "Wild Hevea: An Untapped nomic Maintenanceof Soil Fertility underCon- Source of Germ Plasm." Journal of the Rubber tinuous Cropping with Tapioca." Malayan Ag- Research Institute (Sri Lanka) 54, pp. 227-57. ricultural journal 25:4, pp. 134-45. . 1991. "Ethnobotanical Conservation Leonard, H. J. 1989. "Environment and the Poor: and Plant Diversity in Northwest Amazon." Development Strategies for a Common Diversity 7:1-2, pp. 69-72. Agenda." In H. J. Leonard, M. Yudelman, J. D. Smith, Nigel J. H. 1990. "Strategies for Sustain- Stryker, J. 0. Browder, A. J. De Boer, T. ableAgricultureintheTropics."EcologicalEco- Campbell, and J. Allison, eds., Environment and nomics 2, pp. 311-23. the Poor: Development Strategies for a Common Smith, Nigel J. H., J. T. Williams, Donald L. Agenda,pp.3-45.U.S.-Third World Policy Per- Plucknett, and J. P. Talbot. 1992. Tropical For- spectives 1 1. Brunswick, N.J.: Overseas Devel- ests and Their Crops. Ithaca, N.Y.: Cornell Uni- opment Council. versity Press. 246 Sustainable Agriculture in the Tropics: Issues, Indicators, and Measurement Swaminathan, M. S. 1990. "Changing Nature of Wilson, H. D. 1990. "Gene Flow in Squash Spe- the Food Security Challenge: Implications for cies." BioScience 40, pp. 449-55. Agricultural Research and Policy." Sir John Winkelmann, D. L. 1987. "Diversification, Crawford Memorial Lecture, Consultative Sustainability, and Economics." In T. J. Davis Group on International Agricultural Research and I. A. Schirmer, eds., Sustainability Issues in (CGIAR), World Bank, Washington, D.C. Agricultural Development, pp. 295-303. Wash- Williams, D. E. 1989. "Exploration of Amazonian ington, D.C.: World Bank. Bolivia Yields Rare Peanut Land Races." Diver- York, E. T. 1988. "Improving Sustainability with sity 5:4, pp. 12-13. Agricultural Research." Environment 30:9, pp. 18-20, 36-40. 247 Defining and Measuring Sustainability: The Biogeophysical Foundations Comments areas. Marginal areas do not always-or only- mean areas with eroded and unfertile soils, be- cause even fertile land with a low rate of photo- R. Maria Saleth synthesis can be considered marginal in terms of net primary productivity even though it coin- A few critical issues need to be settled before cides with increased biodiversity and endemism. pursuing the task of defining the concept of sus- An implicit distinction exists between physically tainable development. Just as defining the con- marginal areas and ecologically marginal areas, cept is a precondition for determining a measure and this distinction has implications for of sustainable development, so is determining sustainability indicators. In addition to the the context in and the level at which we need to sustainabilityimplicationsofmarginalareas,pests define sustainable development. Although it is and plant diseases and weeds also pose serious tempting to define sustainable development in a problems to the sustainability of tropical agricul- general fashion tocover everything fromozone to ture. oysters, from a policy and operational point of Smith and Plucknett have suggested a few view, we need to define the concept in the par- important magnitudes or dimensions on which ticular context of a given region or ecosystem or potential biophysical indicators of agricultural sector. sustainability in the tropical regions should fo- Still more important, in my view, is the choice cus: of domain-whether physical or economic-in . The extent of farmers' efforts to slow soil ero- which the concept is to be defined. Many of us sion would agree that the concept of sustainability has a The status of genetic resources available ultimately to be defined and measured in the biophysical domain, but such a definition can * The role of trees in development plans neverbeindependentoftheeconomic,social,and a The extent to which land use is diverse (the even the cultural aspects. In other words, the land use mosaic) economic, social, and cultural criteria will be the * The magnitude of intensification imperative driving forces for the biophysical definition and and measurement. It is well to keep in mind that our * The extent to which indigenous systems of objective is to achieve the sustainability of the knhowledge can be blcnded with modern biophysical systemnot foritsown sakebut for the science. benefit of the people both in the present and in the future. Againstthisbackdrop,Smithand Plucknett The variables selected to reflect each of these have made significant contributions to our un- dimensions may vary depending on the avail- derstanding of the requirements for a proper ability of data as well as the very appropriateness definition and measurement of the sustainability and capacity of the selected variable(s) for repre- of the agricultural systems in tropical regions. senting these magnitudes adequately. For in- Smith and Plucknett have viewed sustainable stance, farmers' efforts to arrest soil erosion can development in the context of the agricultural be reflected by a host of variables both financial systems in the tropical regions as having four and technologicalinnature. These variablesrange dimensions:biophysical,institutional and policy, from quantitative measures like the financial in- sociocultural, and economic. They are careful to vestment in soil control measures to mere quali- distinguish sustainability from the notionsof sta- tative measures like soil control practices or ap- bility, preservation, and low-cost production. Sta- proaches adopted. bilityisa necessary, but nota sufficient, condition Similarly, the role of trees can be reflected by for sustainability because some apparently stable variables like the number and type of trees, the systems can actually be very brittle. Some of the proportion of farm income-both monetary and sufficient conditions for sustainability include real-derived from tree cultivation, the area un- the resilience of the system and its ability to der tree crops as a proportion of the total area absorb and withstand stress and shocks. under farms, and so forth. Likewise, a multiplic- Cross-countryexperienceindicates that within ity of variables reflects both genetic capability thetropics,concemsaboutsustainabilityarecriti- and diversity of land use. Variables such as the cal fortheagricultural systemsbased on marginal extent of area affected or yield loss caused by pest 248 Sustainalelc Agriculture in the Trotics: Issues, Indicators, and Measurement and plant diseasescan reflect indirectly the extent turning the ecologically marginal areas-regions of genetic and land use uniformity present in the withlowerphotosyntheticratesandcloudcover- system besidesindicating thepresenceof climatic into a zone for the conservation of biodiversity. and other factors conducive to the growth of Similarly, weeds can be exploited for their eco- pests. Further, thegenetic diversity of the tropical nomic and genetic benefits. Although weeds al- system also depends on factors exogenous to the low gene flow across cultivated species, they are region to the extent that gene banks and genetic also used both as thatching materials and animal research institutionsmaintained elsewhere contrib- feed. One more fundamental input for improving ute to the genetic potential of tropical agriculture. the sustainability of the tropical agricultural sys- A few indirect indicators implicitly suggested tems, not appreciated enough so far, is the pool of by Smith and Plucknett,such as the rateof photo- indigenous knowledge for resource and genetic synthesis (and cloud cover) and biodiversity, are conservation. Further, biotechnological inputs also evident. Since these two biological or bio- should also be brought down to the village and chemical factors determine biological productiv- farm levels through on-farm research trials that ity (or net primary productivity) essential for integratebothresearchandextensionintoasingle ensuring the ecological sustainability of any eco- process. system, the measures that can reflect these two From a conceptual point of view, the authors' factors will necessarily form part of the set of four-dimensionalconceptionofsustainabilitycan, sustainability indicators suitable for tropical ag- in fact, be reduced to a three-dimensional one riculture. Given the authors' four-dimensional with ecology, economics, and equity dimensions conception of agricultural sustainability, no single insofar as the sociocultural aspects are fixed in the indicator canadequately reflect the sustainability context of a given region or ecosystem and the status of agriculture (or other systems) either in institutional and policy aspects are manifested in tropical or temperate regions. Naturally, we need the economic dimension. However, from a meth- asetofincdicatorsreflectingallfourdimensionsof odological point of view, Smith and Plucknett sustainability noted by Smith and Plucknett. have correctly emphasized the contextual nature These four magnitudes are related only to the of the concept of sustainability both in time and ecological dimension of sustainability. Although space. That is, the definition and measurement of it is true that the ecological system forms the sustainability are specific to a sector or region foundation on which our economic and social since they take stock of the specific conditions systems are based, we also need to concen trate on and requirements of each sector (agriculture, in- theother twodimensionsof sustainability, that is, dustry, and so forth) as well as the requirementof economicsand ethics.Although I agreewithSSmith different regions (also ecosystems), each differ- and Plucknett that sustainability is not a destina- ing in their endowmTient of resources. Likewise, tion but only a journey, I also note that the time factor also plays an important role be- sustainability is only a relative concept. cause what is considered sustainable today may The chapter makes several important contri- not be sustainable tomorrow. I earnestly believe butions. One of them is to note the potential that recognizing thecontextual natureof the con- possibility of turning certain problems into ben- cept of sustainability is a critical first step toward efits. That is, sustainable development of agricul- developing efficient indicators for evaluating ture depends on how apparent problems can be sustainability in any given context of time and converted into potential benefits: for instance, by space. 249 Biophysical Measurement of the Sustainability of Temperate Agriculture C. Lee Campbell, Walter W. Heck, Deborah A. Neher, Michael J. Munster, and Dana L. Hoag There is a growing awareness of the need to have done very little with biological (response) manage agroecosystems in such a way as to as- indicatorsthatmightbetterfacilitatetheinterpre- sure a continuing supply of agricultural com- tation of sustainability of ecosystems. moditiesforhumanusewithoutendangeringour Thus, it is significant that the conference on natural resources.The term sustainabilityisfairly which this volume is based gave focus to both recent, but there has long been a cadre of persons sustainability and the development of biophysi- concerned with natural resources who have been cal measurements that can help interpret the prophets for the concept. In truth, many struggle sustainability of agricultural lands. This chapter to define sustainability for agroecosystems. One addresses sustainability of agroecosystems and reason for this struggle may be that many people identifies biophysical measurements that are ap- are driven more by their view of current agricul- propriate for monitoring programs to assess sta- tural practices as unsustainable and less by what tus and trends of agroecosystem health. they want to seeas future practices in agricultural systems. The challenges, then, are to define the desired Sustainability, agroecosystems, nature or characteristics of a sustainable and monitoring for status and trends agroecosystem and to establish how to measure the level of sustainability of agroecosystems. The The concept of sustainability in temperate agri- biophysical measurements that are chosen may culture has many definitions. As a working defi- reshape the definition of sustainability. Thus, the nition, a sustainable agricultural system is "one types and the spatial and temporal intensity of that, over the long term, enhances environmental measurements must be determined carefully. Do quality and the resource base on which agricul- we need comprehensive measurements taken at ture depends, provides for basic human food and numerous times and over a preselected, limited fiber needs, is economically viable, and enhances geographic area, or do we need carefully selected the quality of life for farmers and society as a measurements that can be taken at selected times whole" (Schaller 1990). over a growing season at a regional, national, or The agroecosystem must be considered as en- international spatial scale in order to assess the compassing more than the cultivated crop and impact of stressors on the system? We have long more than the part of the landscape that produces obtained physical and chemical measurementsof the food, fiber, and shelter normally associated the environment around us to help interpret and with agricultural activities. Agroecosystems are a predict the potential for agricultural production dynamic association among crops, pastures, live- in the international community. However, we stock, and associated biota, atmosphere, soils, Defining and Measuring Sustainability: The Biogeophysical Foundations and water. Agroecosystems are part of a larger of chemical use, but if chemical use must be agricultural landscape that includes adjacent un- excessively high to maintain production, then it is cultivated land, drainage networks, and other likely that both production and resource quality vegetation and wildlife. Complementary to this will not be maintained in the long term. definitionistheconcept that theagricultural land- The measurements that serve as the basis for scape also contains farmers, farm workers, and quantification of assessment end points are indi- rural communities. People are integral to the cators, or measurement end points. Thus, mea- functioning of agroecosystems and agricultural surement end points are characteristics of the landscapes; they make decisions concerning land ecosystemand environment that, when measured, use and selection of crops and livestock, and they quantify the magnitude of stress, habitat charac- impose social and economic values directly onto teristics, degree of exposure to stressors, or de- the agroecosystemand theagricultural landscape gree of ecological response to an exposure (Neher 1992). (Hunsaker and Carpenter 1990). Measurement This definition of agroecosystems recognizes end points, or indicators, canbe divided into four their complexity and emphasizes a holistic ap- classes: response, exposure, habitat, and stressor. proach that considers all components of agricul- Because of the multiple meanings of the word tural landscapes. For an agroecosystem to be indicator in the ecological literature, we use the healthy, a balance must exist between the sustain- terms assessment end point and measurement able production of crops and livestock; mainte- end point. nance of air, soil, and water quality; and assurance A number of natural and human-induced ofdiversityofwildlifeandvegetationinthenoncrop stresses and activities could affect the habitats. A change in anyonecomponent influences sustainability of agroecosystems as well as other the other components in the agroecosystem and in components of agricultural landscapes (see table adjacent, linked ecosystems. 17-1). These include availability and use of water, The selection of measurements that could be cultivation and loss (erosion) of soil, salinization used in a monitoring program to quantify or and alkalization of soil, use and accumulation of indicate sustainability in agroecosystems is criti- chemicals, air pollution, solid waste disposal on cal. Societal values associated with sustainability land, nutrient loss, water-logging, intensive ani- must be determined, and assessment end points mal production, fish farming, and climate. Each must be identified to reflect these social values. of these factors or issues should be addressed in Societalvaluesintegrateindependent,individual a program to monitor the sustainability of values. Therefore, no single answer exists about agroecosystems. Several monitoring programs what is best for society, and these values or con- already incorporate some of the measurements. cepts will be very general. Nevertheless, any Those factors and management strategies that monitoring programmust identify specific,quan- serve to maintain agroecosystem health, as quan- tifiable characteristics that indicate whether the tified by assessment end points, must be identi- perceived societal values are being maintained. fied and encouraged. Many current practices may Assessment end points are "formal expres- prove to be effective, but others may need to be sions of the actual environmental value that is to changed if we are to maintain healthy ecological beprotected,andassuchshouldhaveunambigu- systems. One facet of a monitoring program de- ous operational definitions, as well as social or signed to measure sustainability must be an as- biological relevance" (Knapp and others 1990). sessment of environmental risk. It is expected Assessment end points must be quantitative or that the assessment and measurement end points quantifiableexpressions (Suter 1990). These char- chosen would be useful in risk assessment. Addi- acteristics or assessment end points may be (1) tionally, risk assessment necessitates the assign- single measurements, (2) indexes built from sev- ment of statistical uncertainties to the data, in- eral measurements, or even (3) broader catego- dexes, and conclusions being reported. riesof concern (soil quality) that will beevaluated based on many measurements. Some end points relate to attributes that society wants to sustain, Assessment and research monitoring such as crop productivity. Others are attributes programs that may anticipate changes in the ecosystem (Marten 1988). For example, society may not be It is important to differentiate between assess- particularly interested in sustaining a certain level ment and research monitoring programs. Research 252 Biophysical Measurement of the Sustainability of Temperate Agriculture Table 17-1: Biophysical Measurements Available or Desired for Assessment of Sustainability of Temperate Agroecosystems Issue and measurement Current use Key references Water availability and use (applicable to irrigated, fed, flood-fed, arid, and humid agriculture) Soil moisture EMAP, 26%a Hillel 1982; Barnwell and others 1991 Water release curves None Hillel 1982 Area irrigated EMAP Heck and others 1992 Source of irrigation water used EMAP, NRI Heck and others 1992; Goebel and Schmude 1981 Efficiency of irrigation water used None None Water infiltration rate None Hillel 1982 Hydraulic conductivity None Hillel 1982 Soil texture EMAP, NRI Heck and others 1991; Goebel and Schmude 1981 Soil cultivation and loss (methods of soil tillage, sedimentation, and erosion) Tillage methods employed EMAP Heck and others 1992 Crop residue on soil surface 52% Barnwell and others 1991 Rates of erosion (USLE, WEPP) EMAP, NRI Foster and Lane 1987; Elliot, Foster, and Elliot 1991 Soil loss caused by wind erosion (WEE) NRI Goebel and Schmude 1981 Cesium-137 (retrospective; loss or Canada Kiss, Dejong, and Rostad 1986; Ritchie and accumulation) McHenry 1990 Depth of topsoil None None Presence, quality, and diversity of filter strips None None Outputs measured from drainage basin None None Soil salinization and alkalization (increase in content of sodium and other salts in soils) Electrical conductivity EMAP Heck and others 1992 pH EMAP, 7% Heck and others 1992; Barnwell and others 1991 Exchangeable bases EMAP, 7% Heck and others 1992; Barnwell and others 1991 Aggregate stability 11% Barnwell and others 1991 Land areas of saline and alkali soils NRI Goebel and Schmude 1981 Chemical use and accumulation (agricultural chemicals such as fertilizers, soil amendments, and pesticides) Use of agricultural chemicals EMAP Heck and others 1992 Soil biological health EMAP Heck and others 1992 Nitrates and toxic organics in drinking water None None Number of applications during EMAP, IFS Heck and others 1992; Wijnands and Vereijken 1992 growing season Degree of resistance to pesticides None None Weed populations and communities None Warcholinska 1978 Diversity of pests and beneficials None None Leaching and movement of pesticides None None Earthworm abundance 4% Edwards and others 1990; Barnwell and others 1991 Air pollution (gaseous, solid, or liquid) Indicator plants for ozone (white clover) EMAP Heck and others 1992; Heagle and others 1991, 1992 (Table continues on next page) 253 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 17-1: (continued) Issue and measurement Current use Key references Solid waste disposal on land (contaminants in the air or soil) Concentration of atmospheric pollutants None None Symptoms of abiotic stress in plants None None Heavy metal determinations (Hg, Cd, Pb) EMAP Heck and others 1992 Human health response (infections, exposures) None None Nutrient loss (fertility, soil chemistry, and soil biology) Eutrophication in nearby surface water None None Symptoms of nutrient deficiency in plants None None Phosphorus and potassium status IFS, EMAP, 11% Brussaard and others 1988; Heck and others 1992 Leaching and movement of nutrients IFS Wijnands and Vereijken 1992 Soil nitrate IFS, 4% Wijnands and Vereijken 1992 Total nitrogen 89% Barnwell and others 1991 Mineralizable nitrogen EMAP, 22% Cabrera and Kissel 1988; Stanford and Smith 1972 Total carbon 96% Barnwell and others 1991 Organic carbon None Bornemisza and others 1979; Storer 1984, 1992 Organic matter IFS, EMAP, 15 % Brussaard and others 1988; Barnwell and others 1991 Microbial biomass 22% Lodge and Ingham 1991; Babiuk and Paul 1970 Microbial activity None Schnurer and Rosswall 1982 Nematode communities EMAP Wasilewska 1979 Water-logging (soil tilth and structure) Water-holding capacity EMAP Heck and others 1992 Concentration of oxygen in soil None None Bulk density 70% Barnwell and other 1991 Soil compaction None None Presence of plow layers None None Permeability None Hillel 1982 Effectiveness of drainage systems None None Earthworm abundance 4% Edwards and others 1990; Barnwell and others 1991 Intensive animal production (livestock for market) Production of the herd (milk, meat, wool) IFS Hermans and Vereijken 1992 Size and composition of the herd IFS Hermans and Vereijken 1992 P-output in manure IFS Hermans and Vereijken 1992 Available grassland for grazing IFS Hermans and Vereijken 1992 Amount of purchased feedstuffs IFS Hermans and Vereijken 1992 Pests and diseases None None Fish farming (aquaculture) Accumulation of fish biomass None None Density of fish parasites and pathogens None None Dissolved oxygen content of water None None Turbidity of water None None Aesthetic value of the rural landscape (spatial arrangement and area of agricultural landscape elements) Fragmentation (size, shape, edge, connectivity) EMAP Heck and others 1992 Presence of hedgerows or windbreaks NRI Goebel and Schmude 1981 Availability and quality of wildlife habitat NRI Goebel and Schmude 1981 254 Biophysical Measurement of the Sustainability of Temperate Agriculture Issue and measurement Current use Key references Biodiversity of vegetation NRI Goebel and Schmude 1981 Diversity and vigor of wildlife populations None None Habitat and vegetative cover (riparian, wetland) NRI Goebel and Schmude 1981 Enrollment in conservation reserve programs EMAP, NRI Heck and others 1992 Land use EMAP, NRI Heck and others 1992 Clinate Air temperature 81 percent Barnwell and others 1991 Radiation 52 percent Barnwell and others 1991 Precipitation 81 percent Barnwell and others 1991 Relative humidity 52 percent Barnwell and others 1991 Wind speed 56 percent Barnwell and others 1991 Pan evaporation 33 percent Barnwell and others 1991 Soil temperature 33 percent Barnwell and others 1991 Soil moisture 26 percent Barnwell and others 1991 a. The agroecosystem component of the EMAP; the percentage of twenty-seven U.S. Corn Belt and Great Plains long-term research sites measuring the specified parameter (Barnwell and others 1991). Note: EMAP, Environmental Monitoring and Assessment Program (United States); NRI, National Resources Inventory, conducted by the U.S. Department of Agriculture's Soil Conservation Service; USLE, Universal Soil Loss Equation; WEPP, Water Erosion Prediction Project; WEE, Wind Erosion Equation; IFS, Integrated Arable Farming Systems (the Netherlands). monitoring programs are designed to monitor sampling should be done when the measurement specific sites in great detail. Although often used of interest is most stable (lowest variance) and for extrapolation to larger areas, research moni- most representative of the condition of interest. toring sitesare notdesigned for that purpose, and Although assessment monitoring might be done extrapolation can lead to a biased interpretation over a span of only one to several years, this of what is happening over a larger area. The would provide information only for determining monitoring done at a research site is usually the status of the system. Assessment monitoring continuous over a given period of time. It can be, should be designed to monitor status and trends and often is, done in conjunction with a modeling over a minimum of ten years. Where there is program so the data can be used to predict what continuing interest in the sustainability of vari- might happen in an area under various condi- ous ecological systems, such monitoring should tions. Some research monitoring is done to help become a permanent part of a country's heritage. validate models. In the United States, the U.S. Environmental Assessment monitoring is designed to deter- Protection Agency (EPA) has initiated the Envi- mine the condition of resources within an area ronmental Monitoringand Assessment Program with either political (county, state, national, and (EMAP), which monitors the status and trends of so forth) or ecological (ecotone, ecosystem, wa- sevenmajorecologicalresourceswithintheUnited tershed, and so forth) boundaries. The evaluation States (Kutz and Linthurst 1990). The is based on the chosen assessment and measure- agroecosystem component of EMAP has used ment end points. Such a monitoring program sustainabilityasafocusindevelopingaprogramto should have a statistical sampling design with monitor the status and trends in agroecosystems complete coverage of the area. Sampling inten- of the United States (Heck and others 1991,1992). sity should be appropriate for the area being This is perhaps the most intensive agroecological monitored and for meeting data quality objec- monitoring initiative ever proposed and includes tives (desired ability to detect differences at a a major research component to determine the given level of statistical confidence). Monitoring feasibility of such a large-scale effort. is conducted from only one to several times dur- Although EMAP is the most developed moni- ing the year, so identifying the sample time or toring program for implementation on an exten- index period formonitoring is critical. Generally, sivegeographicscale, several othernational-scale 255 Defining and Measuring Sustainability: The Biogeophysical Foundations monitoring programs have been proposed (see nental-scale models of the hydrological cycle, table 17-2). CSIRO in Australia is implementing assessment of the sensitivity of plant and animal the Land and Water Care Program to monitor species to changes in climate, and application of changes in the extent and severity of land degra- remote-sensing and geographic information sys- dation and production from various systems of tems to predict national impacts of changes in land use on a national, state, and territory level. climate. The European Community is discussing Canada has proposed an Agroecosystem Health the possibility of initiating a large-scale, ecologi- and Management Program to conduct a national cal monitoring program. survey on levels of organic matter, nutrients, and On an intemational scale, the United Nations erosion of soils and to monitor cropping practices Environment Program (UNEP) has initiated the using remote-sensing technology and statistical World Soils and Terrain Project (SOTER; informationfromtheCanadiancensusofagricul- Baumgardner 1990) and the Global Assessment ture. This program and Environment Canada's of Soil Degradation (GLASOD) Program in coop- Indicators Task Force are planning to develop eration with the International Soil Reference and indicators for application on a broad, national Information Center to map digitally and assess scale (Environment Canada 1991; Piekarz 1990). the degradation of global soils and terrain re- In the United Kingdom, the Terrestrial Initiative sources. The primary tasks are (1) to produce a in Global Environmental Research plans to in- general soil degradation map of the world and (2) clude research on the causes and consequences of to develop a soils and terrain digital data base. climatic change across the United Kingdom. TheSOTERProjecthascompletedpilotstudiesin Planned activities include development of conti- Latin America (Argentina, Brazil, and Uruguay), Table 17-2: National-Scale Monitoring Programs for Temperate Agriculture Contact person Mailing address C. Lee Campbell EMAP-Agroecosystems, U.S. Department of Agriculture, Agricultural Research Technical director Service, 1509 Varsity Drive, Raleigh, NC 27606, United States Anastasios Nychas Commission of the European Community, Rue de la Loi 200, B-1049 General director Bruxelles, Belgium David Smiles Division of Soils, CSIRO Land and Water Care Program, Canberra Laboratories, Chief G.P.O. Box 639, Canberra ACT 2601, Australia Geoff Pickup Division of Wildlife and Ecology, CSIRO Land and Water Care Program, Centre Officer in charge for Arid Zone Research, P.O. Box 2111, Alice Springs NT 0871, Australia John Haberern Soil Report Card, Rodale Research Institute, 222 Main Street, Emmaus, PA 18098, President United States M. Beran TIGER Program, Institute of Hydrology, Wallingford, Oxon OX10 8BB, Program manager United Kingdom David Waltner-Toews Agroecosystem Health and Management Department of Population Medicine University of Guelph, Guclph, Ontario NIG 2W1, Canada David Rapport Agroecosystem Health and Management Institute for Research on Environment and Economy, University of Ottawa, Ottawa, Ontario KIN 6N5, Canada G. Philip Robertson LTER Sites, U.S. National Science Foundation, Kellogg Biological Station Principal investigator Michigan State University, Hickory Corners, Ml 49060, United States Gary Barrett Association of Ecosystem Research Centers (AERC), Ecology Research Center Miami University, Oxford, OH 45056, United States M. F. Baumgardner SOTER Project, Agronomy Department, Purdue University Agricultural Experiment Station, West Lafayette, IN 47907, United States 256 Biophysical Measurement of the Sustainability of Temperate Agriculture North America (Montana in the United States Measurements for monitoring sustainability and southern portions of the Canadian provinces of Alberta and Saskatchewan), and West Africa. Three values are of primary importance to hu- The expected outcome of these programs is an man society in determining the condition of operational world data base that can serve as a agroecosystems. These values are (1) supply of model for the design and construction of within- agricultural commodities: the ability of an country data bases with sufficient detail and ac- agroecosystem to provide adequate yield and curacy for local and provincial use. Other interna- quality of crops and livestock over the long term; tional programs that monitor aspects of agricul- (2) quality of natural resources: the freedom of tural systems include (1) the Food and Agricul- natural resources from harmful levels of sub- ture Organization of the United Nations, which stances such as trace metals, pesticides, fertiliz- compiles much of the global information on crop ers, pathogens, salts, and pollutants in one or production and (2) the Famine Early Warning more component(s) of the agroecosystem; and (3) System of the U.S. Agency for International De- conservation of biological resources: the mainte- velopment, which uses remote sensing to moni- nance of the ecological soundness of crop and tor changes in photosynthetic activity and then noncrop components of the agricultural land- targetsareaswherecropproductionmaybeinad- scape as habitat for plant, animal, and microbe equate. An international data base of monitoring species. These three values encompass human efforts is being compiled by the United Nations and ecological values for agroecosystems and Global Environmental Monitoring System. encompass the concept of agroecosystem In the United States, the U.S. Geological Sur- sustainability (see figure 17-1). vey has established the National Water-Quality Assessment Program to monitor water quality and Assessment end points for use related biological end points within designated in a monitoring program watersheds. The U.S. Department of Agriculture, AgriculturalResearchServiceandtheU.S.Geologi- The selection and evaluation of assessment and cal Survey have established the joint Management measurement end points are critical to the success Systems Evaluation Areas Program for research of a monitoring program. Clearly defined criteria monitoring. The U.S. EPA has initiated the Midwest for the identification, selection, and evaluation of Agrichemical Surface/Subsurface Transport and end pointsencourage objectivity and an unbiased Effects Research Program to monitor the quality of evaluation of all important characteristics prior water, including groundwater. to their acceptance or rejection for long-term use Other monitoring programs within the (Knapp and others 1990). Sets of critical and de- United States include those initiated by the sirablecriteriafortheselectionofassessmentand National Science Foundation in Michigan, the measurement end points for an ecosystem moni- Leopold Center in Iowa (Benbrook 1991b), and toring program are given in table 17-3. In any theRodalelnstituteinthestateofPennsylvania monitoring program, the selection and evalua- (Haberern 1991; Shirley 1991). The National tion of both assessment and measurement end Science Foundation has established some Long- points should be ongoing so that the most appro- Term Ecological Research sites that monitor priate selections can be made to assess the health indicators of agroecosystems at the Kellogg of the agroecosystem. As a monitoring program Biological Station in Michigan (table 17-2). The develops, the emphasis will shift from identifica- Leopold Center has outlined a list of bench- tion and evaluation of end points to their selec- mark indicators for monitoring the quality of tion and implementation. soil, the function of hydrogeological cycles, the biotic community, and economic viability, all Biophysical measurements for use for the state of Iowa (Benbrook 1991b). In 1991, in temperate agriculture the Rodale Institute (1991) sponsored an inter- national conference to identify indicators of Ideally, a montoring program must have a suite soil quality and research needs for the produc- or panel of such end points to address the identi- tion ofa national soil report card (Shirley 1991). fied societal values (see table 17-4; figure 17-1). At the conference, measurable properties were The biophysical measurements presented in table identified for soil fertility, hydrology, toxicity, 17-1 represent an extensive list of measurement and temperature. end points that are candidates for agroecological 257 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 17-3: Critical and Desirable Criteria for Selecting Measurement End Points for an Ecosystem Monitoring Program Criteria Characteristic of the measurenent Critical criteria Responsive Must reflect changes in ecosystem condition and respond to stressors of concern or management strategy Regional applicability Must be applicable on a regional basis and to a broad range of regional ecosystems Unambiguous Must be related unambiguously to an end point or relevant exposure or habitat variable Integrate effects Must integrate ecosystem condition over time and space Correlative Must directly measure or correlate with changes in ecosystem processes, including unmeasured ecosystem processes Important Must reflect conditions that are important to overall ecological structure and function Low measurement error Must exhibit low natural temporal and spatial variability at the sampling site during the index period to be able to detect regional patterns and trends Interpretability Must have a clear interpretation or be related through conceptual models to meaningful changes in the ecosystem Desirable criteria Simple quantification Should be quantified by synoptic or cost-effective automated monitoring Standardized method Should have a generally accepted, standardized measurement method Historical data Should be generated from accessible data source Retrospective Should be related to past conditions by way of retrospective analyses Anticipatory Should provide an early warning of widespread changes in ecosystem condition or processes Cost-effective Should have low cost relative to its information value Table 17-4: Association between Agroecosystem Assessment End Points and Societal Values Supply of Quality of Conservation of Assessment end point agricultural commodities natural resourcesa biological resources Crop productivity X Soil quality (physical and chemical) X X Water quality (ponds and existing wells) X X Land use and cover X X Agrichemical use X X X Soil biological health (nematode indexes) X Landscape structure X X Groundwater and well comparisons X X Biological ozone indicator (clones of white clover) X X X Socioeconomic health X X X Pest density X X Foliar symptoms X X X Beneficial insects X X Genetic diversity X X Habitat quality X X Wildlife populations x Livestock productivity X Nonpoint source loading X X Water quantity (irrigation) X Other biomonitor species X X X a. Air, soil, and water, including transport of contaminants into, within, and out of agroccosystems. 258 Figure 17-1: Societal Values to Be Addressed by a Series of Assessment End Points and Biophysical Measurements Societal Values Supply of Agricultural Commodities L ~~~~~~~~~~~~~~~~2> _~~~~~~~~~~~~~~~~~~~~~~~~ Status & Trends in Agroecosystem Health Conservation Quality of of Biological Natural Resources Resources ;3. I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.J~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r Defining and Measuring Sustainability: The Biogeophysical Foundations monitoring programs. The actual measurement All four types of measurement end points are data can come from surveys; field sampling of useful in describing the condition of a natural soil, water, or air; remote sensing; and existing resource, but the most important type is the re- data sets. In practice, more than one societal value sponse end point, which quantifies what is hap- may be addressed by an individual assessment pening to the valued parts of the system. The end point and its associated measurement end other three end points serve supporting roles. points. Forexample, assessment end points listed The difference between an exposure or habitat in table 174 could be monitored to describe col- measurement end point and a stressor measure- lectively the condition of agroecosystems. Some ment end point is that the former indicates how of these assessment end points are physical or much stress is being experienced by a system chemical measures, whereas others are biologi- (concentration of a heavy metal in soil), whereas cal. Some traditionally have been more directly the latter more directly measures the agent caus- associated with agronomic concerns (crop pro- ing the stress (industrial emissions of a heavy ductivity), whereasothers havebeen morebroadly metal). In this context, stressor includes both associated with ecological applications (landscape positive and negative influences. Precipitation is structure). Each assessment end point, however, a natural stressor, because it influences plant is appropriately associated with agroecological growth, sometimes causing increases and some- concerns, times decreases. Tillage is an example of an an- The diversity in assessment end points reflects thropogenic stressor. The relationships among the multicomponent and multipurpose nature of the four types of measurement end points are agroecosystems. The many-faceted character of diagramed in figure 17-2. Any one assessment end point may involve agroecosystems makes it difficult to find a single one or e typeso ment end point s measurement or index with which to answer the (see table 17-5). For example, pest populations question, are the agroecosystems in this region respond totheconditionofanagroecosystembut healthy or unhealthy? A system might be socio- also represent a stressor on other parts of the economically viable but have low genetic diver- system (namelvcropsand livestock).Conversely, sity, for example. Or producers may substitute it may take several types of measurements to labor and capital inputs for lost productivity. Of quantifyasingleassessmentend point. To usecrop course, the focus should be on long-term productivity as an adequate assessment end point sustainability, which we can only surmise. of ecosystem health, adjustments may be needed Asmentioned above, measurementend points for habitat and stressors that influence crop growth can be classified into four types: response, expo- (type of soil, weather, management inputs). The sure, habitat, or stressor. These can be defined as idea behind such an adjustment would be to ac- follows (Messer 1990): count for known large effects so that the subtler 1. Response measurement end point: a biological or aspects of a system's health are discemable. ecological characteristic measured to provide evidence of the condition of a resource at the Use of selected assessment end points organism, population, community, or ecosys- in monitoring susta inability tem level of organization. 2.E.xposuremeasurementendpoint:anenvironmental Assessment and measurement end points were characteristic measured to provide evidence of discussed above in very general terms. To illus- the occurrence or magnitude of contact with a trate these principles in more detail, two assess- physical, chemical, or biological stressor. ment end points will be discussed: soil quality 3. Habitat measurement end point: a physical, chemi- and crop productivity. These are important com- cal, or biological attribute measured to charac- ponents of agroecosystems and, therefore, ought terize the conditions necessary to support an to be considered when monitoring sustainability. organism, population, community, or ecosys- 5OH. QUAIY (P1IYSICALANDCIIEMICALATTRIBUTES) temn in the absence of stressors. Soils function as sinks and sourcesofbiogeochemi- 4. Stressor measurement end point: a characteristic cal elements, as filters for pollutants, and as an measured to quantify a natural process, an environment for growth and development of environmental hazard,ora managementaction plants and other biological communities. Soils that results in changes in exposure or habitat. are liable to change, gradually or abruptly and 260 Biophysical Measurement of the Sustainability of Temperate Agriculture Figure 17-2: Relationship among Measurement End Points HABITAT STRESSOR RESPONSE EXPOSURE Human activity or What happens to natural process valued parts of the system Degree to which stressors reach the individual parts of the system Note: The boundaries are not always clear Table 17-5: Association between Agroecosystem Assessment End Points and Type of Measurement End Point Assessment end point Response Exposure Habitat Stressor Crop productivity X Soil quality (physical and chemical) X X X X Water quality (ponds and existing wells) X X X Land use and cover X X X Agrichemical use X X X Soil biological health (nematode indexes) X Landscape structure X Groundwater and well comparisons X X X X Biological ozone indicator (clones of white clover) X X Socioeconomic health X X X Pest density X X X Foliar symptoms X X Beneficial insects X X X Genetic diversity X X Habitat quality X X Wildlife populations X X Livestock productivity X Nonpoint source loading X X Water quantity (irrigation) X X Other biomonitor species X X 261 Defining and Measuring Sustainability: The Biogeophysical Foundations partly irreversibly, due to human use. The main combine indicator measurements into quantita- activities affecting soils in agroecosystems in- tive indexes so that general statements about soil clude vehicular traffic, tillage, use of agricultural quality on a regional basis can be made. Several chemicals, waste disposal, and land use. Thelong- possible indexes include structure, til th, fertility, term goal of soil quality monitoring in contamination, and productivity. For example, a agroecosystems is to provide a regional assess- tilth index might combine values of bulk density, ment of the cumulative response of the soil to available water capacity, porosity, organic car- these activities and to conservation efforts. bon,andpercentageclaycontent.Third,informa- The focus of soil quality assessment in tion on soil quality iscombined with other moni- agroecosystems is on the presence, extent, and toring data to produce a picture of overall change in soil properties that (1) are important to agroecosystem health. A fourth long-term objec- the functioning of the soil system, (2) are known tive is to integrate information on the health of to be affected by agricultural land management, agricultural soils with information on soils in and (3) can be adequately measured in one an- forests and arid lands to provide an overall pic- nual sampling period at a regional scale. Al- ture of soil quality across terrestrial ecosystems. though they should be sensitive enough to detect Data for assessing soil quality can come from a changes, these properties must also be stable number of sources. The best option is to collect enough that trends can be detected against their soil from agricultural fields that have been se- background variability. Somephysical and chemi- lected from a statistically valid sampling frame. cal measurement end points associated with soil Care must be taken that the samples actually quality are defined in table 17-6. represent what they are intended to represent. An objective in the assessment of soil quality is Another source of data is information published to determine the range and frequency distribu- by governmental soil conservation agencies, such tion (in proportion of land area) of individual as the U.S. Department of Agriculture's Soil Con- measurements and to begin evaluating how well servation Service, which provides the State Soil the chosen measurements and derived indexes Survey Data Base and National Resources Inven- reflect changing conditions. Because standards of tory. To establish which published data are asso- soil quality vary with climate and soil, determin- ciated with the soil found at each sample site, the ing the rate of change of soil quality is one long- soil series at that location could be determined term objective. A second long-term objective is to from soil survey maps or photographs. Table 17-6: Desaiption of Physical and Chemical Soil Quality Measurement End Points Measure Description Organic carbon Total organic carbon in first 20 centimeters of soil (plow layer) Clay content Percentage of clay content in plow layer Available water capacity Water retention between -33 and -1,500 kiloPascals matric potential Porosity Water retention at -5 and -10 kiloPascals matric potential Soil pH Measure of soil acidity Potassium, magnesium, sodium, phosphorus Exchangeable cations for nutrient availability Base saturation Extent to which the cation exchange capacity is occupied by base cations Exchangeable acidity (humid regions) Extent to which the cation exchange capacity is occupied by hydrogen and aluminum Exchangeable sodium percentage (arid regions) Extent to which the cation exchange capacity is occupied by sodium Electrical conductivity Soil salinity Extractable aluminum (humid regions) Extractable aluminum in the plow layer Mercury Total mercury in the plow layer Bulk density (intact core) Mass of dry soil per unit of volume Hydraulic conductivity (intact core) Rate at which soil transmits water while saturated 262 Biophysical Measurement of the Sustainability of Temperate Agriculture The data obtained can be used to evaluate how maintain production, and the overlap of ecologi- well the measurements and derived indexes truly cal and socioeconomic issues in agroecosystems reflect good, poor, or changing conditions. This is is apparent. Either of the last two aspects of crop one of the greatest challenges in monitoring eco- productivity would be appropriate fora monitor- systems: deciding which values are acceptable ing program. In either case, four unresolved is- and which are unacceptable in the context of soil sues are methods to (a) obtain data, (b) account quality. These criteria are known for some soil for management inputs, (c) combine measures parameters (exchangeable sodium percentage, across different crops (if that is even desirable), conductivity). For ob-ter parameters, such as soil and (d) interpret crop productivity in relation to organic carbon or percentage clay content, it may sustainability. be possible to determine if the direction of change The starting point for all measures is the yield is acceptable or unacceptable. Identified ranges of crops on the sampled field. There are methods for indicators and benchmark references of soil of actually taking samples in the field to deter- quality are generally lacking, and evaluation of mine yield or production of dry matter, but it is soil quality measurements is made even more usually simpler to obtain these data by asking the complex by the fact that what is a good or poor farmer. At the same time, information can be rangeorvalue varies withclimate, soil, and man- gathered about fertilizers, rotations, tillage sys- agement scenario. tems, irrigation, conservation practices, and other management variables that may be affecting CROP PRODUCTIVITY sustainability. Complementary material, such as People concerned about agriculture often focus conversion factors for standardizing inputs, mul- on crop production. This concern is embodied in tipliers for calculating primary productivity, the question Will There BeEnoughFood?-the title weather data, and crop models, are needed of the 1981 Yearbook of Agriculture (U.S. Depart- from other sources. Variability and uncertainty ment of Agriculture 1981). In addition to its cru- of conversion factors are expected to make it cial importance to human society, crop plants difficult to assign statistical confidence to the also provide food for soil microbes, plant-eating final indexes. insects, and other organisms. Crop productiv- Crop yields have, of course, been surveyed ity is thus an important ecological parameter and reported for decades, but yield alone is not a and an important assessment end point that is sufficient indicator of health. If one field pro- affected by many elements in agroecosystems duces a greater yield than another because of (see figure 17-3). additional fertilizer, is that first field healthier? Crop productivity as an assessment end point To answer this, it is necessary to account for the has four facets: (1) total production in a region, (2) effects of management inputsand perhaps for the yield (production per unit of land area), (3) yield influence of weather. One way to do this is to as a biological response indicator adjusted for calculate ratios of output to input in which the inputs, and (4) production efficiency (production numerator is some measure of production, and per unit of input). Quantifying either of the last the denominator represents some input (water, two requires a knowledge of inputs as well as nitrogen, and so forth) or combination of inputs. yield, but the two perspectives are subtly differ- The traditional method has been to use prices as ent. To use yield as a biological response variable, the scaling mechanism; however, this does not one must adjust for those factors that contribute have the desired ecological orientation. to yield but are considered extraneous to ecosys- Various types of energy output to input ratios tem health. These may include some natural in- have also been used in agriculture (Fluck and puts (such as rainfall), some human-produced Baird 1980), but the validity of the energy ratio inputs (such as pesticides), or both. The hope is (energy output per energy input) has been ques- that adjusted yield will reflect the subtle differ- tioned, and energy productivity (kilogram of pro- ences in productivity that may be obscured by duction per unit of input energy) has been sug- largemanagementeffects.Thefourthaspect,pro- gested as a better measure (Fluck 1979). Energy ductionefficiency, would quantifyagroecosystem conversions for the various inputs are often diffi- status by comparing production achieved with cult to obtain. A slightly more sophisticated ap- resources expended, whether or not those re- proach than output to input ratios tis o use moni- sourcescontributedirectly toyield. This, of course, toring data to determine the coefficients relating has implications for the ability of a society to yield to inputs (Lin and others 1991). 263 Figure 17-3: Factors That Influence Crop Productivity Natural Factors Management Inputs |Temperature ||Crop Variety Precipitation|| Fertilizers 0 Soil quality | riationH Soaa ato Tillage Disease / PestsA| Pesticides Catastrophe Crop Rotation _ -7~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c- Crop ProductirviRty Cr outants Commodity Pric | Si rosio, Input Costs i SI Compacti |L Commodity Pgms.| |Salinization ||Conserv'n Resev H rb. CarArrPolutant Commditydrice GlobalChn In C orosion Control Anthropogenic Stressors Socio-Economic Factors Biophysical Measurement of the Sustainability of Temperate Agriculture Another way to account for inputs is to esti- gionaldifferencesarestandardizedout.Intheory, mate what the yield on each field would have this type of calculation can be made for any mea- been if a standard set of inputs had been used. surement, once baseline values have been estab- Such adjustments come from existing research lished for each crop. findingsontheresponseofyield toinputsorfrom Net primary productivity provides another models of crop growth. A similar method is to wayof combiningdata from all crops intoa single build a measurement end point from the differ- value. The net accumulation of plant biomass per ence or ratio between each field's yield and the unit of area per unit of time, it is a potentially yield predicted by a statistical or process model. useful measurement end point because it allows Two critical steps are (1) deciding which inputs, comparisons among different types of ecosys- natural and anthropogenic, should be accounted tems, and it can b1> estimated from data on yield. for and (2) finding the means to make those The yield of each crop is expressed in kilograms adjustments. per hectare and then converted from economic Whether to adjust quantifiers of yield for yield to production of dry matter, using conver- fluctuations in the weather is a major question. sion factors from the literature (Sharp and others Such an adjustment may stabilize the variabil- 1976), along with standard moisture contents. If ity inherent in yield data. If all weather varia- calculated this % ay, net primary productivity tions were accounted for, however, it might be should not be reported for individual crops, be- more difficult to detect changes caused by glo- cause it is simply a multiple of the yield. Also, bal changes in climate (except through shifts in comparison among crop species is not meaning- land use). ful because different c.-ops would be expected to A second major question facing a monitoring differ in net primary productivity. Regional pro- program for crop productivity is whether to re- ductivity is a function of both the productivity of port the productivity of individual crop species individual fields and the patterns of land use or to try to integrate values among crops. The (Sharp and others 1976; Turner 1987). latter approach may be a more appropriate bio- A different, but complementary, approach to physical measure at the ecosystem scale. A com- measuring plant productivity in agroecosystems bined index does not, however, account for the is through the use of remote sensing. Greenness fact that individual crops may respond differ- indexes derived from satellite data are related to ently to environmental changes. the primary productivity of the vegetation Two ways in whicha combined index mightbe (Roughgarden,Running, andMatson 1991). These calculated are normalized yield and net primary allow larger scales to be considered than with productivity. Normalized yield (Y' could be cal- field-based measures. Depending on the resolu- culated for each field by using that field's yield tion of the scanner and the heterogeneity of the per hectare (Y = production from field or area landscape, remotely sensed indexes may reflect harvested), the average (Y,rf) over some meaning- not only crops but also pastures, idle land, wood- ful region (such as a county) for an arbitrary land, roads, water, and so forth. Some informa- reference period, and the standard deviation of tion about land use or land cover may be needed that average yield (s). Similar to a standard nor- in order for greenness indexes to be interpreted. mal variate, the calculation could be Again, a primary challenge is to establish a range Y' = [(Y- Y,) / sl + 5 (17-1) of healthy values for a landscape or region. The number 5 is added arbitrarily so that the distribution of Y' has a mean of 5. Because the INTERPLAYOFSOILQUALITYANDCROPPRODUCllVITY standard deviation of Y' is 1 and its mean is 5, The physical and chemical propertiesof soilsand negative values will be conveniently rare. The the productivity of the crops that grow in them values of Y' can be aggregated across crops. The can be treated as two separate aspects of the advantage of this method of standardization is condition of agroecosystems. In some ways, how- that both means and variances of different crops ever, it may be useful to be explicit about the are put on a similar scale. For simplicity, s can be connection between these two assessment end calculated from temporal (year-to-year) varia- points. One way to evaluate the overall condition tioninthecountymeans.Thedisadvantagecomes of soil is to ask how healthy or unhealthy (or in the interpretation, which must be done care- acceptable or unacceptable) each characteristic fully. Trend detection would be an important (orcollectionofcharacteristics)isforplantgrowth. application for such an index, since initial re- Soil productivity indexes may be useful for this, 265 Defining and Measuring Sustainability: The Biogeophysical Foundations but they may be too crop-specific for implemen- the sample into estimates of the measurements tation in national or international programs of interest over the domains of interest; this (Gersmehl and Brown 1990).Viewed anotherway, procedure is known as domain estimation. the soil serves as a habitat for crop growth, so it * The design must permit the assessment of must be taken into consideration among all the precision of the sample results. other inputs. It can either be included explicitly (as with a crop model) or be intentionally left Several examples of existing area frames that among those variables that crop productivity is have been successfully implemented are those of intended to reflect. the U.S. Department of Agriculture: the National The two assessment end points discussed here Agricultural Statistics Service and the Soil Con- are simply illustrative and serve to represent the servation Service, National Resources Inventory many other assessment end points that could (Cotter and Nealon 1987; Nealon 1984; U.S. De- have been discussed. Each one relates to a dif- partment of Agriculture, Soil Conservation Ser- ferent aspect of the agroecosystem, yet each vice 1989). A new area-sampling frame has been presents unique challenges in the collection of proposedinconjunctionwiththeU.S.Environmen- data, calculation of indexes, assurance of quality, tal Protection Agency's EMAP (Overton, White, and interpretation. and Stevens 1991) and is being evaluated for use in agroecosystems by the EMAP-Agroecosystems Statistical basis for a monitoring program Resource Group (Heck and others 1992). for susta inability Ancillary information needed Biophysical measurements intended to assess the in a monitoring program sustainability of agroecosystems should be made with known statistical confidence and should be Biophysical measurements made in ecosystems derived from a sampling frame that gives com- areofteninsufficientalonetocharacteiizepresent plete area coverage for the target populations of or future status and trends in that system. Infor- interest. The issue of statistical confidence is im- mation concerning processes derived from an portant in determining the quality of the mea- understanding of the ecosystem's structure and surements made, which then directly determines function or from modeling efforts may aid in how much emphasis can be given to specific interpreting the meaning and relevance of spe- measurements in making policy and manage- cificmeasurements.Also,informationronspecific mentdecisions.Completecoverageof targetpopu- stressors may help to determine the forces (natu- lations isessential to avoid possiblebias in sample ral or anthropogenic) that are leading to change in selection (for example, finding a problem in a theecosystem'sstatus.Equallyimportanttomea- place that it is expected to occur), which could suring the current conditions or status of the directly affect the interpretation of overall ecosystem is predicting the future condition or sustainability of the system. trendsof the system. Models or trend analysis can A statistically valid sampling design should be used to predict future conditions; however, a have the followproperties(Heck and others 1991): discussion of the extensive literature on this topic is beyond the scope of this chapter. * The sampling frame must cover the complete universe of interest; in agricultural surveys, an ECOSYSTEM PROCESSES AND NATURAL VARIATION area frame is often used. One of the keys to interpreting the resultsderived * A procedure must be available for dividing the from biophysical measurements made as part of frame into identifiable sampling units such amonitoringprogramistounderstandecological that no part of the frame is omitted or included processes in the overall agroecosystern. Such pro- more than once. cesses may include hydrologic cycles and nutri- * The number of sampling units required to ent cycles, as well as the life cycles of the plants achieve a specific level of precision at mini- and animals that constitute the agroecosystem. mum cost must be known. Understanding the structure and functions of the * Each sampling unit must be selected with a ecosystem,asindicatedbyspecificprocessesand known probability. cycles, should allow us to determine the range of specific measurements that are expected as a re- * A procedure must be available for expanding sult of natural variation. For example, for deter- 266 Biophysical Measurement of the Sustainability of Temperate Agriculture mining the assessment end point of crop produc- though this is not a problem when regional esti- tivity,oneofthechallengesinanationalmonitor- mates are desired). Certain ecologically impor- ing program for agricultural sustainability is to tantparameters,suchassolarradiationorevapo- account for inherent natural variation in crop or transpiration, may not be available or may need pest populations. Measurements of soil param- to be estimated from other data. Air pollution has eters also have a certain degree of natural varia- been measured on networks such as the National tion among regions, which also must be taken AtmosphericPrecipitationAssessmentProgram/ into account prior to determining whether a cer- National Trends Network and the National Dry tain region has better agricultural health or Deposition Network in the United States sustainability than another region. (Bromberg 1990), the Canadian Air and Precipita- tion Monitoring Network (RMCC 1990), and the USE OF STRESS INDICATORS TO INTERPRET IMI/PMK network in Sweden (INFORMS 1990). MONITORING RESULTS Bromberg (1990) notes the uncertain future of the As defined earlier, a stressor is an event, process, U.S. networks. He also identifies the high-prior- or activity that causes a change in exposure or ity research indicators for environmental moni- habitat. Thus, stressor measurement end points toring:ozone,sulfurdioxide,nitricacid,andpre- are intimately linked to exposure and habitat cipitation ions. Among the other research indica- measurement end points, as is evident from table tors are carbon dioxide and UV-B radiation. 17-5. Much of the following discussion can be Stressorinformationhasatleastthreeapplica- applied to all three types of measurement end tions in monitoring agroecosystems. First, re- points. Even in the strict sense, stressors are a sponse and stressormeasurement end pointsmay diverse group: air and water pollution, agricul- give different pictures of the condition of the tural chemical inputs, tillage, temperature, and system (Messer 1990). Response end points may rainfall, for example. Although the word stres- indicate the sum of past events in the system, sors has negative connotations, stresses can have whereas stressor information may suggest future either positive or negative effects on the response responses of the system. Second, stressor (or ex- indicators used to take the pulse of the posure or habitat) data can be used to adjust agroecosystem. certain response measurements. The idea behind Some data on stressors can be taken as part of such an adjustment is toaccount forknown,large an agroecosystem monitoring program. Obvious effects so that the subtler aspects of system health examples are management practices such as till- are discernable. An example given above is the age and pesticide applications, which can be de- use of models to account for the known effects of termined by surveys. Others include density of weather, physical and chemical characteristics of pestsorqualityof irrigation water. Exposureand soil, and management on crop productivity, so habitat measures, such as concentration of con- that the effects of other ecosystem components taminantsinsoil,wouldcertainlybeanappropri- can be discerned. Third, connections should be ate part of the monitoring program. Even some of made between stressors and responses; for ex- the response measurement end points can be ample, stressors may be associated with spatial designed to indicate certain stressors or classes of and temporal changes in response measurements. stress. A very specific example of this is the white Studies to identify such associations can be quali- clover system developed by Heagle and others tative (for example, comparing maps of produc- (1991, 1992) as a biomonitor for ozone. tivity to maps of ozone concentrations, compar- Other key information on stresses to the inggraphsoftrendsincropproductivitytographs agroecosystem will most likely come from exter- of specific climatic factors) or quantitative (corre- nal sources. Thebest examplesare data on weather lation analysis). Geographic information systems conditions and pollutant concentrations. In the mayalsobeuseful.Cooterandothers(1991)give United States, for example, temperature and pre- three examples of the use of these systems for cipitation data are compiled from a network of coordinating weather data and for monitoring cooperative weather stations (Cooter and others the health of forests: proximity analysis (to asso- 1991). Use of such data presents several chal- ciateweatherstationswithmonitoringsites),sur- lenges (Cooter and others 1991; Peer 1990). Qual- face models, and overlay analysis. These same ity assurance can be difficult. Data may not be applications should be appropriate for available at the monitoring sites themselves (al- agroecosystems. 267 Defining and Measuring Sustainability: The Biogeophysical Foundations If certain biological or ecological responses conversion factors, and so forth have been men- indicate a degraded ecosystem, we instinctively tioned previously. want to know why. Unfortunately, a monitoring The best examples in the United States of his- program can take us only half-way to an answer. torical data are (1) the land use and crop and Inordertoattributeaproblemtoacertainstressor livestock production estimates produced by the (or exposure or habitat variable), a reasonable U.S. Department of Agriculture's National Agri- mechanismisneeded toexplaintheeffect(Messer cultural Statistics Service, (2) the census of agri- 1990). Even then, we can only talk about associa- culture, which is conducted every five years, and tions, not causes. Messer (1990) warns about sev- (3) the National Resources Inventory of the U.S. eral pitfalls in attributing ecological responses to Departmentof Agriculture'sSoilConservationSer- stressors: correlation does not demonstrate a vice. The National Resources Inventory provides cause-and-effect relationship, degraded condi- information on land use, soil characteristics, soil tion could be the result of multiple stresses, and erosion,irrigation,tillage,andsoforthandhasbeen degraded condition could be caused by one or conducted every fiveyearssince 1977 (U.S. Depart- more stresses that have not been monitored. In mentofAgriculture,SoilConservationServicel989). the end, associations must be interpreted care- Pesticide and fertilizer data are not as uniformly fully; they can be considered suggestive and may collected, but the U.S. Department of Agriculture, provide hypotheses for future research. National Agricultural Statistics Servicehas recently (1990-91) begun an annual pesticide survey in the USE OF HISTORICAL OR OTHER DATA SETS major states producing various field crops, veg- IN A MONITORING PROGRAM etables, and fruits and nuts. There are a number of reasons for a monitoring Some historical data exist in the form of aerial program to use data that have been collected in photography and satellite imagery, which can be the past (or present) by other agencies and pro- used to evaluate land cover, landscape structure, grams. The most tempting prospect is the hope of and perhaps productivity. For example, the U.S. being able to make statements about trends, even Department of Agriculture, Agricultural Stabili- when the monitoring effort is quite recent. Other zation and Conservation Service has low-alti- reasons include avoiding duplication of effort tude, true-color, 35-millimeter slides for North among government agencies, supplementing Carolina from 1984 to the present. There are limi- sampling being done in the monitoring program, tations to this kind of data, such as the need for establishing expected values against which to photointerpretationof aerial photographs. Addi- checkincomingdata.and providing complemen- tionally, acquisition of satellite imagery, such as tary data (on stressors) needed for interpreting from LANDSAT-V, can be expensive even before the monitored data. We focus here on our experi- interpretational analyses are done. ence in the United States in developing the Second, there are various points of view as to agroecosystem component of EMAP, but the dis- how these external data can and cannotbe used in cussion illustrates two general observations. a monitoring program. Some people would like First, other programs are already collecting to be able to merge data sets from existing pro- information relevant to sustainability, but not in grams,atleastincertainareas,withagroecosystem sufficient amounts or with sufficient breadth for monitoring data. This requires that the data be a complete assessment (table 17-1). In general, a comparable. Whether or not sampling designs great deal of information exists on crop and live- must match is an issue that needs to be resolved. stock production in agroecosystems, but it tends For somesoilsapplications, thedata need tocome to beeconomically oriented. Some data arc avail- from the same site as the monitoring data, or, at able on inputs such as pesticides and fertilizer. least, the soils need to have been analyzed by Little information is available on the noncrop, comparable procedures. Certain kinds of data, nonlivestockcomponentsof agroecosystems,and such as weatherdata, can be used after interpola- little national-scale data exist on contaminants. tion between sample points (Messer 1990). Olson, Breckenridge, and Wiersma (1990)discuss The comparability issue also applies to histori- several of the available data bases that are appli- cal data. Methods of data acquisition and analy- cable to assessment of productivity. Olson and ses must be comparable for connections to be Breckenridge (1990) provide information on con- made between the results of past and present taminant monitoringprograms. Data on weather, monitoring efforts (Messer 1990). This is not a 268 Biophysical Measurement of the Sustainability of Temperate Agriculture trivial matter. Forexample, because of advancing Research is needed to identify, evaluate, inter- technology and a continued refinement of goals, pret, and refine biophysical measurements that even the 1977 National Resources Inventory is can be used for monitoring several important not directly comparable to subsequent years of aspects of agroecosystem sustainability. Repre- the same program (U.S. Department of Agricul- sentative examples include the biological health ture, Soil Conservation Service 1989). of soil, density of pests and beneficial insects, A less demanding way of using existing data is quality of habitat for animals other than livestock, to use summaries about the condition of given population density and diversity of wildlife, and regions,based on data from individual programs, quality of irrigation water. The development of and then to incorporate those summaries into the measurements for these vital aspects of assessment of agroecosystem status and trends. agroecosystems will permita morecompleteand In this case, samples do not have to be taken at the realistic assessment of their sustainability than is same locations used by the monitoring program, currently possible. but certain other restrictions might apply. Again, Asa specificexample, measurementend points sampling and analytical methods have to be com- that can be used to assess the biological health of parable. If themain monitoringeffort isbased on soil and can be implemented in a monitoring a probability sample, difCerent approaches are program are generally lacking (Benbrook 1991 a). needed fordata fromprograms that usea probabil- Most research has focused on determiningmicro- ity sample than for those that do not (Messer 1990). bial biomass; however, no one standard method At least three specific examples of the use of of quantifying microbial biomass yet exists that existing data have already been mentioned: the can satisfy all the criLical and desirable criteria use of weather and other stressor data, the asso- given in table 17-3 (Nannipieri, Grego, and ciation of existing soils data with a sampled field Ceccanti 1990; Smith and Paul 1990). Another (the soil series is first determined from a map or proposed measurement end point is the abun- photo, and then its properties are obtained from dance of earthworms, in particular Lumbricus a data base), and the use of yield data from the terrestris,whichmayhelppredicthydraulicprop- U.S. Department of Agriculture-National Agri- erties of soil and potential for movement and cultural Statistics Service as a baseline for calcu- transport of chemicals (Edwards and others 1990; lating normalized yield. In the last example, the Rodale Institute 1991). However, earthworms do county averages could serve as constants for ad- not reside in all soil series and, therefore, are not justing the mean and variance of incoming yield a good choice for implementation and assess- data. At this stage, then, there are several existing ment on a national scale. data sets of interest, several possible ways of An alternative measurement that may serve using them, and several unresolved challenges to assess the biological health of soils in (primarily statistical) in doing so. agroecosystems is the structure of the nema- tode community (Ingham and others 1985; Niblack 1989; Wasilewska 1979). Nematodes Research needed to develop and implement are ubiquitous, have short generation times biophysical measurements of sustainability allowing them to respond quickly to changes in in agroecosystems food supply, are often the last organisms to die, and yet are responsive to disturbances in soil The development and testing of assessment and (Freckman 1988).Trophic,orfunctional,groups measurement end points require (1) long-term can be easily separated, primarily by anterior studies to establish baseline variability, (2) field structures associated with various modes of perturbation experiments of appropriate spatial feeding (Freckman 1988; Yeates and Coleman scale, intensity, and duration to test the sensitiv- 1982), so identifying the species is not neces- ity and specificity of indicators, and (3) compari- sary. Their abundance and size make sampling sons of systems exposed to stresses of different easier and less costly than for other microflora types and magnitudes (Lubchenco and others and fauna (Freckman 1988). The EMAP's 1991). It is essential to know the baseline variabil- Agroecosystem Resource Group tested nema- ity of the physical environment and the selected tode communities as a biological indicator of biological indicatorsin ordertodeterrninewhether soils in their 1992 pilot project (Heck and others undesirable change has occurred. 1992; Neher and others 1992). 269 Defining and Measuring Sustainability: The Biogeophysical Foundations Conclusions Greenhouse Effect, pp. 179-95. New York: John Wiley and Sons. The classification of an agroecosystem as healthy Benbrook, C. M. 1991a. "Natural Resources As- or unhealthy or as sustainable or unsustainable sessment and Policy." In R. Lal and F. J. Pierce, requires the establishment of specific, judgment eds., SoilManagementfor Sustainability,pp. 145- criteria. To develop these criteria, states of eco- 66. Ankeny, Iowa: Soil and Water Conserva- logical system health and properties or character- tion Society. istics of each state must be compiled systemati- . 1991b. "Protecting Iowa's Common cally and supported by experimental or descrip- Wealth: Challenges for the Leopold Center for tive diagnostic procedures (Schaeffer, Herricks, Sustainable Agriculture." Journal of Soil and and Kerster 1988). The criteria on which judg- Water Conservation 46, pp. 89-95. ments are based must also be established from a viewpointthatisecologically, politically,socially, Bomemisza, E., M. Constenla, A. Alvarado, E. J. and economically acceptable to policymakers, Ortega, and A. J. Vasquez. 1979. "Organic Car- citizens, scientists, and farmers. bon Determination by the Walkley-Black and The reference point for determining whether Dry Combustion Methods in Surface Soils and thestatusofaparticularagroecosystemisaccept- Andept Profiles from Costa Rica." Soil Science able or unacceptable remains a paramount ques- Society of America Journal 43, pp. 78-83. tion in the monitoring of agroecosystems. Be- Bromberg, S. 1990. "Indicator Strategy for cause societal values are made up of the values of Atmospheric Stressors." In C. T. Hunsaker individuals who set varying priorities on main- and D. E. Carpenter, eds., Ecological Indica- tainingthecomponentsofanagroecosystem,there tors for the Environmental Monitoring and may exist no single answer about what is best or Assessment Program. EPA 600/3-90/060. most sustainable for a specific agroecosystem. Research Triangle Park, N.C.: U.S. Environ- However, through the presentation of clearly mental Protection Agency,Officeof Research defined monitoring goals, scientifically sound and Development. September. andcomprehensivemonitoringdataderived from Brussaard, L., J. A. van Veen, M. J. Kooistra, and a valid sampling frame, and clear interpretations G. Lebbink. 1988. "The Dutch Programme on of the agroecosystem's condition in relation to Soil Ecology of Arable Farming Systems. l. potential stressors, policymakers and the public Objectives, Approach, and Some Preliminary can make informed decisions based on the bio- Results." Ecological Bulletins 39, pp. 35-40. physical measurements obtained for key assess- Cabrera, M. L., and D. E. Kissel. 1988. "Potentially ment end points. Mineralizable Nitrogen in Disturbed and Un- disturbed Soil Samples." Soil Science Society of References America journal 52, pp. 1010-15. Cooter, E. J., S. K. LeDuc, L. Truppi, and D. R. Babiuk, L. A., and E. A. Paul. 1970. "The Use of Block. 1991. "The Role of Climate in Forest Fluorescein Isothiocyanate in the Determina- Monitoring and Assessment: A New England tion of the Bacterial Biomass of a Grassland Example." EPA 600/3-91/074. U.S. Environ- Soil." Canadian Journal of Microbiology , 6, pp. mental Protection Agency, Office of Research 57-62. C d o o M and Development, Atmospheric Research and Bamwell, T. O., Jr., E. T. Elliott, E. A. Paul, A. S. Exposure Assessment Laboratory, Research Donigian, and A. Rowell. 1991. "Assessment Triangle Park, N.C. November. of Methods, Models, and Databases for Soil Cotter, J., and J. Nealon. 1987. "Area Frame De- Carbon Sequestration Potential for U.S. sign for Agricultural Surveys." U.S. Depart- Agroecosystems." Internal report. U.S. Envi- mentof Agriculture,National AgriculturalSta- ronmental Protection Agency, Office of Re- tistics Service, Research and Applications Di- search and Development, Environmental Re- vision, Area Frame Section, Washington, D.C. search Laboratory, Athens, Ga. Edwards, W. M., M. J. Shipitalo, L. B. Owens, and Baumgardner, M. F. 1990. "A Global Soils and L.D.Norton.1990."EffectofLumbricusterrestris Terrain Database: A Tool to Quantify Global L. Burrows on Hydrology of Continuous No- Change." In A. F. Brunman, ed., Soil and the till Corn Fields." Geoderma 46, pp. 73-84. 270 Biophysical Measurement of the Sustainability of Temperate Agriculture Elliot, W. J., G. R. Foster, and A. V. Elliot. 1991. Heck, W. W., C. L. Campbell, G. R. Hess, J. R. "Soil Erosion: Processes, Impacts, and Predic- Meyer, T. J. Moser, S. L. Peck, J. 0. Rawlings, tion." In R. Lal and F. J. Pierce, eds., Soil Man- and A. L. Finkner. 1991. "Environmental Moni- agement for Sustainability, pp. 25-34. Ankeny, toring and Assessment Program (EMAP): Iowa: Soil and Water Conservation Society. Agroecosystem Monitoringand ResearchStrat- EnvironmentCanada, IndicatorsTask Force. 1991. egy." EPA/600/4-91. U.S. Environmental Pro- "A Report on Canada's Progress towards a tection Agency, Washington, D.C. National Setof Environmental Indicators." SOE Hermans, C., and P. Vereijken. 1992. "Integration Report 91-1. Ottawa, Canada. of Animal Husbandry and Nature Conserva- Fluck, R. C. 1979. "Energy Productivity: A Mea- tion on Grassland Farms." Netherlands journal sure of Energy Utilization in Agriculture Sys- of Agricultural Science 40:3, pp. 301-07. tems." Agricultural Systerns 4, pp. 29-37. Hillel, D. 1982. Introduction to Soil Physics. New Fluck, R. C., and C. D. Baird. 1980. Agricultural York: Academic Press. Energetics. Westport, Conn.: AVI Publishing Hunsaker, C. T., and D. E. Carpenter, eds. 1990. Company. Environmental Monitoring and Assessment Pro- Foster, G. R., and L. J. Lane. 1987. "User Require- gram: Ecologic4l Indicators. EPA/6-/3-90/060. ments USDA-WaterErosion Prediction Project Washington, D.C.: U.S. Environmental Protec- (WEPP)." Report 1. U.S. Department of Agri- tion Agency, Office of Research and Develop- culture, Agricultural Research Service, Na- ment. tional Soil Erosion Lab, West Lafayette, Ind. INFORMS (Swedish Environmental Protection Freckman, D. W. 1988. "Bacterivorous Nema- Agency). 1990. Monitor 1990: Environmental todesand Organic-matterDecomposition."Ag- Monitoring in Sweden. Sweden: Ingvar riculture, Ecosystems, and Environment 24, pp. Bingman. Available fromNaturvArdsverket, 195-217. Information Department S-171 85 Solna, Gersmehl, P. J., and D. A. Brown. 1990. "Geo- Sweden. graphic Differences in the Validity of a Linear Ingham, R. E., J. A. Trofymow, E. R. Ingham, and Scale of Innate Soil Productivity." Journal of D.C. Coleman. 1985. "Interactions of Bacteria, Soil and Water Conservation 45, pp. 379-82. Fungi, and Their Nematode Grazers: Effects Goebel,J.J.,and K.O.Schmude.1981."Planningthe on Nutrient Cycling and Plant Growth." Eco- SCS National Resources Inventory," pp. 148-53. logical Monographs 55, pp. 119-40. General Technical Report WO-28. Arid Land Kiss, J. J., E. Dejong, and H. P. W. Rostad. 1986. Resource Inventories Workshop, U.S. Department "An Assessment of Soil Erosion in West-cen- of Agriculture, Forest Service, Washington, D.C. tral Saskatchewan Using Cesium-137." Cana- Haberern, J. 1991. "A Soil Health Index." Journal dian Journal of Soil Science 66, pp. 591-600. of Soil and Water Conservation 47, p. 6. Knapp, C. M., D. R. Marmorek, J. P. Baker, K. W. Heagle, A. S., M. R. McLaughlin, J. E. Miller, and Thornton, J. M. Klopatek, and D. P. Charles. Le,, ', ' A T 1990. "The Indicator Development Strategy Rl. Joynes. 19 "esnse of Two Wht for the Environmental Monitoring and As- Clover Clones to Peanut Stunt Vrus and sessment Program." Draft report. Environ- Ozone." Phyto pathology 82, J. 254-58. mental Research Laboratory, Corvallis, Oreg. Heagle, A. S., M. R.McLaughlin,J.E. Miller,R. E. Kutz, F. W., and R. A. Linthurst. 1990. "A Sys- Joyner, and S. E. Spruill . 1991. "Adapta tion of tems-level Approach to Environmental Assess- a White Clover Population to Chronic Ozone ment." Toxicology and Environmental Chernistry Stress." New Phytologist 119, pp. 61-68. 28, pp. 105-14. Heck, W. W., C. L. Campbell, A. L. Finkner, C. R. Lin, B-H, L. Hansen, S. Daberkow,and M. Dreitzer. Hayes, G. R. Hess, J. R. Meyer, M. J. Munster, 1991. "Substitutability of Crop Rotations for D. A. Neher, S. L. Peck, J. 0. Rawlings, C. N. Agrichemicals: Preliminary Results." In Agri- Smith, and M. B. Tooley. 1992. "Environmental cultural Resources: Inputs Situation and Outlook MonitoringandAssessmentProgram(EMAP): Report, pp. 24-29. AR-24. Washington, D.C.: Agroecosystem 1992 Pilot Project Plan." EPA/ U.S. Department of Agriculture, Economic Re- 620/R-93/010. U. S. Environmental Protection search Service, Resources and Technology Di- Agency, Washington, D.C. vision. October. 271 Defining and Measuring Sustainability: The Biogeophysical Foundations Lodge, D. J., and E. R. Ingham. 1991. "A Compari- Assessment, Idaho National Engineering Labo- son of Agar FilmTechniques for Estimating Fun- ratory, Idaho Falls, Idaho. gal Biovolumes in Utter and Soil." Agriculture, Olson,G.L.,R.P.Breckenridge,andG.B.Wiersma. Ecosystems, and Environment 34, pp. 131-44. 1990. "Assessment of Federal Databases to Lubchenco, J., A. M. Olson, L. B. Brubaker, S. R. Evaluate Agroecosystem Productivity." EGG- Carpenter, M. M. Holland, S. P. Hubbell, S. A. CEMA-8924. Informal report. Center for Envi- Levin, J. A. MacMahon, P. A. Matson, J. M. ronmental Monitoring and Assessment, Idaho Melillo, H. A. Mooney, C. H. Peterson, H. National Engineering Laboratory, Idaho Falls, Ronald Pulliam, L. A. Real, P. J. Regal, and P. G. Idaho. February. Risser. 1991. "The Sustainable Biosphere Ini- Overton, S. O., D. White, and D. L. Stevens, Jr. tiative: An Ecological Research Agenda." Ecol- 1991. "Design Report for the Environmental ogy 72, pp. 371412. Monitoring and Assessment Program." Draft. Marten, G. G. 1988. "Productivity, Stability, U.S. EPA/600. U.S. Environmental Protection Sustainability, Equitability, and Autonomy as Agency, Washington, D.C. Properties for Agroecosystem Assessment." Peer, R. L. 1990. "An Overview of Climate Infor- Agricultural Systems 26, pp. 291-316. mation Needs for Ecological Effects Models." Messer, J. J. 1990. "EMAP Indicator Concepts." In U.S. Environmental Protection Agency, Atmo- C. T. Hunsaker and D. E. Carpenter, eds., Eco- spheric Research and Exposure Assessment logical Indicators for the Environmental Monitor- Laboratory, Research Triangle Park, N.C. June. ing and Assessment Program. EPA 600/3-90/ Piekarz, D. 1990. "Rapporteur's Report of Work 060. Research Triangle Park, N.C.: U.S. Envi- Group: Indicators and Assessment of Agricul- ronmental Protection Agency, Office of Re- tural Sustainability." Environmental Monitor- search and Development. September. ing and Assessment 15, pp. 307-08. Nannipieri, P., S. Grego, and B. Ceccanti. 1990. Ritchie, J. C., and J. R. McHenry. 1990. "Applica- "Ecological Significance of the Biological Ac- tion of Radioactive Fallout Cesium-137 for tivity in Soil." In Jean-Mar Bollag and G. Measuring Soil Erosion and Sediment Accu- Stotzky, eds., Soil Biochemistry, vol. 6, pp. 293- mulation Rates and Patterns: A Review." Jour- 355. New York: Marcel Dekker, Inc. nal of Environmental Quality 19, pp. 215-33. Nealon, J. P. 1984. "Review of Multiple and Area RMCC (Federal/Provincial Research and Moni- Frame Estimators." SF and SRB Report 80. U.S. toringCoordinatingCommittee). 1990. The 1990 Department of Agriculture, National Agricul- Canadian Long-Range Transport of Air Pollutants tural Statistics Service, Washington, D.C. and Acid Deposition Assessment Report. Part 3: Neher, D. 1992. "Ecological Sustainability in Ag- Atmospheric Sciences. Montreal, Canada. ricultural Systems: Definition and Measure- Rodale Institute. 1991. "International Conference ment."JournalofSustainableAgriculture2:3,pp. on the Assessment and Monitoring of Soil 51-61. Quality, Emmaus, PA, July 11-13, 1991." Neher, D., J. R. Meyer, C. L. Campbell, and W. W. Emmaus, Penn. Heck. 1992. "Monitoring Sustainability in Ag- Roughgarden, J., S. W. Running,and P. A. Matson. ricultural Systems." Presented at the Organi- 1991. "WhatDoesRemoteSensingDoforEcol- zation for Economic Cooperation and Devel- ogy?" Ecology 72, pp. 1918-22. opment workshop Sustainable Agriculture: Schaeffer, D. J., E. E. Herricks, and H. W. Kerster. Technology and Practices, Paris, France, Feb- 1988. "Ecosystem Health. I:MeasuringEcosys- ruary 11-13. tem Health." Environmental Management 12, Niblack, T. L. 1989. "Applications of Nematode pp. 445-55. Community Structure Research to Agricultural Production and Habitat Disturbance." Journal Schaller, N. 1990. "Mainstreaming Low-input of Nematology 21, pp. 437-43. Agriculture." Journal of Soil and Water Conser- Olson, G. L., and R. P. Breckenridge. 1990. "Fed- vation 45, pp. 9-12. eral Contaminant Monitoring Programs and Schnurer, J., and T. Rosswall. 1982. "Fluorescein Databases: A Fish and Wildlife Perspective. Diacetate Hydrolysis as a Measure of Total No. 1990." Informal Report EGG-EST-9341. Microbial Activity in Soil and Litter." Applied Center for Environmental Monitoring and Environmental Microbiology 43, pp. 125661. 272 Biophysical Measurement of the Sustainability of Temperate Agriculture Sharp, D. D., H. Lieth, G. R. Noggle, and H. D. Landscape: 1935-1982." Environmental Man- Gross. 1976. "Agricultural and Forest Primary agement 11, pp. 237-47. Productivity in North Carolina 1972-1973." U. S. Department of Agriculture. 1981. Will There Technical Bulletin 241. North Carolina Agri- Be Enough Food? The 1981 Yearbook of Agricul- cultural Experiment Station, Raleigh, N.C. ture. 0-354-445. Washington, D.C.: U.S. Gov- Shirley, C. 1991. "Experts to Issue 'Soil Report ernment Printing Office. Card."' The New Farn 13, pp. 5-6. U. S. Department of Agriculture, Soil Conserva- Smith, J. L., and E. A. Paul. 1990. "The Signifi- tion Service. 1989. "Summary Report: 1987 cance of Soil Microbial Biomass Estimations." National Resources Inventory." Statistical Bul- In Jean-Marc Bollag and G. Stotzky, eds., Soil letin 790. Iowa State University Statistical Labo- Biochemistry, vol. 6, pp. 357-96. New York: ratory, Ames, Iowa. Marcel Dekker. Warcholinska, A. U. 1978. "Studies on the Use of Stanford, G., and S. J. Smith. 1972. "Nitrogen Weeds as Bioindicators of Habitat Conditions Mineralization Potentials of Soils." Soil Science of Agroecosystems." Ekol. Pol. 26, pp. 391-408. Society of America Proceedings 36, pp. 465-72. Wasilewska, L. 1979. "The Structure and Func- Storer, D. A. 1984. "A Simple High Sample Vol- tionof Soil Ner.atodeCommunitiesinNatural ume Ashing Procedure for Determination of Ecosystems and Agrocenoses." Polish Ecologi- Soil Organic Matter." Communications in Soil cal Studies 5, pp. 97-145. Science Plant Analysis 15, pp. 759-72. Wijnands, F. G., and P. Vereijken. 1992. "Region- .1992. "An Improved High Sample Vol- wise Development of Prototypes of Integrated ume Ashing Procedure for Determination of Arable Farming and Outdoor Horticulture." Soil Organic Matter." Processed. Netherlands Journal of Agricultural Science 40:3, Suter, G. W. Jl. 1990. "End Points for Regional pp. 225-31. Ecological Risk Assessment." Environmental Yeates, G. W., and D. C. Coleman. 1982. "Nema- Management 14, pp. 9-23. todes in Decomposition." In D. W. Freckman, Turner, M. G. 1987. "Land Use Changes and Net eds., Nematodes in Soil Ecosystemns, pp. 55-80. Primary Production in the Georgia, U.S.A., Austin: University of Texas Press. 273 Defining and Measuring Sustainability: The Biogeophysical Foundations coverage, which is statistically representative. Comments The distinction is good, but one mustquestion the statement that intensive, long-term monitoring D. W. Anderson on a few sites has a very limited role in monitor- ing sustainability. The problems of biased ex- Campbell, Heck, Neher, Munster, and Hoag trapolation to more general or larger-scale sys- (henceforth referred to as the authors) have pre- tems can largely be overcome by working within pared a comprehensive report that discusses in a the structures of hierarchically based land classi- practical way the kinds of measurements that can fications based on soil survey, climate, and other be made to assess the sustainability of information (Anderson 1991). Actually, the Ca- agroecosystems in temperate regions. A combi- nadian effort to monitor soil quality (to be dis- nation of text and tabular material has resulted in cussed later) relies heavily on research monitor- a concise presentation and discussion. There are, ing, with extrapolation to complete area coverage however, several points that I offer for consider- based on agricultural resource area maps as- ation. sembled from soil survey and related sources and Does temperate agriculture have special prop- extrapolation to the future based on simulation erties that distinguish it from the agriculture of models(Acton,MacDonald,andPettapiecel992). tropical regions, or grazing systems, and that The authors state that some research monitoring warrant separate discussions? I think that it does is done to help validate models. Our experience, Temperate agriculature generally occurs in regions particularly in western Canada, is that long-term with more resilient, less weathered soils on more experimental sites such as crop rotation studies recent soil parent materials. Soils with reserves of are essential sources of data to understand sys- nutrients in their parent materials can recover tems, develop conceptual and simulation mod- from major disturbance and be productive again, els, and, with other sites, validate the models. albeit at a lower level of productivity. These soii The authors have listed several national-scale characteristics, particularly where combined with monitoring programs for temperate agriculture, moderate climate, impart a high sustainability appearing to rely mainly on published reports index to agriculture in comparison with many and their own personal contacts. From a Cana- agricultural lands of tropical regions. Temperate dian perspective, the list should include the na- agriculture is mostly, but not completely, the tional Soil Quality Evaluation Program that is led large-scale, intensive, high-yielding, high-input by the federal department, Agriculture Canada, farming methods of the industrial countries, and involves its own scientists in cooperative whereas tropical agriculture (as discussed else- studies with universities and provincial agricul- where) is often on smaller scale, less mechanized tural departments. The monitoring is part of the farms. National Soil Conservation Program and is a The resources within agriculture and gener- long-term project with the objective of monitor- ally strong support for agricultural research and ing soil quality in relation to agricultural similar programs indicate that sustainability is sustainability(Acton, MacDonald,and Pettapiece more probable in temperate systems. The opti- 1992). The project involves assessments of soil misticview isconditioned,however, bytheknowl- organic matter, soil salinity, compaction, wind edge that many of the agricultural practices are and watererosion, andmpollutionby organicchemi- strongly driven by economics and that many cals and heavy metals. Canada is a large country agroecosystems in less favorable situations are with about 65 million hectares of agricultural highly dependent on high inputs of fertilizers land.TheCanadianstudyinvolvesdetailedmoni- and pesticides and often irrigation. Those condi- toring ona limited numberof representativesites, tions lead to enhanced environmental risk, plus the development and validation of process mod- the specter of collapse if external circumstances els, and the systematic extrapolation of findings limit or remove the inputs. within the framework of small-scale soil land- The authors have differentiated between re- scape maps or agricultural resource area maps. search monitoring, which is characterized by in- Agricultural resource area maps are generalized tensive, long-term measurements on a limited from larger scale, more detailed soil maps and number of sites, and assessment monitoring that recognize natural boundaries related to physiog- employs a sampling design with complete area raphy and climate. 274 Biophysical Measurement of the Sustainability of Temperate Agriculture Theauthorshaveprovidedcomprehensivelists The authors refer several times to soil quality of measurement end points for evaluating and mention the function of soils in the context of sustainability from the perspectives of societal soil quality but do not present a coherent defini- values, agroecosystem health, and soil quality. tion of it. Larson and Pierce (1991) define soil What is lacking, quite understandably consider- quality as the capacity of a soil to function within ing the scope of their task, is an appreciation of the its ecosystem boundaries and interact positively relative value and applicability of the various with the environment external to that ecosystem. measures. Soil quality, for example, remains Soil qualitycannotbedefined,expressed,orevalu- poorly defined in a quantifiable sense but is a key ated in terms of a single use (yield of a particular element of agroecosystem sustainability. Soil or- crop, for example) and is a key factor in deter- ganic carbon (or organic matter) is a measure mining sustainability. In the words of Larson reasonably responsive to the management of and Pierce (1991) soil quality describes how agroecosystems and often employed as an index effectively soils: of soil quality. The organic carbon content of the Accept, hold, and release nutrients and other cultivated (Ap) horizon has generally declined . ' since temperate soils were cultivated, often by 50 chemical constituents percent, to reach some new equilibrium consis- * Accept, store, and release water to plants, tent with relative inputs of organic residues and streams, and groundwater decomposition rates. Changes in organic carbon * Promote and sustain root growth with cultivation are at first rapid, a consequence * Maintain suitable roil biotic habitat, and of the considerable stress on the system. With Respond to management and resist time, the rate of change slows as a new equilib- rium is approached. The difficulty remains, how- degradation. ever, of understanding the significance of a par- Soil quality can be evaluated and monitored ticular concentration of organic carbon (is or- by determining several soil attributes. Many of ganic carbon decreasing, relatively constant, or the attributes are highly correlated and related to increasing?) and a baseline from which to make the five functions listed above. Soil quality can be comparisons. Comparisons to virgin soils indi- defined quantitatively as the state of existence of catea drastic reduction with cultivation, but com- soil relative to a standard or qualitatively as degree parisons of various management alternatives in- of excellence (Larson and Pierce 1991). Soil quality dicate that well-managed soils (good crop yields, expressed as the sum of individual soil properties applications of fertilizer or manure, rotation of permits comparisons among soils and can handle crops, and minimal erosion) may have enough changes in quality with reference to time. organic matter to provide nutrient reserves, good Another aspect of measuring soil attributes (or soil tilth, intake and storage of water, and so on, measurement end points) is that soil attributes even though the organic carbon is well below the vary in their rate of change or dynamic properties content of virgin soils. Here it is important not so (Stewart and others 1990). Temporal variability much to know the carbon content in relation to a varies with scale, in that small systems tend to be native control, but to know what has occurred in moredynamicorsusceptibletochangethanlarge cultivated soils over the past years or decades. ones. Soil salinity indicates that well. The salt The concentration of organic carbon in the content of an A horizon varies on time scales of plow layer is but one measure of agroecosystem days to months, dependent on recent weather. sustainability. The mass of organic carbon on a The salt content of pedons (to a depth of, say, 1 soil profile basis is critical to evaluating absolute meter) varies over the course of years or decades, losses within the context of contribution to global whereas regional salinity is a longer-term phe- concentrations of carbon dioxide. Several studies nomenon related to hydrogeology (Anderson have indicated that measures of organic carbon 1991). In many cases, rather than monitoring a that evaluate the more readily available energy highly dynamic attribute such as soil salinity, it and nutrient components (mineralizable carbon may be better to monitor the piezometric level of or nitrogen), biomass, or soluble carbon may be the groundwater system that produces the saline much more sensitive indicators of the health of soils in a region. soil ecosystems than total organic carbon (Ander- The authors discuss the statistical basis for a son 1991). The attribute to be measured depends monitoring program, particularly the need for on the objectives of the assessment end points. statisticalconfidenceandcompleteareacoverage 275 Defining and Measuring Sustainability: The Biogeophysical Foundations for the target population. Statistically representa- References tive and valid random sampling is a daunting task, particularly when areas of interest are large Acton, D. F., K. B. MacDonald, and W. Pettapiece. and soil or ecosystem attributes are many. The 1992. "A Program to Assess and Monitor Soil authors are correct in recommending the use of Quality at Regional and National Scales: A existing data sources, particularly census data, CanadianExperience."Proceedingsofthesev- annual reports of yield, and so on. The Canadian enth International Soil Conservation confer- effort in monitoring has related data from the agri- ence, Sydney, Australia. cultural census, done each decade, to agricultural Anderson, D. W. 1991. "Long-term Ecological resource areas rather than administrative divisions. R Data based on natural rather than imposed bound- Risser, ed "Long-Term Ecological Re- aries are more relevant and easier to interpret. Rsr Anotherchapterin thisvolumediscussesevalu- search." Scope 47, pp. 115-34. New York: John ating changes in sustainability at the landscape Wiley and Sons. scale (chapter 9). 1 consider that it will be difficult Larson, W. E., and F. J. Pierce. 1991. "Conserva- and expensive to obtain statistically valid, unbi- tion and Enhancement of Soil Quality." In ased samplings. Goals of monitoring can best be Evaluation for Sustainable Land Managernent in achieved by sampling key ecosystems and at- the Developing World. Vol. 2: Technical Papers, tributes, as defined within a system such as that pp. 175-204. IBSRAM Proceedings12.Bangkok, provided by the hierarchy (from specific to gen- Thailand: International Board forSoil Research eral) soil horizon, pedon-soil landscape (patch) and Management. soil region, or agricultural resource area, as de- fine by sol ph.orahc an'lmt as Odum, Eugene P. 1989. "Input Management of fiined by soil, physiographic, and clinmate maps. Production Systems." Science 243, pp. 177-82. At middle to higlher levels, the number, spatial arrangement, and health of the various patches Stewart,J.W. B.,D.W.Anderson,E.T.Elliott,and (ardassociatedbioticcoTnmunities)becomecriti- C. V. Cole. 1990. "The Use of Models of Soil cal factors in sustainability. Odum (1989) recom- PedogenicProcessesinUnderstandingChang- mends a top-down hierarchical approach. ing Land Use and Climatic Conditions." In H. Finally, the authors are to be commended for W.Scharpenseel,M.Schomaker,and A. Ayoub, their comprehensive treatment of the topic and eds., Soils on a Warmer Earth, pp. 121-31. Pro- their multi-faceted view of the sustainability of ceedings of an international workshop on ef- temperate agroecosystems. I can recommend only fects of expected climate change on soil pro- a more careful discussion of the structure, spatial cesses in the tropics and subtropics, Nairobi, distribution, and regular temporal changes in the February 12-14. Amsterdam: Elsevier. agroecosystems of interest 276 Measuring Sustainability in Tropical Rangelands: A Case Study from Northern Kenya Walter J. Lusigi Sustainability is not a new concept. It originated the word sustainable in several combinations, in man's quest to perpetuate life: each individual such as sustainable development, sustainable wants to survive as well as he can with hisdescen- economy, sustainable society, and sustainable dants. Eternity is, furthermore, an accepted reli- use. It is important for our understanding of gious concept separated from the concept of sustainability to know what these terms pres- sustainability only by the means with which the ently mean. According to the authors of that permanence is achieved. work, if an activity is sustainable, for all practical Innaturalresourcemanagement,sustainability purposes it can continue forever. When people hasaccompanied theuseofresourcesbydifferent define an activity as sustainable, however, they societies. Hunter-gatherer societies lived in per- do so on the basis of what they know at the time. fect harmony with the land, shifting cultivators There can be no long-term guarantee of abandoned their fields when their fertility de- sustainability, because many factors remain un- clined and moved to other locations in order to known or unpredictable. The moral is to be con- allow the land to recover its viability, and servative in actions that could affect the environ- pastoralists moved to balance pressure on re- ment, to study the effects of such actions care- sources as a biological necessity for survival in fully, and to learn quickly from your mistakes. arid environments. The World Commission on Environment and Modernconcernsaboutsustainabilityinnatu- Development (WCED 1987, p. 8) has defined ral resource management seem to have started in sustainable development as "development that Germany in the eighteenth century, when the meets the needs of the present without compro- principle of sustained yield was applied to for- mising the ability of future generations to meet estry production. Since that time, the principle of their own needs." The term has been criticized as sustained yield has been used in resource man- ambiguous and open to a wide range of interpre- agement under various labels like wise use, sus- tations, many of which are contradictory. Ac- tainableuse,optimumsustainableyield,sustained cording to Caring for the Earth, the confusion has regeneration, regenerative capacity, conservation, arisen because sustainable development, sustain- and so forth. able growth, and sustainable use have been used Recent attention to the issue of sustainability interchangeably, as if their meanings were the hasbeen triggered largelyby unprecedented irre- same. They are not. Sustainablegrowth is a contra- versible resource degradation in many ecological diction in terms: nothing physical can grow in- systems through overexploitation causingdeple- definitely. Sustainable use is applicable only to tion of nutrients and erosion of the top soil or renewableresources:itmeansusingthematrates pollution. Caring for the Earth ( IUCN 1991) uses within their capacity for renewal. Defining and Measuring Sustainability: The Biogeophysical Foundations In this chapter, we adopt the meaning of sus- factors. Composition mainly implies that range tainable development defined in Caring for the ecosystems may be mingled with other kinds of Earth: sustainable development means improv- ecosystems, such as forests or cultivated lands, ing the quality of human life while living within that also require manipulation by managers. the carrying capacity of supporting ecosystems. Rangeecosystemsarenatural pasturesorderived The key word here is ecosystem, which brings in pastures managed extensively on the basis of the concept of life renewal processes maintained ecological principles. The optimum combination by ecological systems. Likewise, a sustainable of goods and services is determined by the capa- economy is the product of sustainable develop- bilities of the ecosystem, levels of technology, ment. It maintains its base of natural resources. It economic demands, and social pressures. The can continueto develop throughadaptations and objective may include any of the values that the through improvements in knowledge, organiza- ecosystem is capable of producing. Management tion, technical efficiency, and wisdom. for optimum yield requires a selection of alterna- Natural resources are ecological systems that tives to maximize values and minimize costs or have structure and also function. Understanding negative values. Sustained yield requires a con- the ecological basis of productivity in nature tinuous flow of energy with orderly cycling of means understanding ecosystems. An ecosystem matter. The restrictions imposed by the word results from the integration of all of the living and sustained determine the maximum rate of usage nonliving factors of the environment for a de- under the constraints of the controlling factors fined segment of space and time. It is a complex of (Van Dyne 1969). Range science is the organized organismsand environment forminga functional body of knowledge on which the practice of range whole. Stable ecological conditions are in a func- management is based. If we are to achieve overall tionalequilibriumthatcanbeperpetuatedindefi- sustainability of the range resource-or sustain- nitely by the system's ability to overcome various able development of rangelands-we should seek disturbances. Ecologica!ly stable-or persistent- to obtain a thorough understanding of their struc- systems are sometimes referred to as climax com- ture and functioning. munities, differentiating them from secondary This chapter concentrates on the attributes of communities that are in a stage of succession sustainabilityin tropical rangelandsexploited by towardclimaxaftervariousdisturbancesorforms pastoralists and also inhabited by wildlife. It spe- of use. cifically looks at a case study of the arid and In order to manage ecological systems semi-aridrangelandsofnorthernKenyaandhow sustainably, resource managers have for a long anattemptwasmadetomeasuresustainabilityof time been preoccupied with trying to measure thatecosystemandtounderstanditsfunctioning. attributes of sustainability for various ecosys- A detailed discussion of the results of that study tems. What would be the maximum the system is outside the scope of this chapter, which exam- could be used without damaging it? What is the ines what factors were taken into consideration, threshold point, if one even exists? Measuring what questions were asked, and how the studies sustainabilityof anecological system means mea- were designed to try to understand the function- suring its resilience. This means that some mca- ing and sustainability of that ecological system. surable attributes of sustainability must exist. Although considerable progress has been The complexity of ecological systems dictates achieved in the study of temperate rangelands, that a highly organized and integrated approach systematic studies of tropical rangeland are still be applied to their study and management. This at a relatively infant stage. requires the use of multidisciplinary teams, be- cause no one person can have all the expertise required. Tropical rangeland ecosystems Range management is the management of a renewable resource composed mainly of one or Tropical rangelandsarepartof the total systemof moreecosystems foroptimum,sustained yield of land used by mankind. They are the areas of the the optimum combination of goods and services. world where wild and domestic animals graze or Management means decisionmaking in the pres- browseon natural vegetation. Rangeland vegeta- ence of uncertainty and involves the manipulation tion includes grasslands, savannas or open scat- of one or more of the dependent or controlling tered-tree forests, shrublands, and small grassy 278 Measuring Sustainability in Tropical Rangelands areas within forests. Range vegetation may never The world's tropical rangelands support vast have been disturbed, or it may follow changes in herds of domestic animals; cattle, sheep, goats, land use, such as clearing brush or harvesting water buffaloes, camels, llamoids, donkeys, and timber. Cultivation eliminates rangeland vegeta- horses. About one-third of the world's people live tion, but abandoned cropland returns to range- on these same rangelands both in cities and as land, especially in areas of shifting cultivation producers on the land (see table 18-1). The tropi- (Heady 1982). Varying demands for different cal rangelands support nearly a billion domestic kinds of products from rangelands cause fre- animals and almost as many people (a billion is quent modification of land use. Therefore, bound- 1,000 million). In some tropical lands, most nota- aries between different areas of land use often bly Africa, great numbers of wild animals share change. the ranges with humans and their herds and Rangeland covers nearly half the earth's land flocks. Australian grasslands support kangaroos surface, 47 percent in all. Nearly half of this total and varying numbers of feral rabbits. area lies in the tropics and subtropics, between Principal products from rangelands are meat, 230northandsouthoftheequator.Forgeographi- milk, fiber, and hides. Other rangeland values, cal convenience and because of irregular political which go far beyond grazing by animals, include boundaries, some areas that lie adjacent to, but water, recreation, fuel, and antiquities. Range- outside, these zones are included (see figure 18-1). land management has two sets of goals. (Child For instance, all of Mexico is included as well as 1984). One is the protection, conservation, im- parts of Australia, South America, India, and provement, and continued use of the resources of Saudi Arabia that are not located strictly within the land, water, plants, and animals. T he other is these boundaries. the increased well-being of the rangeland people Figure 18-1: Zone of Tropical Rangelands ____ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~JJ? I- I KK __ _ __ _ _ __ _ __ _ __ i-jl l ii2f1t.14- T T LJ - I I, I o_ L Z" Note: The area between the two dark lines is the zone of tropical rangelands. Source: Heady 1982. 279 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 18-1: Human Population and Number of Domestic Livestock in the Tropical Rangeland Areas of the World (in thousands) Sheep and Horses and Area Hunmns Cattle goats donkeys Buffaloes Camels Africa 206,795 128,484 65,521 8,175 2,150 9,410 Australia 12,755 27,357 162,937 450 - 2 India 547,950 176,750 4,300 1,930 60,000 1,130 Mexico and Central America 86,280 41,648 6,579 12,842 7 0 South America 149,035 56,114 68,694 22,362 150 0 Southeast Asia and Pacific islands 327,550 20,013 7,480 650 67,135 2,469 Total 1,330,365 450,366 315,513 46,429 129,442 13,011 - Unknown population of wild buffaloes. Source: Heady 1982. and others dependent on rangeland production. meters annually, plant growth is mostly desert These aims may be global or local in scale, differ- scrub, grasses, cacti, and still or spiny shrubs. ent in emphasis, and short or long term. They are An annual rainfall of 250 to 500 millimeters, inextricably mixed in complex systems that in- with wet and dry seasons, produces a savanna clude human welfare, market economy, govern- characterized by widely spaced trees or shrubs ment, and conservation. with grasses covering the soil as an understory. An annual rainfall of 500 to 1,000 millimeters Structure of the tropical rangeland usually produces a dry forest ecosystem, with ecosystems large trees and an abundance of scrubby under- growth. Grassesgrow in the scattered open spaces. Climate, patterns of rainfall, topography, type of In areas where the rainfall is more than 1,500 soil, and the relations among vegetation, animal millimeters annually, in some places up to 4,000 life, and humans constitute the structure of tropi- millimeters, forest usually dominates the land- cal rangeland ecosystems. scape. No open spaces remain for grass to grow. Grasslands prevail where man or fire has re- CLIMATE moved the trees and shrubs. The functioningof any rangeecosystem isdepen- Time and frequency of rainfall affect both sea- dentonclimate.Tropicalbioticcommunitiesmust son and height of vegetational growth. Across toleratehigh light intensities, although thelength much of central Africa and in South and South- of daylight does not vary much throughout the east Asia, the prevailing climatic influence is the year. As a consequence, in the drier tropics, sea- monsoonrainyseason.Rainresultsfromthewinds sonal changes in range vegetation are mainly in Africa, India, Southeast Asia, and northern caused by changes in rainfall because the supply Australia when the low-pressure zone near the of radiant energy is more constant. Temperatures equator is invaded by cool air from the Pacific and below freezingseldom or neveroccur except high Indian oceans. on the mountains, which are inhabited by plants As the monsoonsadvance inland from thecoast, and animals adapted to the cold. There is consid- rainfall declines. Topographical features, such as erable variation in the kinds of biotic community mountains, interrupt the monsoonal clouds and found on mountains since differences in rainfall cause them to drop rain on the windward side. On give rise to environments varying from extremely the Kenya coast of Africa, annual rainfall averages dry deserts to the wettest of rain forests. 1,250 millimeters, while 65 kilometers inland, near Tsavo National Park, it decreases to 500millimeters. RAINFALL This occurs also in northern Australia, where the Low rainfall limits the growth of plants and thus average annual coastal rainfall is 1,500 millimeters; limits the population of animals that feed on the 1,600 kilometersinland to the south, annual rainfall plants. In areas of light rainfall, 125 to 250 milli- decreases to 130 millimeters. 280 Measuring Sustainability in Tropical Rangelands Rain falls in varying amounts from one season vegetation. Soils associated with zones of climate to another. Rangelands are dry compared with and vegetation have distinctive properties re- forested areas and may suffer serious droughts, lated to the local climate and vegetation. Forest with lower than normal rainfall, which retard soils are usually very distinct from savanna soils vegetational growth. If drought continues for a of lower rainfall and sparser vegetation. How- number of seasons, grasses, shrubs, and even ever, in some cases, former forest areas have been small trees may die, leading to starvation of both burnt or otherwise changed so that they are now wild and domestic animals and consequent suf- under savanna. In these cases, the original soil fering of the human populations dependent on properties are modified, resulting in transitional these animals. soils between forest and savanna. On the basis of Rangelands inCentral America, Australia, and present knowledge, it is considered that vegeta- the central Indian plains are subject to periodic tional zones match climatic zones more closely and cyclic droughts, which complicate proper than do soil zones. management. The dry rangelands of Africa, south Red soils in the tropics vary widely in fertility, of the Sahara, experience periods of drought last- aridity, and permeability, but they are all charac- ing from three to seven years about three times in terized by the presence of iron oxides. Soil water every century, when the effects of cumulative in the humid tropics contains little organic matter drying on rangelands become severe. In the dry and, in consequence, does not dissolve iron or tropics, the pattern of rainfall varies consider- aluminum hydroxides from the soil. Silica and ably. Where there are two rainy seasons each other minerals are leached out, thus leaving a year, one may fail, while the other brings mois- high concentration of iron and aluminum com- ture; both may fail entirely or rain may fall in pounds. insufficient amounts for proper plant growth. Laterite is an extreme form of this kind of soil, Rain may fall in more than normal amounts for a often characterized by stony concentrations of numberofsuccessiveseasons,increasingtheplant iron ore. These soils contain very little organic cover dramatically. matter and are infertile. Many tropical forests TOPOGRAPHY grow on laterite and support their growth by TOPOGRAPHY and slope exposure are the two critical their own litter. Elevation and slope exposure are the two critical Black soils are moderately fertile soils of volca- elements in topography. The high plateau of East nic origin. They are rich in calcium carbonate and Africa-the highlands of Kenya, Uganda, and other minerals. They occur in semi-arid climates Tanzania-have almost temperate climates be- in an intermediate location between deserts and cause of their high elevation, although they are forests. These dark soils, found for example near or on the equator. Rainfall patterns resemble around Lake Chad and in the highlands of East those of the monsoonal tropics, but lower tem- Africa, support useful rangeland. peratures cause plants to grow at slower rates Vertisols occupy low-lying flat areas. High than in lower elevations, clay content and flatness impede their drainage, In Mexico, Central America, and all tropical and many develop deep cracks during the dry rangelands in the northern hemisphere, south- season. The effective supply of moisture is less facing slopes get more sunlight and are warmer than on many other soils in the same climatic than north-facing slopes. Plants may grow earlier zone, and the natural vegetation thus tends to be on these slopes, but since extra sunlight means more characteristic of drier areas than the amount greater evaporation of moisture, total growth of of rainfall would suggest. plants is less on the south-facing slopes. Steep- ness of the terrain also affects the density and VEGETATION, ANIMAL LIFE, AND HUMANS vigor of plant growth largely due to the capacity Manyclassification schemeshavebeen proposed of the slope to hold moisture and nutrients. The for the world's biotic communities. Perhaps the distance from oceans and other large water sur- mostimportantcharacteristicof thetropicalrange- faces also affects plant growth because large bod- lands is the close relationships among the vegeta- ies of water modify humidity and temperature. tion, the animals, and human activities, which have to some extent maintained them the way TROPICAL SOILS they are. The broad range of animal species from The variety of tropical soils stems from the vari- browsers to grazers and the activities of humans ety of parent material, topography, climate, and through shifting cultivation, burning, and hunt- 281 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 18-2: Structure of a Basic Ecosystem Projects to develop the nalural resources asc _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ -G am e croppin~ - Water Harvestuig Extenral Rangeland Ecosystem A -groforcstry Coritrolling Factors - %rug-- Soil Climate and soil solar inputs chmate ~ ~ Vgebn tersstm Geostructure and irutial organisms Products vield Impacts from populatonons other systems Indigenos people Inputs t Feedbak t LNatonal/regional ser !ce Extemal inputs Extetion Note: Shaded areas constitute areas of potential focus for projects. Source: Child and others 1984. ing have been a major influence on the biotic with scattered trees, and woodlands with nearly structure of the tropical range ecosystem (see complete canopies of deciduous trees. They are figure 18-2 and table 18-2). Although pastoralism vast areas of country characterized by alternating and hunting are still a major use of rangelands in wet and dry seasons and by periodic droughts of the tropics, cultivation is increasing to meet the several years' duration. Frequent fires and re- demand for cereals from the world's growing peated and often heavy browsing from animals human population. Tropical rangelands can be reduce the density of woody plants but by no classified into six broad types: tropical grassland means eliminate them completely. There is a deli- and savanna, tropical forest with seasonal rain- catebalancebetweentheeffectsofanimalsand fire fall, tropical rain forests, desert shrubs and grasses, on the growthof vegetation favoring different types seasonally flooded grasslands, and montane grass- of plants with different intensities of use. lands (Heady and Heady 1984). High grass savanna occurs near dense forest Tropical grassland and savanna occur as broad and is dominated by Pennisetum and Hyparrhenia expanses of grasslands without trees, savannas spp. The area is predominantly used by elephants 282 Measuring Sustainability in Tropical Rangelands Table 18-2: Key Characteristics of Major Tropical Rangeland Systems Length of dry Type of Cattle per Animals per Site Systems season (months) grazing person sq. kilometer potential Agroforestry Savannas with 6-9 Transhumance 6-S15 25, all Low to Semi-arid tall grasses and village based domestic high trees, mostly species legumes Deciduous forests 3-6 Transhumance 1-4 25, all High with Wide range with high grasses and village based domestic seeded of species legumes possibilities Desert shrub and 9-12 Nomadic 0-4 10, mostly Low except Local grasslands camels and with water leguminous goats harvesting shrubs and trees Seasonally flooded and - Seasonal use - - High Slight wetland vegetation possibility Tropical rain forests < 3 Sedentary ± 1 < 5, all High with Local species domestic legumes species Winter rainfall 5-7 Year-long 6-15 10, all High with Coniferous vegetation domestic legumes trees species Montane forests < 5 Year-long - > 25, on High with Coniferous improved legumes trees pastures of temperate species - Not available. Source: Child and others 1984. and buffaloes, which can reach the tall grass. Tall the Amazon basin. Stands of tall deciduous trees grasses I to 3 meters in height and numerous in the seasonal forest zone occupy large areas trees, mainly Acacia spp. and Combretum spp., con- across Zambia and Malawi. Fires are commonly stitute a narrow belt of savanna grassland across used to maintain grassland in these areas. Africa from Senegal to the Sudan and from Kenya Regions in the tropics with large amounts of to Botswana. A third division of grassland sa- rainfall and no lengthy dry periods grow dense vanna is more like desert than the other types of forests and tall trees of many species. Luxurious savanna covering most of West and East Africa. growth of shrubs and herbs may occur beneath Tropical forests with seasonal rainfall are usu- the canopy. Grazing resources for domestic ani- ally partly deciduous and are commonly referred mals depend on destruction of the forest by shift- to as monsoon forests. They occur in Southeast ing cultivation and other practices that permit Asia and from South Pacific to northern Austra- forage plants to grow on the vacated land for a lia. Seasonal forests also occupy large parts of few years. Desert shrubs and grasses occur in Central America, the West Indies, and south of areas too dry for trees and support low-growing 283 Defining and Measuring Sustainability: The Biogeophysical Foundations shrubs with few forage plantsbetween them asin mals, walk on the soil, compacting it and tram- thedesertsofArabia,Pakistan,NorthernMexico, pling plants. Understanding the impact of hu- Somalia, Sahara, and Kalahari. Seasonally flooded mans on rangelands is important in understand- grasslands are scattered throughout the tropical ing how these lands can sustainably be modified rangelands. These may be extensive, like the 11- for their benefit. anos of Colombia and Venezuela and the sudd in the Sudan. Blackcotton soil areas found in Kenya, Measurement and determination of sustain- Tanzania, and Uganda also fall in this category. ability in the rangelands of eastern Africa Montane tropical grasslands, found at high el- The Integrated Projecton Arid Lands(IPAL) study evations in Africa, South America, and on some area (see figures 18-4 and 18-5), which is repre- South Pacific islands, are important for grazing. sentative of rangelands in eastern Africa, covers The human experience on tropical rangelands approximately 22,500 square kilometers. It is suf- hasa longhistory.This history hasproduced a wide ficiently large and covers the major biotic com- varietyofadjustmentsbyrangelandvegetationand munities found in the area, so that all processes soil that are related to local environmental circum- being observed could be investigated. It is stances or human modifications of them and also to bounded, on the west, by Mt. Kulal, a major water different and often changing cultural beliefs and catchment area, and to the east, by Mt. Marsabit. values. In Saudi Arabia and the drier parts of East The study area partly covers the homes of the Africa, Sudan, Ethiopia, the Sahel region of Africa, Rendille, Gabra, Boran, and Samburu, who are partsof Asia,and otherareas, pastoral nomadismis the major nomadic tribes in this area. With vary- oneofthenormalwaysoflife.Apattemroftranshu- ing degrees of facility, the project was able to mance-a combination of seasonal herd migrations study the deterioration of arid land and the com- with subsistencecropping, usuallyof cereals-may bination of causative factors: climatic fluctuation, also be followed ata central or home location where the activities and attributes of the pastoralist soci- the herds return for part of the year. eties and of their cattle, camel, sheep, and goat herds, including such important aspects as the in- Functioning of thje tropical rangeland crease in human population and the practice and ecosystem effects of construction of bomas (night enclosures). Plants are the primary producers of the ecosys- The greater part of the study area consists of a tem, converting through photosynthesis the en- large central plain at less than 700 meters above ergy from the sun into energy usable by humans sea level. The northern part of this plain is sur- and animals. The flow of energy through an eco- rounded by volcanic mountain masses: the Huri system is illustrated by Odum's classic drawing Hills (1,301 meters) to the north; Mt. Marsabit in figure 18-3 (Odum 1959). At each transfer of (1,836 meters) to the east, and Mt. Kulal (2,295 energy, great losses usually occur, and the only meters) to the west. In the south and southwest, way to increase the efficiency of the system is to the area is bounded by the basement complex reduce the energy lost during these transfers. mountains, Mt. Nyiru (2,963 meters), Oldoinyo The range ecosystem also functions through Mara (2,224 meters), the Ndoto Mountains (2,838 the recycling of nutrients. Unlike energy, which meters),and Baio Mountain (1,885 meters). To the can be lost to the atmosphere, nutrients are recy- west of Mt. Kulal lies Lake Turkana. clable, and the most important of these cycles are The main drainage lines originate in the hill the nitrogen and the sulphur cycles, which are masses and are mostly in the form of seasonal usually enhanced by the action of soil microor- rivers, draining into the central plain where their ganisms. Soil nutrients can be lost through leach- waterevaporates or sinks. Water from most of the ing. In order for the system to be sustainable, the land drains into the old saline lake bed of the above processes must be appropriately main- Chalbi desert in the north of the area. There are tained, and this can only be done if we under- four major desert plains, the Chalbi, the Koroli, stand them properly. Measuring sustainability the Hedad, and the Kaisut, each distinct in its means to some extent measuring these processes. degree of soil salinity and vegetation. Primary productivity and cycling of nutrients The soils are derived from the pre-Cambrian in every rangeland ecosystem are affected by basement rocks or from more recent volcanoes, human populations. People build their houses and they are estimated to be roughly equallv and even cities in rangelands. They, like the ani- divided in their area between these two parent 284 Measuring Sustainability in Tropical Rangelands Figure 18-3: Energy-Flow Diagram of a Community Showing Large Respiratory Losses at Each Transfer Import of Community Orgaiuc Matte Source: Ad d fm Om _.Decomposers typs.Th basaltc aas rs utnfrmoca trefth human inluece vaie frm\n SunBBt1 LL \ /Photo- _ Rsiaoy_P _ Hri ai o c ativ sr d th von s 2 o r es 3 ii later i thischapGross ast on organgic \| | ~Production 4 bucoa| /< < ~~~~~~~~Communityr \ ~~~~~~~~~~Respirton e a~ ~ ~~~~t| Note: P, gross primary production; PTh net primary production; Pr, P, P" P secondary producton at the levels indicated. Source: Adapted from Odum 1959. types. The basaltic lavas, resulting from volca- ture of the human influences varies from one nic activi ty, surrou nd the volcanic hills, and the region to another (asdoes the environment), which sedimentary deposits, originating from thebase- makes each local situation unique. Northern ment complex hills, occupy the central plains. Kenya is no exception to this rule. The rains are unreliable and occur in two sea- Three-quarters of Kenya's land surface is and sons, mainly in April/May (long rains) and or semi-arid, and, across this region, several re- November (shortrains).Thclimateisdiscussed cent socioeconomic factors have been identified later in this chapter. as contributing to the processes of range deterio- ration. Through the setting and realigtrnent of Tphe special nature of threats to sustainable political and administrative boundaries, the de- use of range resources in northern Kenya velopment of forest reserves and national parks, and the establishment of commercial ranches, Rangeland degradation hascertain commonchar- and through the influence of missionsand several acteristics wherever it occurs in the world. There other moder institutions, the movement of no- is lossof vegetative cover, followed by soil degra- madic people has been restricted and the area dation throughvarious formsof erosion and com- they formerly occupied reduced. paction. These in turn lead to greatly reduced Traditional antagonisms between tribes had productivity in the eiorattomreconcerned. In the compressed some tribal groups into a fraction of case of grazing systems, there is decreased forage their former ranges until further encroachment for livestock and a consequent reduction in ani- was prevented some forty years ago. The antago- mal products for the subsistence of the humian nisms are still present, and the lack of security population. Although range degradation and its against inter-tribal livestock raiding and ban- consequences can be attributed simply to the ditry further restricts the movement of people combined effects of climatic change and human and the occupation of grazing lands. More than impact, it is important to recognize that the na- 25 percent of the project's studv area is not usecl, 285 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 18-4: Location of the Study Area of the Integrated Project in Arid Land S U DAN i A_ t NL. 0 o 0 00 200 300 * # . , +..* KILOMETRES p. r. Z* @ cx .,d .n Q ETIOI 'ON 4 E T H I . P I A # f * * . . . . . .+o 39~~~~~~~~I ~~~~~~~~~~~~~~ + m _ W N 2 g . . . . . so~. . . t 4z (.G> . . . . . . . . . +Wair r + - + 1 fi \_ 8 |- * * * * ~~~~~~~.* Garisso .. _ +s~~ ~ ~~~ I . . . . . . . T A N ZA N f A ^ ~~+*1*. .R. 4. . . . W I N D I A N 'S.. . . KEY Mombao Arid and semi-arid land High potential land IPAL study area a deSOUza 286 Measuring Sustainability in Tropical Rangelands Figure 18-5: People of the Study Area of the Integrated Project in Arid Land .1k E T H I O P I A KEY Contours at 500 meter interval 1,500 V. International boundary Limit of study area Approximate range of tribes - - - * North Horr \Kolocha K E N Y A \ Gus GABSRA __ TURKANAo _ I< (2 Maikono- Loiengoloni … * 2K_ . \ 0 8MT.KULALL M D _ t \ * G;icbX v Kurkum flI~;f /< t Bo * Kargi AMIBURUI/MXt ° * Bolesa ~ ~ ~ ~ ~ ~ arabt / 0 NrIRU eSouth Mollau eo Srogoi N00TO MOUN7AIN Ngurunit Et-t. 36- G.a2s 287 Defining and Measuring Sustainability: The Biogeophysical Foundations owing to fear of tribal raiding, and the additional livestock are moved to a new area and a new pressure exerted on the more secure areas aggra- boma isconstructed. In the semi-permanent settle- vates their overexploitation. ments, an accumulation of ticks and other para- Another distinctive feature of the problem of sites usually necessitates the periodic burning of use of the Kenyan range is the special nature of the boma and the movement to a site that may be the pastoral economy. Although the people in- aslittleaslOOmetersaway.Thispracticeappears habiting northern Kenya have many similarities to have the greatest impact on the woodlands of tootherpastoralistslivingintheSahelianzone,in northem Kenya and is one of the most serious Kenya,arid zoneagriculturehasneverbeenprac- causes of desert encroachment in the region. In ticed, and thus there is little of the close interac- addition to the changing constraints on pastoral- tion and interdependence between agricultural- ism noted earlier, such as reduced home ranges, ists and pastoralists that are an important feature another important trend of far-reaching conse- of the Sahel. quences for land use is the increasing settlement The problem of accelerated human population of formerly nomadic people. growth in arid areas is perhaps not unique to Centers of human and livestock concentration Kenya, but it is certainly one of the key factors in have arisen and recently have tended to expand thedesertificationprocessesoccurringinthenorth rapidly, around a few springs and wells and of that country. Kenya's human population occu- especially around the boreholes that have been pying arid land has doubled in the last twenty- installed. Although the presence of fresh water is five years and will, if present trends continue, the most important factor in such concentrations, double again in the next ten years. The pressure of several other incentives contribute to their expan- human population is further aggravated by pres- sion. Many of them have become the sites for sure from Kenya's high-potential areas, where shops, schools, medical centers, and famine relief population has exceeded its carrying capacity centers. However, of the greatest significance is and is now moving to the marginal lands. With the fact that these centers offer security from the alarming increase in the human population, inter-tribal livestock raiding. Each of the concen- there has come a corresponding increase in the tration areas becomes a nucleus of denuded land numbers of domestic livestock. that spreads in widening circles as the people are A prominent factor in the deterioration of the obliged to go farther for grazing and for the wood arid lands of northern Kenya through overstock- they need for fuel and for the fences to enclose ing is the drive to satisfy normal ambition for theirlivestockatnight.Thisacceleratestherateof wealth (over and above subsistence needs), com- localized range deterioration. As in many other bined withacertainhelplessnessorfatalismabout arid regions of the world, the northem part of the consequences. In earlier times, when human Kenyasuffersfromperiodicdroughts,whichhave requirements demanded only a moderate and occurred frequently in recent historical times. transitory exploitation of grazing land, minor Formerly, the consequences of these droughts and localized damage could heal. More recently, were not as serious as they are today, since the the pressing, and at times desperate, needs of an arid regions were sparsely populated. With the overgrown and frequently hungry human popu- recent increases in population, droughts have lation have imposed intolerable pressures on the become more serious, claiming the lives of large arid lands, which, with their delicate balances numbers of livestock and, where they shared the and fluctuations, are experiencing, first, the ero- grazing lands, of wildlife as well. After periods of sion of their productive levels and, then, the drought, the systems of production do not re- disintegration of their ecological structures. cover fully. Some families lose the bulk of their As a result of the rising human populations, herds, which may have been small in the first there are increased demands on woody vegeta- place, and many become destitute and reliant on tion. In addition to the usual use of wood for faminerelief.Thesupplyoffaminerelieffoodhas building and for fuel, the pastoralist people in become a permanent feature of the economy in northern Kenya use large quantities in construct- Marsabit District. During serious droughts, such ing bomas to keep their animals together at night as that which occurred in northeastern Africa and andtopreventdepredationsbycarnivores.These the Sahel between 1968 and 1976, pastoralists bomas are built at permanent and temporary become wholly dependent on the few perennial camps. In temporary camps, they may be occu- supplies of water, and the consequent concentra- pied for as short a time as a week before the tionsofpeopleandlivestockdestrovthevegetation. 288 Measuring Sustainability in Tropical Rangelands Although it is a normal traditional practice to tion through misuse. The project was originally disperse from such areas of concentration when set up as a pilot operation to investigate the the rains come, there is a growing tendency for processes and causes of environmental degrada- part of the population, especially women, chil- tion in the arid and semi-arid region inhabited by dren, and the older men, to remain behind, keep- pastoral nomads. ing with them the livestock they need for milk. The IPAL was intended to contribute to the The young men continue to take away into the design of management activities directed toward surrounding country the unproductive animals achieving a sustained balance between produc- (known as the fora herds)-males, castrates, and tion and consumption, taking into account the barren females-following largely traditional requirements of the growing and increasingly nomadic practices. settled population. It was hoped, where possible, The problem of range deterioration in Kenya is to demonstrate practical modifications and alter- therefore serious and complex. It concerns the natives to the traditional livestock--based plight of people who are using the only tradi- economy that could permit rehabilitation of al- tional means they have known to cope with a vast ready degraded lands. Equally important in this problem that has been caused to a great extent by regard was the use of project findings in educa- modem influences. As yet, although these prob- tion and training and the dissemination of infor- lems in northern Kenya are recognized by the mation on rational management. government and are expected to receive consid- Since most of the changes and processes being erable attention in development programs in the investigated needed continuous monitoring, one coming years, administrative and economic con- of the project's objectives was to provide the straints have, until recently, militated against ap- program basis for establishing an institution in propriate corrective measures. Among the most Marsabit District for continued research, moni- important havebeen the lackof funds and commit- toring, and training relevant to the management ment for the development of low-potential areas, of resources in the arid zone. The project devel- the insufficiency of suitable educational facilities, oped and recommended the infrastructural basis the migration of educated people out of the more for the required management. remote and inhospitable regions to the towns, the lack of attraction of such regions for high-caliber Design of the IPAL study civil servants, and above all, the lack of information on human and other resources that is essential for IPAL's research activities in northern Kenya were the rational development of the region. based on the belief or hypothesis that a system of Since the lives of people are at stake, no efforts land use could be adopted in the near future that should be spared to develop the resources of the would reverse the present trends of degradation arid lands of northern Kenya for the well-being of and establish a sustained production within the its people. However, development must take the livestock economy in Marsabit District sufficient form most suited to the sociological, economic, to support the growing population. Such a sys- and ecological circumstances to which the region tem would depend on appropriate management is best adapted. In northern Kenya, owing to the (in the widest sense) of human and natural re- arid and variable climate and the present culture, sources, aided by suitable education at all levels. the greater part of the land must remain range- The IPAL study was therefore designed to land to be used for grazing, and effective use will cover all the major components of the ecological depend on maintaining considerable mobility in system that have a bearing on its productivity for the livestock herds. the welfare of humanity. The complexity of the problems to be tackled necessitated this multidisciplinary approach. Field work included The IPAL study research on the abiotic components of the ecosys- tem-climate, geomorphology, soils, hydrology, The Integrated Project in Arid Lands (IPAL) was and the biotic components of the ecosystem- implemented from 1976 to 1987 in recognition of primary production (vegetation), and secondary the importance of the arid lands of Kenya both production (livestockand wildlife). Humansand because they support the indigenous people and their livestock are one single dominant factor in the economy of the country as a whole and be- the functioning of this ecosystem, and their ac- cause they were gravely threatened by degrada- tivities were investigated under the human ecol- 289 Defining and Measuring Sustainability: The Biogeophysical Foundations ogy and livestock components of the research flects the primary objective of IPAL in seeking program. IPALinvestigationsalsoincluded work practical solutions to a critical human predica- on the contemporary problems of communica- ment. Expanding pastoral populations, respond- tion, education, and training and on the socioeco- ing to the impact of modem influences, occupy nomic and political backgrounds. This case study vulnerable habitats in which the production nec- of the IPAL project looks at how the design of essary for their support is being progressively each project component contributed to the under- reduced by overexploitation. standing of that ecological system and how that The measurement and interpretation of most understanding enhanced our ability to manage ecological parameters are relatively straightfor- that system sustainably under the prevailing cir- ward. Similarly, the basic human biological re- cumstances. In order to understand the function- quirements of food, fuel, and shelter can be as- ing of the system, the correct elements and pro- sessed. Human populations are censused, and cesses must be identified and the appropriate their structures and trends can be described. In research questions formulated. the case of pastoralist societies, their fundamen- tal resource requirements, met mainly by live- Human ecology stock production and woody vegetation, are cal- culable. Beyond these essentially measurable re- Npprotwihstandi terelo eancenof t colog sources lies the relatively unquantifiable com- approach to development planning, a common plex of human attitudes, characteristics, and re- failing, even among ecologically orientated plan- quirements that are not so amenable to inclusion ners, is the scant attention normally given to the in ecosystem models, but that are, nevertheless, so-called human factor. The existence of a human essential considerations in the integrated plan- population presupposes complex ethnic, social, ning of land use programs. The project's addi- and biological influences and interactions. If these tional function of advising the government of are not understood and adequately accommo- Kenya on the rational management of the ard dated in resource management plans, the conse- grazing lands of Marsabit District necessitates quences can be serious, even disastrous. The fail- granands of ai ct necites understanding a wide range of local socioeco- ure of previous attempts to effect appropriate nomic conditions and trends as well as knowl- change in pastoral societies, which has been an edge of the ecological interactions of the important factor in the continuing deterioration pastoralists with their environment. of arid lands in Africa, has been due, at least pastorapits riththIr e r ent. partly, to an ignorance among planners of the Th program of of refore inclde biological and sociological bases of pastoralism. a program of research on human ecology in its Little effort has been made to understand the bodst snse. on tio to augmentinth values, the fears, and the aspirations of the people existing knowledge of the more purelyecological we intend to help. aspects of the human populations, it brought into Although ecologists have long recognized the focus the many socioeconomic factors that have a dominant role of man as a cause of change in the bearing on the process of descrtification in north- biosphere, in practice they have been reluctant to ern Kenya. To do this necessitated adopting a alter the conventional view that man is the last frank and uncompromising approach to several link in the ecological chain and should therefore unpalatable truths and the discussion of matters be considered last. This account of a study of an that are normally avoided because they embar- arid zone ecosystem in northern Kenya reverses rass authority. Nevertheless, unless the realities the customary sequence, considering first the ofthesocioeconomicandthepoliticalbackground human populationsas the main focusof thestudy are squarely confronted, there appears to belittle and working outward to consider the human hope of implementing any land use policies of environment. The ecosystem approach remains significance in the battle against desertification. the guiding principle. The following specific questions were ad- Humans can justifiably be regarded as the dressed in this enquiry: dominant biotic element in the grazing land 1. Whatwerethebasicelementsofthepastoralist ecosystems by virtue of their overwhelming im- tradition that enabled the pastoralists to sur- pact on them, exerted largely through domestic vive in this hostile environment? animals. However, thehuman-orientedapproach 2. What was the structure of the traditional pas- is adopted in this account mainly because it re- toral economy, and how did it work? 290 Measuring Sustainability in Tropical Rangelands 3. How were their basic traditions being influ- important to the people in the area: camels, cattle, enced by modern developments and the mar- sheep, and goats. Livestock are the principal link ket economy? between people and primary production. Their 4. How were the traditional pastoral economies efficiency in that process of energy transfer is an functioning within the modern national mon- important factor in sustainability of the use of etary economy? that system. This efficiency is determined by the 5. Are the traditional pastoral economies sus- number, class, distribution, density, location, dis- tainable under the present circumstances? ease, and offtake of livestock. All these attributes tainale uder he pesentcircmstaces?were the subject of intensive investigation. 6. How could pastoral economies be made to function better? SPECIFIC OBJECTIVES The livestock studies surveyed the study area to Livestock ecology studies determine the number and fluctuation of annual and seasonal livestock and wildlife and to esti- A few observations made in the human ecology mate the rate and distribution of actual stocking studies are worth reiterating here because they in relation to a variety of environmental param- form the background against which the IPAL eters. For the four main species of livestock livestock studies should be viewed. First, how- (namely camels, cattle, sheep, and goats), it deter- ever much we may regret the past and thechanges mined at regular intervals the key food plants in that have been forced on the peoples of northern relation to their availability in the range and the Kenya, life for them will never be quite the same seasonal fluctuations and nutritional value of the again. According to the charter of the Organiza- forage available for livestock. In different areas tion of African Unity, all territorial boundaries and under conditions of varying food availabil- inherited from colonialism must be respected, ity, it monitored milk production and growth of and withinKenya itself theadministrativebound- the four main livestock species, measured the aries have been reaffirmed. When this is coupled quantity and, where appropriate, the quality of with population pressuresfrom theoutside, there food products, and identified inefficiencies and islittlehope that pastoral groupswill everregain recommended solutions. The growth of herds their lost territories, and hence they are forced to was also determined to predict future offtake and accept their present reduced grazing areas. Sec- expected returns. ond,becauseof thelimited land potential coupled The study also monitored theeffects of disease with the firm livestock cultures of the people, on the production of livestock and determined northern Kenya will remain largely rangeland, whether treatment was cost-effective with re- and a degree of mobility will continue to be a spect to increased productivity. This information biological necessity for survival. was used to recommend appropriate veterinary It wasIPAL's intention, therefore,not to set the measures that maybeapplied toeachspecies,sex, clock backwards but to assist in making the lives and age class of livestock. The relations between ofthepeopleofnorthernKenyahappyandmean- livestock and humans were studied to evaluate ingful under the changed circumstances. In their the present use to which livestock were put by terms, this means making their land more pro- theirowners and to determine their effectiveness ductive through their livestock. The major objec- in meeting the needs of the society and to recom- tive of IPAL's livestock studies therefore was to mend appropriate adjustments to fulfill short- develop management strategies for restoring and ages and enable pastoralists to participate in the maintaining ecological stability. This specifically national economy. involved assessing the current importance and Livestock offtake was examined to identify potential economics of livestock production sys- constraints to offtake, in particular marketing, tems for the people; determining livestock popu- and to recommend improvements. Finally, this lation parameters, including their likely trends information was used to develop an integrated and impact on ecological stability; developing management plan for livestock that would lead livestock grazing and production strategies for to (a) recovery of the range and (b) sustained controlled use of the range and long-term, sus- yield of livestock food products to enable the tained production. Rendille people to be self-supporting in food IPAL's livestock studies concenitrated on the and to participate more fully in the national four species culturally and economically most economy. 291 Defining and Measuring Sustainability: The Biogeophysical Foundations METHODS USED the recruitment and mortality of livestock were The principal method used involves systematic recorded with information on matings and abor- reconnaissance flights over a 6 percent sample of tions to assist in interpretation of the gross data. thestudyarea.The method isdescribed inGriffiths While general management of experimental (1975) and Field (1981). The aircraft is used as an flocks and herds followed the traditional pattern observatory from which the number of livestock, of the pastoral herds of the study area, each of the wildlife, and households is recorded and their dis- experimental flocks and herds was divided into tributions related to a variety of environmental two parts. To study disease, one was subjected to parameters, which are monitored simultaneously. a package of inputs, mainly veterinary drugs, The nutrition and intake of livestock were also while the other acted as an untreated control. studied in several types of range using the pro- Differences in productivity between the herds portion of time spent on a given plant species or receiving treatment and those receiving no treat- plant part as an index of the proportion of that mentwerecomparedwiththecostsofthesupple- plant in the diet (Field 1978: Schwartz 1980). This mentary inputs. The chief parameters of veteri- method is adequate for (a) detecting seasonal nary significance that were monitored included changes in food habits and (b) determining key (a) monthly output of fecal worm eggs, compar- food plants. It does not yield proportions of dry- ing untreated animals with those subjected to matter intake. The range section is currently us- regular anthelmintic treatment; (b) volume of ing oesophageally fistulated animals to check the blood-packed cells and the effects of regular method. The consumption of food was deter- anthelmintic,acaricidal, and ad hoc trypanocidal mined for cattle and goats, using the chromic treatment; (c) monthly numbers and species of oxiderationmethod. lntakebycamelswasdeter- ticks feeding on the animals with and without mined by weighing their total fecal output over a weekly spraying with acaricide; (d) bimonthly twenty-four-hour period. Analyses of plants that levels of serum antibodies to trypanosomiasis in camels use for food enabled the calculation of the camels; and (e) causes of clinical disease and digestibility of the diet, the indigestible fraction mortality as they occurred and the effects of of which was represented by the feces. The di- therapy on disease, where appropriate. gestible and indigestible fractions were summed Certain surveys periodically obtained infor- to give total intake of food. Chemical analyses mation on the diseasesin pastoral livestock. Some were carried out on forage plants, which com- data were also acquired on the clinical conditions prise 80 percent of the diet. The analyses include affecting pastoral herds through the operation of dry matter, protein, energy, fiber, minerals, and subsidized veterinary clinical services to the pas- digestibility. The chemical compositions of diets toral communities. were reconstituted from these figures. The vol- The main method used to investigate the inter- ume of water consumed was measured when actions between livestock and humans was the livestock drink at wells. household survey. In addition to asking ques- Experimental animals consisting of camels, tions, measurements were taken on-site of the cattle, sheep, and goats contracted or owned by numbers, weights, milk production, and occa- the project were intensively sampled to deter- sionally blood of livestock. And, finally, the main mine their production. The production of milk method of studying livestock offtake was again was monitored in the morning and evening when the questionnaire survey, with emphasis on trad- half of the udder was milked out and the volume ers as well as households. measured. With camels, the time during which the calf is usually separated from its mother was Vegetation studies recorded. Milk production was then calculated onatwenty-four-hourbasisandestimated forthe The project's first approach to investigating the entire udder. The growth of animals was moni- problems of rangeland use in northern Kenya was tored at regular intervals by taking live weights. to initiate a quantitative ecological study of the These were usually measured on the moming interactions and relationships between the live- before the dehydrated animal was given water. In stock populations and the vegetation. It was envis- the case of camels and cattle, where weighing aged that at least part of the solution to the most equipment is cumbersome, use is made of the obviousproblemn-overgrazing-woulddependon relationship between live-weight and body size, a firm factual basis relating to primary production in particular thoracic girth (Field 1979). Details of and requirements for animal fodder in the region. 292 Measuring Sustainability in Tropical Rangelands The program of research on vegetation main- * What proportion of the biomass and annual tained its main objectives; in the short term, it production of wood is used by the pastoralists sought to identify and describe the processes for building, fencing, and fuel, and what are contributing to land degradation and to deter- their annual requirements? mine the nature, rates, and causes of changes * What is the spatial distribution of use of the taking place in the vegetation. In the long term, vegetation? the main objective was to provide, for the govern- g ment of Kenya, recommendations on the sustain- * What are the tolerance levels of species and able management of the rangeland (within the communities to exploitation? context of a more comprehensive program of * What are the rates of change in plant biomass land and social reform), which would ensure the and productix ity in the different areas in the maximumsustainableproductivityoftheregion, region in response to the impact of humans based on the rational and controlled use of the and animals? vegetation and appropriate rehabilitation mea- sures. The urgency for such recommendations Severalofthesequestions,togetherwiththefinal was evident from the rapid deterioration of the one, which concerns the estimation of carrying ca- vegetation in the inhabited areas with the conse- pacitiesand otheraspectsof management,had tobe quent reduction in livestock productivity, com- answered in collaboration withlivestockecologists. bined with the food requirements of the growing human population. The increase in the number of RANGE MAPPING animals necessary to support the pastoralists, The range mapping and classification done in together with the continuing reduction in no- collaboration with the woodland ecology work, madic movement and the trend toward settlement, which included aerial photo interpretation with had resulted in the gross overstocking of the coun- ground truth checks, identified 144 typesof ranges, try within foraging range of the villages. The pro- which were mapped and described. A range type gram of research on vegetation was designed to is a more or less distinct unit of vegetation, which answer the following fundamental questions: may be delineated on the basis of aspect, compo- * What major plant communities occur in the sition, or density. Based ona minimum mappable region, and what are their botanical and size of 25 square kilometers, which was the aver- rsiogn, age normal daily grazing range of livestock in the study area, and also taking into account similari- * Whataretheirdistributionsand theirrelation- ties of plant composition and infertility of soils ships to altitude, topography, soils, climate, and land forms, the range types were grouped and human and animal influences? into twenty-four range units. The range types * What are the characteristic biomass densities were mapped at 1:100,000 and the range units at of the two main layers (herb and dwarf shrubs 1:250,000. Each range unit was characterized by and trees) in each majorcommunity in relation the three dominant species in the order of their to mean rainfall, drainage, soil conditions, and importance. human and animal influences? * What are the annual levels of primary DESCRIPTION OF TYPE OF RANGE production in each major plant community (in For any recommendation to be made on the use of the herb and tree layers) in relation to recent a specific range type, it was important that all rainfall under different conditions of soil, factors that influence the functioning of that sys- drainage, and use? tem be understood. This was taken into account in the descriptions of each range unit, which The answers to these questions constituted the included the following: baseline information on the vegetation of the region on which future ecological monitoring * A description of all the physical attributes: would depend. A further seriesof questions were general topography, underlyingrock,soil char- posed, the answers to which related more directly acteristics, drainage systems, erosion status, to the management of the vegetation resources: and various climatic factors such as tempera- ture, precipitation and its reliability, and wind * What proportion of the annual primary pro- *Aecitooalteeeainfcos hc duction is available tolivestoc , an wha A descriptionof all thevegetation factors,which duction is available to livestock, and what includes species composition, cover, and bio- proportion IS consumed? mass 293 Defining and Measuring Sustainability: The Biogeophysical Foundations *A description of past and present human influ- plants, bare ground and litter cover, and the state ences through both livestock and settlement of erosion. * A subjective assessment of range condition Rangecondition wasbased on fourcategories: and estimated carrying capacity for that site. excellent,good,fair,andpoor.Withtheexception of one type of mountain range, none of the range Both the range analysis method (U.S. Depart- types had in our judgment a condition better than ment of Agriculture 1970) and the concentric fair, indicating theseriousness of the degradation circle method of vegetation community analysis of the grazing resource. For most of the study (Colorado State University 1970) were used in area, any management of the range resources determnining community structures and compo- would have to be preceded by effective range sition, while standing biomass was determined rehabilitation and improvement programs. by clipped plots along the same paced transects Determination of range condition is perhaps through each type. the weakest part of range science in the tropics. Most methods used were developed for temper- PRIMARY PRODUCTION ate rangelandsgrazed bymonospecies like sheep Knowledge of the primary production of a range or cattle. The requirements of the broad array of site is essential in order to determine carrying animal species and habitats found in tropical range- capacity. Measuring plant production is a com- lands have not yet been adequately investigated. plex process involving the monitoring of plant materials produced by a system that is usually CARRYING CAPACITY protected from most kinds of use, in protected The concept of carrying capacity has arisen as a enclosures over a period of many years, in order result of our recognition that plant communities to reflect seasonal as well as yearly fluctuations. have a limited tolerance to exploitation by ani- In the absenceof such data, and in theurgent need mals. Exceeding this tolerance causes deleterious of advice on which to base decisions on land changes in vegetation, such as a reduction in management in this region, it was decided to use productivity or diversity or both, and may be the aboveground standing crop biomass of the irreversible in practical terms. There are numer- dwarf shrub and herb layer to estimate primary ous difficulties in defining, estimating, and ap- production. The data used were mainly based on plying carrying capacity in actual land use situa- two wet-season surveys. tions. It is generally agreed that carrying capacity is the maximum biomass density of animal life RANGE CONDITION (normally referring to livestock or wildlife) that Range condition is defined as the state of range can be maintained without detriment to the long- health. Like health, condition is relative, and when term productivity of the area. The concept defines a particular piece of range is said to be in good the level of exploitation that will permit the sur- condition or poor condition, the description is vival of the ecosystem. It does not refer to a level always relative to a standard or ideal for that kind of use in which the productivity of the ecosystem of range: its greatest potential lies in soil stability is maximized, such as a properly managed live- and in amount and quality of forage. A stable soil stock ranch. In such circumstances, the maximum is basic to all range types. The primary objective sustainable production of meat and milk will be in range management, then, is to maintain and achieved at a level of biomass density substan- improve those soils that are stable and to stabilize tially lower than the carryingcapacityand would those that are now eroding. be termed the maximal production or optimal A stable soil is a prerequisite to judging as stocking rate. satisfactory any area where a soil mantle has In the definition of carrying capacity given developed. This principle is applicable to ranges above, the term is taken to mean the long-term anywhereand relates toanynormal vegetal cover; sustainable densityof animal life. In arid environ- it doesnotchange with site or with type of vegeta- ments, where climatic variability is high and is tion. If cover is lost, soil soon follows, and once accompanied by proportional variability in plant the soil is gone, the resource is gone. The range production, the long-term carrying capacity is condition for the range sites identified in our determined by the levels of plant production in study area was determined by subjective judg- drier years and not by the average years. Thus, ment based on the following attributes: soil sta- although it is possible to speak of the short-term bility, composition of desirable and undesirable carrying capacity determined by the recent or 294 Measuring Sustainability in Tropical Rangelands current level of plant production, attempting to layer obtained from clipping data. First, a proper maintain stocking rates at this level will result in use factor of 50 percent was deducted, following overstocking when dry years occur. the standard practice in range science. From the While production varies from year to year figure thus obtained for available forage, a fur- with variable rainfall, it also varies spatially ac- ther deduction of 25 percent was made to allow cording to recent distribution of rainfall, local for the errors associated with the use of small drainage and soil conditions, and the history of clipped plots. This gave the quantity of usable land use. Carrying capacity estimates thus have forage as 37.5 percent of the total standing crop to take into account local variation and are gener- biomass of the annual grasses and herbs, which ally applied to relatively large areas in which the was the value used in calculating the carrying variabilityisaccountedforintheaveragingprocess. capacity of these elements in the vegetation. For Despite the possibilities of estimating average car- dwarf shrubs, since the foliage and production of rying capacity across large areas of land, the inter- new growth islow in relation to the standingcrop pretation of such estimates must allow for several biomass, especially in years and seasons of low qualifying factors, which tends to set the actual rainfall, the estimated usable forage biomass was stocking rate attainable at a lower level than the further reduced by half to 18.5 percent of the theoretical carrying capacity. A discussion of these standing crop. h. the special case of the large factors and their implication for determining final dwarf shrub Duosperma eremophilum, which is stockingratesinanyparticularsiteappearsbelow. largely woody, 22 percent of the biomass was Specific recommendations on proper stocking known from previous measurements to be avail- rates for the different range types of a region are able for forage, and therefore only 11 percent was an essential part of any range management plan. used in the calculations of carrying capacity to Therecommendationscanbebasedonlongexpe- allow for low production in dry years. Of the rience in range management under similar cir- standing dead annual grass, only 18.5 percent cumstances, or on published estimates for the was considered to be available. ecoclimatic zone in question, or finally, in the Having estimated the available herb and dwarf absence of adequate prior information on the shrub forage using the above adjustments to the region, on actual measurements of the primary standing crop biomass, the next step was to appor- production of the area in relation to known fod- tion this forage among the different livestock spe- der requirements of the livestock species. cies. In consultation with the human ecologist, it In the IPAL study area the only available data wasagreed thatmultipleuseof therangebythefour to indicate levels of primary production are mea- major livestock species-camels, goats, cattle, and surements from the clipping of the herb and sheep-would have to continue in the immediate dwarf layers of the vegetation, which were ob- future, until greater technical knowledge and a tained in two range surveys in 1982 and 1983, and broaderawarenessenablethe traditional pastoralists the biomass and production values for the tree toacceptanyadvantagestobegainedbyseparating and large shrub layer, obtained from the destruc- the species for the optimal use of the range. Camels tive sampling of the important tree species and overlap goats in their use of the herb layer, while transect observations made in each range type. cattle overlap sheep. This point was resolved by These data were used as the basis for calculating acceptingtheexistingratioof camel togoatbiomass carrying capacity. In order to arrive at realistic and cattle to sheep biomass, which, in both cases, estimates of the forage available for livestock, was approximately 5:1. This ratio was determined several adjustments to the measured values of from the average values of livestock biomass ob- standing crop biomass were made. Since a large served in eleven aerial census surveys carried out proportion of the foliage standing crop biomass over four years. Indigofera spinosa, which is one of in the herb and dwarf shrub layers is annual the most desirable plants for all the livestock spe- production,thedistinctionbetweenstandingcrop cies, was apportioned for use according to the and annual production has not been made here. proportions of biomass of the animal species. Future measurements of these layers will be made The final estimates of carrying capacity based to determine the relative dry weight values of on the herb and dwarf shrub layers were deduced standing crop biomass and annual production in by using the average values of dry matter intake theplantcommunitiespresentindifferentranges. obtained by the livestock ecologist: camels, 4.81 The following steps were taken to determine kilograms a day; cattle, 4.45 kilograms a day; the carrying capacity of the herb and dwarf shrub sheep, 0.76kilogramsa day;goats, 0.84 kilograms 295 Defining and Measuring Sustainability: The Biogeophysical Foundations a day. Carrying capacity was, in all cases, ex- vide firewood, materials for the construction of pressed as the number of animals per square houses and livestock enclosures, and browse for kilometer a year, and an estimated number of the livestock on which the pastoralists are almost animals that could be supported by each range wholly dependent for their subsistence. In addi- unit was also given based on its area. Until more tion to these three essential contributions of the data on primary production become available, pastoral economy, which are the main subjects these figures should be regarded with extreme discussed in this section, trees are valued for their caution. They represent the best estimate based shade(animportantfactorin thesurvivalof cattle on biomass data for the area. and small stock in the arid zone) and for their fruit Theapproach toestimatingcarryingcapacities and other useful products, such as fiber for rope described above is somewhat traditional with and gum for the gum arabic industry. Their roles respect to the tenets of range science and is based in the maintenance of soil fertility and in the largely on the assumption that livestock carrying reductionof soil erosionand desiccation, although capacities are dependent on the productivity of of crucial importance, remain to be investigated the grass, herb, and dwarf shrub layers (the so- systematically and are not discussed here. called range plants). This approach does not take The reduction and eventual loss of the tree into account two circumstances that are particu- layer in areas of exceptionally heavy livestock larly characteristic of the range-livestock ecosys- and direct human exploitation almost invariably tem of the arid zone in northeast Africa. First, take place several years after the destruction of trees and large shrubs constitute a substantial the herb layer through overgrazing and tram- proportion of the range vegetation, and, second, pling. Very large areas of rangeland, which are camels and goats, which are browsers, obtain now in the intermediate stage of degradation, nearly 30 percent of their food from trees and where the soil surface is almost totally denuded large shrubs. Thus, in terms of biomass, 59 per- but where the tree populations are relatively un- cent of the livestock are obtaining 30 percent of harmed, can be seen in the Samburu and Baringo their diet from approximately 8 percent of the districts of Kenya. Indeed, it appears that, in the available forage in the region. It is, therefore, process of degradation of the wooded grasslands evident that considering the region as a whole, of the arid and semi-arid zones of eastern Africa, tree and large shrub forage is a greater limiting a stage is passed through in which the reduction factor in determining the carrying capacity for orremovalofthegrassresultsinvigorousgrowth camels and goats than are the herb and dwarf of small trees, mainly Acacia species. This is due shrub layers. A further essential consideration in to the lack of root completion for the seedling determining thecarryingcapacitiesof woodlands trees and the absence of grass fires, which nor- and shrublands is the impact that browsing ani- mally limit tree growth and effectively maintain mals, particularly goats, have on young trees. In the patterns of scattered trees, or of woodland heavily stocked areas, goats browse young trees and grassland mosaics, characteristic of and large shrubs so intensely that they may pre- undergraded rangelands in East Africa. In the vent their growth and ever.tually kill them. Un- semi-arid grasslands of Kenya, the destruction of der these circumstances, little or no recruitment the herb layer through overgrazing frequently to the larger size classes takes place, and an im- leads to the growth of tree and shrub thickets, portant tree or shrub species may be greatly re- such as the Acacia mellifera thickets of Longido in duced or eliminated in such an area. For this northern Tanzania and Baringo and Samburu in reason, the age class structure of tree and large Kenya. In the arid regions with less than 300 shrub populations has been regarded as a critical millimeters of mean annual rainfall, covering a criterion in determining thestockingratesof goats large proportion of northern and eastern Kenya and, to a lesser extent, of camels. Poor structure, and also substantial areas in Somalia and south- with few young trees present, may necessitate ern Ethiopia, the low and erratic rainfall does not low stocking rates for goats and camels. support a sufficient biomass density of trees to enable thickets to form, despite the absence of Woodland ecology studies grass fires, except along some seasonal water- Trees have three important roles to play in the courses. The most extensive vegetation commu- arid zone in northern Kenya, seen from the stand- nity across more than 200,000 square kilometers point of the pastoralist populations. They pro- in northeastern Africa is the relatively open, dry 296 Measuring Sustainability in Tropical Rangelands shrublands (as defined by Dratt and Gwyne 1977) It is against this background of extremely un- in which the small trees Acacia reficiens and Acacia even use of the land resources that the woodland mellifera are dominant or co-dominant. Within research program was undertaken. Essentially, the IPAL study area, these communities consti- the research sought to discover how woodland tute 26 percent of the area, while the two Acacia exploitation can be managed to ensure the maxi- species, together with several others, occur as the mal sustained production of the tree populations dominant trees in a further 60 percent of the for the benefit of the pastoralist people and the region, which is occupied by annual grassland country as a whole. A second question concerned and dwarf shrubland (Herlocker 1979). the size and disposition of the existing and poten- Over the greater part of Marsabit District, in tial productivity of the woodland and hence its which the IPAL studiestookplace,theherblayer potential contribution to the subsistence of has been affected to varying degrees by overgraz- pastoralist populations and to the wider economy. ing.Inthevicinityofwells,spring,andboreholes, The management of the woodlands clearly had which provide perennial sources of water, the only been undertaken as an integral part of the grasses have been greatly reduced or almost to- coordinated management of the region's land tally removed. In the southwest of the district, and other resources. Detailed information was inhabited by the Rendille people, the denuded needed on the dynamics and productivity of soils extending outward from the settlements for woodlands, on the one hand, and on potential asmuchas30kilometersmaycoalesce,astheydo human use of the woodlands, on the other, to between the villages of Korr and Ilaut. In such ensure rational management of this resource on a areas, the tree populations havealso been greatly sustainablebasiswithin theoverall planning. The reduced but have not been totally destroyed. management of tree populations must be inte- Nevertheless, the process of excessive tree felling grated with the regulation of land use in general, and lopping is continuing, although apparently with water resource management, with settle- at a slower rate than the loss of the herb layer mentpolicy,withtheprovisionofsubstitutebuild- through overgrazing and trampling. ingand fencing materials, and with treeplanting, The direct causes of tree destruction are clear. to help accelerate rehabilitation of the rangeland Pastoralists need wood for fuel and also for the and the provision of fuel and building materials. construction of livestock pens. In the areas of human and livestock concentration, tree felling is Geomorphology and soils studies taking place faster than young trees can be re- Soils with their parent material form the physical cruited to the woodland populations. Although base of any ecological system. In general, soils of many young and repressed trees are normally the IPAL study area reflect (a) climatic zones, (b) available for recruitment to the mature tree popu- toPAphy are geolect (d) geom ones,l(b) lations in such areas, their growth above half a topography, (c) geology,and (d)geomorphologi- meter in height is normally prevented by the cal processes. browsing of goats. The main indirect cause of There is a broad correspondence between the excessive destruction of trees is the increasing major geological and geomorphic units within settlement of formerly nomadic pastoralists, re- the study area. An extensive plain (Heda, Koroli, sulting in the growing impact of tree felling on Kaisut) composed of quaternary sediments de- woodlands close to the settlements. As the sup- rived frompre-uambriangneissoccupiesthecen- plies of wood in and around the villages are ter ofthesstudy area.dTwo major ecological units exhausted, the peopleare obliged to travel farther are present: piedmont plains at higher levels and each year to obtain it. sedimentary plains at lower levels. Soils are gen- In great contrast to the overexploited wood- erally sandy loams but become more finely tex- lands near the settlements, as much as40 percent tured and saline to the northeast, especially in the of the IPAL study area is occupied by woodlands lower drainage lines. An extensive system of old that are rarely used, mainly owing to the lack of stabilized sand dunes fringes the northernmost security from inter-tribal raiding. Such edge of the plains. Thiscentral large plain drains unexploited regions show clearly the potential gently north-northeastward into a large lacus- productivity of the ecoclimatic zones in which trine plain system and eastward down the Milgis they are situated and provide standards against River drainage. The lacustrine plain system com- which the nature and degree of degradation in prises three elevational levels, the lowermost of the inhabited areas can be assessed. whichlieswithinsouthernEthiopia.Thisoverlies 297 Defining and Measuring Sustainability: The Biogeophysical Foundations the old Chalbi Lake bed and is characterized by induced, destabilizing factors like soil erosion poorly drained, excessively saline, clays that ei- contribute to the deterioration of the whole eco- ther crack or are puffy at the surface. The old lake logical system. bed remains uncovered along the perimeter of The different types of vegetation found in the the Chalbi desert and represents the intermediate IPAL study area have significant relationships to levels. These have shallow, well-drained, exces- the present geomorphological processes that are sively calcareous, fine, sandy clay loams over also important in determining ecological stability massive hard pan. Higher levels may have stony in various ways. First, phenological changes in desert pavement. the course of the year follow the patterns of rain- Floodplains and low terrace alluvial plains fall. During the dry season, the grasses dry up occur along the major seasonal drainages such as over the whole study area and are almost com- the Balesa Kulal and Milgis rivers and at the pletelygrazedoffbyanimals.Transectssampling junction of the plains and lava footslope areas. of the whole of the lowland portion of the study Soils vary in texture as well as age. Much of the Milgis flodplain, or instane, is comosed of ust before the beginnng of the rains shows that rlcsilgs floormdplain byr pstariver distriionsy- o bare ground covers more than 45 percent, while relitemsoi Largedmountanst riseronal distridiof the 20 percent is covered by woody or grass herb and tems. Large mountains rise on all sides of the litr whc masttathebgnigote central plain, at 1,500 to 3,000 meters elevation, litter, which means that at the beginning of the and are principally volcanic (Marsabit, Kulal, rainy season at least 65 percent of the ground has Huri Hills), although those in the southwest are no vegetative cover. Results from these transects pre-Cambrian gneisses (Ndotos, Nyiru, Oldoinyo were used to develop a formula for estimating the Mara), from which come the sediments forming risk of erosion in the study area, which was ap- the central plain. plied to different range types. The slope factor The soils of volcanic mountains are related to was not applied to the formula since it was only the major geomorphological units that are used on the lowland plains. When the first rains elevationally defined. Well drained and friable strikeasurfacewithouta closedvegetative over, throughout, they are deep red to dark brown splash erosion occurs, destroying the upper lay- clays at higher elevations, shallow dark reddish ers of the surface. Morphodynamics are deter- brown clay loams with stony and rocky compo- mined mostly by the density of the cover of grass nents at middle levels, and very shallow to mod- and dwarfshruband, toamuchsmallerextent,by erately deep dark brown calcareous stony clay large shrubs and trees. loams with saline-sodic tendencies at lower el- From this observation, it can safely be con- evations. Rocky and bouldery surfaces are also cluded that the factors destroying the herb layer common; similar soilsoccur throughout the stony are farmore significant in causing land degrada- to boulderyolivinebasalt plateau and step-faulted tion thanrin destroying the tree layer in the low- scarps. land portion of the study area. This does not The primary types of vegetation within the apply to the forested areas, the mountains, and IPAL study area reflect a climatic zonation, such the riverine vegelation, where destruction of the as the lack of a shrubland zone between bushland tree layer could lead to serious land degradation. and dwarf shrubland in the southwest, and are often attributable to soils. Soils are also the prin- Climate studies cipal determinants of the smaller, more localized Although drought may be accepted as a natural types of tertiary vegetation within the context of phenomenon in northern Kenya, its frequency a climatic zone. An example would be a topo- and theseriousnessofitseffectsonthelivesofthe graphically related sequence of different soil- pastoralistsweremainconcernsof IPAL'sinves- vegetation units of a hillside catena. tigations. Coetsee (1968) distinguished four types Soil texture plays an important part in the ofdrought:seasonaldroughts,periodicdroughts, distribution of vegetation within the study area. disaster droughts, and droughts caused by man. Soils developed from lavas tend to be loamy to Seasonal droughts are normal, regular phenom- clayey in texture, whereas those developed from ena at certain times of the year. During such gneissic materials tend to be sandier. seasons, rainfall is low, and there is a shortage of Maintenance of an appropriate soil stability is edible material for livestock. Periodic droughts an important factor in the range condition and are unpredictable and may occur at any time of ecological stabilityof rangeland ecosystems.When the year. They are always a menace and require 298 Measuring Sustainability in Tropical Rangelands constant preparedness and continual planning. and bacteriological content of water bodies. The These droughts have neither a fixed pattern of possible means for making maximal use of the occurrence nor regular duration. Disaster surface water resources were also assessed. Si- droughtsareactually protected periodic droughts. multaneously, the processes of land degradation All available feed is used up, and state assistance throughthemovementof surfacewaterwerealso on a large scale is necessary to save livestock. Soil assessed using suspended sediment date., chemi- moisture falls to a very low point where it is not cal leaching data, bed-material data, and other available to plants, and a large number of plants information related to chemical weathering, soil may die. Failure to provide for periodic droughts erosion, and so forth. Preliminary results indi- means that disaster droughts will assume even cated very high seasonal variability in the flow of greater proportions. Droughts caused by humans the river, with long periods with no surface flow. occurwhenashortageoffeedisbroughtaboutby Data on total dissolved solids and suspended overgrazing or injudicious use of the range, even sediments and other attributes are similarly vari- though climatic conditions may not be such as to able during times of flow. cause drought. In areas where droughtsare caused by humans, and this is the situation in northern GROUNDWATER Kenya now, even a seasonal drought is a disaster. In the Chalbi basin, as in most regions of the Although drought has been termed one of the world, groundwater occurs at some depth, but clearest examples of humanity's hopelessness in limitations in its quantity, quality, and the energy the face of the broad-scale phenomena of nature, its required to raise it to the surface may preclude its effectscould be lessened byadequate forward plan- development. Initial results can be summarized ning. Theinformation forsuch planningcould come as follows: from meteorological servicesand intensive weather Records, direct measurements, and resistivity monitoring, such as that carried out by IPAL. measurements indicate that groundwater is The constant necessity of the pastoralists of fairly close to the surface in several areas but northern Kenya to remain almost permanently had notbeen developed toeven a fraction of its on famine relief has clearly demonstrated the full potential, that is, Hedad, Korante, and present inability of the traditional mechanisms to Balessa Kulal rivers. In some areas, the water cope with theproblemof drought. The traditional was very deep and would be costly to develop, system must be improved and supplemented in for example, southeast and southwest of the such a way that the pastoralists are always ready Huri Hills and the Segel and Bora areas. to combat drought. . Most of the groundwater is suitable for live- stock, but not for human consumption or for Water resources studies irrigation. However, in some areas, it is suit- Future development in Marsabit Districtdepends able for all uses, that is, Korr, Kalacha, some on the availability of water, whether for extend- springsnearMaikona,andmuchoftheperched ing the area available for grazing (when effective water in the river beds. grazing control was instituted), for future settle- * The water in the perched or sporadic shallow ment schemes, or for limited irrigation of crops. aquifers (at Ngurunit) maybe limited inquan- Consultants carried out two preliminary recon- tity, and wells in the deeper aquifers may be naissance studies, one on surface water and the limitedinquality.Insomeareas(Korr),thelow other on underground water. These are briefly permeability of the aquifer makes it suitable summarized below. for wellswith only verylow specificcapacities. Generally, the regional water tables are just SURFACE WATER above the bedrock or in the bedrock complex, Over a period of four months in 1981, a prelimi- whose general surface dips toward the Chalbi. nary survey was undertaken of the springs, sea- However,in someareas,thebedrockismorethan sonal watercourses, and dams in the catchment 300 meters below the ground surface, that is, area of southwest Marsabit District; the results beyond the capabilities of the resistivity instru- are now being analyzed. The work included the ment used. Resistivity measurements indicated installation of equipment to monitor thequantity that somegroundwatermustleakoutof theChalbi of water flowing in the major seasonal rivers and basin to the west of North Horr, but the move- collected in the dams and to assess the chemical ment elsewhere is toward the Chalbi. 299 Defining and Measuring Sustainability: The Biogeophysical Foundations Much groundwater is usable in the Chalbi are provided were investigated. A methodology basin. Any future studies could be useful if fo- foranalyzingthedevelopmentpotentialforsimi- cused on the more usable shallow groundwater lar arid areas was also developed. The economic resources because capital and operational costs studies specifically undertook to: are relatively low, and these sources are more * Estimate the annual production of livestock easily managed by local people. and livestock products available for satisfying human wants like food, shelter, clothing, and Pastoral economy studies so forth ThelPALstudiesalreadydiscussedyieldedvalu- * Estimate the quantities of these commodities able information on relationships between the that were necessary at that time for satisfying various components of the ecosystem and the subsistence needs when 70-90 percent of the diet consisted of milk, blood, and meat processes leading to environmental degradation. Although this information is valuable, it is diffi- * Estimate the amount of these commodities cult to translate into practical policy and manage- that was surplus to the people's subsistence ment options to be implemented by planners and needs and hence was available for marketing administrators. The information had to be trans- . Estimate the quantities of livestock and live- formedintomanagementproposals. Aneconomic stock products that were marketed and the analysisand assessment wasinitiated to show the quantities of maize flour, sugar, and tea im- amount of money to be spent, a monetary valua- ported to the area in a given year tion of the expected returns, and a general assess- * Study the marketing channels through which ment of the feasibility of the expected outputs and foodstuffs, and consumersitemducts (exports)w follow-up activities. traded; suggest ways and means by which the Many practices of animal husbandry and land existing system was to be made more efficient use that seem to be grossly inefficient might be in regard to the quality of services rendered to necessary adaptations to existing economic real i- communities and the quan tityof goods moved ties. These might be inherent in the pastoral sys- in and out of the area tem of production or be dictated by develop- * Conduct, using data from ongoing IPAL stud- ments in other parts of the country or even on the ies, a cost-benefit analysis of experimental man- world markets and therefore be outside the con- agementinterventionstodeterminewhetherthey trol of pastoralists. offered good returns for the money expended Without a proper understanding of these eco- * Investigate the potential fordevelopingsources nomic conditions and a conscious attempt to of income other than the sale of unprocessed change them for the better-or to adapt to them- livestock and livestock products, including the any management proposals recommending possibility of developing aloes and Acacia changes to the existing patterns of behavior will senegal as cash crops, the potential for be futile. This observation is in line with the expanding the collection and marketing of conclusions of a recent evaluation by the U.S. honeyusingmodernhivesinsteadof traditional methods, which destroy thebees, thepossibility Agency of locating a small-scale hides and skins tanning livestock sector projects in Africa. The evaluators or curing factory in the area, the possibilities reached a consensus that for those programs to for developing tourism and fishing. have favorable and beneficial impacts on pro- ducer populations, national wealth, and environ- mental conditions, they had to be reoriented to Education, training, and demonstration make them more nearly compatible with the so- "Lackofeducationandenvironmentalconscious- cial, economic, and environmental realities of ness was identified as a major cause of land arid and semi-arid pastoral regions of Africa. Thiseemetojstiftheceatonofnecoom. degradation in Kenya. It was indicated that even studiesemponento just thecreAtion profjaeconomic though we have the technical know-how of the studles compeonent on the IPAL project root causesofdesertification,littleefforthasbeen Broadly, the study aimed to describe the eco- maetedcethmsesndprcurlte nomic relationships that exist within the Rendille made to educate the masses and pardcularly the nomadicpastoral ecosystem. The waysand means young generation on preventive and remedial nofimadovicpathr ecosystem. ithe whyscnd mhean pmeasures to fight against land degradation" (Na- of improing the fficienc with whch the ro tional Environmental Secretariat 1977). This quota- duction and distribution of goods and services tinEvroneaSceaial7)hsqt- dlon, taken from the recommendations of the 300 Measuring Sustainability in Tropical Rangelands Kenya national seminar on desertification, un- in cooperation with the Kenya Ministry of Infor- derlines the importance of the role of education mation and Broadcasting, broadcast a fifteen- and training in the effort to improve and stabilize minuteradioprogram twicea week in theRendille the production systems in arid lands. language. IPAL's education and training activities were based on the belief that the results of scientific Initial project findings research are not useful in achieving the desired goalsunlesstheyaretranslatedintoaformusable From the results of the fertility analyses of the by the population affected by land degradation. soils, it would appear that none of the major The aim was therefore to use the findings of the mineral elements islacking. Although thelevel of project in education and training for the dissemi- nitrogen in the soils is very low, with an average nation of information on rational management. rainfall of only 225 millimeters a year, lack of At the local and national levels, the project was water is the most important factor limiting the integrated into the Kenya government's develop- production of vegetation. After rainfall, the low ment plans for northern Kenya, and the project nitrogen content is possibly the next limiting was a member of the Marsabit District Develop- factor for vegetation other than legumes. ment Committee, where all matters related to the All the range types except the perennial grass- development of the district were discussed. land vegetation on the mountains are either in In addition to making direct inputs at the dis- poor or fair condition. This is surprising since trict planning level, the project, in cooperation some of the areas have not been used for a long with the Kenya Institute of Education, investi- time due to problems of insecurity. A further gated through consultants the best possible way investigation into the history of the use of this of incorporating the project's results and arid vegetation might reveal the reason for such a land ecology in general into the school curricula trend. As demonstrated by the vegetation for the pastoral areas. This effort was designed to exclosures, the recovery potential of the degraded cover primaryand secondary schoolsand later to vegetation is very high, particularly for areas be extended to teacher training colleges and insti- with annual vegetation. In one of the less-grazed tutions of higher learning, like the university. areas, for example, annual grass contributes to Study grants enabled graduate students from the more than three-quarters of the standing crop University of Nairobi and other institutions to be biomasswithsubstantive amountsof litter,which attached to the project and to conduct research on tend to exceed the standing crop biomass. These various components of the IPAL study using the findings are important because they reflect the project's facilities and advice. ecological limits of various plant communities On thebasisof experience gained by the project due to factors of soil fertility and moisture. Annual staff and takingadvantageof the established base grasslands cannot have a succession that leads to for fieldwork, a series of seminars were offered to perennial vegetation, and they should be managed train decisionmakers, administrators, and local to achieve their optimal annual production. people. Training and orientation courses for all Wood requirements in arid environments are levels were offered at the project's new head- known to vary greatly from one region to an- quarters in Marsabit. These courses included other, but the Rendille and Gabbra pastoralists, instruction on all aspects of the project's work, with their traditional diets of uncooked food- especially observation techniques, lectures, and milk, meat, and blood-have a low annual re- demonstrations in arid zone ecology, range quirement for firewood amounting to possibly management, arid zone forestry, and pastoral 0.1 meterperperson.Thismaybecompared with sociology.Thefacilitiesandexperienceof IPAL an assumed annual consumption in agricultural were also made available to visiting scientists areasof Africaof l meterperperson.Now that the and extension officials from other countries in diet of the pastoralists of northern Kenya is in- the region. creasing to include more grains, it can be con- Because communication in the arid areas is cluded that their need for fuelwood will increase. particularly difficult due to an inadequate road In a study of relatively sedentary Rendille people network and theinability of most of thepastoralists in a degraded village environment at Korr, the to read or write, the project investigated the most annual consumption of wood for livestock night effective way of communicating information to enclosures for a single family was approximately pastoralists. Using the project as a test case, IPAL, 432 kilograms. With four people per household, 301 Defining and Measuring S ustainability: The Biogeophysical Foundations the annual consumption per head was estimated economy and policy, in which they still possess at 108 kilograms, and this too isbound to increase marginal status. This trend is motivated by grow- with the increase in the human population. ing awareness that total reliance on pastoral no- The main problem with the Rendille livestock madism will not provide sufficient insurance is that, although in theory their numbers are against drought nor the economic base necessary adequate to supply basic human needs, they are to improve living standards. The desire to im- not very productive. They suffer from numerous prove the standards of living of the households, diseases, and their mortality is very high. For ex- to diversify the region's economic base, and to ample, between 20 and 30 percent of all the cattle in become sedentary are three interrelated processes. Marsabit District died during the 1971-81 drought. Almost400,000 ungulates were in the study area, of which all but 3 percent were domestic livestock. Conclusions Although most of the livestock were sheep and goats-287,040-when expressed in terms of The problem of land degradation in the range- tropical livestock units, they contributed only 18 lands of northern Kenya and eastern Africa in percent to the total animal biomass, while cattle general is serious and needs urgent attention, contributed 40 percent and camels, 36 percent. because not only does it involve the threatened Livestock were present at low densities in 1976 survival of a people but because some of the but increased in 1977 to more than three times processesofdeterioration,ifallowed tocontinue, their former densities. This was mainly due to the may become irreversible. drought of 1976, which did not seem to affect Although livestock production will probably wildlife, which are less dependent than livestock remain the best use of the arid lands of northern on fixed sources of water. Because of the fear of Kenya, the type of livestock operations carried theft, livestock owners are obliged to keep their out need to be considered carefully. Managing animals in night enclosures, which puts them at a livestock for milk production doesnot seempracti- disadvantage when compared with wildlife, cal in an arid environment. The efficiency of con- which graze at night, benefit from moisture con- verting forage into livestock products for use by tained on plants, and retain moisture by staying pastoralistsnmight necessitatecareful consideration in the shade during the day. of the suitability of present breeds of cattle. Theresultsof the study, indeed,indicate thatthe In spite of common belief, the natural ecosys- IPAL study area is capable of supporting twice the tem is not always the most efficient for converting presentlivestockdensities. The key to rehabilitation solar radiant energy in primary production or for of the area does not therefore seem to lie in the transferringenergyintosecondaryproduction.A enforced sale of livestock but ratherin achieving a manipulated system may produce more dry better distributionof livestock.Such a solution would weight of plant material and frequently furnishes also be more acceptable to the people. a diet more conducive to meeting nutritional The nomadic community is expending exorbi- requirements than does the native ecosystem. In tantly high amounts of energy during the dry view of reduced flexibility in the lives of the seasons in walking animals long distances, in pastoralists of northern Kenya, range improve- digging and maintaining wells, and in drawing ments,especiallyinalreadydegraded areas where water for domestic use and for very young ani- natural recovery would otherwise be too slow, mals too weak to walk to water. Results of these should be considered seriously. studies reveal that there are sufficient water re- The Kenya Development Program focuses on sources-both ground and surface-in the study ratifying the rights of pastoral people to their area. The resistivity measurements carried out traditional grazing areas. This is seen as encour- show that water can be obtained almost any- aging the social change necessary to translate where in the study area by the use of shallow hand- subsistence pastoralism in part into commercial dugwells.Thiswouldprobablybethebeststrategy livestock production and to combat the present to adopt since the people can maintain hand-dug overgrazing and deterioration in range condi- wells better than they can the bore holes. tion. Emphasis is placed now on the type of land IPAL's socioeconomic studies have revealed ownership, not because of the moral issues in- that the Rendille and Gabra societies are gradu- volved but because without security of tenure ally moving out of their isolation and are very people have little incentive to improve land re- slowly becoming integrated with the national sources. In Kenya today, most pastoral societies 302 Measuring Sustainability in Tropical Rangelands clearly feel the insecurity of their present posi- Field, C. R. 1979. "A Preliminary Report on the tion. The failure of previous attempts to effect Ecologyand ManagementofCamels,Sheep,and change in pastoral societies is certainly due in Goats in Northern Kenya." [PAL Technical part to ignorance of the biological basis of pasto- Report E-19. UNESCO, Nairobi, Kenya. ralism, but there has also been little attempt to . 1981. "A Summary of Livestock Studies understand the sociological constraints on the withintheMt.KularStudyArea." IPAL Techni- fears and aspirations of the pastoralists. cal Report A-3. UNESCO, Nairobi, Kenya. Local conditions must be taken into account in Griffiths,Norton.1975. "Aerial SurveyTechniques." physical, biological, and hydrological planning of AfricanWildlifeLeadershipFoundation,Nairobi, the arid lands. For optimal productivity of the indi- Kenya. vidual factors that have to be considered in devel- Heady, H. F. 1982. Rangeand WildlifeManagement in opmentplanning,priorityprobablyshouldbegiven the Tropics. London: Longmans. to humnan ecology and ecological approaches to Heady, H. F., and E. B. Heady. 1984. Range and land use. This does not imply that there is a limit Wildlife Management in the Tropics. London: beyond which ecological ideas should always pre- Lonmans. vail, but it does imply that there is a limit beyond Herlocker, D. 1979. "Implementing Forestry which ecological compromise is impossible. In ard Programmes for Local Community Develop- and semi-arid areas, forexample, schemesare some- ment, South-Western Marsabit District, Kenya." times propounded for the settlement of nomadic IPAL Technical Report D-2a. United Nations populations, which, if permitted, not only would Educational, Scientific, and Cultural Organiza- fail to realize any long-term social or economic tion, Nairobi. benefitbutalso would destroy thegrazingresource, IUCN (InterationalUnionfortheConservationof especially during drought. Nature). 1991. Caringfor the Earth: A Strategyfor The initial results of IPAL discussed in this SustainableLiving.WiththeUnitedNationsEnvi- chapter will have some application elsewhere in ronment Program and the World Wildlife Fund. the world. Whetheror not pastoralism will change Gland, Switzerland. to meet the present demands of development L depends on the pastoralist's attitude toward his Lusigi, W. J.1981. "Combating Deseftification and ... . . ~~~~~Rehabilitatine Deeaded Ecosvstems in North- own pastoral activities. A change in practice en- egnKenya." IPAL TechnicalgReportA4.United tails a far-Teaching process of physical, mental, emtKenya." ioal T enti Reor an4 Cunted and spiritual adjustment-it actually amounts to Organization a Nairobi e a reorientation of the pastoralist as a person. But O he will have to be helped through credible scien- Lusigi, W. J., ed. 1983. "Integrated Resource Assess- tific advice to attain his aspirations. mentand ManagementPlanforWesternMarsabit District, Northern Kenya." IPAL Technical Re- port A-6. United Nations Educational, Scientific, References and Cultural Organization, Nairobi. National Environmental Secretariat. 1977. 'Work- Child, D., and others. 1984. Arid and Semi Arid Lands. ingPaperon Soil Conservation." Paperpresented Arkansas: Winrock International. at a seminar on desertification, Nairobi, Kenya. .1987. Arid and Semi-arid Lands: Guidelines Odum, H. T. 1959. Fundamentals of Ecology. U.S.A.: forDevelopment. Arkansas: Winrock International. W. B. Sunders Co. Coetsee, M. J. A. 1968. "Droughts and the Farmer." Schwartz, H. J. 1980. "An Introduction to the Live- Farming in South Africa 44:1, pp. 24-27. stock EcologyProgram." IPAL Technical Report Colorado State University. 1972. Wildland Ecology A.3. UEpm Of agriclue. Handbook. Ft. Collins, Colo. U.S. Departmentof Agriculture.1970. Range Analy- Dratt, D. J., and M. D. Gwyne. 1977. Rangeland sis Handbook. Region2. Washington, D.C. Management and Ecology in East Africa. London: Van Dyne, G. M., ed. 1%9. The Ecosystem Concept in Hodder and Stoughton. Natural Resource Management. New York: Aca- Field, A. C. 1978. "Preliminary Report on the Impact demic Press. of Sheep and Goats on the Vegetation in Arid WCED (World Commission on Environment and Zones of Northern Kenya." IPAL Technical Development). 1987. Our Common Future. Ox- Report E-19. UNESCO, Nairobi, Kenya. ford, England: Oxford University Press. 303 Defining and Measuring Sustainability: The Biogeophysical Foundations Lusigi's work speaks for itself and does it very Comments well indeed. Instead, my comments address the application of his project to the subject of man- Lee M. Talbot aged ecosystems, and, in particular, I identify some of the questions that must be answered By way of introduction, there is a need for some when we are seeking to design ways to identify clarification in the focus of this discussion. Man- and monitor thebiophysicalbasisof sustainability. agement is often considered to refer to conscious, intentional manipulation of some parts of the environment to achieve a particular objective. pplications However, human use of these extensive parts of the globe ranges from casual exploitation by no- The IPAL was a large project, a major undertak- madic grazing and hunting to intensive manipu- ing. It involved many peopleand a broad range of lation. Most of these lands are not managed in the research activities undertaken over a period of usual sense of conscious manipulation of some of about twelve years. It provides a magnificent the factors. Much of the past use of the northern baseline for measuring the biophysical basis of Kenya rangelands, and some of the present use sustainability in that area and that type of area. that Dr. Lusigi studied, is not in this sense man- aged. The principle of the minimum However, rangelands cover vast areas of Af- The IPAL has identified, defined, and measured rica and muchof the restof the world. These lands a comprehensive series of biophysical factors or are used and they provide the resource base for attributes. If we wish to determine the status and the survival of a significant number of people trends in sustainability of these or similar range- worldwide. Ourconcern is with the sustainability lands without duplicating this study, we need to of this use, so, clearly, our interpretation of identify which attributes we need to measure. management is broad. From the standpoint of practicality and expense, Dr. Lusigi's chapter primarily describes a spe- we want to measure as few attributes as possible cific research project, the Integrated Project on to achieve satisfactory results. Consequently, one Arid Lands (IPAL) in order to provide a specific, key issue, from the standpoint of our concern real-lifeexampleofmeasurement of sustainability. with these measurements, is to determine what is The purpose of IPAL was to understand the the minimum number of these parameters and structure and functioning of the range resource measurements that is needed to define involved as a foundation for understanding its sustainabilityand to identifychangesor trends in sustainability and to identify the attributes of the future. sustainability. This information would probablybe most use- In his work, Dr. Lusigi describes the project ful if it could be presented in the form of a model and the attributes and measurements involved, that encompasses these parameters, data, and that is, "how an attempt was made to measure measurements. Such a model would help us un- sustainability of that ecosystem and to under- derstand the system, but it could also be a practi- stand its functioning." A detailed discussion of cal tool that could be applied to this area but that the results of the study was outside the scope of wouldalsoapplytoothercomparableareasof the his work. world's rangelands. A comprehensive research project such as IPAL can address one important part of the objectives The question of periodicity and time scale of the conference that spawned this volume. An- other part involves taking the knowledge from This study identified and measured parameters such a study and identifying from it indexes of or attributes. One of the challenges we face in sustainability, particularly in the form of attributes applying the results is to determine what should or parameters that can be monitored and that be the periodicity of these measurements in the allow the ongoing measurement or determina- future, particularly recognizing that these arid tion of the sustainability of rangelands. lands have great fluctuations in rainfall and re- I will not comment on the specifics of the lated climatic conditions. Among the specific project or the description of them because Dr. questions that come into play are: 304 Measuring Sustainability in Tropical Rangelands * How often do you need to make specific mea- secutive wet years or several consecutive dry surements to determine sustainability? ones. . How do you recognize and take into account These shifts caused profound differences in the radical differences in conditions from wet the composition, appearance, and functioning of to dry periods? What is pertinent here is that the ecosystems involved.Thesechangesinvolved the carrying capacity of these lands, for ex- the percentage composition and, to a large de- ample for herbivores, may be several times gree, the distribution of grasses and other range- higher in wet periods than in dry ones. land plants. In the truly wet or dry periods, this •Therefor,what is the time scale for was true for perennials as well as annuals. For sustainability? example, during prolonged wet periods, some areas were characterized by dense stands of tall In terms of time scale, what is sustainable in a perennial red oats grass (Themeda triandra). Yet in wet year will not be sustainable in a dry one. Are extended dry periods, virtually no Them eda plans we speakingof periodsof one, three, five,or more could be found in the area, and the nearest dense years? For example, if one looks at sustainability stands were more than ten miles away. over a long period, there will be periods of abun- Changes in the seasonal and absolute avail- dant growth (for example, of vegetation and ani- ability of food and in its composition (for ex- mals) balanced by periods of stress, but in the ample, the percentage composition of nitrogen, long run, the system is sustainable. fiber, and water) were reflected by marked In other words, what is sustainable over a changes in the food habits of the wild ungulates, multiyear period may involve severe die-offs with consequent chaages in the dynamics and during the dry periods. Maasai and other behavior of their population. For example, dur- pastoralists in Kenya used to expect about a 15 ing wet periods, the impala in eastern Africa find percentdie-offoftheirlivestockduringadrought optimum food most of the year and maintain year, and they sought to adjust the number of harem herds virtually year-long, breeding theirlivestocktoprovideinsuranceforthatrecur- throughout the year and producing herds that ring situation. However, with increased human contain young of all ages. During dry periods and livestock populations, the die-off in drought when optimum food may only be available for years becomes much more serious. Therefore, if one to three months, the impala maintain their sustainability is calculated on the basis of a long harem herds and breed only during that wet period, there will be losses in dry years, and period, and for some years the herds are charac- livestock and people may die. From their particu- terized by even-aged groups of young. Further, lar points of view, such a situation is not sustain- the breeding success and survival of young are able. The sustainabil ity issue for people is imme- also markedly affected. During dry periods, about diate. Yet from the standpoint of long-term 15percentoftheyoungwildebeestintheSerengeti sustainability, one year is not that important. region survive to become yearlings, while during the prolonged wet periods, that figure may be The special problem of arid areas: The vast close to 100 percent. differences between wet and dry periods These climatic swings are also linked to an- other phenomenon that is often overlooked in Theworld'saridorsemi-arid zonesarecharacter- considering sustainability. This is the fact that a ized by dramatic swings between wet and dry relatively small number of people can exert a periods. These swings have profound effects on disproportionately heavy impact on the land- the ecosystems. There are both short-term and scape during the rare periods of extreme climatic long-term irregular fluctuations in precipitation stress, such as prolonged dry periods when veg- and associated climatic conditions. In my work in etation, soils, and wildlife are particularly vul- the Serengeti region of East Africa, I found that nerable. For example, even though the popula- there had been two really dry periods in the past tion of Masai in the Serengeti region was rela- century. Superimposed on those long-term fluc- tively low, their normal pattern of burning and tuations were short-term ones. These were con- grazinghad theeffectofdramaticallyalteringthe sidered to be normal seven-to-fifteen-year shifts vegetative cover, opening large areas that were between wet and dry periods. There might be a formerly covered by forest or bush, during the period of several years where wet years alter- two periods of extreme drought in the past cen- nated with dry ones, followed by several con- tury. 305 Defining and Measuring Sustainability: The Biogeophysical Foundations Therefore the usual one- or two-year study The human factor represents a view through a window showing conditions that may only be representative of Most areas, even remote ones on the edge of the one-tenth to one-twenty-fifth of what are normal deserts, are under continual and often pro- conditions for the area involved. Where the con- found human impact. In seeking to elucidate ditions can be so radically different from wet to the ecological processes involved and to define dryperiods,a verylongperiod of studyisneeded to the mechanisms of sustainability, and conse- give an adequate representation of overall condi- quently, the parameters that mustbe measured, tions. There is the temptation-indeed the almost it is helpful to be able to separate the human universal habit-of researchers to draw absolute impact from that of other factors. In practice, conclusionsbasedontheirone-ortwo-yearstudies. the best way to do this is to have control or Perhaps this is one of the reasons why our under- benchmark areas where human influence is standing of arid rangelands is still so incomplete. minimal. What does this imply for the design of This situation has profound implications for measurements for sustainability? measurernentsofsustainability.Whenconditions How much of the human role needs to be vary so greatly from one year to the next, and measured? The IPAL study, quite correctly, from one decade to the next, how do you calibrate focused strongly on identifying and measuring sustainability and for what period, and how do the human influences. But humans change, not you determine the frequency and periodicity of only their numbers and their economic situa- measurements? tions but their cultures, aspirations, and ways The boundaries of sustainability of life. The IPAL study chose representative sites for re- The myth of durability search within a large but carefully defined study In considering the sustainability of rangelands, area. However, when seeking to determine there is a persistent problem in the form of the sustainability, how do you determine the bound- p aries of the area involved? The borders are often belief,lI termite myst, tha vn is an permeable. The markets for livestock often exist extremely durable ecosystem, which is sustain- away from the rangelands proper. Equally, the ableregardlessofhumanimpactsandwhich,in sourcesofnewlivestockareoutsidetherangelands. its extreme form, denies that human-caused In areas such as much of the Sahel, many of the land degradation exists or is more than a tran- pastoralists are also cultivators. To determine the sitory phenomenon. This myth has resurfaced sustainability of an area for humans, it is necessary recently in the form of papers based on the last to consider the agricultural lands as well as the twenty years of remote-sensing data for the rangelands. In the same way, both wildlife and Sahe!, which show that the vegetated areas pastoral herdsand flocks are nomadic in arid areas. expand or contract in response to rainfall, but When trying to determine carrying capacity, or not necessarily in response to human use. biomass per unit period, or sustainability, it is nec- It is true that savanna or rangeland is generi- essary to consider the entire area used by the ani- cally durable. To the casual observer, the gross mals. It is relatively meaningless to talk of the physiognomy appears much the same regard- carryingcapacityorsustainabilityofanarea that is less of the shape it is in. But the difference only a part of their annual range. between a severely degraded rangeland and a highly productive one is great. The productiv- Maintenance mechtanisms ity of the rangelands is relatively fragile, in- Sustainability also implies concern with mainte- volving the conditions of soils, vegetation, nance mechanisms, that is, the mechanisms or moisture, and animal life. We are concerned processes that maintain the system involved. with sustainability of the conditions that per- Manyrangelands,particularlythoseinareaswith mit such productivity, not merely the higher rainfall, are at a successional stage main- sustainability of land that has some form of tained by fire and grazing. Considerations and minimal vegetation on it. Dr. Lusigi's chapter measurements of sustainability must take these addresses well the issue of human impact on mechanisms into account. sustainability. 306 Measuring Sustainability in Tropical Rangelands Application to other tropical areas Conclusions A major question that arises is the degree to A comprehensive research project such as IPAL which the results from the IPAL study apply to can address one important part of the objectives other tropical rangeland areas. First, the IPAL of this volume. Another part involves taking the is in a particularly arid portion of the arid knowledge from such a study and identifying tropics. This study was in an area characterized from it indexes of sustainability, particularly in the by very low rainfall amounting to some 9 inches form of attributes or parameters that can be morn- a year. One consequence is that fire plays little or no role in this grassland, unlike the situation determination of the sustainability of rangelands. in areas of higher precipitation. Fire character- teminat thdpruiedinformation helpnus 9 ~~~~~~~~~The IPAL study provided information to help us izes most of the world's grasslands where rain- understand the functioning and the factors that fall is adequate, and indeed, fire-virtually affectthesustainabilityof therangeland ecosystem, entirely caused by humans-is a major factor . . . causig an maitainng sch gasslnds.Sec- and it identified and measured a series of attributes causing and maintaining such grasslands. Sec- or parameters that describe or contribute to that ond, IPAL is in an area where the pastoralists sustainability. This review concentrates on ques- are not also agriculturalists. This is the situa- tions that arise if one wishes to take those param- tion in other areas of Kenya, as with the Masai, eters and apply them to the study area or else- but in most of the rangelands of Africa, for etranaplthmotesudaeares- but i ooh a lwhere, both to describe and to monitor the status example Somalia and most areas around the and trends of sustainability of the system. Sahel, the people are agropastoralists. Among the key questions are those involving Therefore, the conditions are not exactly par- the minimum set of parameters needed, the per- allel with much, and probably with the greater odicity of measurement, the boundaries of the part, of Africa'srangeland. However, themeth- area involved ands t, thime scale for which odology and the definition of the factors in- sustainability is calibrated, and the applicability volved in sustainability and the parameters to other rangelands of the parameters derived that should be measured should, in general, from IPAL. Of particular importance is the need apply to most rangelands. Additional measure- to understand and take account of the impact of ments may be required for areas of higher pre- short- and longer-term climatic fluctuations on cipitation and ones with different patterns of theecosystemsinvolvedandconsequentlyonthe human activities. sustainability of their characteristics under vari- ous types of human use. 307 t 1. Indicators of Grassland Sustainability: A First Approximation Paul G. Risser If ecosystemsaretobemanagedinwaysthatensure climate. The two lessons are that any proposed that the desirable characteristics are maintained indicators of ecosystem sustainability must be cast indefinitely, then managers must know what to in ways useful to society and that they must either measure as indicators of the sustainability of the accommodate or allow for changes in underlying ecosystem. Thisobjectiveof identifyingbiophysical physical conditions. indicatorsof ecosystemsustainabilitywill necessar- ily contain some conjecture because humans have not managed ecosystems using rigorous criteria of Grassland ecosystems sustainability for long periods of time. However, the argument presented here is that there is suffi- Grasslandsarebiologicalcommunitiesthatcontain cient scientific information to offer a first approxi- few trees or shrubs, are characterized by mixed mationof a set of measurements that will predict the herbaceous vegetation, and are usually dominated long-term sustainability of grassland ecosystems. by grasses. They cover about 46 million square The seriousness of the need to predict sustain- kilometers and occur on every continent and large able ecosystems can be seen through observation of island throughout the world (Risser and others human history. This history is replete with ex- 1981; Whyte 1960). Although the climate of grass- amples of human populations who have mistreated lands is characterized by at least one dry season, the ecosystems until sustenance support was no longer presence and behavior of grasslands are controlled forthcoming. One example is the Anasazi (Navajo not simply by total annual precipitation, evapora- for "ancient ones") in the southwestern United tion, or maximum or minimum temperatures but States. This population, once thriving in such places rather by complex relationships including the ratio as Chaco Canyon and Mesa Verde in 1100 A.D., had of precipitation to evaporation and the seasonality of disappeared by 1300 A.D. due to various reasons: precipitation in relation to the temperature regime. exploitation of the nutrients in the soil, the drought Grasslands are used and managed with vari- of 1272 to 1299 and a change toward a drier climate, ous techniques, perhaps usefully arranged along and an increase in the size of cities beyond both the a continuum from native rangeland to highly food sources and social structures necessary to sup- managed pastures and meadows. Throughout port them (Smith 1988). Therefore, the measures of much of the world, many native mesic grasslands ecosystem sustainability must recognize not only have been converted to other cropping systems. the conventional ecological terms such as primary These mesic grasslands produce rich soils that are productivity and species diversity but also the re- easily converted to cultivated crops or managed quirements of human social systems. Moreover, grasslands. Thus, any proposed indicators of eco- any set of biophysical indicators must be suffi- systemsustainability must depend onthedesired ciently conservative to accommodatechangesin the outcomes of management and must aisso be spe- underlying physical conditions, such as a change in cific to the type of ecosystem Defining and Measuring Sustainability: The Biogeophysical Foundations In sumanry,effective indicatorsof sustainability of a grassland ecosystem must be formulated to Box 19-1: System Characteristics Critical meet defined products, be described such that for Maintaining an Ecological System they are useful to society and to managers, be specific to the ecosystem, and allow for changes * Habitat for desired diversity and reproduc- in the physical driving variables. tion of organisms * Phenotypic and genotypic diversity among the organisms Indicators and end points . Robust food chain supporting the desired biota In the search for measurements that predict * Adequate nutrient pool for desired organ- whether an ecosystem will maintain its essential isms properties over long periods of time, it is useful to * Adequate nutrient cycling to perpetuate the distinguish two terms that are frequently en- ecosystem countered in the current literature. First, indica- * Adequate energy flux to maintain the trophic tors refer to specific direct measurements that can structure be made on a property of the ecosystem, such as . Feedback mechanisms for damping undesir- soil moisture. End points refer to synthetic terms able oscillations that also can be measured but that encompass * Capacity to temper toxic effects, including several processes or characteristics, such as spe- the capacity to decompose, transfer, chelate cies diversity. Both concepts are important; that orbind anthropogenic inputs toadegree that is, indicators have the value of specificity but they are no longer toxic within the system. cannot directly measure more integrated pro- cesses; end points presumably capture more inte- Source: Schaeffer and others 1988. grative processes of ecosystems, but interpreting the measurements is more difficult because of their inherently synthetic construction. Both con- there is a reasonable argument that some mea- cepts are included in this discussion, but for con- sure of complexity should be included in any venience indicators are used unless there is a indicator of ecosystem sustainability. special need to discuss the synthetic nature of an Ecosystem propertiesvaryovertimeand across end point measurement. space. Indeed this temporal-spatial variability could be a component of measures of ecosystem Criteria sustainability. Therefore, sustainability indica- tors must have a variability component, either Several schemes have been suggested for criteria explicitly recognizing the variability expected in to be used in determining whether ecosystems the indicator measurement itself or including are sustainableindefinitely(Karrand others 1986; variability as a measure per se. Schaeffer, Herricks, and Kerster 1988) and in re- In making recommendations fordeciding if an storing grasslands (Jordan, Gilpin, and Aber 1987; ecosystem is sustainable, I am not proposing one Werner 1990). Some authors have proposed spe- more slate of characteristics to be included in a cific measurements that can be used in the field monitoringapproach.Theselists, included under (Karr and others 1986), and others have offered the rubric of environmental audits, are certainly more theoretical proposals (Shaeffer and others important for determining whetherenvironmen- 1988). These more theoretical proposals are valu- tal deterioration is occurring and environmental able because they can guide the development of standards are being met. A sample listing of such approaches that can be applied in the field (see variables to be monitored, particularly at the glo- box 19-1). Since ecosystems undergo succession, bal scale, is given in box 19-2 (Cairns 1991). In this either the guidelines for indicators must change discussion, indicatorsof ecosystem sustainability with the successional status of the ecosystem or are designed as a few measurements that predict the indicators must be sufficiently general to ac- simply whether or not the ecosystem, and its commodatesuccessional changes. Also, since eco- essential characteristics and defined products, systems are inherently complex, and this com- will continue indefinitely. These proposed indi- plexity may contribute to the feedback processes cators are designed to be relatively easy to mea- and the stability and resilience of ecosystems, sure and few in number. 310 Indicators of Grassland Sustainability: A First Approximation Grassland indicators of ecosystem Box 19-2: Variables That Should Be In- sustainability cluded in a Global Monitoring System The general concept of grassland sustainability is at the heart of range management, and particu- Species level larly the concept of range condition (Lauenroth * Structural: for example, tissue or organ damage and Laycock 1989). Although the concept may * Functional: respiratory rates or behavior have come from forestry, it has been used in rangeland management for more than half a cen- Community level tury (Korstian 1919; Pendleton 1989). In range * Structural: trophic relations management, a range site is a landscape unit that * Functional: colonization rate or rate of detri- is classified based on its relation to climax plant tus processing communities. These ideas have been modified Ecosystem level over time to include climate and soil conditions * Structural: trophic relationships characteristic more explicitly (Dyksterhuis 1949; Renner and of this particular type of ecosystem in this Allred 1962). Range condition refers to the eco- locale logical or successional status of the vegetation on * Functional: nutrient spiraling or energy cycling the range site as compared with the potential or Landscape level climax plant community. Formally, the Society * Structural: compatible with the landscape for Range Management defines range condition mosaic as "the current productivity of a range relative to Functional: used with appropriate duration what that range is naturally capable of produc- and frequency by species that regularly use ing" and range condition class as "one of a series the larger mosaic of which this is a part. of arbitrary categories used to classify range con- dition and [that] is usually expressed as excellent, Source: Caims 1991. good, fair, or poor." A range in excellent condition would be pro- ductive and would be dominated by species of the climax or potential plant community; a range not so palatable and so increase as the decreasers in poor condition would include few or no climax are selectively grazed), and invaders (species that species and would be dominated by species un- can only invade the grassland community when palatable or of low value to livestock. The con- it has been disturbed by heavy grazing or some cepts of range condition have been criticized on a other event). A grassland in excellent to good number of grounds (Risser 1989). The concerns condition hasahighproportionof decreasersand include the difficulty of defining the climax or few invaders. Table 19-1 demonstrates this re- potential vegetation on all sites, the recognition sponse to grazing in the tallgrass prairie. In this that not all rangelands should be managed to case, big and little bluestem, switchgrass, and approximate climax vegetation, that managed or Indiangrass are decreasers, buffalograss and tall tame pastures cannot be evaluated according to dropseed are increasers, and silver bluestem, this method, and that the process tends to focus tumble windmillgrass, and tumblegrass are in- on plants and livestock and not on the broader vaders. If similar data were collected in the short- array of organisms that might be included in a grass plains, buffalograss and blue grama would complete and sustainable grassland. There is be classified as decreasers. Similarly, because of ample discussion of these criticisms (Lauenroth differences in the potential or climax vegetation, and Laycock 1989), but for this discussion of in the mixed-grass plains between the tallgrass sustainability, the concept remains important. Its prairie and the western shortgrass plains, sideoats importance rests on the assumption that climax grama would be a decreaser species. vegetation is self-perpetuating and as such is Despite these criticisms, this procedure for sustainable. evaluating the condition of a range should be Range condition is estimated on the percent- included as a measure of sustainability of grass- ageofherbaceouscovercontributedbydecreasers land ecosystems. The species to be used in the (species that are palatable and preferentially analysis are well known for most classes of site grazed by livestock), increasers (species that are and are certainly known for most regions of the 311 Defining and Measuring Sustainability: The Biogeophysical Foundations Table 19-1: Composition of Lightly and Heavily Grazed Areas in the Osage Hills in Northern Oklahoma and Eastern Kansas (percentage) Species Lightly grazed Heavily grazed Little bluestem (Schizachyrium scoparius) 74.1 6.1 Big bluestem (Andropogon gerardi) 13.5 6.1 Indiangrass (Sorghastrum nutans) 4.8 Switchgrass (Panicum virgatum) 1.6 Buffalograss (Buchloe dactyloides) 1.6 21.6 Tall dropseed (Sporobolus asper) 0.4 23.4 Silver bluestem (Andropogon saccharoides) - 12.9 Sideoats grama (Bouteloua curtipendula) 1.6 4.0 Scribner panicum (Panicum scribnerianum) 0.4 2.6 Blue grama (Bouteloua gracilis) 0.4 2.9 Purple lovegrass (Eragrostis spectabilis) 0.4 Fringeleaf paspalum (Paspalum ciliatifolium) 0.4 2.0 Tumble windmillgrass (Chloris verticillata) - 6.1 Tumblegrass (Schedonnardus paniculatus) - 0.9 Sedges Carex spp. 0.4 3.0 Other species 0.4 8.4 -Not available. Source: Hazel 1967. United States and in many parts of the world. others 1981). In general, within natural grass- Thus, the first of the five proposed sustainability lands, lowland sites produce more herbage than indicators for grassland ecosystems is that the upland sites; deep soils produce more plant ma- rangecondition be classified within the excellent terial than shallow sites; and with adequate to good categories. With the composition of the amounts of available water in the soil, the burn- species of vegetation characterized as excellent to ing of tallgrass prairie every few years increases good, there is a strong expectation that the grass- the aboveground production (Abrams and land will maintain itself even if the prevailing Hulbert 1987; Abrams and others 1986). weather conditions are modified somewhat. As Since production and loss of plant material discussed below, this first indicator will be omit- occur throughout the growing season, the mea- ted whenconsideringgrassland ecosystemsdomi- surement of biomass at any one time underesti- nated by species other than native ones or when mates the total amount of biomass produced the species composition itself is being managed. throughout the season. For example, in northern Oklahoma in a rangeland dominated by warm-season species, measuring only the peak Aboveground primary production standing crop of the tallgrass prairie probably underestimates the total annual production by Aboveground production of herbage is a func- about 20 percent. Some grasslands have signifi- tion of several conditions, including the actual cant cool-season and warm-season species com- amount of material produced by the plants, the ponents,whichproducemaximumbiomassearly amount that is trampled and lost, the amount of and late in the growing season, respectively. In regrowthaftergrazingand trampling,changesin these cases, a single estimate made at one time in species composition, whether the grassland has the year will underestimate the total annual pro- been burned, and the status of various abiotic duction of herbage even more. variables such as soil depth and soil water In grassland ecosystems, the ratio of (Abrams, Knapp, and Hulbert 1986; Risser and aboveground to belowground biomass ranges 312 Indicators of Grassland Sustainability: A First Approximation from about 1:1 inmesic grasslands to 1:6orlessin Table 19-2: Summary of Peak Standing Crop drygrasslands(Risserandothersl981).Thus,the Values from Several Tallgrass Prairies in the aboveground biomass usually represents less than Central United States half the total biomass in the tallgrass prairie and (grams per square meter) far less than half in many drier grasslands. How- ever, measuring belowground biomass accu- State Peakstandingcrop rately has proven to be difficult and extremely Illinois 302-489 time-consuming (Kucera, Dahlman, and Illinois 328 Koelling 1967). Studies demonstrate that heavy Illinois 280 grazing apparently reduces belowground pro- duction (Weaver 1950), but insufficient data Iowa 364 exist to draw firm conclusions about the rela- Iowa 369 tive reductions that might be expected from Iowa 390 different levels of grazing intensity and from Kansas 180 differences in weather conditions (Hayes and Kansas 325-473 Seastedt 1987). Peak standing crop has been measured from a Michigan 238 number of grasslands. Table 19-2 summarizes Minnesota 447 data from twenty-three sites (see Risser and oth- ers 1981, p. 160, for the original citations). Abrams, Missouri 544 Knapp, and Hulbert (1986) measured Missouri 508 aboveground biomass from several sites in the Missouri 482-570 Kansas tallgrass prairie over a ten-year period Nebraska 344-432 and obtained similar values. Averaged across upland and lowland sites, mid-season live biom- North Dakota 456 ass was 422 grams per square meter on annually North Dakota 430 burned and 364 grams per square meter on un- burned sites for the ten years studied. During Oklahoma 316 1980, which was extremely dry, the comparable Oklahoma 414 measurements were 185 on the lowland site and Oklahoma 348 299 on the upland site. Although there is consid- Oklahoma 592 erable inter-yearvariability within measurements Oklahoma 254335 of peak standing crop, Risser and others (1981) found the variability in this ecosystem compo- South Dakota 500-566 nent far less than other common measures of ecosystem processes. Source: Risser and others 1981, p. 160, for the complete In summarizing aboveground biomass mea- surements of grasslands, the following points Aboveground biomass values demonstrate should be recognized: some variability among years because of * Standing biomass is only an estimate of total different patterns of soil water, temperature, primary production and does not account for and other variables, although plant biomass the material that has been consumed during values are less variable than virtually all other the year. ecosystem processes. * Belowground biomass is usually greater than Tallgrass prairies have peak standing crop aboveground biomass but is difficult to mea- values of about 400 grams per square meter. In sure, and there are insufficient data from which drought conditions, the biomass measurements to define specific relationships between graz- may decrease to below 200 grams per square ing inlensity and belowground productivity. meter, but over periods of nominal changes in * The amount of biomass measured at the time climate conditions, the peak standing crop value of peak standing crop underestimates the total remains above 300. Aboveground biomass does aboveground biomass produced throughout not respond to some types of disturbance the year because not all species produce the (Schaeffer and others 1990), but it does repre- greatest production at the same time. sent an important estimate of ecosystem status, 313 Defining and Measuring Sustainability: The Biogeophysical Foundations especially with respect to changes in climate. Species diversity Therefore, as a first approximation, the second sustainability predictor for tallgrass prairie eco- Although numerous studies describe the plant systems is a seasonal peak standing crop of 300 and animal species composition of tallgrass prai- grams per square meter or greater. This single ries (Abrams and Hulbert 1987, see Risser and predictor, however, can be made more accurate others 1981 for citations), relatively few have for grasslands in the north central Great Plains calculated species diversity. In those cases where and thesoutheasternportionsoftheUnitedStates such calculations have been made, the actual by referring to figure 19-1, taken from Parton and numerical terms have not always been consistent. others (1987). The criterion for any site should be Also, species diversity measurements are end not less than 100 grams per square meter below points, as discussed previously, and it is difficult the average value for the site being considered to provide biological interpretations to the nu- throughout the region, as depicted on this map of merical values. peak standing crops. Although species diversity in grasslands is expected to decrease with severe disturbances, in many cases less severe disturbances result in higher species diversity (Coffin and Lauenroth 1988). This pattern is true in the tallgrass prairie Figure 19-1: Regional Distribution of Peak where light to moderate grazing pressure in- Aboveground Standing Crop creases speciesdiversity. Collins (1987) measured plant species diversity from an Oklahoma tallgrass Above ground plant production prairie that had received different burning and (grams per square meter) grazing treatments. Diversity was calculated on the basis of relative plant cover (pi), H= - sum (pi In p,), and evenness as E = (N2- 1) / (N1 - 1), where N2 = exp(I-) and N, = I / sum p12. As shown in table 19-3, plant species diversity ranged from 9.1 in the grazed and burned grass- land to 5.0 in the ungrazed and burned grassland; ungrazed and unburned was 8.0, and grazed but the not burned was 8.5. The same values for evenness were 0.5, 0.5, 0.6, and 0.6, respectively. A . ,y1 * t ^ t \ In these and other data, the evenness component 00 t e I Qoodoes not appear to be particularly sensitive in grassland ecosystems. Because of the uncertainty of interpreting in- dexes of species diversity, it is tempting to elimi- nate any measure of species diversity from con- siderationof indicatorsof grasslandsustainability. _ \ * * * Y J However, since the final suite of indicators is small, and because the proposed measurements do not includeanydirect measurement of ecosys- A 9 s ,<500 tem complexity, plant species diversity is in- \\ 400 cluded. This term recognizes that the food and < 300 habitat requirements of many invertebrate and vertebrate species involve different and in some cases several plant species and communities. It Source: Parton and others 1987. also recognizes that some levels of disturbance actually increase plant species diversity; how- ever, severe disturbance not only reduces species diversity but, because of the poverty of species, may restrict the ways in which the grassland can recovcr from disturbances or other stresses such as change in climate. 314 Indicators of Grassland Sustainability: A First Approximation Table 19-3: Average Plant Species Diversity and Evenness Values from Oklahoma Tallgrass Prairie Subjected to Burning and Grazing Treatments Treatment Diversity Evenness No burning No grazing 8.0 0.6 Grazing 8.5 0.6 Burning No grazing 5.0 0.5 Grazing 9.1 0.5 Source: Collins 1987. Therecommended indicatorof sustainabilityfor land ecosystem sustainability. However, since tallgrass prairie ecosystems is that the plant species the organic matter content is related to plant diversity Iexp(H')1 must not be less than 5.0. Since productivity, which varies across the region of evenness is not particularly helpful, it is not in- the tallgrass prairie and the central United States, cluded in the recommendation, although the term and because the organic matter is related to the itself can be calculated from the same data set. soil texture, the specificcriterion should be modi- fied according to the location of the grassland. Therefore, the average soil organic matter Soil organic carbon sustainability indicator for the tallgrass prairie is 3.0 kilograms per square meter of carbon in the Organic material accumulates in the soil from top 20 centimeters of the soil for sandy soils and plant production and the activities of soil organ- 5.0 kilograms per square meter of carbon in isms. Thus, organic matter is usually highest in silt-loam soils. However, this indicator should be the soil nearest the surface and decreases as depth modified to match figure 19-2, with the indicator increases (Risser and others 1981). With reduced value being no less than 500 grams less than the plant production, thereisadecreascin theamount predicted carbon content at the site in question. of organic material incorporated into the soil, either from decaying plant roots or from aboveground sources brought belowground by Nitrogen content of the vegetation soil-dwelling invertebrates or vertebrates. Simi- larly, if the top horizons of the soil profile are lost The nitrogen content of the vegetation is an indi- by erosion, then the soil organic matter in the soil catorof the nutritional statusof the plants(Harper, decreases. Finally, in the tallgrass prairie region, Daniel, and Murphy 1933; Spedding 1971). Since sandy soils generally produce less plant produc- plants are the source of nutrition to herbivores, tion than soils with more silt and clay materials. plant nitrogen content is a measure of forage As a result, sandy soils have a lower content of quality. Also, the amount of nitrogen, and the organic material than heavier soils (Schimel, ratio between carbon and nitrogen in the plant Coleman, and Horton 1985). material, affects the rate at which the plant mate- Soil organic material plays several important rial decomposes in the soil (Pastor, Stillwell, and roles in grassland ecosystems. Specifically, it con- Tilman 1987). Therefore, nitrogen content of the tains nutrients that are released on decomposition, plant material is important in determining not thereby becoming available for plants and animals. only the nutritional status of the plants them- Inaddition,itactsasasourceofcarbonformanysoil selves (Jaramillo and Detling 1988) but also the organisms. Also, soil organic material contributes ways in which plants influence organisms that to the water-holding capacity and also to the desir- consume grass (including livestock) and the rates able structural characteristics of the soil. of carbon and nutrient cycling in the soils. Because of its importance, soil organic matter Nitrogen status in the grassland ecosystem is is included as one of the five indicators of grass- controlled by several variables, including graz- 315 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 19-2: Regional Distribution of Soil Carbon ing, buming, and nutrient statusof the soils (Hobbs and others 1991; Parton and Risser 1980). In gen- (kilograms per square meter) eral, when grasslands are burned, a significant (Sandy) portion of the nitrogen in the vegetation is volatil- ized and lost to the ecosystem. Ungrazed grass- 3_5s___ lands frequently have higher concentrations of nitrogen in vegetation than those that are grazed. So, buming ungrazed or lightly grazed grass- lands may result in more loss of nitrogen than 2.5 7 \ \ /\ burning more heavily grazed grasslands (Hobbs 1 _\ + \\4.0 and others 1991). Although several of these variables affect the _ 2.0 7 \ A 9 \ > nitrogen content of vegetation, the nitrogen con- tent of tallgrass herbage is well known because it is frequently used as a measure of forage quality (Risser and others 1981; Spedding 1971). Over- grazing results in a decrease in nitrogen content of the vegetation based in part on changes in speciescomposition.Also,decreased nutrient sta- o.s \ _ ) \ -\>tl tus of the soil results in lower nitrogen concentra- tion in the vegetation. Since maintaining forage quality is important for the plants themselves, for consumers, and for nutrient cycling, nitrogen \ .0\ <3 o content of the vegetation forms the last of the five essential indicators of sustainability of a grass- J1.0 land ecosystem. From the available data, it is possible to propose that total nitrogen in the Soil c aboveground herbage must be maintained at a (kilograms per square meter) minimum of 0.6 gram per 100 grams of dry biom- (Fine) ass (Harper, Daniel, and Murphy 1933; Hobbs and others 1991; Spedding 1971; and other refer- 4.5 J ences cited in Risser and others 1981). Discussion 4.0 \ A. An almost infinitenumber of biophysical character- istics could be proposed as potential indicators of grassland ecosystem sustainability. However, as noted at the beginning of this chapter, the selected indicators should be small in number and relatively easy to measure, should reflect useful characteris- 3.5 | | ) - ytics of the ecosystem as perceived by human values, 3 0 \ tv ( t / and should be quantitatively conservative so as to accommodate changes in driving variables such as changing climate. Table 194 describes the five proposed biophysical measures for determining sustainability of a grassland ecosystem. In other words, if these five conditions are met, the essen- tial grassland properties should persist indefi- .0 na onitely regardless of the use made of the grassland. The test of this proposal will require long-term measurements under many different types of Source: Schirnel, Coleman, and Horton 1985. rsln aaeetadue grassland management and us3. 316 Indicators of Grassland Sustainability: A First Approximation Table 19-4: Proposed Biophysical Measures for Tallgrass Ecosystem Sustainability Indicator Measure Range condition rating Good to excellent Peak standing crop of vegetation More than 300 grams per square meter Plant species diversity [exp (H')] More than 5.0 Soil organic material in top 20 centimeters of soil Sandy soils More than 3.0 kilograms per square meter Clay-loam soils More than 5.0 kilograms per square meter Nitrogen content in vegetation More than 0.6% dry weight basis Source: Collins 1987. The numerical values proposed in table 19-4 species, characteristics of the entire range. There are based on the tallgrass prairie of the central are data that could be used to estimate the habitat United States, although figures 19-1 and 19-2 required to support a native avifauna (Graber demonstrate how these quantitative terms can be and Graber 1963; Wiens 1974) and small and large modified for conditions across the region. Similar mammal fauna (Coppock and others 1983; Grant values need to be developed for grassland eco- and Birney 1979; McNaughton, Ruess, and Seagle systems throughout the world; however, because 1988). Thus, if additional indicators were added thecentral UnitedStatesincludesa rangeofenvi- to the five currently proposed, the most likely ronments that encompasses conditions found in candidate would be a minimum size required to most grasslands worldwide, only small modifi- maintain a representative complement of bird cations should be required to make the proposed species. indicators useful elsewhere. The focus of this discussion has been on grass- Since grassland ecosystems include both ani- lands that are used for their native species. In the malsand microorganisms,itcanbeasked whether management of tame grasslands or highly modi- the proposed plant and soil indicators adequately fied grasslands, the species composition is man- represent these essential components. Both range aged, in some cases for just a single plant species, condition and plant species diversity indicate such as bermudagrass (Cynodon dactylon; that a variety of food materials is available for McMurphy and Tucker 1975). Under these condi- consumers and decomposers; these two indica- tions, the first and third indicators (range condi- tors and the production of herbage indicate the tion and plant species diversity) are not appli- quality of cover for animals; soil organic material cable. However, the other three criteria-produc- and nitrogen content of plant material indicate tion of herbage, amount of soil organic material, both the quality of food materials aboveground and concentration of nitrogen in the plant tis- and belowground as well as the potential rates of sue-are important, and perhaps even adequate, decomposition and nutrient cycling; and finally, indicators of grassland ecosystem sustainability. the content of organic material in the soil and concentration of nitrogen in tissue are indirect measures of the ability of the ecosystem to retain References nutrients. If these five indicators are sufficient to capture the essence of the entire ecosystem Abrams, M. D., and L. C. Hulbert. 1987. "Effect of sustainability, the field measurements will be Topographic Position and Fire on Species more easily accomplished than if animal census Composition in Tallgrass Prairie in North- or soil sampling are required. east Kansas." American Midland Naturalist One component of grassland sustainability has 117, pp. 442-45. been omitted from this analysis, namely, the up- Abrams, M. D., A. K. Knapp, and L. C. Hulbert. per trophic levels of birds and mammals. If these 1986. "A Ten-year Record of Aboveground two components were included, the indicators Biomass in a Kansas Tallgrass Prairie: Effects would include the size of grasslands, heteroge- of Fire and Topographic Position." American neity of habitat, and, in the case of migratory Journal of Botany 73, pp. 1509-15. 317 Defining and Measuring Sustainability: The Biogeophysical Foundations Cairns, J., Jr. 1991. "Environmental Auditing for Karr, J. R., K. D. Fausch, P. L. Angermeier, P. R. Global Effects." Environmental Auditor 2, pp. Yant, and 1. J. Schlosser. 1986. "Assessing 187-95. Biological Integrity in Running Waters: A Coffin, D. P., and W. K. Lauenroth. 1988. "The Method and Its Rationale." Special Publica- Effectsof Disturbance Size and Frequency on a tion 5. Illinois Natural History Survey, Shortgrass Plant Community." Ecology 69, pp. Champaign, Ill. 1609-17. Korstian, C. F. 1919. "Native Vegetation as a Collins,S. L. 1987. "Interaction of Disturbancesin Criterion of Site." Plant World 22, pp. 253-61. Tallgrass Prairie: A Field Experiment." Ecology Kucera, C. L., R. C. Dahlman, and M. R. Koelling. 68, pp. 1243-50. 1967. "Total Net Productivity and Turnover on Coppock, D. L., J. K. Detling, J. E. Ellis, and M. an Energy Basis for Tallgrass Prairie." Ecology 1. Dyer. 1983. "Plant-herbivore Interactions 48, pp. 536-41. in a North American Mixed-grass Prairie. ll. Lauenroth, W. K., and W. A. Laycock, eds. 1989. Responses of Bison to Modification of Veg- SecondarySuccessionand theEvaluationofRange- etation by Prairie Dogs." Oecologia (Berlin) landCondition.Boulder,Colo.:WestviewPress. 56, pp. 10-15. McMurphy, W. E., and B. B. Tucker. 1975. "Mid- Dyksterhuis, E. J. 1949. "Condition and Manage- land Bermudagrass Pasture Research." Okla- ment of Rangeland Based on Quantitative Ecol- homa Agricultural Station ProgressReport 715, ogy."JournalofRangeManagernent2,pp.104-15. pp. 14-20. Oklahoma State University, Graber, R. R., and J. W. Graber. 1963. "A Com- Stillwater. parative Study of Bird Populations in Illinois, McNaughton, S. J., R. W. Ruess, and S. W. Seagle. 1906-1909 and 1956-1958." Illinois Nat ural His- 1988. "Large Mammals and Process Dynamics tory Survey Bulletin 28, pp. 383-528. in African Ecosystems." BioScience 38, pp. 794- Grant, W. E., and E. C. Birney. 1979. "Small 800. Mammal Community Structure in North Parton, W. J., and P. G. Risser. 1980. "Impact of American Grasslands." Journal of Mammal- Management Practices on the Tallgrass Prai- ogy 60, pp. 23-36. rie." Oecologia 46, pp. 223-34. Harper, H. J., H. A. Daniel, and H. F. Murphy. Parton, W. J., D. S. Schimel, C. V. Cole, and D. S. 1933. "The Total Nitrogen, Phosphorous, and Ojima. 1987. "Analysis of Factors Controlling Calcium Content of Common Weeds and Na- Soil Organic Matter Levels in Great Plains tive Grasses of Oklahoma." Proceedings of the Grasslands." Journal of the Soil Science Society of Oklahona Academy of Science 14, pp. 36-44. America 51, pp. 1173-79. Hayes, D. C., and T. R. Seastedt. 1987. "Root Pastor, J., M. A. Stillwell, and D. Tilman. 1987. Dynamics of Taligrass Prairie in Wet and Dry "Little Bluestem Dynamics in Minnesota Old Years." Canadian Journal of Botany65, pp. 787-91. Fields." Oecologia 72, pp. 327-30. Hazel, D. B. 1967. "Effect of Grazing Intensity on Pendleton, D. T. 1989. "Range Condition as Used Plant Composition, Vigor, and Production." in the Soil Conservation Service." In W. K. Journal of Range Management 20, pp. 249-53. Lauenroth and W. A. Laycock, eds., Secondary Hobbs, N. T., D. S. Schimel, C. E. Owensby, and Succession and the Evaluation of Rangeland Con- D. S. Ojima. 1991. "Fire and Grazing in the dition, pp. 17-34. Boulder, Colo.: Westview Tallgrass Prairie: Contingent Effects on Nitro- Press. gen Budgets." Ecology 72, pp. 1374-82. Renner, F. G.,and B. W. Allred. 1962. "Classifying Jaramillo, J. V., and J. K. Detling. 1988. "Grazing Rangeland for Conservation Planning." U.S. History, Defoliation, and Competition: Effects Department of Agriculture Handbook 253. on Shortgrass Production and Nitrogen Accu- Washington, D.C. mulation." Ecology 69, pp. 1599-1608. Risser, P. G. 1989. "Range Condition Analysis: Jordan, W. R., M. E. Gilpin, and J. D. Aber, eds. Past, Present, and Future." In W. K. Lauenroth 1987. Restoration Ecology: A Synthetic Approach and W. A. Laycock, eds., Secondary Succession to Ecological Research. Cambridge, England: and the Evaluation of Rangeland Condition, pp. Cambridge University Press. 143-56. Boulder, Colo.: Westview Press. 318 Indicators of Grassland Sustainability: A First Approximation Risser, P. G., E. C. Birney, H. D. Blocker, S. W. Rangeland and Cropland Toposequences in May, W. J. Parton, and J. A. Wiens. 1981. The North Dakota." Geoderma 36, pp. 201-14. True Prairie Ecosystem. Stroudsburgh, Penn.: Smith, D. A. 1988. Mesa Verde National Park Shad- Hutchinson Ross Publishing Company. ows of the Centuries. Lawrence: University of Schaeffer, D. J., E. E. Herricks, and H. W. Kerster. Kansas Press. 1988. "EcosystemHealth. l. MeasuringEcosys- Spedding, C. R. W. 1971. Grassland Ecology. Ox- tem Health. Environmental Audit. VI." Envi- ford, England: Clarendon Press. ronmental Management 12, pp. 445-55. Weaver,J. E. 1950. "Effectsof Differentlntensities Schaeffer, D. J.,J. A. Perry, H. W. Kester, and D. K. of Grazing on Depth and Quantity of Roots or Cox. 1988. "The Environmental Audit. 1. Con- Grasses." Journal of Range Management 2, pp. cepts." EnvironmentalManagement 9, pp.191-98. 100-13. Schaeffer, D. J., T. R. Seastedt, D. J. Gibson, D. C. Werner, P. A. 1990. "Principles of Restoration Hartnett, B. A. D. Hetrick, S. W. James, D. W. Ecology Relevant to Degraded Rangelands." Kaufmann, A. P. Schwab, E. E. Herricks, and E. Australian Rangeland Journal 12, pp. 34-39. W. Novak. 1990. "Field Bioassessments for Wiens, J. A. 1974. "Habitat Heterogeneity and Selecting Test Systems to Evaluate Military Avian Community Structure in North Ameri- Training Lands inTaligrass Prairie. Ecosystem can Grasslands." American Midland Naturalist Health. V." Environmental Management 14, pp. 91, pp. 195-213. 81-93. Schimel, D. S., D. C. Coleman, and K. A. Horton. Whyte, R. 0. 1960. Production and Environment. 1985. "Soil Organic Matter Dynamics in Paired London: Faber and Faber. 319 Sustainability in Tropical Inland Fisheries: The Manager's Dilemma and a Proposed Solution Peter B. Bayley The artisanal, multispecies fisheries typical in tropical lakes and river floodplains present unique problems forthe manager. The manager's dilemma is to try to optimize yields fora few large, individual species or increase the total yield at the expense of some species that are more valuable. Manyfisheries also require the resolution of conflicts between urban-based commercial operations and local, rural demands. Traditional management options are limited because of the high cost and complexity of enforcing regulations such as restricted types of gear, minimum size of fish, closed seasons, or limited entry. A progressive pulsefishing paradigm is proposed to aim for a sustainable multispecies yield consistent with local socioeconomic realities and persistence of species. This scheme permits high yields with compositions of diverse species near cities or major ports combined with increased control of exploitation through periodically closed areas as distancefrom markets increases. Increasing control lowers yields but permits optimal harvestingfor larger, more valuable species. A long-term experimental management approach is essential, so that different levels of sustainability can be monitored at different points in the gradient of fishing restrictions. Other vital roles of thefishery manager are stressed, including the publicizing of the value of commercialand subsistencefisheries and, with thecooperation of otheragencies, the maintenance or restoration of critical support factors (hydrology, water quality, higher vegetation) in the drainage basin. The 1989 FAO Yearbook indicates that of the 99.5 appropriate forapproachinglevels of sustainable million tons of nominal fish yield worldwide, 13.8 yields in extensive, multispeciesfisheries. Broader million tons (14 percent) were from freshwaters, issues of sustainability are addressed for these which in turn mostly originated from capture fish- and other fisheries by Regier, Bocking, and eries in river floodplains and lakes in tropical Asia, Henderson (chapter 22 of this volume). Africa, and Latin America (FAO 1989). In the past two decades, the published litera- ture on freshwater tropical ecology has blossomed Approaches to predicting yield (see, for example, Lowe-McConnell 1975, 1987; Welcomme 1979, 1985), but the same cannot be Traditional fish population models explicitly or said of publishable information on fish yields, implicitly incorporate birth, death, recruitment, fishing intensity, and socioeconomic factors. In and growth rates on a stock of fish and can be view of these data limitations and the effects that used to predict yields given a constant environ- rapidly increasing levels of human population ment (Beverton and Holt 1957; Graham 1935; have on exploitation rates, this chapter identifies Gulland 1969; Ricker 1975; Schaefer 1954). Data the problems facing the managers of tropical requirements are considerable even for the fisheries and provides a management paradigm single-species temperate fisheries for which they Defining and Measuring Sustainability: The Biogeophysical Foundations were developed. Extensions of such models to perficially resembles the Graham-Schaefer multispecies fisheries are hampered by numer- logistic-derivedparabola forasingle-speciesfish- ous theoretical and practical problems that will ery (see figure 20-1; Schaefer 1954). However, the not be solved in a timely manner for most, if any, generalized effort value appropriate for compar- tropical freshwater systems (Pauly and Murphy ing fisheries (number of fishermen per unit of 1982). area) changes qualitatively at higher intensities, Studies that link production-related factors involving changes in gearand reductions in mesh with fishery yield can be useful to obtain first- size more appropriate for catching smaller spe- order estimates of yield (Henderson and cies. Also, the drop in multispecies yield at high Welcomme 1974; Melack 1976; Schlesinger and effort values reflects as much the technical or Regier 1982; Welcomme 1974, 1976) or to under- economic lirnitations of harvesting smaller but stand the environmental factors that support the more productive fish or invertebrates as it does fishery (Bayley 1989) but have limited predictive the overexploitation of some species. power (despite significant correlation coeffi- Many managers and conservationists worry cients). that overexploitation can cause biological extinc- Two empirical approaches offer more hope. tion, and commercial extinction is often incor- First, comparative models based on ecologically rectly reported to imply biological extinction. and socioeconomically similar fisheries that ac- There is no evidence that an intensified fishery in count for fishing effort improve our ability to a system unaltered by anthropogenic environ- predict and assess thecurrent state of multispecies mental change (of hydrology, water quality, or yield (Bayley 1988; Henderson and Welcomme species introduction) has caused the biological 1974). However, these are limited by the avail- extinction of a fish species. As a more extreme ability of data on yield and effort and the lack of example, even with extensive degradation of the additional explanatory variables, such as those hydrologic regime and pollution in addition to describing the nature of the flood regime in large exploitation, no fish species in the Upper Missis- rivers(Junk,Bayley,andSparksl989;Welcomme sippi River has become extinct (Fremling and 1985). Second, yield predictions based on the others 1989). However, it is conceivable that an influence on recruitment, and hence future yield, intensive fishery on an isolated species in a small of a key environmental variable such as flood stage have shown promise in river floodplains where time seriesare available(Welcomme 1985). Figure 20-1: Generalized Multispecies Yield and Such models would be much improved if fishing Effort Curve from a Single Fishery Based on a effort and the effect of water level on catchability Model of Lake and River Floodplain Fisheries were known. Results from these comparative models and descriptions of trends in the species composition of multispecies fisheries are currently the best points of departure to define, albeit approxi- mately, what the manager can do and avoid in order to sustain a tropical fishery. The multispecies fishery and the manager's dilemma Fishing mtensity and change in gear .. - Mean fish size The diverse fish assemblages of tropical systems are reflected in most fishery yields, even though Note: The model predicted yields with 95 percent a few large species are typically the most valu- confidence ranges of 114-188 kilograms per hectare per vear frffen rver floodplains (on the basis of maxmumur able. The relationship between multispecies yield flooded area) and 76-122 kilograms per hectare per vear for and fishing effort follows a unimodal relation- thirty-one lakes. As effort (number of fishermen pei rut of area) increases, methods change to exploit smaller, but ship, such as the curves developed for tropical usually more productive, species. Small curves indicate lakes, river floodplains, and coastal lagoons hypothesized single-species yield curves. (Bayley 1988). The multispecies yield curve su- Source: Bayley 1988. 322 Sustainability in Tropical Inland Fisheries: The Manager's Dilemma and a Proposed Solution pond or stream could result in local extinction. ship among some fishermen, and can deprive the Therefore, the following arguments refer to large human population of a sustained supply of pro- lakes or river floodplains, which supply the bulk tein. Therefore, in some fisheries it is fortunate of the worldwide yield of tropical inland fish. I that the regulationscannotbe enforced. Inothers, also assume that the multispecies curve approxi- some form of practical regulation is required. mates an equilibrium yield for a given fishing The dilemma is made more acute when differ- intensity, which has not yet been proven. ent socioeconomic interests are exploiting the Ineconomicterms,themultispeciesyieldcurve same fishery resource, as exemplified by com- corresponds to the Graham-Schaefer curve. Yield mercial operations supplying cities versus local (income) increases with increasing investment or fishermen satisfying subsistence requirements cost (effort) until the maximum sustainable yield and local markets. For example, in the Amazon is reached, but profitability (yield / effort = slope basin, migration from rural areas to cities has not from origin to any point on thecurve in figure 20-1) decreased economic dependency on fish as the for individual fishermen usually decreases with major source of animal protein but has resulted in increasing effort. Therefore, a fishery may be- a more capital-intensive system for preserving come unprofitable at any point on the curve. and marketing fish (Bayley and Petrere 1989). Within the range of profitability lie consider- Although such fishermen travel hundreds of ki- ations of employment and optimizing the supply lometers for valuable fish species, most fish are of animal protein to the human population, both caught in a smaller radius around urban areas important factors in the socioeconomic circum- and comprise smaller, cheaper species that the stances of developing countries (Pauly, Silvestre, majority of city dwellers can afford (Petrere 1978, and Smith 1989). 1985). Therefore, fisheries near cities usuallyyield Therefore, the manager of a multispecies fish- more species and provide a higher yield than ery on a large lake or river floodplain has a range those far away. Conversely, rural, part-time fish- of optionson the left side of the mul tispecies yield ermen are accustomed to higher catches per unit curve. Significant departures from this curve of effort, corresponding to the higher biomasses through theoptimal harvesting of individual spe- of lesser exploited areas (lower left limb in figure cies (summing the maximum yields of individual 20-1). Rural fishermen resent a competing fishery species shown in figure 20-1 to exceed the pre- that requires them to spend more time fishing or dicted multispecies yield)are severelyconstrained to change to other fish species. by interactions between fishing gear and various species and by ecological interactions, of which predation is particularly prevalent in the tropics A proposed solution: Progressive pulse (Lowe-McConnell 1975). fishing The manager's dilemma is to try to optimize yields fora few large, individual species (left end In large systems, it is possible to accommodate of figure 20-1), thus making the fishery more the interests of all groups of fishermen and con- profitable, but with less production of protein, or sumers to some extent, if managers recognize that to increase the total yield at the expense of some enforceable regulations should be adapted to so- species that are more valuable, but not necessar- cioeconomic realities in different parts of the sys- ily profitable. It is difficult for a manager to admit tem and if they accept that no single sustainable that stocks of some species are being overex- yield or species composition is optimum (Bayley ploited, that is, that increased effort is reducing and others 1992). An approach, termed here pro- the yield of those species, even though the gressive pulse fishing, is proposed for large, dis- multispecies yield is well below the maximum persed, multispecies fisheriesin which total yield indicatecl by similar fisheries elsewhere. At the per unit of area is allowed to remain relatively same time, it has proven to be economically, high near cities and is reduced by a combination socially, and technically impossible to manage a of inherently higher fishing costs and regulation multispecies fishery for a few large species under to optimize exploitation of progressively larger the conditions of most developing countries. andmorevaluablespeciesasdistancefrommajor Therefore, the typical manager implements regu- markets increases (see figure 20-2). Therefore, lations (close seasons, minimum sizes for mesh higher yields based on more productive, cheaper and fish) that are difficult to enforce, cause hard- fish can be maintained profitably to supply urban 323 Defining and Measuring Sustainability: The Biogeophysical Foundations areas because the fisheries are closer to market. be exploitable somewhere. Many managers may Conversely, lower yields of larger, high-valued find the proposed increase in restrictions farther species can be exploited profitably at greater dis- from urban centers counterintuitive,because they tance, and through more extensive, periodic area are based in urban centers and are influenced by closures such species will be conserved along local information and fishermen. with the rest of the fish community at a higher Because the spatial distributions, dynamics, biomass.Exceptfornationalparksandareasclose interactions, sustainability, and other ecological to urban centers, valuable species would always aspects of fish stocks are poorly known, the de- gree and size of area closures should be treated as a long-term management experiment, with a com- Figure 20-2: Progressive Pulse Fishing Paradigm parable amount of money spent on data collec- Proposed for Large River Floodplains and Some Lakes tion and analysis as on enforcement. Frequency of closure should initially be at least twice the loo longest generation among the key species. Equi- librium levels (or approximations thereof) of multispecies yields and their compositions should be determined by monitoring their resilience following changes in fishing effort. :0R\ /Vrlous Gear regulations are impossible in dispersed pulsefishng fisheries (including many large lakes as well as " 0 \options rivers), and closed seasons are costly to enforce Y @ Ł \ because they attempt to prohibit fishing every- > | 6k, \ where. Neither approach guarantees the stocks against rapid depletion throughrecruitmentover- o ' fishing. However, pulse fishing-in which easily Park Reserve City controlled zones are periodically rested (except 4-- DLstance from major market for local subsistence fishing) to allow stocks to recover-is a less-efficient but more feasible and safer approach. Specialized commercial opera- tions that have to disembark at a limited number of ports, such as on some large lakes, may be amenable to regulations on gear, closed seasons, or catch quotas. Limi ted entry is rarely a practical option, except on the small scale of communal fishing property rights. Discussion Two elements of fishery management are implied but rarely stressed sufficiently: quality of data and stewardship of the environment. Poor qual- ity of data is almost universal among artisanal ___________________________________ ,fisheries and is also problematic in lakes with Fishing intensity + chiange in gear less-dispersed fisheries (Bailey 1992). Many man- agers and some scientists do not even recognize that their data may omit a large proportion of the Note: Percentage of the area protected from pulse fishing theldattn ositence ort mar- versus distance from fish market (upper graph) is compared e subsstence or locally mar- with the multispecies yelid versus fishmg intensity (lower keted fish. For example, 60 percent of Amazon graph). The multispec6es yield is shown as a broad lme to yield is from subsistence or local markets which emphasize the uncertainty in sustainabilitv until more data , are generated by the expenmental management process. cannot be effectively monitored using traditional The graphs are aligned to mnoicate that fishing mtensity (and survey techniques (Bayley and Petrere 1989). The self-re ulated chan es mi gear) should usually be inverselv related to distance hom market. official statistics on landings forwarded to the 324 Sustainability in Tropical Inland Fisheries: The Manager's Dilemma and a Proposed Solution Food and Agriculture Organization from Para- Therefore, although in many socioeconomic guay in 1984 totaled 5,000 tons, whereas a conser- situations exclusive rights to exploit a given area vative estimate based largely on per capita con- of water may be appropriate, many fish popula- sumptionoflocalfishindicatedl8,000tons(Bayley tions must remain as common property with 1985).Therefore, yield and effort from most tropi- management being applied on an appropriate cal fisheries, in particular riverine ones, are un- scale. Such management, however, should take derestimated by varying degrees. Catch per ef- advantage of any communal regulatory tradi- fort obtainable from part of the fishery cannot, on tions that serve levels of sustainable exploitation its own, be used to assess the status of the fishery. appropriate for the projected human population, Also, demographic and social changes in devel- which is increasing rapidly in rural as well as oping countries result in changes in the contribu- urban areas. The long-term responsibility of man- tion of different components of the fishery. agement agencies is to obtain the confidence and Stewardship of the environment is self-evi- cooperation of all fishermen groups so that they dent but is rarely within the purview of the fish- can assist in enforcement and data collection and ery manager. Protection or restoration of the hy- actasaconstituencytodefendtheresourceagainst drology, soil, riparian or floodplain vegetation, harmful environmental change (Pinkerton 1989). and water quality of a drainage system is usually Data on yield versus effort obtained from lakes the responsibility of another agency or of no and river floodplains indicated significant (P < agency at all. Fishery managers can at least moni- 0.01) modes, implying that some fisheries were at tor and publicize the current or potential eco- or beyond a maximum multispecies yield (Bayley nomic value of the fishery, and with the help of 1988). Considering that those data are mostly ecologists, publicize the probable consequences from the 1970s and human population has typi- of alternativedevelopments. Environmental deg- cally grown 2.5 to 3.5 percent a year since then, radation occurs gradually in concert with in- morefisheriesmaynowbebeyond thosemaxima creasing human population but also can result in the sense that increasing effort under current from grandiose development plans, such as the technology and socioeconomics is decreasing proposal toconvert theRiverOrinoco-Apureinto multispecies yield. The tragedy is that the col- a shipping lane, remove the floodplains (which lection of reliable data on yield and effort has currently support 45,000 tons per year of fish deteriorated since the 1970s even though there yield),andintroduceavarietyofexoticfishknown was much room for improvement then. Also, as tilapias (PROA 1991). major changes, such as the effect of introducing The issue of part-time rural versus large com- nile perch into Lake Victoria (Reynolds and mercial fishermen introduced above has Greboval 1988), require management decisions prompted manyconservationistsand anthropolo- that are conjectural because of insufficient data gists to take the side of rural fishermen and to on fisheries. promote propertyrights legislation in riverflood- The definition of sustainability for a fishery is plains (Chapman 1989). It is essential to foster the as illusive as that for any other resource, espe- cooperation of fishermen in the management and cially for single species (see chapter 4 of this data acquisition process in order to obtain accept- volume). There is no single sustainableyield for a able rates of compliance with regulations. How- given multispecies fishery, and even optima ever, widespread granting of fishing property should vary according to socioeconomic and eco- rights at the community level would heighten the logical differences within an extensive fishery. conflict and deny essential supplies of protein to Attempting to optimize the management of an urban areas. Chapman's (1989) claim that prop- extensive tropical fishery for a few, high-valued erty rights will remove the "tragedy of the com- species may seem ecologically conservative but mons" problem associated with a common prop- eventually leads to high-cost, engineering solu- erty resource is incorrect, because she does not tions (such as rearing in large-scale hatcheries to account for the fact that most of the fish catch stabilize recruitment, which is highly variable in derives from migratory fish populationsthatcover a natural system). There is growing evidence that areas many orders of magnitude greater than such optimization increases the risk and cost of those of proposed or traditional property bound- failure, at least in the socioeconomic sense (chap- aries (Welcomme 1985). This is true of river flood- ter4 of this volume). Ironically, thecurrent lack of plains in general and also of some fish associated money and efficiency in tropical fishery manage- with lakes (Lowe-McConnell 1987). ment is an advantage in this respect. 325 Defining and Measuring Sustainability: The Biogeophysical Foundations Conversely, the use of a variety of fish spe- To summarize, the manager of a multispecies cies offers more stability, because a more di- fishery resource should realize that levels of verse, adaptable market will better accommo- sustainability are very poorly known and are date natural fluctuations in population. The probably not realistic for single species. There- main question is whether an intensified fore, the management process must allow the multispecies fishery will cause an irreversible resource to be tested and evaluated for change-beyond such changes that occur natu- sustainability while providing as much of it as rally-such as the extinction of a species. This is possible to satisfy long-term demands for protein. unlikely in the large systems that currently supply most of thecatch or have most potential, providing that the environment is maintained. References Because the environment is often changed, the proposed gradient of protection within a fishery Bailey, R. G. 1992. "Inland Fishery Resources of provides some assurance in the event of misman- Sub-Saharan Africa." In K. T. O'Grady, A. J. B. agement or natural disruption (figure 20-2). Butterworth, P. B. Spillett, and J. C. J. Domaniewski, eds., Fisheries in the Year 2000, pp. 21-28. Proceedings of the twenty-first an- Conclusions: The fishery manager's niversary conference of the Institute of Fisher- responsibilities ies Management, September 10-14,1990, Royal Holloway and Bedford New College, England The following list summarizes what the manager Nottingham, England: Institute of Fisheries of a typical multispecies fishery should do. Management. * Invest sufficient resources in regular data collec- Bayley, P. B. 1985. "Fish Resources." In Environ- tion (yield, effort, species composition, environ- mental Profile of Paraguay, pp. 147-52. Wash- mental variables) in cooperation with scientists. ington, D.C.: International Institute for Envi- ronment and Development. * Devise regulations that can be enforced (pro- -______ 198 "Accounti gressive pulse fishing in extensive systems) . 1988. "Accounting for Effort When and avoid legislation that causes widespread Comparing Tropical Fisheries in Lakes, River cheating. Hoodplains, and Lagoons. Limnologyand Ocean- * Recognize different socioeconomic circum- ography 33, pp. 963-72. stancesand different managementoptions that . 1989. "Aquatic Environments in the can aim for different levels of sustainability Amazon Basin, with an Analysis of Carbon within extensive fisheries. Sources, Fish Production, and Yield." Special publication of the Canadian Journal of Fisheries * Treat management decisions as long-term and Aquatic Sciences 106, pp. 399-408. • Bexpre of overiments.izatio . when subsidiz- Bayley, P. B., and M. Petrere Jr. 1989. "Amazon . Beware of overcapitalizationl when subsidiz- Fihre:AssmnMtod,CrntSa ing T pTmotng icresed ear r pocesing Fisheries: Assessment Methods, Current Sta- ng or promoting creased gear or processig tus, and Management Options." Special publi- capacity. cation of the Canadian Journal of Fisheries and . Avoid concentrating management resources Aquatic Sciences 106, pp. 385-98. on a few high-valued species. Bayley, P. B., P. Vazquez, F. Ghersi, P. Soini, and * Broaden the demand for different species by M. Pinedo. 1992. "Environmental Review of supporting product and market development the Pacaya-Samiria National Reserve in Peru (which is very limited in Asian and African and Assessment of Project (527-0341)." Report fisheries). for the Nature Conservancy. April. * Obtain the confidence and cooperation of all Beverton, R. J. H., and S. J. Holt. 1957. "On the fishermen groups so that they can assist in Dvnamicsof Exploited FishPopulations."Fish- enforcement and data collection. eries Investigations (United Kingdom Ministry * Publicize the importance of the fishery to the of Agriculture and Fisheries) 2:19. region and, in cooperation with other agencies Chapman, M. D. 1989. "The Political Ecology of and fishermen, the importance of conservation Fisheries in Amazonia." Environmental Conser- or restoration of the environment. vation 16, pp. 331-37. 326 Sustainability in Tropical Inland Fisheries: The Manager's Dilemma and a Proposed Solution FAO (Food and Agriculture Organization). 1989. Petrere, M., Jr. 1978. "Pesca e esforco da pesca no FAO Yearbook. Rome. Estado do Amazonas. II. Locais, aparelhos de Fremling, C. R.,J. L. Rasmussen, R. E. Sparks, S. captura e estatisticas de desembarque." Acta P. Cobb, C. F. Bryan, and T. 0. Caflin. 1989. Amazonica 8, pp. 1-54. "TheMississippiRiverFisheries:ACaseHis- -. 1985. "A pesca comercial no Rio tory." Special publication of the Canadian Solimoes-Amazonas e seus afluentes: Analise Journal of Fisheries and Aquatic Sciences 106, dos informes do pescado desembarcado no pp. 309-51. Mercado Municipal de Manaus (1976-1978)." Graham, M. 1935. "Modem Theory of Exploiting Ciencia e Cultura 37:12, pp. 1987-99. a Fishery and Application to North Sea Trawl- Pinkerton, E., ed. 1989. Co-operative Management of ing." Journal du Conseil International pour Local Fisheries: New Directions for Improved Man- l'Exploration de la Mer 10, pp. 264-74. agementandCommunityDevelopment.Vancouver, Gulland, J. A. 1969. "Manual of Methods for Fish Canada: University of British Columbia Press. Stock Assessment: Part 1. Fish Population PROA (Programa Orinoco-Apure). 1991. Analysis." FAO Manuals in Fisheries Science "Programa Orinoco-Apure." Publicaci6n Es- 4. Food and Agriculture Organization of the pecial DGSPROA/PE/01. Ministerio del United Nations, Rome. Ambiente y de los Recursos Naturales Henderson, H. F., and R. L. Welcomme. 1974. Renovables, Caracas, Venezuela. "The Relationship of Yield to Morpho-edaphic Reynolds, J. E., and D. F. Greboval. 1988. "Socio- Index and Numbers of Fishermen in African economicEffectsoftheEvolutionofNilePerch Inland Fisheries." CIFA Occasional Paper 1. Fisheries in Lake Victoria: A Review." CIFA Food and Agriculture Organization of the Technical Paper 17. Food and Agriculture Or- United Nations, Rome. ganization of the United Nations, Committee Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. for Inland Fisheries of Africa, Rome. "The Flood Pulse Concept in River Floodplain Ricker, W. E. 1975. "Computation and Interpreta- Systems." Special publication of the Canadian tion of Biological Statistics of Fish Popula- Journal of Fisheries and Aquatic Sciences 106, pp. tions." Canadian Bulletin of Fisheries and Aquatic 110-27. Sciences 191, p. 382. Lowe-McConnell, R. H. 1975. Fish Communities in Schaefer, M. B. 1954. "Some Aspects of the Dy- TropicalFreshwaters: Their Distribution, Ecology, namics of Populations Important to the Man- and Evolution. London: Longmans. agement of the Commercial Marine Fisheries." . 1987. Ecological Studies in Tropical Fish Bulletin of theInter-American Tropical Tuna Com- Communities. Cambridge Tropical Biology Se- mission 1, pp. 27-56. ries. Cambridge, England: Cambridge Uni- Schlesinger, D. A., and H. A. Regier. 1982. "Cli- versity Press. matic and Morphoedaphic Indices of Fish Melack, J. M. 1976. "Primary Productivity and Yields from Natural Lakes." Transactions of the Fish Yields in Tropical Lakes." Transactions of American Fisheries Society 111, pp. 141-50. the American Fisheries Society 105, pp. 575-80. Welcomme, R. L. 1974. "Some General and Theo- Pauly, D., and G. I. Murphy, eds. 1982. Theory and retical Considerations on the Fish Production Management of Tropical Fisheries. Proceedings of African Rivers." CIFA Occasional Paper 3. of the ICLARM/CSIRO Workshop on the Food and Agriculture Organization of the Theory and Management of Tropical Fisher- United Nations, Committee for Inland Fisher- ies, January 12-21, 1981. Manila, Philippines: ies of Africa, Rome. International Center for Living Aquatic Re- . 1976. "Some General and Theoretical sources Management; Cronulla, Australia: Considerations on the Fish Yield of African Commonwealth Scientific and Industrial Re- Rivers." Journal of Fish Biology 8, pp. 351-64. search Organization. . 1979. The Fisheries Ecology of Floodplain Pauly, D., G. Silvestre, and 1. R. Smith. 1989. "On Rivers. London: Longmans. Development,Fisheries,and Dynamite: ABrief .1985. "River Fisheries." FAO Fisheries Review of Tropical Fisheries Management." Technical Paper 262. Food and Agriculture Natural Resource Modeling 3, pp. 307-29. Organization of the United Nations, Rome. 327 Sustainable Development of Fisheries in Southeast Asia Aprilani Soegiarto The marine and coastal waters of the Southeast Southeast Asia will have to pay dearly in order Asian region comprise one of the world's most to restore the base of resources that have been productive areas, in which shallow-water ma- destroyed. rine plants and animals reach their peak of species diversity. This diversity is associated with very high production of organic matter, The physical setting which in turn is converted into high fishery yields. Coastal ecosystems, such as upwelling ThewatersandislandsbetweenAsiaand Austra- areas, are capable of producing more than ten lia and between the Pacific and Indian oceans times as much organic matter per unit of area form one geographical unit. Geographically, the per unit of time as offshore waters. This very region consists of highly fragmented land area high production of organic matter is trans- interspersed among wide stretches of sea surface formed into a tremendous variety of economi- and extremely long coastlines. Physically, it is cally valuable products used by the people in divided into a continental part of mainland Asia, the region. which consists of Myanmar, Thailand, and the The region produces about 8.4 million metric Indo Chinese states of Laos, Kampuchea, and tons of fish. Due to the economic benefits that Viet Nam, and into the archipelago of Southeast could be derived from these rich and diverse Asia, including peninsular Malaysia, Brunei ecosystems, the coastal zones of Southeast Asia Darussalam, Singapore, Indonesia, and the Phil- are densely populated. More than 70 percent of ippines (Chia Lin Sien and MacAndrews 1979). the population in the region lives in coastal In oceanographic terms, however, the waters of areas, resulting in rather high levels of exploi- this region are part of the Pacific Ocean, which is tation of the natural resources and degradation separated from the Indian Ocean by the islands of of theenvironment. Fish and otheredible coastal Sumatra, Java, and the Lesser Sunda (Nusa products are consumed locally or exported. Tenggara). The Southeast Asian waters consist of Beaches, such as coral reefs, attract tourists in the Andaman Sea, the straits of Malacca, the Straits growing numbers. Coastal habitats are more of Singapore, the South China Sea, the Gulf of and more coveted for aquaculture of shrimps Thailand, the Java Sea, the Florest Sea, the Banda and fish, a booming industry of economic im- Sea,theArafuruSea,theTimorSea,theCelebesSea, portance in the region. the Sulu Sea, and the Philippines Sea. The whole The problem facing the nations of Southeast body of these waters covers approximately 8.94 Asia is how to maintain the integrity of the base million square kilometers, which represents about of marine and coastal resources for sustainable 2.5 percent of the world's ocean surface (Soegiarto use. An overview has indicated that resources 1978,1985). Figure 21-1 and table 21-1 present geo- are being overused, and the basis of their pro- graphic information on Southeast Asia, in particu- duction is eroded. If this trend continues, it will lar the members of the Association of the Southeast lead to a point of no return, and the countries of Asian Nations (Soegiarto 1991). Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 21-1: Map of Southeast Asian Nations Mongolia ac ' EAST ASIA AND ro-aun ' A 1 PACIFIC REGION China Lao People's fDen. Rep. Myonmo,, PACIFIC OCEAN lThailan lsinrIds Moo~~~~~~~~~~o Meldivus 9"0F ' IGribofrl Ifndermd /Ne lZun 00e I INDAN . .... Idcnebn WNn X Islands ~waslern OCECAN *unN L,a t Soa Vanuc ~ N ano Fishery production with networks extending to the interior are based on these fisheries. Fisheries are a valuable resource in the region. The early mechanized fishing efforts of Ja- Indeed, the Southeast Asian seas support one of pan have been joined in recent years by fishing the world's most productive marine fisheries. fleets from Taiwan (China) and South Korea. Total annual catch from the region in recent years Modernization and expansion of the region's has been approximately 8 million metric tons, coastal fishing fleets began in earnest some with certain fisheries capable of providing still twentyyears ago with the introduction of trawl greateryields. Table 21-2 illustrates annual catch, fishing in Thailand, followed by other coun- imports and exports, as well as consumption of tries. Coastal country trawling fleets have now fish in some countries in Southeast Asia for the become established in much of the region, fish- year 1988 (FAO 1988). ing the inshore and coastal waters for demersal Excluding the Gulf of Thailand, the Chinese fish, crustaceans, and molluscs. Coastal coun- and Vietnamese continental shelf in the South tries have also established purse seining fleets China Sea yields the greatest total annual catch in in recent years, providing increased catches of the region, nearly a million tons. Historically, coastal and oceanic pelagic fish. Mechanized these abundant fishery resources have been har- fishing for export constitutes a significant source vested in inshore and coastal waters with a vari- of foreign exchange, and the infrastructure sup- ety of traditional fishing gears and have been an porting these fisheries is a further source of important source of food, animal protein, and income and employment, for example, freez- employment for many of the region's coastal ing, cold storage, processing, boat building, populations. Whole market and barter systems and net making and mending. 330 Sustainable Development of Fisheries in Southeast Asia Table 21-1: Characteristics of the Marine and Coastal Zone of Members of the Association of the Southeast Asian Nations, 1988 Brunei Characteristic Darussalam Indonesia Malaysia Philippines Singapore Thailand Land area (thousands of square kilometers) 5.8 1,904 303.8 297 0.6 513 Marine area, including 200 miles of extended jurisdiction (1,000 square kilometers) 0.7 6,841.7 138.7 551 0.1 94.7 Length of coastline (kilometers) 130 81,000 4,800 18,417 193 3,219 Area of mangrove (thousands of hectares) 18.4 4,250 113.3 (west) 106.1 1.8 287.3 538.9 (east) Population (millions) 0.2 175 16.9 58.7 2.6 54.5 Percentage of the population living in the coastal zone 86 70 60 67 100a 60a Gross national product per capita (U.S. dollars) 17,000 550 1,775 630 7,550 860 a. Approximate figure. Source: Soegiarto 1991. Table 21-2: Catch and Consumption of Fish in Southeast Asia, 1988 Imports Exports Consumption Catch (millions of (millions of (Idlograms Country (metric tons) U.S. dollars) U.S. dollars) per capita) Members of the Association of the Southeast Asian Nations Brunei Darussalam 3,279 8.50 0.30 42.7 Indonesia 2,703,260 19.38 664.48 13.6 Malaysia 604,128 146.98 190.28 36.6 Philippines 2,041,920 63.06 406.50 33.7 Singapore 15,240 370.31 356.19 40.7 Thailand 2,350,000 537.92 1,630.90 21.6 Non members Kampuchea 70,000 - - 9.3 Laos 20,000 - - 5.6 Viet Nam 874,000 - 182.24 12.5 - Not available. Source: FAO 1988; International Centre for Ocean Development 1988. Problems of sustainable development overexploitation of several important fisheries. Concurrent with these events and adding to these The marked increase in regional fishing effort difficulties hasbeen the loss of important spawn- during the past twenty years, encouraged by the ing and nursery grounds of many valued species rapidly increasing local and international demand due to increased coastal pollution and the wide- for fishery products, has subjected many of the spread development of coastal lands. region's inshore and coastal fisheries to intense Thus, many of the region's fisheries are al- fishing pressure and has resulted in the readyunder stress. Spilled oil can havedirectand 331 Defining and Measuring Sustainability: The Biogeophysical Foundations indirect lethal and sublethal effects on eggs and humans, as measured by both primary produc- juveniles of species as well as on adults. Impor- tivity and biomass yields (Polunin and Soegiarto tant fish may be tainted, and fishing gear may be 1980). Production and export of organic materials fouled by oil, depressing the fishing industry and in estuaries contribute to estuarine, coral reefs, thus indirectly affecting all those who depend on and offshore fishery nutrition. The leaf litter pro- it for their food or livelihood. duction of mangroves is increasingly correlated T he variability in fishing pressure is directly with fisheries production within the mangrove reflected in the total fishery catch of a given area. system, represented by fish, bivalves, and crusta- In general, the areas with the highest total catch ceans and by diverse fisheries production in the are also the areas most intensely fished, such as nearby estuaries and coral reef sea grass com- those where the fish and invertebrate popula- munities. Coral reefs, very efficient in recycling tions are under the greatest stress from fishing nutrients and using nutrients from adjacent sys- activities. Anyadditional stresson these intensely tems, support sizable fisheries for numerous or- fished areas, such as oil pollution, may have an ganismsthatareharvestedbyandprovideasignifi- immediate, negative impact on existing fishery cant portion of the protein consumed by coastal catches from the affected areas. In less intensely people (Gomez 1980; Murdy and Ferraris 1980). fished areas, the detrimental effects of marine The impact of oil on tropical estuaries is not yet pollution on catches is likely to be less obvious. well known, althoughitmaybesurmised that the Very heavy fishing intensity (larger than 1,000 most significant effect of high toxicity in the water kilograms per square kilometer) is concentrated column would be on shoreline fauna and flora. in the Gulf of Thailand, particularly in the Thai Since estuaries receive up to 50 percent of their portion of the Gulf, the central Malacca Strait, in organic matter from mangrove systems, the im- the Andaman Sea immediately west of the Isth- pact on mangroves is of importance. Mangroves mus of Kra and northwest through most of the (and coastal marshes) have been ranked as the MerguiArchipelago,offthemouthsoftheMekong ecosystem most sensitive, or vulnerable, to oil, and Pasig rivers, and in the central Philippines. owing to the persistence of oil in that environ- Heavyfishingactivities(100to999.9kilograms ment and the slow recovery time of the ecosys- per square kilometer) occur in the northern and tem, estimated at twenty years or more (Odum southern Malacca Strait, along the east coast of and Johannes 1975). The vulnerability of coral the Malay Peninsula, in the southern Gulf of reefs to oil depends on the level of toxicity in the Thailand and over parts of the central Sunda water column (Ray 1980), presence and degree of Shelf, along the entire coast of Viet Nam, and mixing, and degree of direct exposure of corals northward along the entire Mainland Shelf, in- and other organisms to the oil. Beach systems, cluding around Taiwan (China), in most Philip- although not too productive, provide habitat for pine waters, off East Malaysia and Brunei, around certain organisms vulnerable to oil. southern Sulawesi and in the central and south- ern Makasar Strait, in the Bali Strait, along the north coast of Java, around the Riau archipela- Conclusions goes off eastern Sumatra, and off northwest Sumatra. In conclusion, Southeast Asian marine and coastal The important ecosystems for consideration in waters are one of the most productive areas in the the Southeast Asian region are world. The fisheries are an important resource in . Estuaries, common within the mouthsof larger the region, in particular for the livelihood of the river systems. traditional coastal communities. However, due • Mangroves, associated with low coastlines, to population pressures and the marked increase t and rivers. in regional fishing efforts in the past twenty years, estuaries, coupled with the increase in pollution and en- * Coral reefs, associated with most smaller vironment degradation, these waters have experi- islands and those coasts on larger islands enced tremendous pressures. If this trend contin- lacking large inputs of freshwater or sediment ues, countries of the Southeast Asian region will from river systems. have toundertake heavy tasks andburdensin order Estuaries, mangroves, and coral reefs are torestorethedestroyed resourcebase. Itisprobably among themostproductiveecosystemsknown to not too late to make a concerted and coordinated 332 Sustainable Development of Fisheries in Southeast Asia effort for this purpose. Regional cooperation, such International Centerfor Living Aquatic Resources as that of the Association of the Southeast Asian Research Newsletter, 3:1, pp. 21-22. Nations, Indo Pacific Fishery Council, or Southeast Odum, W. E., and R. E. Johannes. 1975. "The Asia Fishery Development Council, could serve as Response of Mangroves to Man-induced Envi- a vehicle for such an effort. ronmental Stress." In E. J. Ferguson-Wood and R. E. Johannes, eds., Tropical Marine Pollution, pp. 52-62. Elsevier Oceanographic Series 12. Note Amsterdam: Elsevier. Polunin, N., and Aprilani Soegiarto. 1980. "Ma- This chapter was also published in 1993 by the rine Ecosysterms of Indonesia: A Basis for Con- Southeast Asia Programme on Ocean Law, servation." Report of the International Union Policy, and Management in the following col- for the Conservation of Nature and World lection of papers: SEAPOL International Work- fordlie CondvIon ogre Bogor. shop on Challenges to Fishery Policy and Diplo- Wildlife Fund Indonesia Programme, Bogor. macy in South East Asia, edited by K. I. Matics Ray, J. P.1980. "The Effects of Petroleum Hydro- and Ted L. McDormnan. carbon on Corals." Paper presented at the Pe- troleum and tie Marine Environmental Inter- national Conference and Exhibition, Monaco, References May 27-30. Soegiarto, Aprilani. 1978. "Introduction to the Chia Lin Sien and C. MacAndrews, eds. 1979. RegionalOceanographicoftheSoutheastAsian Southeast Asian Seas: Frontiers for Development. Waters." Paper presented at the fifth FAO/ Southeast Asian Series. New York: McGraw- SIDA Workshop on Aquatic Pollution in Rela- Hill. tion to the Protection of Living Resources, FAO (Food and Agriculture Organization of the Manila, the Philippines, January 17-February United Nations). 1988. Fishery Statistics. Vol. 27.1977. 66: Catches and Landings; vol. 67: Commodities. .1985. "Oceanographic Assessment of the Rome. East Asian Seas." In A. L. Dahl and J. Carew- Gomez, E. D. 1980. "Status Report on Research Reid, eds., Environment and Resources in the and Degradation Problems of the Coral Reefs Pacific,pp. 173-84. RegionalSeasProgramStud- of the East Asian Seas." South China Sea Fish- ies 68. Nairobi, Kenya: United Nations Envi- eries Development and Coordinating ronment Program. Programme, Manila. May. . 1991. "The South China Sea: Its Eco- Interational Centre for Ocean Development. logical Features and Potentials for Developing 1988. WorldnFisheries: InternationlMap. Canada. Cooperation in Marine Scientific Research and 1988yF. WOr,and Fisheri.: J.ernational Map. Cana- Environmental Protection." Jurnal Luar Negeri Murdy, F. O., and C. J. Ferraris. 1980. "The Con- (Department of Foreign Affairs, Indonesia) 18, tribution of Coral Reef Fisheries Production." pp. 28-47. 333 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Henry A. Regier, Stephen A. Bocking, and H. Francis Henderson With all of nature's renewable phenomena that year in particular waters. During the reproduc- have been used by humans, concern about tiveprocess,manyfishspeciesenterwaterswhere sustainability must be about as old as anycultural localized and species-specific spawning sites are husbandry practices. Many traditional cultures found and where they are particularly vulnerable had effective systems for managing human uses to capture by humans. Barring access to these of fish and their habitats. Empirical understand- sites, or disrupting the behavioral rituals that ing of how technologically innovative humans precede spawning, may curtail effective repro- threatened the biophysical foundations of these duction. An awareness of such biophysical foun- valued phenomena began to cumulate several dations has long been part of fisheries traditions, centuriesago.Wenowunderstandsomethingabout especially with anadromous species in the tem- thethreatsposedbynumeroushumancausestothe perate zones. In consequence, spawners were to sustainability of various kinds of fisheries. be permitted unhindered, unmolested access to In this review essay, the term fisheries is used their preferred spawning area and to be left alone in abroad sense to include finfish, shellfish, aquatic during their spawning activities. When the repro- mammals such as whales and seals, and reptiles ductiveprocesswasaffected,attemptsweremade such as turtles. Our primary focus is on finfish; to mitigate the direct effects: fish hatcheries to the story of other fisheries generally resembles compensateforsacrificedspawninggrounds,fish that of finfish. For the scientific, technical, and ladders to bypass dams that barred spawners' policy aspects of the whole set of fisheries, there migrations, and artificial spawning locations. is effectively one collegial disciplinary domain These have seldom been fully successful. with a number of major and many minor special- The harmful effects on valued fish of the pollu- izations. This applies particularly to the issue of tionof fish habitatswerewidelyrecognized in the biophysical foundations of sustainability. nineteenth century. The evidence was sometimes Early concern about sustainability of fisheries quite compelling: for example, fish flesh smelled focused on reproduction, apparently because of of phenols that were being released into the wa- the belief that this was the major factor limiting ters by early petroleum-based industry. By 1875 the abundance of fish (see, for example, Milner industrial wastes, sawdust from sawmills, and 1874; Nettle 1857). As a result, attention focused animal and human wastes from farms and cities on preventing interference with the reproductive were recognized as damaging to the Great Lakes process or on supplementing it artificially. The fisheries (Pisani 1984). The human and industrial reproductive period of temperate fish stocks has wastesgenerallycamefromsettlementsthatwere long been known to be predictable, from other built near rivers and estuaries that in turn were seasonal and ecological processes, from year to ecologically quite productive of preferred fish. Defining and Measuring Sustainability: The Biogeophysical Foundations Ecologically unsustainable practices of voiding had dwindled. Depending on which attributes of human and industrial wastes into nearby waters an ecosystem are used to specify a goal of have not yet been effectively corrected in many sustainability, some introductions have threat- locales. Large-scale voiding into the atmosphere ened and some have fostered sustainability. of acidic gases due to combustion has acidified With hindsight, we note the beginnings a cen- vulnerable waters far downwind. tury ago of awareness of another type of threat: Informed observers have long recognized the the globalization of harvest and trade, for ex- harmful effects that destructive practices in the ample, with seal furs and whale oils. The catchment basin-for example, clear-cutting in nonadaptive dynamics of the commercial pro- forestry,overgrazingoflands,hillsideplowingin cesses involved were such that it made good agriculture, and filling in of shallow waters- short-term economic sense to mine-that is, de- have on aquatic systems (Marsh 1857). Major liberatelytooverharvest-differentstockssequen- effects on fish have been the result of increased tially, starting with those nearest a commercial variability of the water flow, with increased fre- headquarters and then expanding farther along quency of harmful episodes of abnormally high shore and offshore, and eventually reaching the and low flows. Such abuses havebeen addressed most distant ocean. Limits on catches for the repeatedly in many locales for many decades and purpose of sustaining the resource made no eco- still occur as new surprises for the ecologically nomic sense to some commercial interests. innocent. Some informed observers recognized these A fourth class of concerns about sustainability threats to sustainability of preferred fish species hasrelated to overfishing. The storyof the extinc- by late in the nineteenth century. Attempts to tion of the passenger pigeon and American bison correct the more flagrant abuses are apparent in in North America in the nineteenth century alerted the nineteenth-century legislation ofvariouscoun- people interested in marine mammals and then in tries. Some attempts proceeded to the interna- fish of the potential danger of uncontrolled ex- tional level; those that relate to fisheries interests ploitation. The earlier presupposition, often sup- included the 1892 Bregenzer Ubereinkunft for the ported by respected scientists, was that oceans, Bodensee (Regier and Applegate 1972), the 1902 large lakes, and large rivers as systems were International Council for the Exploration of the simply too large to be threatened in any practical Seas(Went 1972), the 1909 Boundary WaterTreaty sense by humans. Eventually a risk of overfishing between the United States and Canada (Rawson was perceived even with large stocks of marine Academy of Aquatic Science 1989), and the Dec- finfish, perhaps mostly with respect to effective laration of Principles of the 1908 North American new methods of fishing by fishers who used less- Conservation Conference (Van Hise 1911). effective older methods. The myth that the shelf Some piecemeal corrective measures then fol- seas and the open ocean could not be overfished, lowed. But by the mid-twentieth century, it was a myth that was under attack beginning in the late widely apparent that abusesof all six classes were nineteenthcentury, was finallydispatched in 1972 expanding in spite of earlier reform efforts. New (FAO 1973). rounds of emphasis on corrective measures fol- The introduction of foreign species occurred lowed; apparently none of these measures has on a grand scale in the nineteenth century and has been fullyeffective in addressing the root causeof continued to the present. Many introductions an abuse. weredeliberateand others wereaccidental;many The second half of the twentieth century introductions of both types were eventually found brought with it a seventh threat to sustainability: to be undesirable. Some introductions, such as hazardous contaminants produced through ad- the entry of sea lamprey into the Great Lakes vancedtechnology.Nuclearbombsbroughtwide- through ship canals, were disastrous: lamprey spread contamination with radionuclides. The exhibited no prudence whatever in preying on pesticideindustryfirstproducedpersistentpesti- fish species highly preferred by humans. Other cides (such as DDT) and then persistent wastes introductions served useful roles in ecosystems (such as dioxinsasby-products in the production that had been altered culturally in ways that of pesticides such as 24D). The threat that haz- could not be exploited efficiently by preferred ardouscontaminantsposetosustainabilityoffish nativespecies,forexample,Europeanbrowntrout stocks relates to the risk of sterilization of the thrived in moderately altered cold-water streams adults, mortality through poisoning of the em- in North America in which stocks of brook trout bryos, or crippling of the surviving young. 336 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement An eighth class was predicted late in the nine- values, such as equity, is largely to waste one's teenthcenturybutonlytakenseriouslylateinthe efforts (WCED 1987). Any measure related to twentieth century: widespread pollution of the the biophysical foundations of ecological atmosphere with radiatively active gases that sustainability must also make sense with re- cause the atmosphere to trap heat and thus in- spect to other major dimensions of the cultural crease average temperature (American Fisheries mind-set, and especially to equity consider- Society 1990). ations. Our present emphasis on sustainability Much of the second half of the twentieth cen- and equity should not be interpreted to imply tury has involved see-saw-like oscillations be- that we dismiss as irrelevant other values, such tween intensifying abuses and more effective, as aesthetic and inherent values. though piecemeal, counteractive measures, at least with respect to fish generally. On balance, the Small-scale artisanalfisheries threats to sustainability of fish, worldwide, may Throughout the world innumerable small-scale stillbeexpandingandintensifying,althoughthey ghro vide w ous ierable gener- may be abating in some locales. fisheries provide nutritious yields that are gener- Husbandmen have known about harmful lo- ally availableevento the localmpoor people. About cal effects from local abuses for centuries. The 40 million tons ef fish-almost half of the total political activities that flowed from the relevant production-are taken annually by artisanal awareness of the 1890s show that people had fisherfolk, mostly in warmer climates. An un- come to recognize that humans were capable of known but also large quantity is taken by occa- adverselyaffectingregionalphenomenathatthen sional foragers, usually for food, but also for had a bearing reflexively on the sustainability of recreational purposes. preferred stocks in particular locales. During the The rights and responsibilities of subsistence second half of the twentieth century, we have andartisanalfishersareorderednindifferentways become aware that we now affect the global bio- in different cultures. Under normal circumstances, sphere and so risk the sustainability of some thesetraditionsrelateinsomeappropriatewayto preferred regional and local phenomena of direct the ecological sustainability of the yields and to interest to some of us. social equity in the sharing of those yields. Exter- nally imposed efforts to develop these fisheries have sometimes ignored the indigenous conven- Six kinds of fisheries, with an emphasis tions, or suppressed them for no good reason, on sustainability and equity with adverse effects on sustainability. Small-scale fishing may be the occupation of last resort in places and in times of poverty and Distinctions are commonly drawn between sev- sca ioinain ne h meaieo eral types of fisheries development. Six types are immedisurivaloncer abu susaiabivity considered here: small-scale subsistence and com- andity may notrb apaet locally mercial artisanal fisheries on wild stocks; com- and equlity may not be apparent locally. The mercial, industrial, and incidental fisheries on erepollution to then follows may be marine mammals; large-scale commercial and termed pollution due to poverty. An unusual industrial fisheries on wild stocks of fish, other schenotsin so cultur to destroyithe foc than marne mammls; modeate-siz commer such events in some cultures: to destroy the sanc- than marine mammals; moderate-size commer- tity of the place. cial fisheries on stocks of wild fish augmented bv y Artisan er enhanced~~~~~~~ ~ sokadhbttcomrilndsub Artisanal fisheries can and do intercept migra- enhanced stock and h tat; ommiadosub- tory stocks of interest to other groups in the broad sistence husbandry of stocks of semi-domesti- fseissco.Mgaoyautsanr n cated species in artificial habitats under intensive culture; and recreational fisheries with variants juvenile recruits to stocks may be removed from that bear some resemblance to each of the other rivers and coastal areas to an unsustainable de- gee, In conditions of open access to such stocks. types. They are presented here in historical se- gr Some artisanal fishers remove large riparian trees quence, that is, in the order in which governance for making boats, use firewood for smoking fish, procedures related to sustainability were devel- ao forth, us create prolms-the con- oped for each type. ~~~and so forth, and thus create problems-the con- oped for each type. sequence of increased erosion-for other users of We share the current consensus that n the forests and for themselves. But the great ma- jority of artisanal fisheries do not pose this kind of 337 Defining and Measuring Sustainability: The Biogeophysical Foundations threat to other fishery interests nor to other inter- tures is a fundamental defectof much formal devel- ests in society. In contrast, artisanal fisheries are opment work and a threat to the eventual sustain- oftenharmedbycompetitionfrompolitically pow- able existence of such groups in a modern society. erful, large-scale commercial and intensive aquac- What might be an appropriate measure for the ultural fisheries, especially in the marine coastal biophysical foundations of finely dispersed small- zone. Also artisanal fisheries suffer adverse con- scale fisheries in a region? Perhaps experts in sequences of other practices in the river basins integratedruraldevelopmenthaveaddressedthis and along the coasts. These aquatic systems inte- issue. The leakiness of locales with respect to grate environmental consequences of many de- linmitingplantnutrients,suchasphosphorus,rmay velopmental practices on land and in the water, be an integrative measure. for example in bioaccumulating hazardous con- taminants. The harmful consequences of various Marine mammals: Whaling and sealing abuses are particularly apparent with the large aquatic organisms, including fish, of the rivers, Generally conceas about excessive levels of di- lakes, and estuaries. rect harvests and about significant incidental Relatively few of the innumerable artisanal catches in other fisheries have been more appar- fisherieshavebeendevelopedintensivelythrough ent politically with marine mammals than with external political processes and funds. Fisheries othervfish.l ntenseconcernaboutthesustainability in somelargelakes and new reservoirs havebeen of harvestsof someseal and whalestocksemerged developed with motors for boats, synthetic fibers relatively early in the Northern Hemisphere, more for nets, and so forth, but usually along artisanal than acentury ago. Some populationsand species lines, had been suppressed to low levels. The interna- Generally, artisanal fisheries do not now have tional convention on the Pribilof fur seals of the much potential for increased yields and do not North Pacific created important precedents. The pose major direct threats to other developmental total number to be taken and procedures for how opportunities. Where the local system of gover- the removals were to be allocated between coun- nanceisworking well, it maylbe best to leaveitbe tries were established. Thus sustainability and (Berkes 1989). Where the governance system has equity were both addressed in an operationally broken down, it might be fostered back to health direct way. Good results followed from this con- rather than supplanted byan introduced system. vention, although further difficulties have had to Sometimcsahybridbetweenlocalcommrunitarian be resolved from time to time. and central bureaucratic or market-based gover- Sustainability and equity issues were much nandcentral burkwlPeaucraticor market-based gove more difficult to resolve with whales than with nance works well (Pinkerton 1989). the Pribilof seals. The cost of achieving credible Many subsistence or forage fisheries not only scientific information on which to base formal have been severely affected by the threats to scis ion on which dynas forma sustainability outlined above, but have also been decisions was much higher. The dynamics of the disturbed by the consequences of adopting mod- marketplace, of commercial interests, and of in- ernization as the principal aim of development. ternational politics were not compatible with the This has been particularly serious among indig- ecological dynamics of the whale stocks. For ex- enous communities in the process of conversion ample, the interest rate onmoney was higher than from a subsistence to a market economy. Because the growth rate of the whales, hence it was eco- a community often lacks the time to train its nomically rational to mine the resource. Because young people in the skills of foraging, and forag- the issue had become strongly politicized inter- ing itself is often regarded as backward, tradi- nationally, different nations were rarely willing tionaitselfis knowlenredgeai rther, bckwause theadie to provide full and accurate information. tional knowledge is lost. Further, because thediet About a century after the issue of excessive tends to become less diverse until a fully devel- direct harvests of some marine mammals was oped food supply and marketing system Is first raised, it was solved politically, at best tem- reached, or because wild resources are less acces- porarily, by the imposition of very severe con- sible during periods of adversity, both individual straints and some complete bans on commercial health and community resilience may suffer. In- and industrial whaling and sealing. A long legacy deed, failure to address issues of short-term of political concern about excessively intense sustainability during the period of transition from whaling, which severely threatened the survival traditional to modem economies and social struc- of some species, was combined with the new 338 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement politics of animal protection to limit drastically be more acceptable politically. The main the harvests of marine mammals generally. countervailing argument advanced by animal More recently, some fisheries that use long protection interests may then be that people drifting gill nets on the high seas for some species should generally change to vegetarian diets. (It offinfishandshellfish(tunaandsquid)expanded may only be a matter of time until political furtherwithinopenoceanicwaters.Marinemam- interests related to plant rights become more mals (and fish-eating birds) were quite vulner- prominent than they are currently.) able to capture in such nets, because of the man- ner in which the nets vere actually fished in some Large-scale fisheries forfinfish and shellfish waters. Because of widespread public concerns about the welfare of those mammals, political The main general concepts and methods cur- action has already included instances of a shift of rently used in research and management related onus in which the drift net fishers have to demon- to capture fisheries were mostly developed first strate that the fishing method to be used does not with respect to finfish populations. From here, involve risks for the marine mammals (plus turtles they have often been adapted and applied to and birds, perhaps) of those waters. Unless appro- some of the other types of fisheries, such as shell- priate information is marshaled and found to be fisheries, whaling, and sealing. convincing, access to the fishery may be denied. From relatively small beginnings in the late The example above is a case in which the nineteenth century, large-scale fisheries (which processoftechnologyassessmentisalegalcondi- no longer include whaling) have expanded pro- tion of access. For decades, the calculation of gressively to all parts of the ocean. Inventiveness maximum sustainable yields or of total allowable and ingenuity played key roles with respect to catches of stocks has been a stock assessment vessels, gear, processing, and marketing activi- process. A shift of onus in which fishermen have ties. The main emphasis has always been on ap- to demonstrate responsibility for the relevant propriate, adaptive technology, at least to some measure of sustainability or performance has general extent: large-scale, nonselective use of implications for the nature of information ser- barriers,explosives,poisons,andentanglingtwine vices useful with such a regime. has seldom been condoned socially. Marine mammals going about their business Thenmainmotivation for theexpansionof mod- in coastal waters have become major tourist at- ern, large-scale fisheries was to satisfy demands tractions in various parts of the world's oceans. for fish (and fish meal) as reflected in the open Because tourist-related activities can and do in- markets, or in the bureaucratic processes of cen- terfere with the normal behavior of those mam- trally planned economies, of the countries al- mals, progressively more restrictive constraints ready well along the path of conventional devel- are also being placed on whale-watching and opment. Feeding the poor in any direct way was seal-watching activities. seldom a motivation, although less-favored or Thepoliticalsuccessesoftheanimalprotection poor-quality fish did become available to the interests have shifted the locus and type of com- poor. Fisheries were intended to produce valu- mercial activities related to marine mammals. In ablecommodities, directly and often in theireven- some parts of the world, those animals are not tual contribution of hard currency to a country's now intercepted and killed for their skins, flesh, balance of accounts. or fat but are rather intercepted and watched for The overall process of large-scale fisheries de- personal pleasure. Enterprises and communities velopment involved sequential cropping down that serve tourists-often involving redirected of themore-valued stocksneara fishingportwith fishingand whalinginterests-havecometoben- a subsequent shift to less-valued stocks nearby efit economically. and to more-valued stocks farther away. Some Burgeoningpopulationsofsomemarinemam- stocks were so plentiful, such as the cod of the mals and birds are competing with fisheries for Grand Banks, that they already attracted fishers some finfish and shellfish stocks. Reduction in from distant ports some five centuries ago. The accumulated natural capital of valued fish in catches and an increase in price of the relevant stocks in a pristine fishery seldom yielded great in ccsn nnee piotprofit to fishers, at least during the present cen- fishery products. Control programs on some tury. The exploratory fishing process was costly. marine mammal populations may then come to The rushed, disorderly competition in open- 339 Defining and Measuring Sustainability: The Biogeophysical Foundations access situations generally led to overcapitaliza- Networks of scientific researchers routinely tion, inefficient practices, and the dissipation of shared information, sometimes even when such profits. sharing was discouraged by a scientist's country. Because the fish and fisheries habitat of large On occasion, pro-sustainability researchers infor- waters were deemed to be unowned and access mally provided relatively accurate data to their was effectively open, there was no rent to be paid foreign peers for scientific purposes, while their for harvesting the fish. Unpaid rents that accrued countries were officially declaring the data to be to the fishers were eventually dissipated when inaccurate for political purposes. the levels of fishing effort overshot what was The information collected and the publication necessary to achieve optimal yields. Rather than services of the many fisheries commissions, to- charge rent for use of such resources, countries gether with the very extensive information ser- often subsidized their fleets in a competitive in- vicesattheFoodand AgricultureOrganizationof ternational race for resources and thus contrib- the United Nations (FAO) have played indis- uted to the overfishing. pensable roles in the general, if slow, evolution of Over the past hundred years, most countries international cooperation toward sustainable and with large-scale international fisheries (and also equitable fisheries regimes. Through moral sua- whaling) haveemployed two often distinctgroups sion, the international agencies sought improve- of fisheriesexpertswithcontrastingcommitments. ments in the relevance, accuracy, and precision of A pro-growth group, closely affiliated with the research and management data. Effective coop- commercialfishersandfinancialinstitutions,was eration seldom came in time to prevent some committed to increasing landings progressively, degree of overfishing. usually with an emphasis on finding more effec- Pro-sustainability scientific researchers and tive ways to locate fish stocks and to catch fish. A advisersoften assumed the responsibility of serv- pro-sustainability group was intent on prevent- ing as stewards of sustainability and equity. This ing the overexploitation of the stocks. Greater is to be expected with officials of international political power in the short run usually lay with commissions and the FAO, which were set up for the pro-growth group, which often enjoyed di- thatpurpose,amongothers.Butitalsofrequently rect access to senior politicians as patrons, while was the case with scientists of competing coun- thepro-sustainabilitygroupserved withinahier- tries. In part, this may have been because scien- archical bureaucracy or in an advisory role. tists serving ostensibly quite different political Starting in the 1950s, social scientists came to ideologies shared scientific paradigms and mo- participate increasingly in the information ser- res; scientific collaboration became an indirect vices for fisheries. Through their bioeconomic form of diplomacy. model withitsemphasisonefficiency,theycrafted In retrospect, informal and then formal com- a partial synthesis to transcend the antithesis of mitments to sustainability and equity have acted the pro-growth and pro-sustainability factions. to limit excesses within the pro-growth develop- Eventually the social scientists-now including ment of large-scale fisheries. The open-access fea- anthropologists, human ecologists, and political ture was progressively constrained and super- scientists-addressed issues of rights and respon- seded by regimes of limited or closed access, as in sibilities, hence social equity. In the abstract, it the 1984 United Nations Convention of the Law may be difficult to plan fishery practices so that of the Sea (UNCLOS; United Nations 1984). Al- economicefficiencyand socialequityarecompat- though most agreed that these regimes were a ible. A two-stage process may be appropriate. step in the right direction, closure of access at the The relevant government first specifies clearly national level was not necessaTily accompanied the social equity considerations to be met: subject by comparable measures within a country's fish- to those considerations, the issue of economic eries. Many of the most valued large-scale stocks efficiency can then be addressed through the are now exploited too intensely. market, the bureaucracy, or other secondary in- From relatively small-scale beginnings several sti tu tional mechanisms. Sometimes a govemment centuries ago, large-scale fisheries now land some does not possess the will to make the primary 50 million tons a year. Of those, about 30 million decisions and implicitly delegates this responsi- tons are for human food and 20 million tons are bility to thesecondaryinstitutions,generally with for fish meal to be fed to domesticated animals, the result that the problem then worsens. fish inaquacultural pensand ponds,and so forth. 340 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement No major new stocks of large-scale fisheries orevenextinguished.Thecostsincurredinassur- are likely to be found. As demand for fish for ing appropriate spawning conditions justify the human food will rise, prices paid to fishers will priorityclaimedbythecountriessponsoringsuch rise. Progressively, the small pelagic fish that husbandry to the recruits from it, as indicated in are now rendered into fish meal and the inci- United Nations (1984). dentally caught low-valued species that are In efforts to rehabilitate valued fisheries in now discarded offshore will be processed for degraded habitats or to enhance new fisheries in human food, because it will pay to do so. Pro- reservoirs, species not native to the aquatic sys- cessing and marketing all small pelagic fish tem have often been introduced. Some introduc- and all discarded by-catches for human food tions have proven successful-ecologically, eco- will not necessarily improve the lot of the poor nomically, and socially-but some species be- consumer. The prices will likely be too high, in have differently and less desirably than expected part because preventing those catches from in habitats new to them. Fish diseases and para- spoiling in warm climates, which may occur in sites may be inadvertently brought in with the a matter of minutes in the absence of preserva- introduced fish. Major problems of this sort have tive means, will be expensive. occurred in some waters. Prior impact assess- Appropriate changes in fishing technology, mentsonanyproposedintroductionsarecoming regulations, and practices will lead to a recovery to be expected, as in the Laurentian Great Lakes. of some stocks and an increase in the overall value Coastal marine wetlands that are periodically of the landings, as has been demonstrated in inundated and floodplains bordering rivers and some fisheries. It is clearly not technical knowl- lakesaremajorspawningandnurseryareas.These edge, but rather the necessary political compre- are being drained and diked to an increasing hension, that will remain illusive. extent. What was once an interactive ecosystem In summary, manyinternationaldisputes,con- of rivers, floodplain, estuary, and coastal plain frontations, and some occasional skirmishes (the becomesdismembered withcommensuratelosses Icelandic cod wars) have occurred in the large- in fish productivity. Often the value of the lost scale fisheries. But progress has also been made fish yields is unknown or ignored. Gradually, the toward regional and global regimes in which importanceofwetlandsorwastelandsto fishand sustainability and equity appear as prominent otherspecieshascometoberecognized.Someareas guideposts. The positive achievements in fisher- are being granted protection, in the form of re- ies and environmental (fish habitat) issues de- serves,becauseoftheirecologicalimportance,which serve to be celebrated. extendsbeyond the immediate locale in which they occur. Recently, issues relating to the protection of Moderate-scale fisheries, with enhancement fisheries have emerged within the larger policy contextof safeguardingbiodiversity. Consequently, Moderate-scale fish stocks of the coastal zone, h ou sepnigfo h utiaiiyo large rivers, and large lakes have been subject to individual stocks to that of ecosystems. intense fisheries development, but their habitats have also been subject to many effects of other Semi-domesticated species cultured in developments on land. With many stocks, some . . b. adverse effects of other developments have been artificial habitats corrected and some habitat factors have been Aquacultureresemblesanimalhusbandrywithin deliberately enhanced to benefit particularly val- agriculturemore than it resemblescapture fisher- ued stocks, such as salmon species. ies on wild stocks. The size of the productive With anadromous species, many distinctive habitat and the environmental impact of a par- spawning runs or races have been extinguished ticular enterprise are usually quite small and due mostly to physical alteration and pollution of local. In aquaculture, sustainability and equity rivers. Attempts to create sustainable new runsin considerations usually relate directly to the re- recent years have had some success. sponsibilities that come with individual or com- Certain development uses of a river must be munity ownership rights. The relevant regime of severely constrained or managed if the river is to rights and responsibilities is not well-developed beappropriateforspawningby valued species. If in countries where aquaculture is not an old the costs of environmental husbandry were not tradition; parts of Asia and Europe have old incurred, the spawning stock would be depressed traditions. 341 Defining and Measuring Sustainability: The Biogeophysical Foundations As with small-scale artisanal fisheries, aquac- featuresof hybridsof thewild and cultural stocks, ulture should fit within a regime of integrated that is, by gene pollution. rural development. By comparison, moderate- Aquaculturists are turning more and more to scale fisheries such as those described in the pre- genetic enhancement for improved production. ceding section tend to fit within river basin and Access to wild gene pools is recognized as neces- coastal zone development. The oceanic regime sary for long-term success. Preservation of the for large-scale fishing is not yet as well articu- natural biodiversity in appropriate nature re- lated, but the 1984 UNCLOS has clarified major serves is therefore important to aquacultural in- principles (United Nations 1984). At all three terests. scales, the relevant institutions are designed to resolve problems of equity and sustainability, Recreationalfisheries among others. It is unlikely that small-scale artisanal and Recreationalfishers(mostlymales)mayruse very large-scale commercial fisheries-capture fisher- little equipment and resemble some artisanal sub- ies-can be developed further to increase sus- sistence fisheries in this respect. Or they may tained yields by more than a fraction of their employscostlymgear comparable tofthat ofEmoder- current levels. Hopes that increased productive ate-sizecommercialcapture fisheries. Enterprises employment will produce greatly increased fish that cater to anglers' wishes may lease use of yields to help feed the vast growth in numbers of vessels, sell access to cultured fish in ponds or to people are pinned mostly on intensive aquacul- wild fish in streams, and so forth. peopl are inne mosty on ntenive aua - Recreational anglers are often as interested in ture. But even then, the costs of production will Rereglely wild, intereste- often put the price at a level too high for the poor. experiencing relatively wild, intact natural phe- In theforeseeable future, if thepoorest of the poor nomena as they are in actually catching valued of the world will have some fish in their diets fish species. But there must be some reasonable much of it will come from protected and en- likelihood of the latter. Many of the fish caught hanced artisanal fisheries. contribute to the diet of the angler's family. Intensive aquaculture, as with intensive agri- In industrial countries, anglers, when orga- culture, can be ecologically disruptive if prac- nized, are often more powerful politically than ticed carelessly. Aquaculture ponds, as in a man- small-scale to moderate-scale capture fisheries. grove coastal zone, may interfere with the func- Progressively, anglers limit the access of com- tion of this zone as spawning and nursery areas mercial fishermen from heavily angled water and for marine species. Pens in open waters lead to eventually may have them excluded entirely. Not organic pollution throughlost foodand the physi- infrequently, commercial fishermen then trans- ological wastes of the cultured organisms. form their enterprises to serve anglers and tour- Aquaculture,intenshveandextensive,nowuses ists, as has also been the case with commercial fish meal and partially processed low-value spe- whaling communities, which have adapted to cies caught incidentally, for example, in fisheries serve tourists. To compensate displaced commer- targeted on shrimp. If the fish meal and incidental cial fishers for the loss of access to fish, some catches are processed for direct human consump- governments puwrchase the fisheries, expunge the tion, the productivity of some types of aquacul- rights, and write off the assets. ture is reduced. Competition among anglers in turn leads to Fish-eating birds and mammals become seri- progressively severe limitations on the catch that ous pests with aquaculture enterprises. Attempts a particular angler can land. Increasingly, more to control them may bring the fishers in conflict angler fisheries are now being managed, through with animal protection interests. Introduced spe- mutual consent, so that few fish are killed, al- cies that may have been artificially selected for though many may be caught and released. Then particulargenetictraitsareoftenusedinaquacul- the large fish of valued species remain in those ture. When some of these escape and become watersforfutureanglingopportunities.(lnsome feral, they may become pests; for example, they waters individual fish that are caught repeatedly may pollute the gene pool of the well-adapted and relatively harmlessly are given distinctive natural stocks. If a natural stock is tuned geneti- names.) Anglers may help to protect healthy cally to particular spawning circumstances, then aquatic ecosystems and to rehabilitate degraded spawning may be disrupted by inappropriate waters. These are ways in which sustainability and equity considerations are being addressed. 342 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Angling is generally a sport for the more pros- fish were diminishing, probably due to destruc- perous people of the more industrial parts of the tion of immature fish. In 1893, a British committee world, although the progression from foraging considered measures to preserve and improve and subsistence fishing to recreational angling British sea fisheries; no action was taken on its appears to be a common means of legitimizing recommended controls (Select Committee on Sea such activity even in developing countries. An- Fisheries 1893). glers are usually prepared to pay the owners of In the 1890s, Petersen examined plaice stocks the fish and their habitat for the opportunity to in the Kattegat, focusing particularly on growth fish; wealthy anglers pay well for exceptional rates (Petersen 1903). He noted the possibility opportunities. Many anglers belong to clubs that fish were being taken before they had grown within which there is ritualized competition for to full size and that this growth overfishing re- relative status. Increasingly, practice of proper duced thetotalvalueof catches.Thebestsolution angler ethics (to safeguard sustainability) is a was education: fishermen could avoid growth prerequisite for high status. overfishing themselves if the problem were ex- plained to them. He also suggested a size limit for salable fish to remedy overfishing. He considered Sustainability concerns in marine f isheries: recruitment overfishing, which would reduce re- A case study productive effectiveness and might threaten the survival of the stock, as only a distant prospect, Concern about sustainability has long been ap- because even if greatly depleted, there were still parent with the North Sea fishery. During the enough mature fish to provide sufficient eggs. nineteenth century, fishermen spread north and Garstang (1900) formulated catch per unit of west across the North Sea, as coastal stocks be- effort as an index of stock and used it to show that came depleted (Alward 1911). Technological between 1889 and 1898 stock density of the En- change accompanied this movement. Instead of glish North Sea fishery had declined almost 50 waiting for fish to school near shore or migrate percent. into rivers, fishermen devised gear and vessels to The International Council for the Exploration catch the fish farther offshore. By approximately of the Sea was founded in 1902 to promote ratio- 1880, expansion had reached the limit of the sea, nal, scientific exploitation of the fisheries (Went and catch rates in the more distant areas began to 1972). Overfishing was of major concern from the decline.ThisdeclineoccurredastheBritishNorth start. From 1905 to 1914, the North Sea plaice Sea fisheries became more and more industrial- dominated the debate over whether overfishing ized after 1880, as fishermen replaced sails with was primarily an economic problem, as Petersen steam power, and passive with active gear. Plaice and others believed, or a biological problem, in declined most, followed by haddock. With de- which fish stocks were actuallybeing depleted, as cline in yield, reduction in the size of individual Heincke argued. Petersen advised no controls on fish, and -hanges in species composition, fisher- the fishery;Heinckeargued thattheywereneeded. men had to work harder for less gain (Cushing No action was then taken. 1988; Meyer 1947). Similar events occurred in Pacific Northwest In 1863 a British royal commission asked halibutfisheries,includingashifttomoredistant whether the supply of fish was increasing or fisheries as those closer were depleted and, be- decreasing(Commissioners Appointed to Inquire tween 1910 and 1930, a great decline in stock into the Sea Fisheries 1866). Although some evi- density and catches while effort rose three times. dence of overfishing was presented, most wit- As in the North Sea, the fishery was marked by nesses argued that fish remained abundant. The industrialization and technological innovation commission concluded that supply had not di- (Thompson and Freeman 1930). Beginning in 1917, minished and that further expansion was pos- Thompson proposed that the observed decline in sible. Some felt that the oceans' living resources stock density be corrected by restricting fishing were inexhaustible (Huxley 1884). Others sug- effort. His objective was to ensure a stable, eco- gested that there were limits to the productivity nomically efficient fishery. Importantly, he pro- of fish and warned of destructive exploitation posed a limit on entry to the fishery, to permit and overfishing (Meyer 1947). Since 1875, empiri- stock density to recover and, eventually, catches cal evidence accumulated that the stocks of flat- to increase. 343 Defining and Measuring Sustainability: The Biogeophysical Foundations Russell (1931) used a mathematical mass bal- in weight versus fishing mortality can be defined ance expression as a basis for achieving a maxi- and that there is a greatest catch that can be safely mum sustainableyield and avoidinggrowthover- taken for a long time. Such formulations could be fishing. He considered recruitment overfishing explained and justified to persons who are not to be unlikely. Graham (1935), like Russell, pur- experts, could be translated into management sued mesh regulations and developed a logistic advice, and could form the basis for negotiations model to predict the most profitable level of fish- and agreements. Their utility in arenas outside ing effort. Before the 1939-45 war, empirical evi- science is reflected in their use in areas beyond dence became available showing that fish stocks those originally intended (Holt and Talbot 1978); heavily exploitedbefore the 1914-18 war recovered in part, it was the demands of negotiations and when hostilities prevented fishing (Russell 1939). agreements that led to sustainability being de- Catch statistics gathered in 1946 indicated a simi- fined in terms of maximum sustainable yield. lar phenomenon had occurred in the 1939-45 Changingdefinitionsofsustainabilitymayalso war. At the 1946 Overfishing Convention in Lon- be identified by comparing the objectives of in- don, most European countries agreed to mini- ternational agreements. In 1946, the International mum landing sizes for demersal fishes and mini- Fisheries Convention defined the central concern mum mesh sizes for trawls, except those used for as overfishing; in 1949, the International Conven- herring and shrimp, but did not agree to limit tion for North-West Atlantic Fisheries sought, entry toa fishery. The major problem was seenas after Ricker (1946), to achieve a maximum sus- growthoverfishing,and thesolutionwasbuilton tained yield; the Permanent Commission in 1954 Russell and Graham's work. (which in 1959 became the North East Atlantic In short, before the 1950s, in the North Sea (the Fisheries Convention) sought to achieve conser- most intensively studied marine fishery in the vation through rational exploitation; and in 1984, world), sustainability was defined as achieving the Law of the Sea Conference defined conserva- maximum biological yield and maximum eco- tionasoptimumsustainableyield(Cushingl988). nomic yield by avoiding the capture of young A major preoccupation of these international fish, that is, growth overfishing. Sustainability agreements was growth overfishing. Beverton was not defined as maintaining the stocks them- and Holt (1957) justified dealing with growth selves, that is, avoiding recruitment overfishing; overfishing by increasing mesh size. The Perma- given the scale of the fishery and the state of nent Commission's first task was to extend the technology, recruitment overfishing was not con- principle of mesh regulation,but during the 1950s sidered likely (Cushing 1988). This reflected the and 1960s, larger fleets, new technology, and new biology of the dominant North Sea fisheries: be- markets, such as those for fish meal and frozen ing based on "r species," with little parental in- fish, led to a new stage of overfishing (Cushing vestment in offspring other than broadcasting 1988). Fish production increased from 22 million large quantities of eggs, recruitment overfishing tons annually between 1948 and 1952 to 72 mil- is unlikely. Influential work by Beverton and lion tons between 1978 and 1983 (Gulland 1983). Holt (1957) was based on experience with such Some scientists, on occasion, haveresisted this fisheries and was done before cases of recruit- emphasisontheconservationofexistingfisheries ment overfishing had been convincingly docu- solelythroughapplicationofpopulationdynam- mented with such stocks. Contemporary workby ics. Kasahara (1961) argued that fisheriesresearch other researchers (Ricker 1954, 1958) that empha- should not focus on conservation but on develop- sized "K species" such as centrachids and salmo- ment and exploitation of new resources (McHugh nids, with strong parental involvement in caring 1970). This reflected the interest in other coun- for fewer eggs, inferred a greater risk of recruit- tries, especially Japan and Russia, in the discov- ment overfishing. ery and exploitation of distant resources. Some Simple equations describing the dynamics of North American scientists also emphasized at fish populations, and relating stock density to that time that global production of fish was po- fishing effort, were offered asa basis for manage- tentially almost unlimited; Larkin (1965), for ex- ment and international agreement. This required ample, noted that the then world catch of 50 simple, understandable formulations emphasiz- million metric tons was "certainly less than one- ing economic consequences. Ricker (1946) de- tenthof thepotential catchand perhapsaslittleas fined maximum sustained yield to express the one-fortieth." This judgment was based on a no- principle that the maximum of the curve of yield tion that large fish were ecologically inefficient 344 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement and could be sacrificed in favor of trophically fisheries. There was also some consensus on a moreefficient small fish, or invertebrates, or even more complex view of sustained yield, going plankton; the greatly increased yield would more beyond maximum biological yield to include so- than compensate for the lower value per unit of cioeconomic factors such as equity (allocation of mass of the catches. This notion is not now held in resources) and a diversity of management objec- high regard, for reasons that, apparently, have tives (FAO 1973). Critical reappraisal of assess- not been fully explained. mentsof stock-based maximum sustainableyield, In the 1960s, a different view of sustainability as the primary kind of information on which to took hold. With the second industrialization of base decisions of total allowable catches to ensure marine fisheries, and associated widespread evi- sustainability, then followed, for example, by dence of recruitment overfishing, sustainability Larkin (1977) and Holt and Talbot (1978). became defined not simply as avoiding growth A tacit consensus was codified to some extent overfishing, but as maintaining the viability of in United Nations (1984), which emphasized the each valued stock, that is, avoiding recruitment conservation of existing living resources, not the overfishing. Markets sold many particular kinds development of new resources. Conservation of fish; they did not sell fish as homogeneous within an exclusive economic zone came to be substance. Thus, different methods of control, in primarily the responsibility of the coastal state. addition to mesh regulation, were needed. Only Conservation requires ensuring that living re- catch and effort quotas were politically feasible sources are not endangered by overexploitation with large-scale fisheries; because effort statistics (avoiding recruitment overfishing) and that popu- were poor, partly because effort measures could lations of harvested species are maintained or not be standardized sufficiently, the catch quota restored at a level that can produce the maximum or total allowable catch became the focus of man- sustainable yield (article 61.2, 61.3). All produc- agement (Cushing 1988). tion in excess of that required to eliminate risk of Since the 1940s, the definition of sustainability overexploitation, and to maintain maximum has gradually become more complicated, as ob- sustainable yield, should be included within the jectives other than maximumbiological yield have allowable catch and harvested. If the coastal state become important. Graham (1943) argued that cannot fully use the allowable catch, then it shall unlimited fishing must become unprofitable and give other states access to the surplus (article 62.2). inefficient; therefore, entry mustbe limited. There The definition of maximum sustainable yield is a greatest yield, but he thought that to limit in UNCLOS is modified by consideration of effort for greater value was much more impor- several factors: tant. In 1943, Herrington and Nesbit debated Environmental factors, such as the interdepen- whether management should seek to achieve . ' s maximum biological yield or maximum economic dence of stocks (article 61.3) or the mainte- yield. Nesbit, who argued for maximum eco- nance ofrspcie a te ha s nomic yield, proposed reducing entry into the species (article 61.4). fishery to ensure healthy returns for those fisher- * Economic factors, such as the needs of coastal men who remained (McHugh 1970); the tradeoff fishingcommunitiesanddevelopingstates(ar- was between achieving maximum yield and ticle 61.3) and the needs of those who have avoiding loss of jobs. In summary, the objectives habitually fished in areas to which other states of large-scale commercial fisheries management are given access (article 62.3). became a combination of maximum yield, value, * Equity, such as the right of participation of and jobs: in short, the greatest yield for the great- land-locked states (article69) and geographically est value and the least loss of jobs (Cushing 1988). disadvantaged states (article 70). By 1973, a consensus had emerged among These considerations imply that indicators of fishery scientists that global production fromcap- sustainabilit limited to standard biophysical ture fisheries with conventional types of fish could y na y not increase far beyond its then current level measures, or emphasizing only economic effi- n icrauseofte farbseyo theunecurrentolve ciency, are insufficient. Management must also beasefhebenefarenepoiestcs be guided by the conditions particular to each and, consequently, that the chief task of fishery coastal te cludingoth require to in- management wasnot to develop unexploited fish coastal state, including the requirements of in- matagcms, nt wantto devntainsust blo y unexplistedish shore artisanal fisheries traditionally neglected in stocks, but to maintain sustainability of existing resource development programs (Troadec 1983). 345 Defining and Measuring Sustainability: The Biogeophysical Foundations Management must relate to the sociopolitical and Regier 1990;Francisand othersl979;GLWQA behavior of individual fishers and aggregated 1978 with protocol of 1987; Loftus and Regier fisheries. Individual transferablequotashavebeen 1972). The general approach has been pragmatic. introduced to give market forces and commercial The Great Lakes Water Quality Agreement of management processes roles in sustaining eco- 1978, administered by the United States-Canada nomically efficient levels of production. But with International Joint Commission, committed the such a reorientation of management, new inequi- parties to a goal of attaining ecosystem integrity ties may arise, as in the initial assignment of in the basin (Regier 1992a). Soon after that, the quotas. Also entrepreneurship may be fostered at United States-Canada Great Lakes Fishery Com- the expense of husbandry. mission effectively committed itself to the same In retrospect, sustainability has always been a goal, consistent with the Strategic Great Lakes major concern of fisheries scientists and manag- Fishery Management Plan of 1981, which was ers. That it has not been achieved in any major negotiated among all the states, the province, and way may be partly because entrepreneurial pro- the federal governments under the aegis of the growth interests enjoy political strength and have commission. Subsequently, the states and prov- some scientific support on their side and partly ince also committed themselves to this goal with because efforts to achieve stable fisheries have respecttoissuesofthequantityandqualityofwater been focused too heavily on the short-term tech- (Rawson Academy of Aquatic Science 1989). In nical aspects of the problems as they are per- recent years, various municipal and metropolitan ceived at the time. As fisheries evolve along with governments have joined the consensus, for ex- many other societal changes, the issues change ample, in the greater Toronto area (RCFTW 1992). and earlier partial solutions become obsolete. A quick study of the practical political conse- Planning and management for sustainability in quences of commitments to ecosystem integrity fisheries arecoming tobe focused moreclearlyon in the Great Lakes Basin shows that a variety of the processes of their development and practices objectives are now accepted as subsidiary to the and less on particular end states as defined in goal, as indicated in table 22-1. All of these objec- numerical measures of sustained yields of a few tives, taken together, are intended to ensure the valued species. sustainability of the ecosystem's integrity. For currently degraded parts of the lakes, a satisfac- tory state of ecosystem integrity must first be The fisheries of the Great Lakes: attained and then sustained subsequently. Another case study Only some of the key features of a vision of ecosystem integrity have as yet been specified for In the first section of this chapter, we sketched these lakes. The most comprehensive specifica- eight major kinds of threats to the sustainability tion was included in the 1987 protocol to the 1978 of fisheries. By 1900, the first six had been recog- Great Lakes Water Quality Agreement for Lake nized within lheGreat LakesBasin (Bocking 1987). Superior. The shared vision was that Lake Supe- Although marine fisheries research was until re- rior should be rehabilitated to a state reasonably cently based on single-species approaches, in the close to that of its pristine state. Implicitly, this Great Lakes Basin (as in the Bodensee and Rhine meant that all the classes of threats or cultural River), awareness of the impact of ecosystem stresses sketched above be relaxed and some factors on fisheries, such as changes in the physi- active rehabilitative ecosystemic therapy be un- cal environment, developed relatively early. dertaken. All that is involved in the 1987 commit- Throughout much of the twentieth century, how- ment is only gradually being made explicit. ever, efforts to cope with these threats were hin- A restoration to near-pristine states is unlikely dered by divergent views on which threat was to be undertaken with anyof theother fourGreat most urgent(overfishing versus watershed modi- Lakes or the four major rivers that thread the fication; Egerton 1985). Morerecently,Great Lkes lakes to each other and to the Gulf of St. Lawrence. ficaties reg ert , portici ngtykeholaers, The preferred vision for each of these lakes has fishrierearcers,pariciptin stk s E yet to be specified in any detail. The cultural and government administrators have assumed a stresses are to be relaxed and their adverse conse- more comprehensive approach to these threats quences to be remediated, but what the rehabili- (American Fisheries Society 1990; Cairns, tated ecosystem's dominant structures and pro- McCormick, and Niederlehner 1991; Edwards cesses should be has not yet been agreed fully. 346 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Table 22-1: Four Domains of Emphasis in the Current Politics in the Great Lakes Basin Concerning Sustainability and Integrity of the Environment Focus of interest Good features Bad features Relevant professionals Quality of materials Valued abiotic resources, Chemical and physical, Engineers, geologists, and substances clean sand, pure water pollutantsgarbage and environmental chemists, wastes in dumps, hot- hydrologists spots in sediments Abundance of species Valued living resources, Human pathogens, Forestry, fisheries, and in their habitats naturalists' friends, rare unwanted exotics, wildlife managers, species harmful pests epidemiologists, public health officers Local ecosystems in Healthy centers of ecological Pathological centers of Parks and preserves landscape networks organization and connecting disorganization and pro- remedial action planners, links in a dvnamic network liferation channels landscape designers Natural and cultural Caring healthy humans living Disoriented ailing humans The new regional planners, interactions in with adapted nature in a degrading with debased participation facilitators for bioregions regional mosaic of nature in a regional self- redemocratization self-organizing ecosystems reinforcing slum Note: The normative distinction between good and bad is simplistic, of course. Figure 22-1 shows three general states of eco- system integrity that are currently being clarified. Figure 22-1: Three States of Ecosystem Integrity The vertical ovoid on the left denotes a reasonable approximation to the integrity of the primeval HIGH state, as with the formal commitments for the upper lake, Superior. The diagonal ovoid depicts / degraded ecosystems, that is, states of disintegrity or self-reinforcing pathological integrity, which State still persistin thetwo lowerlakes,ErieandCOntario, of and in large parts of the two middle lakes, Michi- OrgaSyszateon gan and Huron. The horizontal ovoid relates to \r1a--1!-on human-dominated ecosystems of the future that will exhibit an acceptable measure of partially designed cultural and natural integrity. LOW AlltheGreat LakeBasin's jurisdictionsatwhat- LOW ever level apparently concur that the parts of the basin ecosystem that now exist in a degraded, disintegratec state should be rehabilitated or re- LOW S Cultural _ HIGH stored to one of the two general domains that Influences possess desirable integrity. Few of the degraded ecosystems, other than locales within Lake Supe- Note: Ovoids refer to three multidimensional ecosystemic rior, will likely be restored to approximate a centralslantng ovoid reflects thediffr natlural state;the primeval state. So they will be transformed into degradation due to cultural abuse; the right ovoid reflects a cultual an natual inegrit, as n the healthv natural and cultural ecosystem or landscape mosaic. stateswithculturaand naturalintegrity,asinthe of healthy' self-integration is high at the top of the horizontal ovoid. Comprehensive features of this figure and low at the bottom. kind of integrity are only now coming to be addressed explicitly in some parts of the basin. The Great Lakes Fishery Commission (1992) com- mitted the various lake committees that serve 347 Defining and Measuring Sustainability: The Biogeophysical Foundations under its aegis to formulate by 1993 quantitative The accepted goal of ecosystem integrity in the targets for the fish species associations in each of Great Lakes Basin has not yet been specified the lakes consistent with the quantitative envi- sufficientlytoaddressthequestionofwhatmight ronmental commitments under the Great Lakes be the best single measure related to the Water Quality Agreement, as amended progres- sustainability of a state of ecosystem integrity. A sively. judiciously selected set of variables for a single For Lake Superior, the ecosystem objective integrative indicator species may be as close as was specified with respect to numerous water we can get to a single measure of sustainability, quality criteria, but also with respect to species but such a set would also be insufficient (Regier that were deemed to be integratively or cumula- 1992b). It seems unlikely that any single measure, tively vulnerable to the entire mix of cultural or some single numerical aggregation of interre- stresses being visited on that lake. The two spe- lated measures, will suffice for all practical pur- cies were the lake trout and an amphipod, poses. A particular measure of sustainability re- Pontoporeia affinis. In this chapter, the suitability lates mainly to a subset of the pluralistic values of of the lake trout as an integrative indicator of the a democratic community of people. Different in- state of ecosystem integrity is sketched briefly, terests that focus on different legitimate values following Ryder and Edwards (1985) and Regier need information relevant to their various interests: (1992b). this implies that sustainability will not likely be Lake trout thriveinappropriatenatural waters defined monistically nor measured in a single way. that are not severely affected by moderate levels of any of the classes of stresses sketched above. Each stress triggers a diagnostic response and Sustainability,science,andtheenvironment also contributes to an overall syndrome of harm of the population (Rapport, Regier, and Definition and measurement of sustainability Hutchinson 1985).The abundanceof lake trout (a depend not solelyon understandingitsbiophysi- salmonid) as an integrative indicator may pro- calfoundations. Atleast threeaspectsof theinter- vide the best measure of sustainability of ecosys- action among science, society, and the natural tem integrity in the deep, cold, nutrient-poor environment should also be considered: the di- parts of the Great Lakes waters. The necessary versity of forms of science considered relevant to research has been completed to demonstrate that sustainability, the divergent perspectives on the the walleye (a percid fish) could serve a similar roleof science in sustainability development, and role for moderately shallow, cool, nutrient-rich the definition of a sustainable ecosystem in a state waters in the Great Lakes (Edwards and Ryder of desired integrity. 1989). Although the necessary work has not been done, a similar case can presumably be made for Sciettce relevant to sustainability: smallmouth bass (a ceritrarchid fish) in shallow Academic considerations inshore waters of bays, at least where the degra- dation of such waters is minimal or has been Maruyama (1974) has identified three kinds of remediated (GLSAB 1989). mind-sets or broad predispositions that relate to The useof ecologically sensitive speciesof fish contemporary science(see table22-2). Eachmind- as integrative indicatorsof the stateof integrity of set has mutually compatibleelements fromontol- the water and land ecosystem of a watershed has ogy (regarding the nature of reality or what is), been applied to tributary basins to the Great epistemology (regarding ways of knowing or Lakes, especially those in the Toronto area enquiry),ethics(regardingwhatoughttobedone), (Kauffman and others 1992). Both an index of and perhaps aesthetics (regarding what is beauti- biotic integrity and habitat suitability indexes- ful orpleasing). Clearly thecharacterizationspre- each of which is strongly rooted in ecological sented in table 22-2 are abstractions-presum- understanding (Karr and others 1986; Raleigh ably few people (except perhaps some disciplined 1982)-have been tested for the streams in the academics!) would fit neatly into only one of the Toronto area (Beak Consultants Ltd. 1991; classes. Steedman 1988). They have been judged less use- It is possible that the characterization of vari- ful, on balance, than the integrative indicator ous features within a particular class-that is, the species approach (Kauffman and others 1992; contents of a particular column in table 22-2- Regier 1992b). may be only one set of a number of possible sets 348 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Table 22-2: Characteristics of Three Mind-sets Related to Sustainability of Renewable Resources and the Natural Environment Characteristic Unidirectional causal Random process Mutual causal Science Traditional cause-and-effect Classic thermodynamics, Open system thermodynamics, model Shannon information theory post-Shannon information theory Information Past and future inferable form Information decays and Information can be generated, gets lost, blueprint must nonredundant complexity can contain more information be generated without than finished product preestablished blueprint Cosmology Predetermined universe Decaying universe Self-generating and self- organizing universe Social organization Hierarchical Individualistic Nonhierarchical interactionist, holarchic Social policy Homogenistic Decentralized Heterogenistic coordination Ideology Authoritarian Anarchistic Cooperative Philosophy Universalism Nominalism Network Ethics Competitive Isolationist Symbiotic Aesthetics Unity by similarity and Haphazard H-larmony of diversity repetition Religion Monotheism Freedom of religion Polytheistic harmonism Decision process Dictatorship, majority rule, Do your own thing Elimination of hardship on or consensus individuals, communitarian Logic Deductive, axiomatic Inductive, empirical Complementary Perception Categorical Atomistic Contextual Knowledge Believe in one truth, if people Why bother to learn beyond Polyocular: learn are informed, they will agree one's own interest different views and take them into consideration Methodology Classificational, taxonomic Statistical Relational, contextual analysis, network analysis Research hypothesis Dissimilar results must have There is probability Dissimilar results may come from and strategy been caused by dissimilar distribution; find out similar conditions due to mutually conditions; differences must probability distribution amplifying network; network be traced to conditions analysis instead of tracing the producing them diffcrence back to initial conditions in such cases Assessment Impact analysis What does it do to me? Look for feedback loops for self- cancellation or self-reinforcement Analysis lPreset categories used for Limited categories for Changeable categories all situations one's own use depending on the situation View of community Ignorant, poorly informed, Egocentric Most direct source of information, people as lacking expertise, limited in articulate in their own view, essen- scope tial in determining relevance Planning By experts; either keep Laissez-faire Generated by community community people uninformed, people, learning by doing or inform them so that they will agree Source: Maruyama 1974. in each. Thus the mind-set of the second column the table: the second may relate to regional (as may have a second variant relevant to bureau- opposed to local) systemsofatypenowemerging cratically dominated centralized economies that in the European Economic Community or in the are (or were) managed primarily for the benefit of Great Lakes Basin bioregion. a favored class or nomenklatura. Also the third Maruyama's schema presupposes that the idea mind-set may have another variant besides the of fully objective science is an extreme abstrac- traditionial communitarian approach shown in tion. That sustainability could usefully be ad- 349 Defining and Measuring Sustainability: The Biogeophysical Foundations dressed objectively with respect only to its bio- Figure 22-2: Illustration of How Three Mind-sets physical foundations is similarly unrealistic, as is Relate to Different Academic Disciplines implied in previous sections above. Generally it is now recognized thatan ethics of improved equity Dasciphnes and Protessiorns must go hand-in-hand with an ethics of Phvsical Environniental Health Social Pohtcal sustainability (WCED 1987) and that both will and Manager," and Economic and influence the selection of biophysical measures of Engineenng Mecdcimn Science, CGoveniance sustainability. If Maruyama's schema has some relevance, then the epistemology (the dominant enquiry methods) and aesthetics (what people Emergesi enjoy)-both relevant to a primary mind-set- evolutionary will also influence the choice of measures of opensvstem, new SLIenLe sustainability. Decisionmakingin such issuesin westerncoun- - tries is ostensiblyquite pragmatic: a goal or vision E Chan that is widely shared in society is identified and re endorsed through a participatory or representa- c, > chast sterii, tive poli tical process, and means are then created to work toward achieving thegoal orrealizing the -z Causal,% vision. Academic emphases on abstract features predetermuned of different mind-sets within the policv may fos- closed system, ter discord and hinder achievement of the shared purpose, at least in the short run (Norton 1991). We grant Norton's pragmatic argument Abiotic BIots. Psschosocial Cultur sketched above. Nevertheless scientists tend gen- Types, of Phenomeiu erally to be disciplined into particular mind-sets, paradigms, methodologies, rituals, and so forth. Note: Scientific and teclhnical professionals commonly Scientists tend to ignore the advice of the oracle to balablnce their work or Iterate between three different m snd- Scieniststend o igore he adice f theorace to sets Few have ani explicit strategy orp)rocedure for doing so. know thyself! They tend to select items from The ecosvstem approach is Just one o a X arietv uf initiatives that inivolve some mixture of the three, with ah) emiphasis on within their received disci plinary mind-set, add a elaborating conicepts anid mieLhods in the new science modicum of intellectual or practical value to the items, and then offer them to their peers or sell them to some client. This general process may be Presumably a set of measures of sustainability discernible within a quasi-competitive process will need to serve all major interests as reflected such as a conference to identifv appropriate bio- in the contemporary political balance concerning physical measures of sustainabilitv. Unless the the natural phenomenon to be sustained, for ex- processby which such measuresaretobeselected ample, with respect to a particular fishery. Also if is intended to be entirely pragmatic, some further only one measure of sustainability is sought for a attention to these mind-sets may help to under- particular natural phenomenon, then that mea- stand the proposals of different scientists. sure should be relevant within all the mind-sets, Figure 22-2 is an attempt to show how three if a workable, pragmatic consensus is needed to mind-sets, very approximately as expanded realize sustainability. Further, the actions pro- from the sketch by Maruyama (1974) given in posed to achieve sustainability of a particular table 22-2, relatecurrently todifferent academic natural feature must take into account, and per- disciplines that may have something to contrib- haps help to modify, the evolving nature of hu- ute to the issue of sustainability. Compared to man societies, that is, they must be oriented to- physicists, on the one hand, and political scien- ward process. tists, on the other, ecologists as a set, at about It is obvious that the analysis sketched above themiddleoftheleft-rightspectrum,maycom- does not lead directly to the identification of a prise three subgroups of roughly equivalent particular measure of sustainability. But it may strength. An ecological set, as it relates to help to explain why different scientists prefer sustainability in fisheries, is examined further differentmeasuresorwhyparticularinterestgroups in table 22-3. in society favor one measure over another. 350 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Table 22-3: Different Schools within Ecology, by Broad Mind-sets and Types of Fisheries Representative Type of Mind-set Key concepts researchersa fisheries Causally predetermined Environmental determinism with G. A. Gulland, Whaling and sealing; large reality, closed system respect to spatial occurrence of M. B. Schaefer finfisheries and shellfisheries particular species as fixed entities in shelf seas or ecological production at a particular trophic level Chance-driven reality, Ecological association due entirely P. A. Larkin, Moderate-scale artisanal stochastic system to the adaptive capabilities of C. J. Walters fisheries, especially on individual organisms of species anadromous stocks; populations subject to natural aquaculture selection through unpredictable environmental fluctuations Emergent evolutionary Ecosystem integrity or harmony H. A. Regier, Small-scale artisanal reality, open system due to self-organizing capabilities R. A. Ryder fisheries; regional fisheries of many living components of the and bays holarchically nested open system a. Fisheries scientists do not usually limit their involvement to a particular paradigm, hence the authors listed here have been involved with, but are not limited to, the relevant mind-set. Role of science in sustainable development prominent in the late 1960s and early 1970s, with efforts to reform reflected in the incorporation of Friedmann (1987), as discussed by Dorcey (1991), a concept of integrity in legislation, such as the has identified four traditions concerning the role of 1 U.S. WaterP tin C ontrol A n e science (ormorebroadly, knowledge) on issues like 1972 U.S. Water Pollutilon Control Amendments. sustainability. Theseinclude policy analysis, which inoe science related tofisiestin ioregio envisages science as capable of providing the best contextsc Kauftman and others 992;RCFiW 1992). solutionswithinexisting social and economic struc- It is, at least, evident that viewson appropriate tures; social reform, which envisages the applica- measuresofsustainabilityareaffectedbypercep- tion of science to making government action more tions of the tappropriate re ofece in te effective, with planning as a preeminently scientific nteraction between society and the natural envi- activity,andmakingconventionalpoliticssubordi- ronment. Understanding these perceptions may nate; social learning, in which science, and social aonhelp explan differe scentists oie experimentation, can contribute to incremental so- est gops eprefe different measures of cietal change; and social mobilization, which, in est groups prefer dafferent measures of contrast to the preceding three, asserts the primacy of direct collective action from below to transform . . . society, with science playing some supportive role in this transformation. The concept of sustainability presupposes that It is unlikely that any formulation of science in the living systems of interest can accommodate sustainabledevelopment fits neatly within a single some use and abuse without collapsing or trans- of these traditions. Nevertheless, certain patterns forming into a less desirable kind of living sys- areapparent.Strictrelianceonabiophysicalmea- tem. Thus the living system is expected to have sure of sustainability, for example, incorporates some adaptive capabilities to accommodate our elements of the traditions of policy analysis and interventions and presumably also some recu- social reform. Considerations of equity, and of perative capabilities to reconstitute itself in some the significant political and social factors within way following some interval of inevitable excess each of the six kinds of fisheries described above, on our part. That there are limits and thresholds may be grounded primarily within the tradition to such accommodation, in the sense of Holling of social learning. This tradition became more (1986), has been explored briefly for the Great 351 Defining and Measuring Sustainability: The Biogeophysical Foundations Lakes (Steedman and Regier 1987), with an infer- References ence that such discontinuities can and do occur with these ecosystems and with fish populations Alward, G. L. 1911. The Development of the British in them. Fisheries during the Nineteenth Century with Spe- Weintervenewithlivingphenomenonatmany cial Reference to the North Sea. Grimsby, En- levels from the suborganismal to the biosphere. gland: Grimsby News. Presumably all levels contribute to the American Fisheries Society. 1990. "Proceedings sustainability of each level, and all levels of such of theSymposiumon Effectsof ClimateChange a holarchicsystemhave roughlyequivalent value on Fish." Transactions of the American Fisheries for us and deserve the necessary protection and Society 119:2, pp. 173-389. sustenance. In all of the holarchic levels, a capability for Beak Consultants Ltd. 1991. Duffin Creek Water- normal self-organization istheessential biophysi- shed Study: Fisheres Component (Phaseo). Final cal foundation. But what is normal self-organiza- report for the Metropolitan Toronto and Re tion for the increasing proportion of ecosystems gion Conservation Authority. Downsview, in which cultural forces rival in strength the natu- Ont. Beak Consultants, Ltd. ral forces?Thisquestion maydeserve more atten- Berkes, F., ed. 1989. Common Property Resources: tion before we try to decide what a sufficient set Ecology and Community-Based Sustainable De- of measures of sustainability would be. For the velopment. London: Bellhaven Press. Great Lakes, and especially for the fish and fish- Beverton, R. J. H., and S. J. Holt. 1957. On the eries of the Great Lakes, it may now be timely to Dynamics of Exploited Fish Populations. London: stipulate, through a trans-jurisdictional political Her Majesty's Stationery Office. process and to a practically sufficient degree, Bocking, S. 1987. "The Origins of Aquatic Eco- what and how much of it should and can be logical Research in the Great Lakes Region." sustained followingappropriate remediation and M.A. thesis, University of Toronto, Toronto. rehabilitation. A constraint to keep transactional costs within bounds would spur careful consider- Cairns, J., Jr., p. McCormick, and B. Niederlehner. ation of choices of measures. Then all the statisti- 1991. "AtProposed Framework for Developing cal questions related to sampling and estimation Indicators of Ecosystem Health for the Great can be addressed effectively. Lakes Region." Council of Great Lakes Re- Wecannot say whatmanifestationsof sustain- search Managers, International JointCommis- able states of fisheries will be considered most sion, Ottawa, Canada, and Washington, D.C. desirable by future generations and suspect that Commissioners Appointed to Inquire into the Sea preferences will in any case continue to evolve Fisheries of the United Kingdom. 1866. Report with time. Currently the characteristics of fisher- of the Commissioners Appointed to Inquire into the ies that are seen to be desirable on a sustainable Sea Fisheries of the United Kingdom. London: Her basis differ considerably from one societal grou p Majesty's Stationery Office. to another and from one culture to another, ow- Cushing, D. H. 1988. The Provident Sea. Cam- ing to differing economic and social conditions. bridge, England: Cambridge University Press. Just as we expect component living systems to be Dorcey, A. H. J. 1991. "Towards Agreement in able to accommodate some degree of external Water Management: An Evolving Sustainable intervention, we also expect a range of states to Development Model." In A. H. J. Dorsey, ed., exist in which fisheries might be sustained ac- Perspectives on SustainableDevelopment in Water cording todifferingmixes of levels at which other Management: Towards Agreement in the Fraser human activities are being sustained. Achieving River Basin, pp. 555-86. Vancouver, B.C.: and managing sustainability are thus likely to Westwater Research Centre. continue to be a dynamic process and to evolve as humanneedsandvalueschange,bothlocallyand Edwards, C. J., and H. A. Regier, eds. 1990. An globally. It is in this sense that we view Ecosystein Aproach to the Integrity of the Great sustainability and development, however incon- Lakes in Turbulent Times. Special Publication sistent they may appear to be with each other, as 90-4. Ann Arbor, Mich.: Great Lakes Fishery inextricably mixed. Commission. 352 Sustainability of Temperate Zone Fisheries: Biophysical Foundations for Its Definition and Measurement Edwards, C. J., and R. A. Ryder. 1989. "Biological Holt, S. J., and L. M. Talbot. 1978. New Principles Surrogates of Mesotrophic Ecosystem Health for the Conservation of Wild Living Resources. in the Laurentian Great Lakes." Great Lakes WildlifeMonographs59.]ournalofWildlifeMan- Science Advisory Board, International Joint agement 43 (supplement). Commission, Windsor, Ont. Huxley, T. H. 1884. "Inaugural Address of the Egerton, F. N. 1985. "Overfishing or Pollution? Fishery Conferences." Fisheries Exhibition Lit- Case History of a Controversy on the Great erature 4, pp. 1-19. Lakes."TechnicalReport4l.AnnArbor,Mich.: Karr, J. R., K. D. Fausch, P. L. Angermeier, P. R. Great Lakes Fishery Commission. Yant, and I. J. Schlosser. 1986. "Assessing Bio- FAO (Food and Agriculture Organization of the logicalIntegrityofRunningWaters:A Method United Nations). 1973. "Proceedings of the and Its Rationale." Illinois Natural History Food and Agriculture Organization of the Survey Special Publication 5. Champaign, Ill. United Nations, Technical Conference on Fish- Kasahara, H. 1961. "Fisheries Resources of the ery Management, Vancouver, B.C., A. W. H. North Pacific Ocean, Part 1." H. R. MacMillan Needler (Chair)." Journal of the Fisheries Research Lectures in Fisheries. Vancouver, B.C.: Univer- Board of Canada 30:12 (part 2), pp. 1925-2537. sity of British Columbia. Francis, G. R., J. J. Magnusson, H. A. Regier, and Kauffman, J., P. Rennick, H. A. Regier, J. A. D. R. Halhelm. 1979. "Rehabilitating Great Holmes, and G. A. Wichert. 1992. "Metro Wa- Lakes Ecosystems." Technical Report 37. Great terfront Environmental Study." Metropolitan Lakes Fishery Commission, Ann Arbor, Mich. Toronto Planning Departments, Toronto, Ont. Friedmann, J. 1987. Planning in the Public Domain: Larkin, P. A. 1965. "North American Fishery Po- From Knowledge to Action. Princeton, N.J.: tential." In The Fisheries of North America: The Princeton University Press. First North American Fisheries Conference, April Garstang, W. 1900. "The Impoverishment of the 30-May 5, 1965. Washington, D.C.: Govern- Sea." Journal of Marine Biological Association ment Printing Office. (United Kingdom) 6, pp. 1-69. . 1977. "An Epitaph for the Concept of GLSAB (Great Lakes Science Advisory Board). Maximum Sustainable Yield." Transactions of 1989. Report of the Great Lakes Science Advisory the American Fisheries Society 106, pp. 1-11. Board to the International joint Commission. Loftus, K. H., and H. A. Regier, eds. 1972. "Proceed- Ottawa, Ont., and Washington, D.C. ingsof the 1971 Symposium onSalmonid Com- GLWQA (Great LakesWaterQuality Agreement). murities in Oligotrophic Lakes." Journal of Fish- 1987. "Revised Great Lakes Water Quality eries Research Board of Canada 29, pp. 613-986. Agreement of 1978." International Joint Com- Marsh,G.P. 1857. "ReportMadeunderAuthority mission, Ottawa, Ont., and Washington, D.C. of the Legislature of Vermont on the Artificial Graham, M. 1935. "Modem Theory of Exploiting Propagation of Fish." Free Press Print, a Fishery and Application to North Sea Trawl- Burlington, Vt. ing." Journal du Conseil International pour Maruyama, M. 1974. "Paradigmatology and Its l'Exploration de la Mer 10, pp. 264-74. Application to Cross-disciplinary, Cross-pro- . 1943. The Fish Gate. London: Faber. fessional, and Cross-cultural Communication." Great Lakes Fishery Commission. 1992. "Strate- Cybernetics 17, pp. 136-56. gic Vision of the Great Lakes Fishery Commis- McHugh,J. L. 1970. "Trendsin Fishery Research." sion for the Decade of the 1990s." Ann Arbor, In N. G. Benson, ed., A Century of Fisheries in Mich. North America,pp.25-56. Bethesda, Md.: Ameri- Gulland, J. A. 1983. "World Resources of Fisher- can Fisheries Society. ies and Their Management." In 0. Kinne, ed., Meyer, P. F. 1947. "Raubbau in Meer? Ein Beitrag Marine Ecology, vol. 5, part 2, pp. 839-1060. zur Uberfischungsfrage unserer Meere." Hans London: John Wiley and Sons. A. Kuene Verlag, Hamburg, Germany. Holling, C. S. 1986. "The Resilience of Terrestrial Milner, J. W. 1874. "Report on the Fisheries of the Ecosystems: Local Surprise and Global GreatLakes:TheResultoflnquiriesProsecuted Change." In W. C. Clark and R. E. Munn, eds., in 1871 and 1872." In Report, 1872-73 (U.S. Sustainable Development of the Biosphere, pp. Fish Commission). Washington, D.C.: Govern- 292-317. Cambridge, England: Cambridge ment Printing Office. University Press. 353 Defining and Measuring Sustainability: The Biogeophysical Foundations Nettle, R. 1857. The Salmon Fisheries of the St. Russell, E. S. 1931. "Some Theoretical Consider- Lawrence and Its Tributaries. Montreal, Quebec: ations on the Overfishing Problem." Journal du John Lovell. Conseil International pour l'Exploration de la Mer 6, Norton, B. G. 1991. Toward Unityamong Environmen- pp. 3-20. tzlists. New York: Oxford University Press. . 1939. "An Elementary Treatment of the Petersen, C. B. J. 1903. "What Is Overfishing?" Jour- Overfishing Problem." Rapport Proces-Verbaux nal of Marine Biological Association (United King- des Reunions (Conseil International pour dom) 6, pp. 587-94. I'Exploration de la Mer) 110, pp. 5-14. Pinkerton, E., ed. 1989. Co-Operative Management of Ryder, R. A., and C. J. Edwards, eds. 1985. A Local Fisheries: New Directions for Improved Man- Conceptual Approach for the Application of Bio- agement and Community Development. Vancouver, logical Indicators of Ecosystem Quality in the Great B.C.: University of British Columbia Press. Lakes Basin. International Joint Commission, Windsor, Ont., and Great Lakes Fishery Com- Pisani, D. 1984. "Fish Culture and the Dawn of mission Ann Arbor Mich. Concern over Water Pollution in the United ' ' States." Environmental Review 8, pp. 117-31. SelectCommitteeonSea Fisheries. 1893. Reportof the Raleigh, R. F. 1982. "Habitat Suitability Index Mod- Select Committee on Sea Fisheries. London: Her els: Brook Trout." FWS/OBS-82/10:24. U.S. - Majesty's Stationery Office. partrnent of the Interior, Washington, D.C. Steedman, R. J. 1988. "Modification and Assess- Rapot,D..C. Hutchinson. ment of an Index of Biotic Integrity to Quantify Rapport, D. J., H. A. Regier, and T. C.Hthno. Stream Quality in Southern Ontario." Cana- 1985. "Ecosystem BehaviorunderStress." Amen- . . . . C can Naturalist 125, pp. 617-40. dian Journal of Fisheries and Aquatic Sciences 45, Rawson Academy of Aquatic Science. 1989. To- p 492-501. wards an Ecosystem Charterfor the Great Lakes-St. Steedman, R. J., and H. A. Regier. 1987. "Ecosys- Lawrence. Occasional Paper 1. Ottawa, Ont. tem Science for the Great Lakes: Perspectives on Degradative Transformations." Canadian RCFTIW (Royal Commission on the Future of the Journal of Fisheries and Aquatic Sciences 44 Toronto Waterfront). 1992. Regeneration,Toronto's (supplement 2), pp. 95-103. Waterfront and the Sustainable City: Final Report. Thompson, W. F., and N. Freeman. 1930. "History David Crombie, Commissioner. Toronto, Ont. of the Halibut Fishery." Report of the International Regier, H. A. 1992a. "Ecosystem Integrity in the Fishery Conmission (Seattle, Wash.) 5. Great Lakes Basin and Historical Sketch of Troadec, J. P. 1983. "Practices and Prospects for Ideas and Actions." Journal of Aquatic Ecosys- Fisheries Development and Management: The tem Health 1, pl. 25-37. Case of Northwest African Fisheries." In B. J. .1992b. "Indicators of Ecosystem Integ- Rothschild, ed., Global Fisheries: Perspectives for rity." In D. Mackenzie, ed., Proceedings of the the 1980s, pp. 97-122. New York: Springer- International Symposiutn on Ecological Indica- Verlag. tors, Ft. Lauderdale, Florida, September 1990, pp. United Nations. 1984. United Nations Convention on 183-200. Barking, England: Elsevier Applied the Law of the Sea. New York: United Nations. Science Publishers. Regier,-H. A.,andV.C. Applegate. 1972. "Histori- Van Hise, C. R. 1911. The Conservation of Natural caglReview of the.Managemenate Approach torl- Resources in the United States. New York: cal Review of the Management Approach to MacMillan Company Exploitation and Introduction in SCOL Lakes." P Y Journal of Fisheries Research Board of Canada 29, WCED (World Commission on Environment and pp. 183-692. Development). 1987. Our Common Future. Re- Ricker, W. E. 1946. "Production and Utilization of port of the World Commission on Environ- . ' . . ,< ~~~~~~ment and Development, chaired by G. H. Fish Populations." EcologicalMonograph 16, pp. metadDvlpet chie by G.H 374F9i P Brundtland. Oxford, England: Oxford Univer- 374-91. ~~~~~~~~~~~sity Press. - .1954. "Stock and Recruitment." Journal of Wet, A. Fisheries RLsairch Board of Canada I1, pp. 559-623. Went, A. E. J. 1972. "Sevent,y Years Agrowing: A FseiseerhodofCaa1, 55 2 History of the Intemational Council for the Ex- 1958. "Handbook of Computations for ploration of the Sea, 1902--1972." Rapport Proces- Biological Statisticsof Fish Populations." Bulletin Verbauxdes Reunions (Conseil International pour of the Fisheries Research Board of Canada 119. I'Exploration de la Mer) 165. 354 Sustainability of Managed Temperate Forest Ecosystems Jerry F Franklin We would expect foresters to know a lot about some suggestions for a minimal monitoring pro- sustainable management. Forests take long peri- gram. An extensive section on alternative man- ods of time to develop, and foresters have been agement approaches follows; development and managing temperate forest ecosystems for sev- application of alternative silvicultural and land- eral centuries. They are used to taking a long scape practices offer immense potential for inte- view, planning for forests and planting trees that grating sustainable production of environmental are not likely to be harvested within their profes- goods and services with commodities. A section sional, and often personal, life span. At the same describingexistingapproaches, models,and data time, the concept of sustainability in forestry has sets follows. The review concludes with a series often been narrow and limited to continued pro- of proposals for managed temperate forests to (a) duction of wood fiber. Similarly, our scientific develop critical scientific information, (b) imple- understanding of the basis for forest productivity ment and test new management systems, and (c) and techniques for assessing trends in productiv- assess long-term productivity. ity, especially of the entire ecosystem and over long periods of time, is not very robust. This chapter reviews what we know about the Definition of sustainability sustainability of managed temperate forest eco- systems. Since this is such an immense topic, the Sustainability refers to the maintenanceof thepoten- review is primarily an overview with an empha- tialforourforestand associated aquaticecosystems sis on recent knowledge and emerging concepts to produce the same quantity and quality of goods of the productivity and maintenance of forest and services in perpetuity. Potential isemphasized ecosystems rather than a comprehensive review since it makes implicit the option to return to alter- of the last 100 years of forest science. native conditions rather than focusing exclusively The chapter begins with a definition of oncurrentconditions.Thisconceptofsustainability sustainability, taking a broad view of forest pro- considers a broad range of goods and services. It ductivity and sustainability rather than simply includes, for example, retaining the forest's capac- focusing on the production of wood products. itytoprovidefunctionalservices,suchasregulating The first major section considers the status of our the flow of streams and minimizing the loss of knowledge of major ecosystem processes related nutrients and soil as a result of erosion. It means an to sustainability; it will be clear that much critical abilitytoprovidehabitat,eithercurrentlyoratsome informationonlong-termproductivit;vislacking, future time, for the full array of animal and plant particularly on the soil ecological subsystem, and organismson the site. And, of course, sustainability other data, such as on respiration, are nearly means the continuing capacity to provide the same impossible to gather. Next, biophysical measure- quantity and quality of products for human con- ments of sustainability are considered along with sumption. Defining and Measuring Sustainability: 'The Biogeophysical Foundations The basis for sustainability lies in maintaining paradigm is of relatively recent origin. Major the physical and biological elements of produc- boosts to our understanding of forests have come tivity. Hence, sustainability requires that we pre- as a result of recent research programs that have vent the following: focused on forests as ecosystems. The contribu- • Degradation of the productive capacity of our tions of the International Biological Programme forest lands and the associated water bodies, are particularly notable at both the international that iand national levels (Reichle 1981; Edmonds 1981). that is, net loss of productivity, and Programs centered on individual sites, such as * Loss of genetic diversity, including extirpation Coweeta Hydrologic Laboratory (North Caro- of species, that is, net loss of genetic potential. lina; Swank and Crossley 1988), the Hubbard Each of these principles has both an ecological Brook (New Hampshire; Bormann and Likens and an ethical basis; even though they are human 1981), and H. J. Andrews (Oregon; Edmonds 1981) constructs, they can be objectively defined in experimental forests, have also contributed major ecological terms. Principle two-no net loss of advances to the knowledge of forest ecosystems. genetic potential-is probably the most funda- A general review of the major components and mental, since we can sometimes restore produc- processes of ecosystems that underline produc- tive capacity to degraded ecosystems but have tivity is provided in this section, which considers only very limited capacity to restore lost genetic the physical and biotic elements that are essential potential. No principle, in my view, is absolute or to the productivity of temperateforests.The physi- inviolate.Therewillbetimeswhenrational,even cal variables that act directly on the biota (the ecologically sensitive, human beings will violate operational environment) aredistinguished from either principle. But when such violations occur, second-and third-orderenvironmental variables, they should be done with society's full knowl- such as elevation and aspect, which are indirect edge of the act and its consequences, not as a influences. Although a secondary variable, soils result of ignorance and not in secrecy. are considered in some detail because of their Sustainability absolutely should not be viewed influence on moisture and nutrient regimes, their exclusively or primarily in terms of the short- importance to sustainability, and their suscepti- term production of specific commodities, such as bility to human influences, both positive and sawlogs or trophy ungulates, although such con- negative. The biotic components of productivity, cerns are an appropriate component of a concept including the "ecosystem support staff" of smaller of sustainable forestry. Assuming the above, sus- organisms, such as decomposers, are considered tainable practices for managed temperate forest along with the photosynthetically active primary ecosystems should place a very high priority on producers, such as trees. practices that meet the dual standards of main- Identification and discussion of importanteco- taining (a) productive capacity and (b) genetic system processes, such as productivity and de- diversity. It is essential to maintain a broad view composition, are also covered. Ecological defini- of productivity-goods and services-and of the tionsof productivityare presented and contrasted spatial and temporal scales to which it is to be withtraditionalforestrydefinitions.Naturalvaria- appl ied rather than to adopt a narrow construct that tion in rates of ecosystem processes and recovery focuses solely on the production of wood fiber. rates are also considered. Phlysical comnponents of productivity Status of knowledge of major Productivity comprises physical elements (the ecosystem processes operational environment) and physical variables (such as soil and elevation). There is a substantial base of knowledge on the components and processes that are the basis of Ti IE OPERATIONAL ENVIRONMENT productivity in temperate forest ecosystems. For- At the most fundamental level, the biota respond esters and forest scientists have contributed sub- to a relatively small set of physical factors: light, stantially to this base during the last century, carbondioxide,temperature,moisture,nutrients, although their contributions have tended to be mechanical forces, and toxic chemicals. This in- narrowly focused on trees, wood production, and c1Ludes that part of the biota responsible for pro- managed forests. Furthermore, the ecosystem viding the energy base for the whole ecosystem, 356 Sustainability of Managed Temperate Forest Ecosystems for example, organisms with chlorophyll that are endemicoperational environment arecatastrophic capable of photosynthetically capturing the sun's mechanical disturbances that essentially destroy energy. These physical elements are sometimes the existing ecosystem and initiate a new one, referred to as the operational environment in an such as an intense wildfire or volcanic eruption. effort to distinguish them from physical vari- Toxic materials provide a chemical equivalent ables, such as soil, elevation, or aspect, that indi- to the physical forces in the operational environ- rectlyinfluencebiotic activity through their effect ment of the forest. These can be natural materials on these operational factors, the ones that the that are found in the soil or atmosphere of a biota are actually sensing and to which they are region. For example, excessive magnesium levels responding (Waring and Major 1964). in the ultrabasic metamorphosed type of rock Several of these variables can be quickly dis- known as serpentine effectively exclude many posed of in this review since they are generally organisms and retard growth of many others. viewed as constants in temperate forest regions. High atmospheric levels of sulphur compounds Light, for example, is typically not a limiting associated with hydrothermal vents might be factor in temperate forest regions. Sufficient light another example of a natural chemical toxin. is generally available throughout the year even However, most of the toxic chemicals that are though the intensity is obviously greater in the discussed todayareofanthropogenicorigin(Aber summer than in the winter. Hence, light does not and others 1989), such as concentrations of atmo- seriously constrain photosynthetic activity even spheric ozone and acid fogs. As with elevated during winter months, assuming that tempera- levels of carbon dioxide, an analysis of the effects ture and moisture conditions are suitable. Tem- of various atmospheric and soil pollutants of an- perate forests in northwestern North America thropogenic origin is beyond the scope of this re- provide a good example of this, since mild, wet view and are not considered here. winters allow a substantial amount of the annual photosynthesis to occur outside the growing sea- MAJOR OPERATIONAL VARIABLES son (Edmonds 1981, chapter 10; Waring and The operational environmental elements of tem- Franklin 1979). perature, moisture, and nutrients stand out as The carbon dioxide content of the atmosphere variables responsible for mostof the variability in is, in contrast, often viewed asa limiting factor for productivityamong temperateforestecosystems. productivity and, over the short term, as a con- Levelsofthesevariablesvarywidelyamongtem- stant. The atmospheric concentration of carbon perate forest sites on continental, regional, and dioxide is, of course, gradually increasing, and even local scales. Multivariate analyses of the the potential impact of this increase on productiv- operational environment invariably identify some ity of green plants is a controversial topic cur- combinationoftemperature,moisture,andnutri- rently discussed and debated by scientists and ents as major controllers of distribution and pro- participants attempting to assess impacts of glo- ductivity of the forest community (see Gholz bal changeon productivity. Predictions varyfrom 1982; Zobel and others 1976). little or no response to the increased levels of Temperature is typically the primary opera- carbon dioxide to predictions of significant in- tional physical factor controlling the distribution creases in plant productivity. An analysis of how and productivity of forest ecosystems over re- productivity responds to increases in carbon di- gions and elevational gradients. Local variations oxideisbeyond thescopeof this reviewand is not in temperature regimes can also be substantial, considered further here; the reader is referred to however, such as on steep slopes of contrasting Adams and others 1990; Bazzaz 1990; Easmus aspect or as a result of topographic conditions, and Jarvis 1989; and Jarvis 1989 for discussions of such as a depression that accumulates and forms this topic. a frost pocket. Temperature directly controls vari- Mechanical forces actually cover a variety of ous chemical and physical processes (such as mechanical effects that can limit productivity photosynthesis, decomposition, and water uptake) through theirdirectimpacton the green plantsor and, indirectly, through its effects on moisture re- other ecosystem processes. Examples include gime, controls potential evapotranspiration. heavy snow or ice loads that damage or break Moisture, as an operational element, is sensed trees; powerful winds that break twigs, branches, by trees and otherplants in termsof internal plant and boles of trees; and floods. Excluded from the moisture stress (Waring and Schlesinger 1985). 357 Defining and Measuring Sustainability: The Biogeophysical Foundations The moisture regime is, however, the result of a can be very long lasting. complexinteractioninvolvingthebalancebetween As a physical medium, soils provide several water uptake, which is generally from water stored functions: storage of moisture (a portion of which in the soil, and water loss, which is primarily the isavailabletoplants),sourceandstorageofnutri- result of loss to the atmosphere through stomata ents, site for anchorage by plants, and habitat for or openings in the leaves. Hence, daily and sea- critical plant symbionts, such as mycorrhizal- sonal patterns in the intensity of the gradient forming fungi and other organisms essential to from soil to plant to atmosphere are critical. Many ecosystem processes. Both physical and chemical indirect measures are used to provide an inte- aspects of the soil are important. Physical aspects grated index to overall moisture conditions of a include such variables as depth, drainage, bulk forested site. Direct measurements typically mark density, porosity (especially macropores), tex- internal plant moisture stress at selected times of ture, and temperature; these variables, in turn, year using pressurized chambers. influence conditions critical to the biota, such as Nutrients are a third operational variable that aeration (oxygen content), capacity to hold mois- typically controls forest productivity. Indeed, ture, and availability of moisture. Important nutrientshavereceivedanextraordinaryamount chemical aspects of the soil include quantities of attention from foresters because this is often and qualities of the various macro- and micro- the only environmental variable that is readily nutrients and the rates at which they are made subject to human manipulation. Included here available. would be all of the macro- and micronutrients that green plants require, although some, such as Biotic components of productivity nitrogen, have received much more attention than others. There is substantial regional variance in Froma perusal of any soil textbook, itis clearthat therelativeimportanceofnutrientsandmoisture we know quite a bit about the chemical and as limiting factors on productivity. In much of the physical processes of soil, but not nearly as much world's temperate mesic forests, moisture isavail- about theirbiota and biological functioning(Jenny able throughout most of the summer; analyses of 1980). Biota provide the other essential compo- such areas, such as eastern North America and nents of productivity. This includes the primary eastern Asia, including Japan, typically identify producers of theecosystem-theorganisms with nutrients as a more important variable than mois- chlorophyll that arecapable of capturing the sun's ture. In contrast, moisture is typically ranked as energy through photosynthesis-and the most more important than nutrients in northwestem important of these in forest ecosystems are the North America, where there is substantial mois- trees. It also includes many other essential organ- ture deficit in summer (Edmonds 1981; Waring isms that support the ecosystem: plants, animals, and Franklin 1979; Zobel and others 1976) fungi, monera (such as bacteria), algae, and pro- In conclusion, the physical operational envi- tozoa that decompose organic substances, make ronment of a forest can be defined in terms of nutrients available, and assist the primary pro- relatively few variables. The most important of ducers. Fungi that form mycorrhizae with vascu- these in influencing productivity are tempera - larplants,therebyfacilitatingmoistureandnutri- tures moisturen and nutrients. ent uptake from the soil, are a classic example of the latter, although other relationships may be of SOIL AS AN ELEMENT OF PRODUCTIVITY comparable importance, as in the case of the Foresters often focus on soil as the physical basis endophytic communities found on leaves and for productivity, which is reasonable given the needles (Carroll 1980). direct relationship that soil has to two of the three The importance of the biotic components to sus- operational variables: moisture and nutrients. It tained productivity of the ecosystem should be is theamountandconditionof thesoil that largely obvious; nevertheless, it seems to be absent from control the moisture and nutritional regime to manydiscussionsofproductivityand sustainability which the tree issubjected. Consequently, human and is not made explicit in many others. The prin- impacts on the soil's ability to provide water and ciple that sustainability requires maintenance of nutrients can have a dramatic impact on forest genetic diversity explicitly recognizes the impor- productivity and may be either positive or nega- tance of the biotic components and may be, in fact, tive (Grier and others 1989; Harvey and the most important practical reason for conserving Neuenschwander 1991). Furthermore, impacts biological diversity in all of its forms. 358 Sustainability of Managed Temperate Forest Ecosystems PRIMARY PRODUCERS Besidesdetermining the rateof energy fixation Relatively little needs to be said about the impor- or production for the site, primary producers also tance of green plants to the productivity of our have the potential to alter significantly the soil's forest ecosystems. Green plants, and specifically physical and chemical conditions. This can have trees, are the basic agents for capturing energy, either positive or negative effects on the long- the energetic base on which the entire ecosystem term potential of a site and obviously should be operates. Forest ecologists generally recognize considered in selecting genotypes, species, or that different species of trees, either singly or in combinations of species for management pur- combination, have different capabilities to cap- poses. The ability of some families or genera of ture the potential productive capacity of a site. vascular plants to support nitrogen-fixing sym- The fact that different genotypes of the same bionts in root nodules is a well-known example; species maydiffermarkedly in their productivity representative tree genera with this ability are on the same site is also generally understood and Acacia, Alnus, and Robinia. Some tree species, is the basis for various tree breeding programs to such as members of the Cupressaceae, as well as develop genotypes that have improved capabili- many deciduous hardwoods, produce a base- ties, such as more rapid growth or greater resis- rich litter that, among other things, reduces acid- tance to disease. ity, increases levels of nutrients, and results in Nevertheless, there are many important gaps richerand moreactivecommunitiesof organisms inourunderstandingofhowspeciescomposition in the soil. Other tree species produce litter that affects either short- or long-term forest produc- increases soil acidityand decreasesavailabilityof tivity. For example, definitive theoretical and soil nutrients. Picea and Tsuga are well-known empirical information is still lacking on the rela- coniferous examples. An extreme example of the tive yield of mixtures of species versus a single- negative effect of specific tree species on soil species monoculture. This is also true of contrast- properties is Eucalyptus, which, over time, gener- ing forest structures: yields from an even-aged ates beneath it a bleached, nutrient-poor zone stand with a single canopy layer versus an un- sometimes referred to as an eggcup podzol. even-aged stand with multiple layers of canopy. Clearly, we need to recognize explicitly (a) the The fact that trees and their production are importance of the genetic component of the pri- typically used to assess potential productivity of mary producers at both the specific and intraspe- a forest site creates further complications. For cific levels in influencing attainable short- and example, trees and genotypes indigenous to a long-term productivity and (b) the circular prob- locality may not be as capable of exploiting the lemsinherentinusingtheproductivityoftreesas productive resources of a site as exotic species. the measure of productivity of a site. New Zealand provides some outstanding ex- amples of this phenomenon. Pinus radiata-a TREE SYMBIONTS AND DECOMPOSERS pineendemictoasmallareainCalifornia-grows The support staff of an ecosystem include many very rapidly in New Zealand and is highly pro- other organisms that carry out important func- ductive of commercial wood products. Many of tions, such as facilitating primary producers (the these exotic pine forests are grown on sites that fungi that form mycorrhizae) or participating in wereoriginally grasslands. AnotherNorth Ameri- the decomposition of organic materials and re- can pine, Pinus contorta, hasescaped fromcultiva- lease of the nutrients they contain (many inverte- tion in New Zealand and is forming forests at brates,bacteria,andfungi).Theseorganismsmake elevations substantially above the original tim- up the bulk of the biological diversity found in berline formed by native trees; this ability to grow forest ecosystems but are rarely explicitly recog- at lower temperatures obviously has the poten- nized; hence, they are sometimes referred to as tial to alter drastically the structure and function the invisibleorhidden biodiversity of ecosystems of these previously alpine habitats. Again, the (Franklin 1992). point is that trees or other plants native to a Many of these groups are represented by sev- habitat may not be the genotypes capable of eral species, which may provide some functional achieving maximum short- or long-term produc- redundancy, but also an array of genetic types tivity on the site. At the same time, local species or that are closely adapted to specific niches. As a genotypes may well be optimal for other ecosys- result, dominance among these organisms can tem functions, including the provision of habitat shift seasonally or over longer periods in re- for native species. sponse to environmental changes and maintain a 359 Defining and Measuring Sustainability: The Biogeophysical Foundations high level of functioning. For example, we know Herbivores and pathogens are clearly impor- how fungi capable of forming mycorrhizae with tant biotic elements that influence the productiv- trees shift dominance seasonally with soil mois- ity and sustainabilityof forestecosystems. Asub- ture and temperature conditions, thereby main- stantial base of information exists on the negative taining optimal mycorrhizal function for the tree impactsof pathogensand herbivores, and it is clear symbionts. that, in someplacesand at some times, thesemaybe Although we have begun to appreciate the a dominant influence. Very little quantitative data importance of such 'lesser" organisms to ecosys- are available on the positive contributions these tem function, detailed knowledge of their distri- organisms may make to the productivity and bution, community structure, ecology, function- sustainability of ecosystem. ing and, importantly, response to disturbances that disrupt or destroy the forest is not available. Major ecosystem processes associated Developing this information is a high priority for with productivity scientists; in the meantime, applying manage- ment practices that are likely to conserve this A great deal of research has been conducted on functionallyimportantdiversityisa high priority the function and structure of forest ecosystems for foresters. during the last three decades, with the Interna- tional Biological Programme providing much of HERBIVORES AND PATIOGENS the impetus. Much of this work hasdealt with the Herbivores and pathogens are a group of organ- captureand fixationof energy throughphotosyn- isms that feed on and sometimes damage or kill thesis (primary productivity) and with cycles of the primary producers. Although foresters have material (carbon, nutrients,and water): pathways, tended to focuson the negativeimpactsthatthese rates, and controls. An important result of this organisms have on short-term productivity, her- research has been a new appreciation of the im- bivoresand pathogensalso make important func- portance of structure and structural complexity tional contributions to the ecosystem. to ecosystem function, including long-term pro- Herbivorescan have very important influences ductivity and the provision of habitat for a variety on productivity over either short- or long-term of forest-dwelling organisms. We now recognize, periods. For example, epidemic-level outbreaks for example, that dead trees and tree parts are as of moths can defoliate forests for one or more important to the functioning of the forest as live seasons, drastically reducing tree growth and trees (Franklin, Shugart, and Harmon 1987). increasing mortality. Bark beetles are another common cause of death in trees. Grazing by un- DEFINITION OF ECOSYSTEM PRODUCTIVITY gulates, such as deer or elk, can alter the compo- Various measures of ecosystem productivity and sitionandstructureofforests,significantlyaffect- associated formulas are used by ecologists in ing their ability to regenerate and produce. assessingecosystem productivity. These are sub- Disease organisms (rusts and various fungi stantiallydifferent from measuresused byforest- capableofinfectinglivingtrtes)aswellasdecom- ers, as will be discussed in a following section. posers and herbivores can also reduce growth Unrecognized, these differences are frequently and cause the decay and death of trees and other the basis for significant, often public, disagree- plants. Yet many of the effects of herbivores and ments among ecologists and foresters about the pathogensareessential to the long-term function- productivity of natural forests. Productivity is a ing of the ecosystem. They contribute to the natu- rate and is typically measured on a yearlybasisas ral thinning process, for example, by reducing mass per unit of area per year. In these formulas, vigor or killing individual trees. A continuing autotrophs are organisms that capture energy flow of dead trees is essential to provide the fromprimarysources(primaryproducersorgreen coarse woody debris essential to a variety of plants), while heterotrophs are organisms that ecosystem functions (Franklin, Shugart, and use organic compounds created by the primary Harmon 1987). Similarly, decay organisms may producers as their source of energy (all animals). create cavities and other habitat niches in living The most common measures of productivity trees. There issome suggestion thatmoderate levels used byecologists are gross primary productivity of herbivory may actually contribute to overall (GPP), net primary productivity (NPP), and net productivityofan ecosystemby increasing the avail- ecosystem productivity (NEP; Kimmins 1987). ability of nutrients and reducing competition. Gross primary productivity encompasses all of 360 Sustainability of Managed Temperate Forest Ecosystems theproductivityof (energycaptured by)anecosys- are only for the aboveground portion because of tem, forest or otherwise. The formula for GPP is the immense technical difficultiesassociated with GPP = NPP + Ra estimating belowground productivity. Not only where Ra is the respiration of the autotrophs in is there no easy method of observing and measur- the ecosystem. Net primary productivity is ing belowground, but there is also considerable controversy about the accuracy of the labor-inten- NPP = B + L + C siveapproachescurrentlyunderuse(Kimminsl987). where AB is the change in biomass, L is total litter The difficulty of measuring belowground pro- production (including tree mortality), and C is ductivityisextremelyunfortunatebecauseof the consumption of green plants by herbivores. Net supposed overall importance of belowground ecosystem productivity is calculated as productivity to the carbon budget of the forest. NEP = NPP - Re Recent research has shown that the belowground where Re is the respiration of the entire ecosystem portion of the ecosystem is very dynamic, with bothauttrophsan yeterotrohsValusfort high rates of turnover in fine roots and mycor- formhautorps typica erep hort hese rhizal fungal hyphae. Only 20 percent of the formulas are typically reported as grams per square biomass is found belowground in a typical forest, meter per year or metric tons per hectare per year. and ie stud belowground These formulas contrast markedly wi thforest- and earlier studies assumed that belowground Terseformulationsontraductiry, whith forest- productivity was proportional to the mass. Un- ers' calculations of productivity, which typically fortuna tely, energy demands may be as high as 50 Involve only the production of bole or wood fotntl,nryemnsabahga5 nvolvmeanly thebase prodonmea of boeegrorwod to 70 percent of the photosynthate produced by vortalume,andarebasomedtomeasubires ofegrowth, T the forest due to the high turnover of roots and mortality, and, sometimes, birth (imgrowth). The hyae differing viewpoints on productivity have pro- hyphae. found consequences for examining older forest bdIt is also known that the energy requirements ecosytemswher bot GPP nd NP ma re- belowground increase on sites that are deficient ecosystems where both GPP and NPP may re- intinsrae,easmrfnrosn main high even though increments of additional innutrientsorwaterbecausemorefinerootsand wood mass have fallen to low or negative levels. hyphae must be produced to exploit the soil mass Although the concepts and formula may be for the required materials. One very important calculating the productivity implication of this finding is that some orall of the qut clar acual increases in productivity associated with forest of forest ecosystems is extremely difficult. Esti- fertilization may represent shifts in the allocation mating respiration is one of the serious problems fergy fro beoround to abovetron in determining either GPP or NEP; although hene, teobereinresnd to aboveground; respiration can beestimated for individual com- hence, the observed Increases in aboveground ponents, developing reasonable estimates for productivity may not represent increases in total the entire ecosystem is impossible with existing productivity of the ecosystem. techeologies. These discoveries about the energetic require- ments and productivity of the belowground por- PRIMARY PRODUCTIVITY tions of terrestrial ecosystems, including forests, ,,,, environmental factors controlling primary are forcing drastic reassessments. First, they have Theducenviroinmfrenta facosysto s controll prir e made clear that basing conclusions about total productivity in forest ecosystems have been ex- NPonynabvgudmesretsi tensively studied (Waring and Schlesinger 1985). . . Y As noted earlier, temperature, moisture,and avail- highly questionable, if not dangerous. Today, ability of nutrients are the key variables. any estimates of ecosystem productivity that do Estimates of NPP have been calculated for not include the belowground portion of the forest numerous forest ecosystems (see,forexample, are open to challenge. This specifically includes numerous forest ecosystems (see, for example, ayassmnoflgtemrndinpduiv Cannell 1982; Reichle 1981). The NPP variables of any assessment of long-term trends In productmv- .. . . . . ~~~~~~~ity and responses to experimental treatments, living biomass increment, litter production, and such as thinning and fertilization. Consequently, consumption by herbivores aboveground are all valid observational and experimental studies are susceptible to measurement, albeit with some relatively rare, and almost all of the older litera- difficulty in the case of litter production and ture on productivity of the forest ecosystem is consumption. But perhaps the most difficult as- open to question. pect of measuring NPP is the productivity that A second important implication is that the occurs belowground. trees and other green plants are critical sources of Most estimates of NPP for forest ecosystems energy to sustain the extremely dynamic 361 Defining and Measuring Sustainability: The Biogeophysical Foundations belowground ecosystem. In effect, the tree has feed on organic litter), and most of the secondary been shown to be as important to the vitality of productivity in forest ecosystems is associated the soil as thesoil hastraditionallybeen viewed to with decomposition or detrivory. Much informa- be to the tree. Loss of this source of energy as a tionhasbeendevelopedonratesandpathwaysof result of forest removal, even for short periods, is and controls on decomposition during recent hypothesized to cause the failure of reforestation years. Important environmental variables include efforts and long-term loss of forestland to vegeta- the moisture and temperate conditions found on tion other than forests (Perry and others 1988). the site; both can limit rates of decomposition. Calculating NPP requires an estimation of lit- Chemical attributes of the detritus or litter have a ter production over the period of measurement. major influence on rates of decomposition. Lig- This is generally done by periodically collecting nin and nitrogen contents of leaf litter, for ex- and weighing litterfall within the forest stand of ample, both have been shown to be important interest. Numerous well-documented techniques variables and are used in general equations for involve litter traps placed on the forest floor that predicting rates of decomposition. The available collect insect frass, flowering parts, leaves, twigs, biota is another critical variable. For example, soil and branches. arthropods play critical roles in fragmenting larger Few litterfall studies and calculations include organic materials while feeding on and consum- the largest pieces of litter: dead trees. Current tree ing portions of them, providing large surface mortality is technically part of the litter factor in areas for colonization by other decomposer or- the NPP equation. Long-term studies of tree popu- ganisms. Consequently, excluding or eliminating lations are necessary to obtain accurate data on segments of the soil fauna can have significant annual rates of mortality because of high year-to- impacts on rates of decomposition. year variability, which often includes a major Decomposition of large or coarse woody de- stochastic, or random, component (Franklin, bris, such as large standing dead trees (known as Shugart, and Harmon 1987). Forthisreason, many snags) and logs on the forest floor, is particularly studies of ecosystem productivity ignore tree complexandhasonlyrecentlybecomethesubject mortality even though tree death may contribute of intensive study (see, for example, Harmon and as much organic material as the smaller, tradi- others 1986; Harmon and Chen 1992). tional components of litterfall (Sollins 1982). Obtaining accurate measures of consumption OTHER CRITICAL ECOSYSTEM PROCESSES by herbivores, the third element of the NPP equa- Primaryproduction and decompositionhavebeen tion, is very difficult although some techniques singled out for attention in this review because of provide an approximation. Fortunately, their importance to the sustained productivity of aboveground herbivory is relatively insignificant all ecosystems, including forests. Many other eco- in healthy forest ecosystems (Kimmins 1987). system processes are important, however, some Hence, assumed values are unlikely to produce of which have already been identified, such as major errors in calculating NPP. Herbivory consumption by herbivores. The identification belowground is much more poorly understood, and elaboration of theseprocessesalonecould fill however, and could be a major factor in any several pages. calculation of NPP. Nitrogen fixation is one additional process that requires mention, however, because of its DECOMPOSITION AND SECONDARY PRODUCTIVITY importance to fertility of the soil and site. It has Decomposition is probably second only to pri- also been the subject of important recent discov- mary production as the most important ecosys- eries. Nitrogen fixation involves the conversion tem process. Decomposition is carried out by a of elemental nitrogen in the atmosphere to the variety of organisms that break down organic biologically useful forms of ammonia or nitrate. materials to release energy and nutrients: Although physical processes such as lightning organic compounds + decomposition = energy discharges can produce this conversion, much, if + carbon dioxide + water + nutrients, not most, of the nitrogen is fixed biologically. Relatively few organisms are capable of nitro- Most of the secondary producers (organisms gen fixation (cyanobacteria). Although some of that use existing organic carbon compounds as these are free-living organisms, many of the most their base of energy) found in forest ecosystems important nitrogen fixers live in association with are decomposers or detritivores (organisms that other organisms. Well-known examples are ni- 362 Defining and Measuring Sustainability: The Biogeophysical Foundations centuries because they decay or disappear from ment and high productivity of target tree spe- ecosystems slowly (Harmon and others 1986). cies-in the belief that much of the structural Furthermore, functions change throughout the complexity found in natural stands is not essen- lifetime or gradual decay of a snag or log. tial to sustained productivity of the site. The recent research, briefly reported here, on the role OVERALL STAND STRUCTURE of structural diversity in maintaining the pro- Overall structural heterogeneity is an important cesses and organisms essential to forest feature of almost all natural forests. The forest as sustainability is a major challenge to those as- a whole cannot be reduced simply to individual sumptions. structures and aggregated into a whole. Hetero- geneity inboth thehorizontal and vertical dimen- Disturbances and ecosystem recovery rates sions is a hallmark of natural and, especially, Responses of forest ecosystems to disturbances, older forests.Repneoffrsecssestdiubae, Variations in the density of the overstory including the pattern and rate of recovery, are Varitincluding complete gaps in the canopy, highly dependent on the intensity and type of caoy nldn opeegp ntecanopy, disturbance, which, in turn, determine the are an important element in stand-level struc- carryover whical matern, frmine the tural diversity. A natural stand typically has lo- carryover of biological materials from the old or catls where lcvels of light are higher and vegeta- disturbed to the new or recovering ecosystem. tionontheforest floor isgbetterdeveloped than in This section briefly reviews disturbances, bio- other areas, where dense tree foliage, a logical legacies, and rates of ecosystem recovery. of shade-tolerant species, produces especially A great deal of literature is available on distur- shaded environment from which da heavily bances and their effects, including effects on at plants may be absent or nearly so. understory least some aspects of productivity. The variability in light conditions, as well as belowground competition for moisture and nu- DISTURBANCES trients, contributes to the complexity and rich- Forestecosystems are subject toa widemvariety of ness of understories in many late-successional disturbances that influence both immediate and forests. These diverse understories can be critical long-term productivity. The important variables for some organisms; for example, the old-growth in determining impacts on sustainability are the Picea sitchensis-Tsuga heterophylla forests of the type, intensity, size, and frequency of distur- Alaskan panhandle provide essential habitat for bance. Among the important types of natural Sitka black-tailed deer (Odocoileus hemionus disturbances are fires, windstorms, floods, land- sitkensis; Alaback 1984; Schoen and Kirchoff 1990). slides, epidemic outbreaks of insects or disease, Researchthroughoutthetemperateforestregions and volcanism. Forest cutting by humans is of the world is showing that developing and probablythemostimportantsingledisturbance maintaining diverse understory plant communi- globally. tiesinforeststandsareanimportantandcomplex Each of these types of disturbance does, of undertaking, not simply a mratter of manipulat- course, display a range of intensities. For ex- ing crown density or levels of light, ample, wildfires can be intense, stand-consum- ing crown fires, such as the 1989 fires in STRUCTURFS IN TRADITIONAL MANAGED STANDS Yellowstone National Park, or low-intensity, Structural attributes of temperate forest stands creepinggroundfiresthatleavemostof theforest subject to traditional management are typically intact. Wind displays a similarly wide range of very different from those of natural stands. The behaviors, often generating intense damage at a most common managerial system has been the regional level in the form of a hurricane or ty- creation of even-aged, even-sized stands using phoon or, at a more local level, in the form of a clear-cutting and artificial reforestation (Oliver tornado; however, wind most often disturbs and Larson 1990). Such stands are highly simpli- chronically and at the smaller spatial scale, blow- fied and lack many structural components, such ing over or breaking individual or small groups as snagsand logs, aswellasstand-level structural of trees. Other natural disturbances illustrate a complexity, such as multiple levels of canopy, similar gradient from slight to intense effects on the chaotic tree spacing, and gaps. forest ecosystem. This gradient is typically found Managed stands have been simplified in re- within a single disturbance, such as a fire or sponse to economic criteria-efficient manage- windthrow, especially if it occurs on a larger scale. 364 Sustainability of Managed Temperate Forest Ecosystems Disturbances occur over a very wide spatial rubber or nuts. Sizes can range from the small scale. Wildfiresand windstorms can range froma patch to thousands of hectares, as in the case of few squaremeters to thousandsof hectares. Some some forest cutting. Intensity, as noted later, can types of disturbances, such as floods and land- vary from intense clear-cutting followed by slash slides, are constrained to certain scales by land- burning to selective cutting of individual trees. forms; these may have an extensivelinear (down- Finally, human disturbances can recur each year streanm) dimension, however, even if limited in or after many decades or even centuries. width. Again, thelarger thedisturbance, themore Disturbances do have contrasting impacts on heterogeneous it will be in terms of intensity; for the productivity of a site, and even a single type example, larger wildfires almost always include of disturbance can have either positive or nega- areas of intense and very light burning. tive effects, depending on the nature of the forest Disturbances are sometimes characterized as ecosystem and the intensity of the disturbance. being either stand regenerating or intrinsic to the For example, wildfire negatively affects ecosys- within-stand dynamic of a forest. Such a catego- tem productivity by volatilizing significant rization includes consideration of both size and amounts of nitrogen as the organic matter is intensity of a disturbance. A very low-intensity consumed. Positive effects include short-term disturbance,suchasagroundfire,isoftenconsid- release of soil nutrients, particularly basic ele- ered to be an integral part of the environment of ments, as the organic matter is consumed; on a forest stand, even if it is of largeextent. Creation some sites, accumulations of organic matter may of a gap in a forest canopyby the uprooting of one be excessive from the standpoint of site nutrition. or several trees is also typically considered to be Wind-driven disturbances rarely result in short- partof a stand dynamicdespite itsintensitywithin term losses or gains in nutrients. Geomorphic a small area. disturbances, such as floods or landslides, can Frequency of disturbance is a fourth and ex- have positive or negative benefits, depending on tremely important variable. Many disturbances whether nutrient-rich materials are removed, arehighlyepisodic,occurringatinfrequentinter- added, or buried by erosional or depositional vals. In the case of some disturbances, such as processes. wildfires, longintervalsbetween occurrences tend to result in much more intense events than where BIOLOGICAL LEGACIES short intervals are involved; this is typically re- Studies of early-successional recovery of ecosys- lated to the period available for fuels to accumu- tems following disturbances generally give little late. Frequent disturbances can also have very attention to the influence of the ecosystem before negative effects on productivity and the process the disturbance (Franklin 1990; Franklin, Frenzen, of recovery, however, where they result in the and Swanson 1988). The role of migration or loss of nutrients, soil organic matter, or organ- reinvasion of organisms is typically emphasized, isms; repeated disturbances can, for example, while surviving organisms and structures are eliminate or dramatically reduce the level of bio- largely ignored. However, disturbances are in- logical legacies, such as sources of mature tree creasingly recognized as processes that leave be- seed, at each iteration. Hence, repeated intensive hind varying levels of organisms, structures, and crown fires can produce large areas that are very patterns. These biotically derived legacies from slow to reforest. One of the reasons for the rapid predisturbanceecosystems have importantinflu- recoveryof ecosystemsatMountSt. Helens(Wash- ences on the paths and rates of recovery. ington State) in 1980 was the absence of a second As defined here, biological legacies are living major eruption over most of the area; therefore, organisms that survive a disturbance, particu- the legacy of surviving organisms was not subject larly a catastrophic or stand-regenerating distur- to further death and burial (Franklin, Frenzen, bance, organic debris, particularly the large or- and Swanson 1988). ganically derived structures, and biotically de- Human-induced disturbancesexhibitall of the rived patterns in soils and understories. The liv- same variablesas natural disturbances: type, size, ing legacies may take a variety of forms, includ- intensity, and frequency. Indeed, forest harvest ing intact plants and animals, perennating struc- activities can be considered and scaled with re- tures (rhizomes), and dormant spores and seeds. gard to each of these. Types of activities can be as Important biotically derived structures include variable as felling and removal of timber and dead trees (snags) and fallen logs, large soil ag- nondestructive removal of forest crops, such as gregates,anddensematsoffungalhyphae.These 365 Defining and Measuring Sustainability: The Biogeophysical Foundations structures are appreciated more and more for As a result, the young forests that develop their role in ecosystem functioning, such as the following traditional clear-cutting practices are importance of large woody structures as wildlife typically much simpler in composition and struc- habitat (Harmon and others 1986; Maser and ture than those that develop following natural others 1988). Pattern legacies include those cre- disturbances. ated in soil properties-chemical, physical, and The types and relative levels of biological lega- microbiological-through theaction of plantsand cies following a catastrophic disturbance, then, their litter, and patterns in understory vegetation are extremely important in determining the rate associated with variations in the conditions of at which the new forest ecosystem will recover canopy light. These patterns can be either posi- and, perhaps even more important, the diversity tive or negative; for example, patches of soil asso- of organisms, processes, and structures that it ciated with some tree species may be enriched in will contain. Many of these have direct signifi- nitrogen or various bases, while others may be cance for sustained levels of productivity. An out- leached of nutrients and acidified. standing example is the retention of organisms Disturbances of various types, intensities, spa- capable of fixing nitrogen and providing appropri- tial scales, and frequencies produce different types ate habitat for their propagation and functioning. In and levels of biological legacies. Some of the forest ecosystems, this retention may encompass a relationships are obvious. More intense or fre- wide range of forms; in old-growth Douglas fir quent disturbances tend to have lower levels of forests, for example, it includes canopy-dwelling livinglegacies;however,disturbancesvarywidely lichens with cyanobacteria elements and microor- in the types of living legacies they leave behind. ganisms that live in decaying wood, such as fallen For example, wildfires are most likely to kill logs and snags (Franklin 1992). smaller and thin-barked trees and spare large, The types and quantities of biologically de- thick-barked dominant tiees. Windthrow, how- rived materials persisting through a disturbance ever, typicallyeliminates dominant trees, leaving generated by either natural or human causes have behind the largely intact understory of tolerant a powerful influence on the levels of nutrients tree seedlings and saplings. In northwestern North and organic matter present in the recovering eco- America, fire and wind differ dramatically in system. Nature generally provides for high levels theircompositionalorsuccessionalconsequences; of legacies and for other mechanisms that retain wildfire favorstheshade-intolerantpioneer, Dou- nutrients. However, traditional forest harvest glas fir, while wind favors survival and subse- practices, such as clear-cutting, tend to minimize quent dominance of the shade-tolerant western biological legacies and maximize nutrient losses, hemlock and western red cedar. as in the volatilization of nitrogen that occurs Almost all intense disturbances in forest eco- during slash burning. systems tend to leave behind large legacies of dead organicmaterial,iincludingstructures(snags ECOSYSTEM RECOVERY TIMES and logs); this is because most natural forest There have been numerous studies of succession disturbances, such as wildtire and windthrow, in forest ecosystems, but very few actually inves- kill trees but consume or remove relatively little tigateorpredict compositional, structural, or func- of the material. This legacy provides a continuity tional recovery, except as it relates to production of wildlife habitat, bridging the two generations and standing crops of wood. Models and data of ecosystems as well as providing long-term related to production of wood are considered transfer of organic material and nutrients. later in this report. Traditional intensive harvest of forests by hu- Recovery rates in forests are actually consid- mans has typically left a much smaller biological ered to be quite slow compared with other major legacy than have natural disturbances. Although types of ecosystems, such as grasslands, deserts, many of the original plant and animal species or tundra (MacMahon 1981). This relates in large may survive, the intensity of management prac- measure to the structural complexity of forests ticeshasastronglynegativeinfluenceonthelevel and the long period of time requirecd to reestab- of living legacies (Halpern 1988,1989). Legacies lish a diverse and fully functional forest ecosystem. of large organic structures, such as snags and Foresters have focused heavily on regenera- fallen logs, are also drastically reduced under tionoftreesandreestablishmentofaforestcanopy most current silvicultural practices, which in- (forest dominance) on a site. Regeneration of clude both harvest and slash disposal operations. trees can occur immediately under managed con- 366 Sustainability of Managed Temperate Forest Ecosystems ditions as a result of planting but is highly vari- and most structural and functional features to re- able under natural succession; it may be essen- cover. In the temperate hardwood forests of north- tially instantaneous where an abundant source of easternNorth America, 150 to 200yearsmaysuffice tree seeds is present or may require many de- for recovery, but in the coniferous forests of north- cades where environmental condi tions are severe western North America, as many as 250 to 450years or seed sources are distant. Growth of the regen- appear to be necessary to fully achieve late-succes- erated trees to the point where the tree canopy sional forest conditions (Franklin and others 1981; becomes continuous is also highly variable, de- Franklin and Spies 1991). pending on the productivity of the site. In the case of temperate hardwood forests, rapid growth of RESTORATION OF SOIL PROPERTIES pioneers, such as Prunus or Alnus, may produce Restoration of soil properties almost certainly canopy closure in two or three years (Reiners requires even longer time periods than does re- 1992). Among the coniferous forests, moist and coveryofthebiologicalelementsoftheecosystem warm regions dominated by Pinus (such as the (Grier and others 1989; Jenny 1980). Very little southeastern United States or the exotic planta- good information is available on the rates of soil tions of New Zealand) are the fastest to return to formation, or even on the rate at which organic tree dominance. In northwestern North America, matter typically accumulates in the soil. Never- closure of the forest canopy may require a decade theless, it is clear that soil typically develops at a for completion, even on productive sites; twenty very slow rate. to thirty years of succession may be required on In fact, much (and possibly most) accumula- typical sites following either logging or natural tion of soil parent material on a site results from disturbance (Halpern 1988, 1989). episodic depositions of materials from adjacent Much more is involved in ecosystem recovery, sites and not the weathering of parent materials however, than simply tree dominance or even in place. The majority of deep forest soils are achievement of some level of biomass. A diverse composed of alluvial, colluvial, glacial, aeolian, array of structures, processes, and organisms must and volcanic materials that were moved to the reestablish themselves at some level approximat- site by water, gravity, ice, wind, or eruptions. ing the original forest. Significant biological lega- Hence, the frequency and type of episodic events cies, such as snags and fallen logs, largely deter- (primarily geomorphic) are extremely influential mine how rapidly recovery of the full functional in determining both the depth of existing parent ecosystem will take place. If such legacies are material and the probability for replenishment. absent so that new structures have to be grown to This is an important point: replacement of soil desired sizes (and, in the case of dead wood parentmaterialsonmanysitesmaydependprima- structures,killedanddecayedtoparticularstates), rily on the recurrence of infrequent and highly the recovery process can be extremely slow, per- episodic geomorphic process, such as a volca- haps involving many centuries in some types of nic eruption; hence, soil conservation should forest. If such legacies are retained on the dis- have a high priority among forest management turbed sites, recovery can be much more rapid. considerations. There is increasing evidence that some ele- Once in place, biological processes are critical ments of the ecosystem are very slow to recover. in the evolution of the soil parent materials into One study in the Appalachian Mountains of east- an organically and nutritionally rich medium for ern North America, for example, has shown that growth. As noted earlier, this can be a slow pro- some understory plant species-mosses, herbs, cess. It is probable that most forest soils are con- and shrubs-may not have recovered to their tinuously and gradually accumulating soil or- natural levels even 100 years after logging. In the ganic matter under natural successional regimes. temperate rain forests of southeastern Alaska, Although the available information is inconclu- development of a compositionally diverse under- sive, soil organic matter is probably not accumu- story of the type required as winter range by Si tka lating-and may be declining-under many for- black-tailed deer (Odocoileus hemionus sitkensis) est management regimes currently in use typically requires 200 years following logging (Kimmins 1987). (Alaback 1984). Perhaps the most difficult problem in soil res- If a late-successional forest is taken as the end toration is the reintroduction and establishment point of successional recovery, it appears that of critical soil organisms, such as fungi, inverte- several centuries are required for composition brates, and bacteria, once they have been elimi- 367 Defining and Measuring Sustainability: The Biogeophysical Foundations nated. There is good evidence that significant effect would probably be negative; conversely, elements of the soil biota can be lost with the increased levels of atmospheric carbon dioxide eliminationofhosttreespeciesfromthesite(Perry may result in increased productivity and more and others 1989); this can lead, in turn, to serious efficient use of available moisture. problems in the reestablishment of forest cover. Episodic and stochastic processes Natural variation in ecosystem productivity and thresholds Significant variation in both space and time exists Forest ecosystems are subject to many important in the productivity of forest ecosystems. Spatial processes that are either episodic or stochastic or (site-to-site)variabilityhasalreadybeendiscussed both. Some of these have already been discussed and can effectively span two-and-a-half orders of in earlier sections, particularly with regard to magnitude in production of wood, from less than disturbances. Disturbances are among the most 10 to more than 250 square meters per hectare a important of the processes that are, in the major- year. ity of cases, both episodic and stochastic, or ran- A forest ecosystem on a specific site can also of cases, bt episodic nan sohatior ran- experience substantial year-to-year variation in dom. Vaniaton in environmental conditions (cli- productivity, quite aside from long-term trends mate) is another example. associated with successional development of the two po rtant aects ofoth dynaicso forest. Thegreatestvariability-certainlyin terms tree populations central to forests and forest of percentages and, often, in absolute values as productivity-birthanddeath-canbeepisodic well-occurs on sites that are subject to major or stochastic processes or both. Birth, the suc- environmental stresses. These include marginal cessful establishment of new tree seedlings, forests on hot, droughty sites, such as those found may require a major (stand-regenerating) dis- at lower timberline, and on cold arctic and alpine turbance, which is typically both episodic and timberline. Productivity is typically low on these stochastic. It may also depend on the produc- sites, and growth is responsive to variations (ei- tion of a bumper seed crop, another process ther positive or negative) in climatic conditions. that is at least episodic, and possibly stochastic. Dcndrochronology, the analysis and interpreta- Finally, successful regeneration of trees on sites tion of tree rings, is based on the sensitivity of tree with severe environmental conditions may de- productivity to climatic fluxes, especially on se- pend on the occurrence of one or two years with vere sites. Productivity might easily span two an unusually favorable climate. Regeneration orders of magnitude on sites with major environ- of Pinus ponderosa forests in central Arizona, for mental limitations. Although annual and peri- example, requires the combination of two un- odic variability in the productivity of temperate usually moist springs and a bumper seed crop; forest ecosystems is present on moderate sites as these conditions occur only every two or three well, it is smaller in magnitude than year-to-year decades. variation. Scientific knowledge of tree death or mortality Changes in local or global environmental and is surprisingly poor, considering its importance climatic conditions can be expected to produce in forest ecology and productivity (Franklin, majorchangesin theproductivityofforesteco- Shugart, and Harmon 1987). Some mortality is systems. Effects of local pollutants, such as quite regular and predictable, particularly the emissions from smelters, have been well docu- natural thinning that occurs early in develop- mented. Effects of regional changes in pollut- ment of the stand. Much mortality, however, ants or in climatic conditions are less clear but includingsucheventsasanoutbreakof pestsand are currently the subject of intense scientific pathogens or a major windstorm, is highly epi- interest (Franklin and others 1991). Such sodic. Generally, mortality of established trees changes clearly have the potential to produce (above the seedling and sapling stage) in forest major changes in the productivity of forest eco- stands is both episodic and stochastic, impeding systems; the direction of the change will, of our ability to predict rates, causes, and spatial course, depend on the current circumstance. If patterns of mortality in mature and late-succes- changes in environmental and climatic condi- sional stands. tions produce additional drying in a forest eco- One of the major needs in forest manage- system already moisture-limited, the resulting ment is to recognize the highly stochastic na- 368 Sustainability of Managed Temperate Forest Ecosystems ture of most natural ecosystems. More manage- measures show a high rate of error when applied ment plans need to consider the potential for to specific sites and stands. This natural variabil- stochastic disturbances. More management ity makes it extremely difficult to identify long- decisions need to consider probabilities, such term signals or trends in site productivity, a com- as in the potential for successful natural regen- mon problem with many ecological phenomena eration. Far too many management decisions (Likens 1989). are based on a deterministic view of forest ecosystem responses and an unwillingness to Biological measurements accept outcomes with less than 100 percent prob- Biological measures used in assessing forest pro- ability of success. Bityoicl measures of forest pro- Thresholds have not been a major topic in ductivityincludedirectmeasuresofforestyields considerationsof theproductivityof temperate and tree growth, measures of total ecosystem forest ecosystems. Nevertheless, the occurrence productivity, rates of key ecosystem processes, of thresholds is implicit in many discussions, and vegetational associatesor plant communities since so many processes fit the traditional lo- as indicators. gistic curve. However, few interpretations of physiological, population, community, or eco- TREE PRODUCTIVITY AND WOOD PRODUCTION system phenomena have been explicitly made The yield of a fully stocked forest stand over a in terms of thresholds; they could be, however, given time period is the ultimate measure of and resource managers often assume that there arboreal productivity for a site (Daniel, Helms, are such points beyond which responses accel- and Baker 1979). Since such a measure is rarely era te or decelerate. Threshold phenomena have possible, traditional approaches to predicting for- been explicitly recognized in the area of land- est site potential and forest growth have been scape ecology. One example is the effect of combined in a tree-based growth measurement dispersed patch clear-cutting on various land- called siteindex withyield tablesof varioustypes. scape measures, such as mean patch size (see, Site index is the height to which a tree of a for example, Franklin and Forman 1987). given species will grow within a specific time period. Site index is typically based on height growth curves developed using empirical data Biophysical measurements on cumulative height attained by dominant and for temperate forests codominant trees overtime. The index age varies with tree species; 50 years is a common age for fast-growing species, such as Pinus sp. in the Many approaches to predicting the productivity southeastern United States, while 100 years is a of forests and forest sites have been developed common index inawestern Not Aerica over the last century. These include direct mea- common index age r h western North Amernca, surements of tree and stand growth and many whereinitial growthratesareslower. Asiteindex indirect approaches, such as those using soils, on a sample of dominant and codominant trees landforms, and plant communities. Most of the om t site and poetnhigto theex approachesfocusultimatelyon thearboreal com- from the sete and provectg helghtt tethe rdex ponient as the measureof productivity and,often, age using a set of sitecurves; the selected treesare ponlyont asoe measureoof productivi,atyr thand on typically assumed to have grown naturally. Only on bole (wood) production, rather than onHihtrwhaeofre- wntesws all ecosystemcomponents.Thisisacceptabletoat g grgro g least smdr,icmyselected and has been defended by foresters on least some degree, sice many capabilities of a the basis that it is relatively unaffected by stock- forested site (such as processes and organisms) iglvl rte est,weesohrmaue , , . , , . .~~ Ing level sor tree density, whereas other measures are related to or indexed by the ability of the site of growth, such as diameter, are. Hence, it is to grow a tree to a maximum size at a particular considered to be a direct indicator of site poten- rate. However, focusing exclusively on trees in tial irrespectiveofstandconditions.Studieshave assessing productivity ignores many elements tias,iirresp ectiof standc dns.tudieha essen tial to sustainable forestry. shown signi ficant effectsof stand density on height Techniques for direct assessment or measure- growth, however, with considerable variability Techni s fr damong species (Daniel, Helms, and Baker 1979). ment of long-term trends in site productivity, as Furthermore, growth curves (patterns of height opposed to the modeling approaches discussed growth over time) may differ among sites for the later, are, not well advanced. Most of the existing same species and site index; that is, even where 369 Defining and Measuring Sustainability: The Biogeophysical Foundations tree heights on two sites are identical at the index into doubt the accuracy of most available esti- age, patterns of height growth both prior to and mates of gross primary productivity and net eco- after the index age may differ. This has encour- system productivity for forest ecosystems. aged the development of more localized, or poly- Estimates of net primary productivity of forest morphic, site curves as an alternative to the cre- ecosystems can and have been made for research ation of a generalized set of site curves for a large sites. However, obtaining the necessary measure- region (Daniel, Helms, and Baker 1979). ments requires heroic physical, as well as concep- Yield tables are the other half of the traditional tual and financial efforts, so as a routine measure approach to predicting forest productivity. Most of ecosystem productivity, estimating net pri- yield tables, including all of the older ones, are mary productivity is impractical. The conceptual constructed by sampling fully stocked stands of contribution of ecosystem productivity is prob- one (usually) or two or more (rarely) tree species ably the aspect most relevant to this review. It that representdifferentagesand levels of produc- recognizes explicitly the productivity of all parts of tivity as measured by site index. The empirical the ecosystem rather than focusing exclusively on data are used to develop comprehensive tables treesandvolumeofwood.Asaresult,itprovidesus that predict the volumes of wood (cubic meters withaverydifferentpointofviewonthehealthand per hectare or board-feet per acre) to be expected productivity of an older forest ecosystem than do from fully stocked stands of those species at vari- measures of productivity that are based on addi- ous ages and on sites with different site indexes. tional increments of wood or biornass. Yield bulletins typically include much other tabu- lar and graphical information as well, such as PROCESSES AS INDEXES calculations of increment per unit of time and Although not widely accepted, rates of key eco- changes in tree density. Yield tables can be viewed logical processes, such as decomposition or min- fundamentallyas treepopulation ordemographic eralization of organic matter, are sometimes pro- models. Early yield tables were almost entirely posed as indexes to overall health or productivity for stands of natural origin, but more recently, of an ecosystem. One variable that seems to have many have been developed for managed stands, a high level of sensitivity to pollutants and some such as plantations, and include effects of thin- other stresses is the time that evergreen trees ning and other management activities. retain needles or leaves. Yield tables have been superseded in many regions by computerized growth models that SITE COMMUNITY CLASSIFICATIONS have the ability to incorporate substantially more Plantcommunitiesand vegetative indicatorplants variables, including effects of variable stand den- have been proposed and are sometimes used to sities.Thesearediscussed later, since theyare not assess the productivity of a forest site. Such ap- direct measurementsof productivity. Most direct proaches are based on the concept that specific measurements of the productivity of a forest site plants-singly, in sets, or as communities-are are based on site index, which is then coupled indicative of specific environmental conditions, with yield tables or yield models. Other tech- such as moisture, temperature, and nutrient re- niques, such as projecting growth on the basis of gimes. Hence, inferences about site conditions recent patterns of tree growth, are occasionally and overall productivity can bedrawn from their used for estimating growth over the near future. presence or abundance on a site. The Finnish types of forest site represent the ECOSYSTEM PRODUCTIVITY earliest development of this concept. These were Ecological measures of productivity-gross pri- based on thebelief that empirical relationshipsexist mary productivity, net primary productivity, and between plant cover and tree growth. The presence netecosystemproductivity-aremuchmorecom- of a certain plant species in the understory was prehensive than the traditional forestry measures. assumed to indicate a particular quality of site. At least conceptually, they include all compo- Many vegetation-based approaches havebeen nents of the ecosystem. Unfortunately, accurate developed and are widely applied throughout estimates of total ecosystem productivity are ex- the world. In Scandinavia, most vegetational ap- tremely difficult, if not impossible, to obtain, proaches follow the original model pioneered by particularly the respiration component and al- Cajander. The vegetation classification system most all measurements belowground. This calls developed by Braun-Blanquet dominates in cen- 370 Sustainability of Managed Temperate Forest Ecosystems tral Europe and many other parts of the world. In The need to predict yields on sites lacking the western North America, Daubenmire pioneered trees or the tree species of interest has been a theapproachwithhishabitat-typeconcept,which particular stimulus to the use of soils. Soils are is now widely applied to national forestlands. also viewed as a permanent feature of the site, in Many other approaches have used plants as contrast to vegetative cover. indicators of site environment and productivity, Soil taxonomic units or types have been one including the use of vegetative indicators to de- basis for predicting forest yields. Exclusive de- fine the operational environment (see, for ex- pendence on soil types has had limited success, ample, Waring and Major 1964). Most plant com- however. This is related, in part, to variability in munity or plant indicator approaches ultimately the soil mapping units. Site curves are typically return to traditional forestry measures-site index used as a basis for rating soil productivity, but and stand yields-to rate productivity of a site. using inappropriate curves may create another problem, although one of the major uses of soil Phlysical measurementts types has been in developing stratified manage- ment plans. Finally, soils are only one part of the Many scientists have proposed that measure- environment to which the forest is responding. ments of physical site conditions, rather than Predictiveequationsbasedoncombinationsof trees or other biota, be used to rate potential soil properties-soil-site indexes-have been productivity of a site. One advantage is that such widely used. Theseoare typically localized, em- approaches do not require the presence of any pircly relaTihis. are .cllectedo av- partiularrgansmortand ondiion nordr to pirical relationships. Data are collected on a vari- particularorganism ortstand condition n order to ety of physical and chemical propertiesof soil and rate productivity; this also can (using some ap- then subjected to multiple regression analysis proaches) avoid biasing productivity estimates with a site index as the response variable. Repre- toward any particular genotype, species, or life- sentative properties that have been used in equa- form. Biotic productivity is, however, the ulti- tions include thickness of the A horizon, mois- mate measure of sustainability, so that most physi- ture-holdingcapacity, and total soil depth (Daniel, cal measures are, in fact, referenced back to plant Helms, and Baker 1979). Currently, soil-site index production. approaches do not appear to be in wide use. ENVIRONMENTAL REGIME Site classification systems Occasional proposals have been made to assess potential productivity of a forest site using direct Varioussiteclassification schemesattempttocom- measures of environmental variables, such as bine elements of geology, physiography (land- mean temperature, frost-free days, precipitation, forms), soils, and vegetation for predicting the and so forth. As noted earlier, there are strong management potential, including productivity, correlations between overall productivity and of forest sites. These include the physiographic environmental variables, such as moisture and types of site developed in Ontario, Canada, by temperature. Most of these studies are based on Hills, the classification of biogeocoenoses devel- measures of the operational environment, how- oped in British Columbia by Krajina, classifica- ever, and not on measurements of the regional tions developed for the Great Lakes region of the climate. United States by Barnes and associates, and forest Although climatological indexes have been site classifications developed for New England developed for temperate regions, they are not by Leak and his associates. currently used in assessing forest productivity. Indeed, the life zone approaches used in the United States early in this century might fall into this Alternative management options category. Holdridge's life zone concept is one environmental indexing scheme that is widely Traditional approaches to forest management in used for tropical forest areas. temperate regions focus on economically effi- cient production and harvest of wood products SOIL PROPERTIES and reforestation of the site with a new genera- Foresters have made considerable use of soils as tion of trees. The dominant paradigm is produc- a basis for predicting forest productivity (Daniel, tion of even-aged plantations of a single species Helms, and Baker 1979; Pritchett and Fisher 1987). with final harvest accomplished by clear-cutting. 371 Defining and Measuring Sustainability: The Biogeophysical Foundations Precommercial thinning and, in some cases, her- approaches at the level of both the stand and the bicide treatments are used to maintain rapid landscape.The stand level isconsidered firstwith growth in crop trees and to free them from com- the focus on management approaches in stands peting herbs, shrubs, or trees viewed as weeds. thataretobemanagedforsomelevelofcommod- Some management regimes use fertilization to ity production. A consideration follows of land- stimulategrowth.Commercialthinningsmayalso scape-level approaches that involve both man- be carried out. Rotation ages are determined pri- aged and reserved areas. marily by economic analysis or, in the case of The central concept of alternative approaches some government forests, by some biological cri- formanaged standsof temperate forest is tomain- terion, suchas culmination of mean annual incre- tain or recreate stands that are structurally and ment. Levels of use vary but typically do not compositionally diverse. That is, within the con- involve the removal of needles and twigs from straints of objectives and stand conditions, the the harvest site, although subsequent slash dis- effort is to maintain as much of the structural and posal activities, such as broadcast slash burning, compositional diversity as possible rather than to may consume much of this material. Tradition- simplify the stand. Structural diversity is usually ally, all structural material that can be used is the goal, because structure is normally closely removed, and theremainderisdisposed.The site correlated to organisms and processes; that is, isassumed tobecapableofsustainingproductiv- structure provides the necessary conditions or ity levels under this regime. habitat for desired organisms and processes. The Such an approach to forest management is general principle of maintaining structural diver- fundamentally agricultural: it aims to maximize sity should be kept in mind during the following the desired output by simplifying the ecosystem discussion. Theexact setof silvicultural practices of interest and subsidizing it with energy inputs, the treatments developed to create or maintain such as fertilizers. Implicit in this approach is a structuraldiversity-willvarywiththetype,condi- belief that what is good for wood production is tion, and environment of the forest and, of course, good for other resource values. The tendency with the specific set of management objectives. toward simplification is of particular concern because it traditionally occurs at many levels in Creation of young stands temperate forest management in terms of geno- type,species,product,standstructurejlandscape Young stands provide many opportunities for pattern,and successional stage (Franklin and oth- developing elements orattributes that are impor- ers 1986). Foresters sometimes persist in simpli- tant in enhancing ecosystem processes and fying forest ecosystemseven whendoing soisnot biodiversity. Aggressive efforts to create stands essential to managementobjectives and is done at of mixed compositionareoneinitial step.Plantings substantial expense. of single species and weeding by either chemical Development of alternative management para- or mechanical means strongly direct the man- digms is clearly appropriate, with the increased aged forest toward a monoculture. Plantings of emphasis placed on the sustained production of multiple speciesand efforts to retain speciesmix- all forest values, including wood products, and tures in precommercial thinning or weeding ex- with our vastly greater knowledge of forest eco- ercises will, however, create compositionally di- systems and their functioning (Hopwood 1991). verse forests. Maintaining a mixture of species Traditional practices, which had their genesis 50 can greatlyenhancea varietyof ecological values, to 100 years ago, do not reflect either the broad- such as the ability to provide habitat for a broad ened societal objectives for forest land or the array of organisms. For example, an occasional scientific findings of the last twenty to thirty hardwood can add significant structural and spe- years. Returning to my initial comnmentary defin- cies diversity (as host to a variety of plant epi- ing sustainability and its basis, it seems that sus- phytes and animal species) to a conifer-domi- tainable forest management practices should nated stand; this would include substantial het- emphasize the maintenance of the productive ca- erogeneity in microclimatic and edaphic condi- pacityoftheforestland(principlel)andofthebiota tions. Hardwoods such as Alnus sp. and Robinwa that are the engines of the ecosystem (principle 2). bring an additional benefit of nitrogen fixation. The question, then, is wlhat form the new alter- Richer mixtures of conifers can also provide natives to traditional practices should take. The valuable diversity. For example, species belong- answer involves considerations and altemative ing to the Cupressaceae (cypress or false cedar 372 Sustainability of Managed Temperate Forest Ecosystems family), such as Thuja and Chamaecyparis, im- since stands have traditionally been clear-cut. Re- prove the quality of soil in addition to producing taining some of the structures from the old stand at valuable wood products. All Cupressaceae accu- the time of final harvest is one of the best ways to mulate calcium and other bases in their foliage, providethenewstandwithahighlevelofstructural which produces high-quality litter (Kiilsgaard, diversity, including larger structures. Greene, and Stafford 1987). This li ttercontributes, in turn, to higherbase saturation, higher rates of nitro- COARSE WOODY DEBRIS gen mineralization, reduced acidity, and produc- Coarse woody debris, including large standing tion of more biologically active mull humus condi- dead trees and fallen boles, are extremely impor- tions (Alban 1969; Turner and Franz 1985). tant to ecosystem function. They provide habitat Delaying the process of canopy closure can for many elements of biological diversity and also have environmental benefits in some young essential processes. Rotting wood is also impor- stands. Canopy closure is probably the most dra- tant to maintenance of site productivity by con- matic and, for some organisms and processes, the tributing nutrients and organic matter to the soil; most traumatic single event in the life of the even the identifiable wood fragments incorpo- stand, other than its ultimate destruction by some rated in to the soil-soil wood-play a distinctive catastrophe. Many aspectsof the forest, including role (see, for example, several papers in Harvey its composition and functioning, change rapidly and Neuenschwander 1991). Practices that con- and significantly at the time of canopy closure. tribute to maintenance of coarse woody debris Intensive forest management has traditionally include the retention of such material at the time sought to achieve rapid closure of the canopy of harvest cutting and the creation of snags and (early full occupancy of the site by commercial logs from trees reserved for that purpose. trees) following a disturbance, such as clear-cut- Retention of snags and logs is particularly ting. Yet the open conditions prior to closure of effective for maintaining coarse woody debris the tree canopy are important ecologically. The whenharvestingtreesinyoungandmaturestands stage prior to canopy closure is rich in plant and of natural origin and old-growth forests. Natural animal species, including many game species, standstypicallyhavesignificantamountsof coarse thatarevaluedbyhumans(Hunterl990;Thomas woody debris that can be used as a biological 1979). Vascular plant species with nitrogen-fix- legacy. Retention of snags is more controversial ing symbionts are most common during this pe- and less effective than maintenance of logs for a riod. Hence, maintaining open conditions farther variety of reasons. There are concerns for the into the rotation-delaying full canopy closure- safety of forest workers because of the potential can provide ecological benefits. for structural failure in snags. Retention of snags Canopyclosurecanbe delayed by maintaining increases logging costs. Snags also create poten- wider spacingsbetween trees. Reducing the plant- tial problems in fire protection because they tend ingdensitiesbetweentreesandundertakingheavy to produce firebrands once ignited. Snags are a precommercial thinning achieve this objective. potential hazard to aircraft involved in manage- Furthermnore, studies show that wide spacing can ment activities. Furthermore, many snags have be maintained in young stands with little or no relatively short life spans, especially if they are sacrifice in the volume of commercial wood pro- heavily rotted. Nevertheless, efforts to maintain duced (for example, Reukema 1979; Reukema snags on cutover forestlands are increasing be- and Snmith 1987). Trees can be pruned to produce cause of their importance to many animal species. high-quality wood in the open-grown stands. Two approaches used to reduce hazards associ- Such prescriptions are used in management of ated with snag retention are (a) clustering of Pinus radiala plantations in New Zealand. snags in small groups or patches and (b) creating snags from living trees following harvest cutting. Structural retention at harvest Silvicultural prescriptions can be designed to maintain agiven numberand distribution of snags Structural diversity is emerging from forest eco- rotation.ba nd objtive in system researchasa critical attributein providing oveud otatfors i n nrh estern n for a diversity of processes and organisms. Fur- v thermore, it is some of the large structures-large America, forexample, isthecontinuousavailabil- living trees, large snags, and large fallen boles- ity of five large (more than 50 centimeters in that are typically absent from managed stands, diameter) snags per hectare. 373 Defining and Measuring Sustainability: The Biogeophysical Foundations Living trees can be used as sources for coarse by retaining living trees on harvested areas while woody debris in stands that lack either large simultaneously producing and removing wood snags or logs, as is the case for stands currently products. First, living treescanbe used as sources under intensive management. This is a particu- of coarse woody debris-snags and logs-espe- larly valuable practice for restoring structure to cially where safety concerns or logging methods stands and landscapes that have been simplified make retention of snagsdifficult. Living trees can by past practices. also be retained to provide wildlife habitat. Maintenance of appropriate quantities and Living trees can function as refugia and in- qualities of coarse woody debris in managed ocula for many of the smaller organisms or hid- stands is, of course, much more complex than den diversity mentioned early in this review. For simply providing periodically for a few dead example, many species of the rich invertebrate trees. Different tree species provide snags and fauna found in forests have poor dispersal capa- logs with substantially different characteristics bili ties (Lattin 1990). Such organisms typically do and ecological potential. All tree species are not not recolonize areas once their habitat has been equal in terms of their behavior as coarse woody eliminated by clear-cutting. Refugia for these debris! Furthermore, coarse woody debris needs kinds of organisms can be provided by leaving to be present in various stages of decay. Material host trees, which then become an inoculum or of greaterstructuralsoundnessmaybeimnportant source of seeds for the new stand. The same for geomorphic and some animal habitat roles, concept is applicable to mycorrhizae-forming for example, while highly decayed wood is of fungal species. At least some of these fungi can greater value as a component of soil. Numerous disappear from cutover areas if all potential host questions exist as to the quantities and spatial species are eliminated (Perry and others 1988). distribution of coarse woody debris required to When some of their hosts are left behind, the achieve specific management objectives; how fungal communities are conserved and can in- much is enough? Developing the specific data oculate the young stands. This concept is a coun- ultimately needed for silvicultural prescriptions terpoint to the forester's common complaint that will be a challenge to scientists and managers for living trees cannot be retained because they are many years to come (Maser and others 1988). sources of pests and pathogens, such as inverte- brates and fungi. Most invertebrates and fungi in RETENTION AND PARTIAL CUTTING OF GREEN TREES forest stands are, in fact, essential components Retention of green trees on cutover areas is an- that should be retained, and maintaining living other practice that can create higher levels of trees on the si te is one important tool for achiev- structural diversity on managed stands. This ap- ing this objective. proach canbereferred toas partial cutting, partial Retention of livi ng trees,especiallydominants, retention, or green tree retention, in an effort to also alters the microclimate of the cutover area. distinguish it from both clear-cutting and selec- That is what traditional shelterwood cutting is all tive cutting (Franklin 1990). Retention of green about: theoverstorymoderatesthemicroclimate, trees involves reserving a significant percentage encouraging regeneration of trees where the en- of the living trees, typically including some of the vironment on a clear-cut would be too severedue larger or dominant individuals, at the time of to heat or frost (Daniel, Helms, and Baker 1979). harvest for retention through the next rotation. Obviously, what works for tree seedlings will The density, composition, condition, and distribu- work for other forest organisms as well: they tionof thereserved treesvary widely,dependingon would also be expected to survive better on partial management objectives, initial stand conditions, cuttings than on clear-cuts. Perhaps as important, and otherconstraints. The general objective is, how- many organisms will move more readily through a ever, to sustain a more structurally diverse stand patch or landscape that hasat least some living trees than could be obtained through even-aged man- than through a clear-cut environment because of agement. Partial cutting has not been widely used theameliorated climateorprotectivecoverorboth. or recognized in forestry. Some silvicultural text- Replacementofclear-cuttingwithpartialcuttingon books briefly discuss related concepts such as a managed landscape matrix could dramatically shelterwood with reserve, but these approaches improve overall connectivity and reduee the isola- have not been widely taught or applied in forestry. tion of islands of natural habitat. Many ecological objectives, including several Retention of green trees can be used as a strat- that contribute to sustaiinability, can be achieved egy to grow large, high-quality wood during the 374 Sustainability of Managed Temperate Forest Ecosystems next rotation. For example, mature (80- to 250- centrate them in patches-is another important year-old) Pseudotsuga menziesii are still capable of issue in partial cutting. The answer depends par- substantial growth (Williamson 1973). Hence, tially on objectives and constraints. For example, large living trees could provide both economic when small patches of living trees are retained, and ecologic benefits in scenarios involving man- they may be more effective as refugia for inverte- agement of a mixed stand with a low density of brates; aggregating them may also minimize im- large, slower growing trees of one species on a pacts on logging and other forestry operations. At long rotation, while simul taneously growi ng sev- the same time, well-distributed snags and logs are eral rotations of a rapidly growing second species. desirablebecause they maintain productivityof the Partial cutting could be used to create mixed- soil and provide habitat for some wildlife species. structure stands that providecritical wildlifehabi- tat. For example, in northwestern North America, SELECTIVE CUTrING numerous natural stands represent mixtures of Selective cutting involves the removal of indi- young and old structures; the stand may be rela- vidual or small groups of trees at relatively fre- tively young (for example, eighty years) but also quent intervals (every ten years). This system is containasignificantcomponentof large,old trees, typically aimed at creating or perpetuating un- large snags, and large logs. Such stands are typi- even-aged stands (more than three age classes) cally the consequence of wildfires or windstorms and always maintains a protective cover at the that left behind a large legacy from the original site. In the last several decades, large industrial stand. Forests of this type often provide suitable forest landowners and government forest agen- habitat for animal species generally associated cies have rarely used selective cutting as a major with old-growth forests (Ruggiero and others approach to forest management, although there 1991). Partial cutting systems could be used to have been exceptions. Its unpopularity is due to create comparable mixed-structure stands, which the high costs, inefficiency, technical difficulty in would provide late-successional forest habitat application, and potential damage to stands and conditions in one-quarter to half of the time that sites when applied to large trees on steep moun- would be required following traditional clear- tainous topography. cutting. Selective cutting can, however, be an effective Partial cutting could also be used to reduce the technique for maintaining compositional and impact that harvesting the forest has on hydro- structural diversity in stands that are managed logic and geomorphic processes. For example, for low to moderate levels of wood production retention of living trees can reduce the potential (Daniel, Helms, and Baker 1979). Of all thecutting forlandslidesby maintaining root strength, which systems, it provides the highest level of biological is critical to maintaining the stability of soils on legacies of all types and, under some circum- steep slopes. Retention could also be used to stances, minimizes impacts on the long-term pro- reduce theimpact of cutting where harvesting the ductive potential of the site. forest contributes to the frequency and intensity Selective cutting is not a panacea, however. It of flood flows. In northwestern North America, is a difficult system to apply and requires the for example, maintaining a sufficient number of forester to have a high level of technical compe- large trees to intercept snow and maintain the tence, particularly if a high-grading approach, in thermal balance on cutover areas will reduce the which only the most valuable trees are removed, intensity of the rain-on-snow floods that are com- is to be avoided. It can be very difficult to apply mon in this region. when the preferred species of tree crop is intoler- Prescriptions for partial cutting vary widely, ant of shading, and potential shade-tolerant com- depending on many factors, including manage- petitors are present. The potential is also very ment objectives and stand conditions. For ex- high for damage to residual stands and for accel- ample, providing a minimal number of snags and erated erosion and soil degradation when ap- logs may require retention of as few as ten to eigh- plied to stands on steep mountain topography; teen trees per hectare. Creation of mixed-aged, this is because selective cutting requires frequent mixed-structureforestssuitableforlate-successional entries to the stand, which, in turn, may necessi- forest species may require twenty to forty retained tate creation and maintenance of a dense system trees per hectare, depending on age of the forest. of roads and skid trails. Use of aerial logging The appropriate spatial distribution of retained techniques can reduce some of these impacts, but living trees-whether to disperse them or con- their high cost may be prohibitive. 375 Defining and Measuring Sustainability: The Biogeophysical Foundations It also appears that selective cutting may not tions. Overall, long rotations help to maintain a maintain conditions suitable for many interior or greater diversity of organisms and processes. late-successional animal species, despite the high Most large industrial forest landowners view levels of structural retention. For example, at least long rotations as an anathema, but only from an some of the neotropical migrant songbirds that economic standpoint. In their view, the return use the eastern North American deciduous for- on their investment is simply not acceptable in ests respond negatively to the creation of even managed forests with rotationslonger than fifty smallopeningswithinthelargeintactforestareas to sixty years. Such strictly economic criteria they require(Terborgh 1992). In anotherexample are rarely applied to government forest lands, from the tropics, selective logging has had signifi- however. cant and pervasive impacts on animal species of the interior forest. Hence, it is not safe to assume Amelioration and restoration practices that selective cutting is the best approach to inte- grating forest harvest and environmental values, The potential for restoring structures, organisms, including sustainability. and processes to forest stands that have been Much moreextensive use of selective cutting is simpaified or degraded warrants mention. There appropriate in future efforts to develop forest are many situations where silviculture can con- management approaches that are more ecologi- tribute to the restoraton of degraded sites, the cally sensitive. It will be most applicable on areas creation of habitat (such as snags and other struc- where there is less emphasis on commodity pro- tures), and the reintroduction of organisms. For duction and where the species and topography example, intheNorthwestern UnitedStates, work- are appropriate to frequent light-harvest entries. shops, experiments, and pilot tests are underway In contrast, partial cutting, as presented earlier, is aimed at restoration of structural complexity in typically designed around asingle-harvest entry simplified young stands developed following per rotation. earlier logging. Specific objectives typically in- clude provision of habitat for species associated LONG ROTATIONS with mature and old-growth forests. The practice of using long rotations has a high potential as an alternative management approach that would reduce the impacts of harvest cutting Existing materials for assessing on environmental values, including site produc- sustained productivity tivity and biodiversity. This might involve in- creasing rotation ages by a factor of 1.5 to 2.0 over Productivity is a topic that has concerned forest- current rotations, which have been based on eco- ers and forest scientists for several centuries. nomic factors or biological maturity of crop trees Numerous approaches have been developed to (theculmination of mean annual increment). Asa assess productivity and numerous datasets have specific example, the rotation age (frequency of been collected that vary widely in terms of their final harvest cut) on national forests in the north- formality, levels of sophistication, effort, etc. Not western United States might be shifted from the surprisingly, there are great differences in ap- current 80 to 100 years to 160 to 200 years. proaches, both between different countries and Such shifts in rotation age can have numer- regions within countries. In general, uniform ap- ous environmental benefits, including reduced proaches have not been developed and adopted, impacts on soils and biological diversity. They except in countries where a national forestry or- can drastically reduce the proportion of a man- ganization has had the ability to define and push aged landscape that is in a recently cutover con- adoption of a countrywide approach. There have dition, which, in turn, can reduce the risk to soils been attemps through organizations like the and water quality since, for example, recently International Unionof Forest Research Organi- cutover areas are much more subject to erosion zations (IUFRO) to at least develop and adopt and landslides. A higher percentage of the land- standardized terminology and explore com- scape would be in forest cover under long rota- mon methodological approaches (AAAS 1967 tions, and some of this forest would include later and Newbould 1967). However,uniform methods stages of forest succession that would not be do not generally exist even among adjoining presentin a landscape managed undershort rota- regions, let alone countries. 376 Sustainability of Managed Temperate Forest Ecosystems Numerousother problems areassociated with Such traditional inventories typically have existing methodologies, predictive tools, and data numerous deficiencies that drastically limit their sets in addition to this general lack of uniformity. potential value as models for a system to assess For example, the vast majority deal only with the long-term productivity. Foremost among these is production of wood or biomass and do not ad- that they are typically designed to provide a dress total productivity of the ecosystem. Fur- statistical sample of an entire forest ownership, ther, even when a methodology purports to ad- state, region, or country. They do not provide dress productivity of the forest ecosystem, sig- estimates of standing crops (let alone productiv- nificant components, such as the productivity ity) at the level of an individual stand; the data belowground, are ignored (Harris, Sanantonio, cannotbe related to some spatial data base. Hence, and Mc Ginty 1980). Hence, a wealth of material it is possible to infer that there are x hectares of exists for assessing productivity of temperate stands of y age and z volume, but there is no way forests, but little of it can be directly adapted to to determine where they are located within the address the primary issue in this volume: mea- sampling area. suring the productivity of ecosystems over long Because forest sampling is conducted at very periods of time. low densities and is not stand-specific, interpret- ing the causes of changes in stand volumes and Measurement programs growth rates between remeasurement periods is very difficult. The phenomenon of declining pro- Sample plots placed In forest stands have been a ductivity in the third generation of forests in the primary tool of foresters almost since the incep- southeastern United States provides an excellent tion of the forestry profession. Although they example; in the 1970s and 1980s, growth rates vary widely in size, layout, permanence, and appeared to have declined, leading to sugges- almost all other features, sampling plots have tions that site productivity might also be declin- been around for a long time and have provided ing. Several alternative hypotheses were pro- the bulk of the empirical data for estimating for- posed, however, and resolution of this issue has est productivity. been rendered difficult by the low density and Most countries, agencies, orcorporationsrthat are geographical resolution of the sample. involved in management of significant forest pirp- Stand-based inventory systemsdoexistin some ertics have some kind of continuing forest inven- ownerships and provide a better opportunity to tory program that uses a system of sample plots, identify changes, and their causes, in the produc- Continuous forest inventory is a central concept in tivity of a forest over time. Clearly, none of these forestrythatgenerallyinvolvestheestablishmentof systems is designed to measure changes in the permanent sample plots over the forest property, inherent productivity of a site. They are simply usually using some systematic sampling design. approaches to estimating standing volumes of The specifics of these plots vary widely with the wood or aboveground organic material and to organization,buttheus ofplotclustrs iscomrmon. calculating rates of accumulation or loss; that is, Mcasurement intervals also vary; however, five- growth or productivity of the current stand. year remeasurements are common. A protocol or set of protocols to assess changes Inrsomencountries, national organizationshave in long-term forest productivity has not yet been responsibility for conducting forest inventories, designed nor, obviously, implemented. The for- In theAUnited States, forvexample, therDpartment estry organization that has gone the furthest in of Agriculture Forest Service takes periodic in- this direction, so far as I know, is the group in ventories of resources on all publicly owned for- New Zealand concerned with management of the csts. Regional experiment stations design and exotic Pinus radiata forests. conduct this inventory, except on national forest Long-term experiments are being developed lands, where it is carried out by the National andimplementedbyforestresearchandmanage- Forcst Management Organization. This inven- ment organizations to address the specific issue tory system,as with many directed to forest lands, of long-term forest ecosystem productivity under IS begining to broaden its focus beyond the varying management regimes. Examples exist in counting of standing live trees and calculation of many temperate forestregions. Theseareresearch wood volumes to the acknowledgment of other projects, not routine operational activities; the forest attributes (structures such as snags and investment in data collection is high, and the fallen logs) and resources. 377 Defining and Measuring Sustainability: The Biogeophysical Foundations geographic scope is typically low. Nevertheless, of the International Biological Programme(Reichle these experiments will provide the most defini- 1981). This volume summarizes most of the re- tive information on long-term productivity and search on forest productivity conducted during a will almost certainly provide useful insights into global ten-year effort. Included are numerous the design of broader schemes for assessing the data sets from intensive study sites throughout productivity of forest ecosystems. (primarily, but not exclusively) the northern tem- perate region. Although incomplete for some Existing data sets variables, the focus is on ecosystem productivity. Numerous research data sets have been gener- An immense number of data sets address the ated since the 1960s, when studies of ecosystem issue of forest productivity over a variety of spa- productivity firstbecame popular. Cannell (1982) tial and temporal scales. The forestry literature is has compiled many of these in a book, World full of data on productivity, as are the files of Forest Biomass and Primary Production Data. Other forest research and management agencies. As Foretantompilatimanductionsothere noted earlier, the concept of continuous forest important compilations and discussions of forest inventoted isarlerally follonept i o countries productivity are Eckardt (1968, a volume pub- inventory is generally followed in most. lished as part of a UNESCO series), the proceed- that have significant forestry programs. ings of several conferences (International Union It would be impossible to describe or list this of Forest Research Organizations 1971,1973), and immense body of data in a short review. Many a summary of research, Productivity of the World reviews and directories are already available, Ecosystems (Reichle, Franklin, and Goodall 1975). several of which are cited below. Unfortunately, No list of major publications on forest productiv- much of this information has limited value. As ity would be complete without mention of the noted earlier, these data sets typically focus only Russian classic, Productivity and Mineral Cycling on the tree and often only on the wood compo- in Terrestrial Vegetation, by Rodin and Bazilevich nent. They do not address productivity of the (1967), which includes many tables of data on for- ecosystem as it is currently defined. ests as well as other types of terrestrial ecosystems. The fact that these data sets consider only the On an operational rather than a research level, aboveground portion of the tree or forest is a the most comprehensive data sets on forest pro- serious deficiency. When monitoring changes in ductivity are probably those associated with very forest productivity, this approach is unaccept- intensive management of plantations. Examples able, particularly since shifts in the relative pro- include the exotic Pinus radiata plantations in ductivityofabovegroundandbelowgroundcom- New Zealand, Australia, and South Africa and ponents can occur with changes in site condi- Now Zealustpla,tand Sorldw ica and tions. In temperate regions, the proportion of hybrid Populus plantations worldwide. Forestry photosy nthate usedi belowground tends to in- agencies in at least some of these locales have given photosnthat usedbelowgound ends o in- substantial attention to long-term productivity and crease substantially with increased nutrient or its measurement, including belowground compo- water stress (Kimmins 1987). Finally, these data nents in a few casecs sets have rarely been designed to assess long- Asnoted, innumerabledatasetsonforestpro- term changes in productivity. Not only are essen- dAuctivity address the production of wood. Al- tial elements missing, but the methods used have though essentially no data sets were designed often been modified over decades of specifically to address forest ecosystem produc- remeasurement,makingcomparisonsofdatadif- tivity over long time periods, some major forest ficurit. research projectshave been or currently arebeing During the last few decades, various data sets established throughout thenorthemn temperatezone on the productivity of temperate forest ecosys- that should correct many of these deficiencies. tems have been developed that do attempt to address overall, rather than just wood, produc- Predictive models tivity. Even theseare far too numerous to compile in this review. There are, however, some major It is not surprising that many models of various references that provide access to many of these types are aimed at predicting forest productivity; data sets and a great deal of the literature. this is, after all, one of the major concerns of One of the best compilations of data on pro- foresters and forest management organizations. ductivity of temperate forest ecosystems is the Many approaches have been taken, including synthesis volumeon forestsgenerated by studies traditional forest yield tables and computerized 378 Sustainability of Managed Temperate Forest Ecosystems growth and yield models of highly varied con- oped at Oregon State University, is a good ex- structs. A growing number of forest succession ample of a state-of-the-art model of growth and models include predictions of overall changes in yield (Hann, Olsen, and Hester 1992). FORTNITE forest structure as well as biomass. is one of the few examples of a growth model designed to look at forest productivity as it is WOOD YIELD TABLES AND MODELS affectedbyvariousmanipulationsoverlongtime Yield tables have been the most traditional form periods (Kimmins 1987). of forest production models (see, for example, The limited empirical data bases from which Society of American Foresters, Forestry Handbook, yield tables and models have been constructed 1984). These models typically have been devel- have been one of their major limitations. Mortal- oped using data collected from a large number of ity, the most difficult variable to estimate, creates forest plots located in forest stands of relative the greatest degree of uncertainty in predicting species composition, stocking levels, geography, yields. and so on. Through a variety of mathematical, statistical, and subjective analyses, such data are ECOSYSTEM MODELS used as a basis for constructing tables that indi- During the last twenty years, a new class of forest cate the wood yields that should be expected at growth model has emerged that focuses on pre- various time intervalson sites of varying produc- dicting successional changes in forests over very tive potential. Productive potential is typically long timeperiods. Thesemodelsarebasedon the indexed by the site index criterion discussed ear- dynamics of tree populations, but they also pro- lier (heights achieved by dominant trees by some vide output on stand-level attributes, such as index age). Most yield table publications also accumulations of organic matter. Many of these include many other tables relating stocking den- forest ecosystem models are conceptually related sity, periodic growth and mortality, and other and are sometimes referred to as the FORET stand variables to stand age. family of forest growth models (Shugart 1984). Forest yield tables have most commonly been The first of these models was JABOWA, which developed for well-stocked, even-aged stands of was developed for the Hubbard Brook Experi- asinglespecies.Suchtablesexistforalmostevery mental Forest (New Hampshire). Its primary fo- important species and type of forest in the north- cusison theprocesses of birth, growth, and death ern temperate zone, and, typically, there are sev- of the tree population. These processesare driven, eralyield tables (often represenltingdifferentgeo- in turn, by environmental conditions, including graphic regions or management intensities) for temperature and moisture at the site and light very important commercial tree species. As very within the stand. The birth and death processes general models to predict forest growth, yield are based on probability functions. A variety of models have major limitations, particularly in model outputs-tree density, biomass, leaf area, their ability to predict accurately the growth of and so forth-is possible, dependingon theinter- specific stands. ests of the scientist or manager. The prediction of forest growth has shifted Extensive work is under way to improve the toward computer-based growth and yield mod- capabilitiesof FORET-typemodels, including the els with the development of computer technolo- developmentof spatially explicit versions, which gies. These tend to be much more sophisticated keep track of the location of individual trees, and deal with a larger set of variables than was incorporate thedynamics of coarsewoody debris possible with traditional yield tables. However, as (standing dead trees, fallen logs, and so forth), with yield tables, they are normally based on1 em- and incorporate more realistic probability func- pirical data sampled from some forest population. tions, such as for tree mortality. Versions have Many such growth and yield models use a been developed that focus on the nutrient status wide variety of approaches and assumptions. of the site as well as forest structure. Early yield tables and growth simulation models Current models of this type have some limita- focused on natural stands, but development of tions. Because they are designed to simulate growth simulators for managed stands has been changes in a variety of ecosystem attributes over particularly popular with forestry agencies dur- verylong bmeperiodsand fordiversesites,predic- ing the last several decades. This has allowed tionsoftreegrowthorwoodproductionforspecific managers to consider the effects of various man- stands may not be as good as for the traditional, and agement regimes. ORGANON, a model devel- much more specific, growth and yield models. 379 Defining and Measuring Sustainability: The Biogeophysical Foundations An important attribute of these models is that step tor a program that truly intends to monitor they are stochastic or probabilistic rather than sustainability.Fortunately,suchastrategyiscon- deterministic.Mostforestgrowthandyieldmod- sistent with the emerging interest in adaptive els are deterministic: only one solution is possible management of resources.Thisapproachrequires with an initial set of conditions. But, with the comprehensivemonitoringtoprovidetheresource FORET family of models, an infinite number of management system with corrective feedback. solutions is possible. Hence, numerous, even Once a decisionhasbeen made toproceed with hundreds of, simulations mav be run for a given a multi-factor monitoring program, the specific stand to produce an array of predicted outcomes: variables can be chosen and protocols developed in effect, a probability distribution for future con- by working with appropriate scientific teams. ditions of the stand. There will undoubtedly be substantial variation among forested regions, both in terms of vari- ables and sampling techniques--another blow, Conclusions and recommendations unfortunately, to the notion of a singular global scheme. In the following sections, some candi- Assessing the long-term productivity or date variablesareproposedforaminimalmoni- sustainabilityoftemperateforestecosystemsrep- toring program to assess productivity of the resents a major challenge. Currently, this is not forest ecosystem. being adequately accomplished anywhere, nor does a suitable model or prototype exist for such Minimnal program for monitoring a program. Hence, development of a protocol or, susta inab ility of the forest ecosys tern better still, a series of protocols for measuring the sustainability of forest ecosystems should have A minimal program should incorporate the fol- very high priority. Existing research and man- lowing measures: forest cover and condition at agement programs can provide useful informa- therlandscape level, systemilosses and hydrologic tion and guidance in this effort. controls, bilological condition of the forest,condi- The most important points in designing the tion of the soil, and biological diversity. assessment program are (a) recognizing the ne- cessi ty of assessing several variables and (b) iden- PrioESiCOVesrANDCDitsON AThe1 IExANIDtAIfELEVce tifying those variables. Clearly, no single variable Periodic assessments of the extent of forest cover, will adequately assess sustainability. Nlonitoring and some interpretation ofvitscondition(age and sustainability of forest lands and associated wa- stocking density), are the variables of interest in ters will require periodlic assessments of a broad this segment of the monitoring program. These arrav of variables from the landscape to popula- assessments would probably be made at five- lions of specific organisms. Specifically, a moni- year intervals and use various types of data ob- toring ogram org assesspefollowini: tained by remote imagery, satellites, or aircraft. toriig programn shouldassess the following: The assessments would probably be made at the * Forest cover and condition at the landscape level of regions, in the case of large coUntries, or level at the level of small countries. * Flow and quality of water * Structural conditionis, including live and dead SYSTEM LOSSES AND I IYDROL(X],C CONTROL trees, of the forest stand Aquatic systems-streams, rivers, and lakes-- • Physical, chei-nical, ai biological conditin of are probably the best integrators of the effects of * Physical, chemical, and biological condition of human activities on terrestrial landscapes. The the soil and system losses referred to here are primarily losses • Populationsand trendsiTnindicatororganisms. of soil and nutrients that can be measured within A program that covers such a broad range of aquatic systems as suspended or dissolved sedi- parameters will require programs of highlv var- ments and materials. Hence, a monitoring ele- ied spatialand temporalscale. Anapproachbased ment that addresses the production oraccumula- on a single measurement, index, or sampling tion of sediments and the quality of water is strategy is notgoing to besuccessful. Admittedlv, important. Note that this monitoring occurs in giving up the notion of such a simple monitorinlg the a(luatic environment rather than in upland program creattes gre ater comiiplexities and much forest area>. The prodtictionl of water-the total higher costs; nevertheless, it is an essential first anmounit,seasona.l distribu tioni,anld frequercyand 380 Sustainability of Managed Temnperate Forest Ecosystems level of flood flows-is also extremely important, biological measure of site productivity negates especially when water is recognized as one of the the need to monitorsoil parametersdirectly, there major products of a forest ecosystem. Both sys- is the possibility of doing irreversible harm to tem losses and water production are probably forest soils before that harm is reflected in declin- best monitored by creating a system of bench- ing productivity of the site. There is also the mark watersheds in forested regions where the possibility of biological compensations for de- flow and quality of water are sampled on a more- clining soil condition. or-lesscontinuousbasis.Techniquesforsuchmea- A soil monitoring program should include surement programs are well known and should assessments of loss (due to erosional processes), be easily adapted. The biggest problems are the physical conditions (bulk density and physical initial cost of such installations and the continu- conditions), chemistry (primarily levels of critical ing costs in funds and technical personnel of nutrients with consideration of trace and toxic maintaining and analyzing data generated by elements), and biota of the soil. One specific ele- such a monitoring program. However, government ment of the soil biota that should receive attention organizations such as the U.S. Geological Survey in a monitoring program is the diversity of myc- have extensive experience with these activities. orrhizae-forming fungal species that are present. BIOLOGICAL CONDITION OF THE FOREST BIOLOGICAL DIVERSITY The primary focus of this portion of the monitor- General measuresof biological diversity, in termsof ing program is on some measure of productivity overall species richness, are probably not of much at the level of the individual forest stand. Tree value in a monitoring program. Diversity indexes growth per unit of time under some specified tend to assume that all species are of equal interest conditions, such as dominant free-grown indi- and that the richer the ecosystem, the better. This is viduals, still appears to be the best measure for clearly not the case with forested and, probably, integrating the overall effect of all variables influ- most other natural or semi-natural ecosystems. encing productivity. A direct measure of net pri- It will often beappropriate to include monitor- mary productivity is beyond the scope of a rou- ing of selected organisms, however, because they tine monitoring program. have intrinsic importance to the ecosystem as The use of tree growth per unit of time is indicator and keystone species or because they conceptually the same approach that is used with have high interest and significance to Homo sapi- the site index concept reviewed earlier. An alter- ens. Species chosen for monitoring will have to be native approach might be some measure of over- carefully selected based on scientific and societal all productivity of the stand; however, this can be considerations; however, monitoring at the level very strongly influenced by stand conditions, so of individual species, guilds, functional groups, the problem of standardization is greater than and so forth must almost inevitably be part of a where individual trees are used. comprehensive monitoring program. In addition to a measure of site productivity, it This is probably themost difficult and, in terms may also be important to monitor the structural of criteria for selecting organisms or organismal diversity found within forest stands. Levels of groups, the most poorly developed assessment standing dead trees (snags) and fallen logs on the approach. Although techniques exist for moni- forest floor are an example of an important struc- toring many vertebrate groups, such as birds, tural element that has often been ignored in pro- mammals, and amphibians, they may require grams to monitor the condition of a forest. Be- high levels of technical expertise. More critical is cause such material is extremely important as the lack of developed approaches to the monitor- animal habitat, and can be important in maintain- ing of functionally important groups such as in- ing site productivity, it should be an element of vertebrates, including insects, and fungi. any scheme for monitoring forest ecosystems. Any monitoring program that purports to ad- dress sustainability of the forest ecosystem must, CONDITION OF THE SOIL of necessity, address organisms as species or The soil is, in terms of human life spans, a largely groups of species; it cannot be based totally on a nonrenewable resource. It is important, there- single measure or integrated index of ecosystem fore, to monitor specifically the physical, chemi- function, such as productivity. Biodiversity is cal, and biological state of this basic resource. basic to long-term sustainability or productive Although it can be argued that the use of some potential. 381 Defining and Measuring Sustainability: The Biogeophysical Foundations Implementing a monitoring program For other suggestions on research programs, see Forestry Research: A Mandate for Change (Na- Generic plans for monitoring forest sustainability tional Research Council 1990). can be developed at the global and continental levels, but details of parameters and sampling techniques will have to be adapted to the particu- References lar conditions of regions and individual coun- tries. Establishment of a global advisory body to AAAS (American Association for the Advance- developgeneral guidelinesand assist in planning mentof Science). 1967 Primary Productivityand and implementing monitoring programs at the Mento Scinc 16 Primay Productivity Nd level of countries, regions, and continents would Mnral certainly be useful, so long as the equally critical York. elements of flexibility in design and scientific Aber,J. D., K. J. Nadelhoffer, P. Steudler, and J. M. integrity and credibility are maintained. Melillo. 1989. "Nitrogen Saturation in North- Residentsof rural environments in and around ern Forest Ecosystems." BioScience 39, pp. 378- forests should be given special consideration for 86. employment in the monitoring program. Tradi- Adams, R. M., C. Rosensweig, R. M. Peart, J. T. tionally, monitoring programs are assigned to Ritchie, B. A. McCarl, J. D. Glyer, R. B. Curry, professional and technical personnel within es- J.W. Fones, K. J. Boote,and L. H. Allen,Jr. 1990. tablished agencies, who often live outside the "GlobalClimateChangeand U.S. Agriculture." affected region and are subject to frequent trans- Nature 345, pp. 219-24. fers. Resident populations have long-term famil- Alaback, P. B. 1984. "A Comparison of Old-growth iari ty with the region, includi ng appropriate work Forest Structure i n the Western Hemlock-Si tka experience, and they intend to reside in the locale. Spruce ForestsofSoutheast Alaska." In W. R. Necessary scientific and technical training could Meehan, T. R. Merrell, Jr., and T. A. Hanley, be provided for selected residents who could eds., Fish and Wildlife Relationships in Old-Growth then be incorporated into the long-term monitor- Forests, pp. 219-25. Bronx, N.Y.: American In- ing program. stitute of Fisheries Research Biologists. Alban, D. H. 1969. "The Influence of Western Accelerated research program Hemlockand Western Red Cedaron Soil Prop- It should be clear from this review that research erties." Proceedings of the Soil Science Society of based on the productivity of forest ecosystems America 33, pp. 453-57. and their maintenance needs to be drastically Bazzaz, F. A. 1990. "The Response of Natural expanded. Forest science, in particular, needs to Ecosystems to the Rising Global CO2 Levels." broaden its view from the level of the tree (or just Annual Review of Ecologyand Systematics 21, pp. the bole) to that of the whole ecosystem. The 167-96. following categories of research are critically in Bormann, F. H., and G.E. Likens. 1981. Pattern and need of attention: Process in a Forested Ecosystem. New York: * Productivity of belowground portions of forest Springer-Verlag. ecosystems Cannell, M. G. R. 1982. World Forest Biomass and * Canopy architecture and its effect on produc- Primary Production Data. New York: Academic tivity, particularly the relative effectiveness of Press. multilayered and multispecies canopies Carroll, G. C. 1980. "Forest Canopies: Complex * Dynamicsof soil organic matterand chemistry and Independent Subsystems." In R. H. War- over long time periods, including rates of soil ing, ed., Forests: Fresh Perspectives from Ecosys- development under natural regimes tem Analysis, pp. 87-107. Proceedings of the * Ecological role and dynamics of coarse woody fortieth annual Biological Colloquium. debris across a full range of ecosystems and Corvallis,Oreg.:OregonStateUniversity Press. * Improved understanding of causes and Daniel, T. W., J. A. Helms, and F. S. Baker. 1979. patterns of tree mortality and other stochastic Principles of Silviculture. 2d ed. New York: processes in forest ecosystems. McGraw-Hill Book Company. 382 Sustainability of Managed Temperate Forest Ecosystems Denison, W. C. 1979. "Lobaria oregana, A Nitro- Douglas Fir Forests, pp. 71-80. PNW-GTR-185. gen-fixing Lichen in Old-growth Douglas Fir Portland, Oreg.: U.S. Forest Service, Pacific Forests." In J. C. Gordon, C. T. Wheeler, and D. Northwest Research Station. A. Perry, eds., Symbiotic Nitrogen Fixation in the Franklin, J. F., and others. 1981. "Ecological Char- Management of Temperate Forests, pp. 266-75. acteristics of Old-growth Douglas Fir Forests." Corvallis, Oreg.: Oregon State University For- General Technical Report PNW-118. U.S. De- est Research Laboratory. partment of Agriculture, Forest Service, North- Easmus, D., and P. G. Jarvis. 1989. "The Direct west Forest Range and Experiment Station, Effects of Increase in the Global Atmospheric Portland, Oreg. CO2 Concentration on Natural and Commer- .1986. "ModifyingDouglas Fir Manage- cial Temperate Trees and Forests." Advances in ment Regimes for Nontimber Objectives." In Ecological Research 19, pp. 1-55. C. D. Oliver, D. P. Hanley, and J. A. Johnson, Eckardt, F. E., ed. 1968. Functioning of Terrestrial eds., Douglas Fir: Stand Management for the Fu- Ecosystems at the Primary Production Level. Pro- ture, pp.373-79. Contribution 55. Seattle, Wash.: ceedingsof theCopenhagen Symposium. New University of Washington, College of Forest York: United Nations Educational, Scientific, Resources. and Cultural Organization. .1991. "Effectsof GlobalClimate Change Edmonds, R. L., ed. 1981. Analysis of Coniferous on Forests in Northwestern North America." Forest Ecosystems in the Western United States. Northwest Environmental Journal 7, pp. 233-54. United States/International Biosphere Pro- Gholz, H. L. 1982. "Environmental Limits on gram Synthesis Series 14. Stroudsburg, Penn.: Aboveground Net Primary Production, Leaf Hutchinson Ross. Area, and Biomass in Vegetation Zones of the Franklin, J. F. 1990. "Biological Legacies: A Criti- Pacific Northwest." Ecology 63:2, pp. 469-81. cal Management Concept from Mount St. Grier,G.C.,andothers. 1989."Productivityof the Helens." In Transactions of the Fifty-fifth North Forests of the United States and Its Relation to American Wildlife and Natural Resources Confer- Soil and Site Factors and Management Prac- ence, pp. 216-19. Washington, D.C.: Wildlife tices: A Review." General Technical Report Management Institute. PNW-222. U.S. Department of Agriculture, .1992. "Scientific Basis for New Perspec- Forest Service, Pacific Northwest Research Sta- tives in Forests and Streams." In R. J. Naiman, tion, Portland, Oreg. ed., Watershed Management: Balancing Halpern, C. B. 1988. "Early Successional Path- Sustainability and Environmnental Change, pp. ways and the Resistance and Resilience of For- 25-72. New York: Springer-Verlag. est Communities." Ecology 69, pp. 1703-15. Franklin, J. F., and R. T. T. Forman. 1987. "Creat- _. 1989. "Early Successional Patterns of ingLandscapePatternsby ForestCutting: Eco- Forest Species: Interactions of Life History logical Consequences and Principals." Land- Traits and Disturbance." Ecology 70:3, pp. scape Ecology 1, pp. 5-18. 704-20. Franklin, J. F., P. M. Frenzen, and F. J. Swanson. Hann, D. W., C. L. Olsen, and A. S. Hester. 1992. 1988. "Re-creation of Ecosystems at Mount St. ORGANONI User's Manual. Oregon State Uni- Helens: Contrastsin Artificial and Natural Ap- versity, Department of Forest Resources, proaches." In J. Cairns, Jr., ed., Rehabilitating Corvallis, Oreg. Damaged Ecosystems, vol. 2, pp. 1-37. Boca Harmon, M. E., and H. Chen. 1992. "A Compari- Raton, Fla.: CRC Press.HamnM.E,adHCh.19."Cop- Raton, Fla.: CRC Press. son of Coarse Woody Debris Dynamics in Two Franklin, J. F., H. H. Shugart, and M. E. Harmon. Old-growth ForestEcosystems:Chanbai Moun- 1987. "Tree Death as an Ecological Process." tain, PRC, and H. J. Andrews Experimental BioScience 37, pp. 550-56. Forest, U.S.A." BioScience 41, pp. 604-10. Franklin, J. F., and T. A. Spies. 1991. "Composi- Harmon, M. E., and others. 1986. "Ecology of tion, Function, and Structure of Old-growth Coarse Woody Debris in Temperate Ecosys- Douglas Fir Forests." In L. F. Ruggerio, K. B. tems." In A. MacFadyen and E. D. Ford, eds., Aubry, A. B. Carey, and M. H. Huff, tech. Advances in Ecological Research 15, pp. 133-302. coords., Wildlife and Vegetation of Unrmanaged Academic Press. 383 Defining and Measuring Sustainability: The Biogeophysical Foundations Harris, W. F., D. Santantonio, and D. McGinty. Maser, C., R. F. Tarrant, J. M. Trappe, and J. F. 1980."TheDynamicBelowgroundEcosystem." Franklin, eds. 1988. "From the Forest to the In R. H. Waring, ed., Forests: Fresh Perspectives Sea: A Story of Fallen Trees." General Techni- from Ecosystem Analysis, pp. 119-29. Corvallis, cal Report PNW-GTR-229. U.S. Department of Oreg.: Oregon State University Press. Agriculture, Forest Service, Pacific Northwest Harvey, A. E., and L. F. Neuenschwander, eds. Research Station, Portland, Oreg. 1991. "Proceedings: Management and Produc- National ResearchCouncil. 1990. Forestry Research: tivity of Western-montane Forest Soils." Gen- A Mandate for Change. Washington, D.C. eral Technical Report INT-280. U.S. Depart- Newbould, P. J. 1967. Methods for Estimating the ment of Agriculture, Forest Service, Intermoun- Primary Production of Forests. Oxford, England: tain Research Station, Ogden, Utah. Blackwell Scientific Publications. Heath, B., P. Sollins, D. A. Perry,and K. Cromack, Oliver, C. D., and B. C. Larson. 1990. Forest Stand Jr. 1987. "Asymbiotic Nitrogen Fixation in Lit- Dynamics. New York: McGraw-Hill Book Com- ter from Pacific Northwest Forests." Canadian pany. Journal of Forest Research 18, pp. 68-74. Perry, D. A., and others. 1988. Maintaining the Hopwood, D. 1991. Principles and Practices of New Long-Termn Productivity of Pacific Northwest For- Forestry. LandManagement Report71.Victoria, est Ecosystems. Portland, Oreg.: Timber Press. B.C.: British Columbia Ministry of Forests. _ 1989. "Bootstrapping in Ecosystems." Hunter, J. L., Jr. 1990. Principles of Managing For- BioScience 39, pp. 230-37. ests for Biological Diversity. Englewood Cliffs, Pritchett, W. L., and R. F. Fisher. 1987. Properties N.J.: Prentice-Hall. and Management of Forest Soils. 2d ed. New International Union of Forest Research Organiza- York: John Wiley and Sons. tions. 1971. Working Party on the Forest Biomass Reichie, D. E., ed. 1981. Dynatic Properties of Studies.Section25,GrowthandYield.Gainesville, Forest Ecosystems. International Biological Fla.: University of Florida. Programme 23. Cambridge, England: Cam- .1973. Working Party on the Mensuration bridge University Press. of Forest Biomass. 54.01 Mensuration, Growth, Reichle, D. E., J. F. Franklin, and D. W. Goodall, and Yield. Vancouver, B.C., Canada. eds. 1975. Proceedings of a Symposium on Produc- Jarvis, P. G. 1989. "Atmospheric Carbon Dioxide tivity of World Ecosystems. Washington, D.C.: and Forests." Philosophical Transcripts of the National Academy of Sciences. Royal Society of London B 324, pp. 369-92. Reincrs, W. A. 1992. "Twenty Years of Ecosystem Jenny, H. 1980. The Soil Resource Origin and Reorganization Following Experimental De- Behavior. New York: Springer-Verlag. forestation and Regrowth Suppression." Eco- Kiilsgaard, C. W., S. E. Greene, and S. G. Stafford. logical Monographs 63:4, pp. 503-23. 1987. "Nutrient Concentrations in Litterfall Reukema, D. L. 1979. "Fifty-year Development of from Some Western Conifers withSpecial Refer- Douglas Fir Stands Planted at Various Spac- ence to Calcium." Plant and Soil 102, pp. 223-27. ings." Research Paper PNW-254. U.S. Depart- Kimmins, J. P. 1987. Forest Ecology. New York: ment of Agriculture, Forest Service, Pacific Macmillan Publishing Company. Northwest Forest Range and Experiment Sta- Lattin, J. D. 1990. "Arthropod Diversity in North- tion, Portland, Oreg. west Old-growth Forests." Wings 15:2, pp.7-10. Reukema, D. L., and J. H. G. Smith. 1987. "Devel- Likens, G. E., ed. 1989. Long-Term Studies in Ecol- oppment over Twenty-five Years of Douglas ogy: Approaches and Alternatives. New York: Fir,WestcrnHemlock,andWesternRedCedar Springer-Verlag. Planted at Various Spacings on a Very Good MacMahon,J. A. 1981. "Successional Processes: Site in British Columbia." Research Paper Comparisons among Biomes with Special PNW-381. U.S. Department of Agriculture, Reference to Probable Roles of and Influ- ForestService,PacificNorthwestResearchSta- ences on Animals." In D. C. West, H. H. tion, Portland, Oreg. Shugart, and D. B. Botkin, eds., Forest Succes- Rodin, L. E., and N. I. Bazilevich. 1967. Production sion: Concepts and Application, pp. 277-304. and Mineral Cycling in Terrestrial Vegetation. New York: Springer-Verlag. London: Oliver and Boyd. 384 Sustainability of Managed Temperate Forest Ecosystems Ruggiero, L. F., K. B. Aubry, A. B. Carey, and Trappe, J. M., J. F. Franklin, R. F. Tarrant, and G. M. H. Huff, eds. 1991. "Wildlife and Vegeta- M. Hansen. 1967. Biology of Alder. U.S. Depart- tion of Unmanaged Douglas Fir Forests." ment of Agriculture, Forest Service, Pacific General Technical Report PNW-285. U.S. Northwest Forest and Range Experiment Sta- Department of Agriculture, Forest Service, tion, Portland, Oreg. Pacific Northwest ResearchStation, Portland, Turner, D. P., and E. H. Franz. 1985. "The Influ- Oreg. ence of Western Hemlock and Western Red Schoen, J. W., and M. D. Kirchoff. 1990. "Seasonal Cedar on Microbial Numbers, Nitrogen Min- Habitat Use by Sitka Black-tailed Deer on Ad- eralization, and Nitrification." Plant and Soil miralty Island, Alaska." ]ournalof WildlifeMan- 88, pp. 259-67. agemnent 54, pp. 371-78. Waring, R. H., and J. F. Franklin. 1979. "Evergreen Shugart, H. H. 1984. A Theory of Forest Dynamics: Coniferous Forests of the Pacific Northwest." The Ecological Implications of Forest Succession Science 204:4400, pp. 1380-86. Models. New York: Springer-Verlag. Waring, R. H., and J. Major. 1964. "Some Vegeta- Societyof American Foresters. 1984. ForestryHand- tion of the California Coastal Redwood Region book. New York: John Wiley and Sons. in Relation to Gradients of Moisture, Nutri- Sollins, P. 1982. "Input and Decay of Coarse ents,Light,andTemperature."EcologicalMono- Woody Debris in Coniferous Stands in West- graphs 34, pp. 167-215. ern Oregon and Washington." Canadian Jour- Waring, R. H., and W. H. Schlesinger. 1985. Forest nal of Forest Research 12, pp. 18-28. Ecosystems Concepts and Management. Orlando, Swank, W. T., and D. A. Crossley, Jr., eds. 1988. Fla.: Academic Press, Inc. Forest Hydrology and Ecology at Coweeta. New Williamson, R. L. 1973. "Results of Shelterwood York: Springer-Verlag. Harvesting of Douglas Fir in the Cascades of Terborgh, J. W. 1992. Diversity and the Tropical Western Oregon." Research Paper PNW-161. Rain Forest. New York: Scientific American U.S. Department of Agriculture, Forest Ser- Library. vice, Pacific Northwest Forest and Range Ex- Thomas, J. W., ed. 1979. Wildlife Habitats in Man- periment Station, Portland, Oreg. aged Forests: The Blue Mountains of Oregon and Zobel, D. B., A. McKee, G. M. Hawk, and C. T. Washington. USDA Agricultural Handbook553. Dyrness. 1976. "Relationships of Environment Washington, D.C.: U.S. Department of Agri- to Composition, Structure, and Diversity of culture. Forest Communities of the Central Western Cascades of Oregon." Ecological Monographs 46, pp. 135-56. 385 Defining and Measuring Sustainability: The Biogeophysical Foundations Much more serious is the loss of soil or genetic Comments resources. As theauthornotes,although soilrnaybe restored in the longer term, loss of genetic potential Ian J. Payton isessentiallyperrnanent. What wefrequentlyfail to realize is that the interdependent nature of much of the biota means that the loss of genetic diversity has The literature associated with the management of ongoing and frequently unforeseen consequences. temperate forests is, as the author acknowledges, ArecentNewZealandexampleisthediscoverythat immense. In bringing together those elements a rare root parasite (Dactylanthus taylori), which is that relate to sustainable management, Professor under threat from introduced herbivores and col- Franklin draws on his very extensive knowledge lectors, is pollinated by a species of endemic bat, of forest management in its broadest sense. In this which is also considered endangered. Simply pro- review, I offer comment on several aspects of the tecting Dactylanthus against predators will not en- chapter, particularly as they relate to New sure its long-term survival. Zealand's attempts to manage temperate forests Harvestinginvolvesdisturbanceandcanincrease in a sustainable manner. the risk of subsequent loss of nutrients and soil. The author comments that as temperate forests Disturbance, however, is also a part of the natural have been managed for several centuries, and be- forest cycle by which individuals and communities cause forestry crops have lengthy rotation periods, arcreplaced. Harvestingregimes,be they for timber foresters are used to taking a long view. Although orother forest products, need to takecognizanceof this may now be true of temperate forest manage- the natural processes by which individuals and ment practices in those parts of the world with a communitiesrespond todisturbance.Thisneedsto long history of human settlement, elsewhere the include consideration of whether replacement pat- emphasis continues to be placed on the harvesting terns are typically large or small scale and the of virgin old-gmrowth forests with little thought for processesby which nutrientsare recycled. Manage- the longer term. Only when the finite nature of the ment practices thatdo not mirror those of the natu- resource is recognized are serious attempts made to ral forest ecosystem need to be closely examined for manage it on a sustainablebasis. Or to put it another their potential to lead to a lossof net productivity. In way, moves toward sustainable forestry manage- thisrespect,timberharvestingsystemsthatinvolve ment practices appear to be born of necessity rather large-scale clear-cuts, particularly where these are than an inherent desire for sustainability. In New followed by fire to remove the slash, create a huge Zealand, where European settlement during the potential for loss of not only soil and nutrients but nineteenth century was followed by large-scaleclear- also of genetic diversity. ance of forests for agriculture and timber produc- The author makes the interesting assertion that tion, it was recognition of the finite nature of the indigenous genotypes may not be as capable as indigenous timber resource that led to the develop- exotic species of exploiting the productive resources ment of theexoticpineplantationsthat now supply of a site. In support of this argument, he cites the much of the country's timber requirements. Even rapid growth and high productivity of two North today, serious attempts to manage indigenous spe- American pine species (Pinus contorta and P. radiata) cies for timber production are hampered by the inNewZealand. Althoughitiscertainlytruethatin perception that rotation periods are long, when the short-term these exotic conifer species outper- compared with those of the exotic conifers, and the form their indigenous counterparts, the long-term emphasis on short-term financial returns. sustainability of these fast-growing, short-rotation Implicit in our concepts of managing forest eco- forests is still open to question. Few are past their systems for other than purely conservation pur- second twenty-five- toforty-yearrotation,and,par- poses is the recognition that this involves the re- ticularly on nutrient-poor sites, fertilizer inputsare moval or harvesting of at least some parts of the necessarytomaintaingrowthrates.Bycontrast,the ecosystem. Using the author's pragmatic definition indigenousspecies they have replaced,by adopting that sustainability requires that there be no net loss a somewhat more conservative growth strategy, of productivity or genetic potential, discussions on have weathered the vagaries of millenia of ecosys- sustainable management center on the ability to tem change. replace that which has been removed from the The treatment of herbivores and pathogens ecosystem. For much of the biota, sustainable har- seeks to balance their negatively perceived im- vest is a viable option and may even improve pro- pacts on forest productivity with their positive ductivity. Similarly, loss of nutrients through har- contribution to nutrient cycling and productivity vesting may be amenable to human manipulation. within forest ecosystems. In addition we also 386 Sustainability of Managed Temperate Forest Ecosystems need to distinguish between those herbivores When it comes to the measurement of and pathogens that have evolved with the par- sustainability, the author quite rightly points out ticular forest ecosystem and are part of the natu- thatnowhereisthisbeingadequatelyaccomplished, ral cycle and those that have been deliberately or norare theresuitablemodelsorprototypesforsuch inadvertentlyintroduced,frequentlywithout their a program. However, most of the elements speci- natural regulatory agents. While deer and elk fied as necessary for assessing long-term species have evolved as part of many forest eco- sustainabilityarealreadyavailableforatleastsome systems in the northern hemisphere, forests in temperate forest ecosystems, albeit not in a coordi- New Zealand evolved both in the absence of nated fashion. Although we may not, at least in the mammalian herbivores and their predators. To- short to medium term, be able to measure ad- gether with the Australian brushtail possum equatelyallaspectsofecosystemproductivity,key- (Trichosurus vulpecula), which was introduced to stone elements of productive capacity do exist and can be used to assess changes in ecosystem produc- establish a fur trade, introduced ungulate species tanity. usimlry althoughswecannote pe tod- have dramatically altered, and in some cases tvy. Siial,atog.e anthp oma havsedthe d tcallyaltered, andle it e ysoeme c sure the full spectrum of genetic diversity, we can caused the collapse of, whole forest ecosystems in identify keystone speciesas indicatorsof thecontin- parts of New Zealand (King 1990). Once forest ued presence of the guilds of organisms they de- understory vegetation has been depleted, even pend on for their survival. Clearly a range of physi- low numbers of these introduced browsers are cal and biological variables needs tobemeasured in sufficient to maintain the depleted state (Allen, order to assess the sustainability of temperate forest Payton, and Knowlton 1984). Similarly, fungal management practices.Thechallenge is to get wide- pathogens such as Dutch Elm disease and defoli- spread agreement on the variables that are impor- ating insects like the Asian gypsy moth threaten tant for determining changes in productivity and the sustainability of genetic potential in temper- genetic diversity and that are amenable to routine ate forests of which they are not naturally a part. measurement. Any such program should seek to Temperate forest management systems that fo- build onexistingmeasurementprograms,establish cus on wood production have moved toward even- a minimum set of variables to be measured, and aged monocultures of the dominant tree species, make full use of computer-based technologies to with a clear-cut harvesting practice. New Zealand collate and analyze the data gathered. pineplantationsareaprimeexampleofthisagricul- Professor Franklin has put together a compre- tural approach to forestry, with their single-canopy hensive and well-balanced account of ecosystem tree species, uniform single-layer canopy, and typi- processes and management practices as they relate callyclear-cutharvestatintervalsasshortastwenty- to the question of sustainability in temperate for- five years. Yet even these forests retain a consider- ests. The challenge of measuring sustainability and able indigenous biodiversity, and for some ground managing our temperate forests in a sustainable orchid species they are nlow the major strongholds manner remains. Like many other fields of humnan (Johns and Molloy 1983). Even the kiwi, New endeavor, success will hinge as much on the com- Zealand's flightless rati te, can occasionally be found mi tment of people and the coordinationof resources within pine plantation areas. as on the technologies required to assess The author ad vancesa range of potential benefits susta inability. for moving away from the simplification of man- aged forest structures, all of which require forest managers to see the object of management in much References broader terms than the efficient production of wood and the maximization of financial returns. Forest- Allen, R. B., I. J. Payton, and J. E. Knowlton. 1984. ers, however, are not the only group who have "Effects of Ungulates on Structure and Species championed the cause of single-purpose manage- Composition in the Urewera Forests as Shown ment of temperate forests. Of recent years, the con- by Exclosures." New1 Zealand Journal of Ecology 7, servation lobby has been equally forthright in its . 119-30. advocacy of locking up natural forests against most forms of harvesting. Changing these perceptions Johns, J., and B. Molloy. 1983. Native Orchids of New and gaining community sanction for multiple-use Zealand. Wellington,NewZealand: A. H.&A.W. management will be critical to attempts to manage Reed, Ltd. temperate forests in an ecologically sustainable King, C. M. 1990. The Handbook of New Zealand manner that recognizes the need to retain both Mammals. Auckland,NewZealand:OxfordUni- productive capacity and genetic potential. versity Press. 387 I 49r Sustainable Management of Temperate Wildlife: A Conceptual Model Richard L. Knight My thinking on this subject has benefitedfrom discussions with D. R. Anderson, I. A. Bailey, T. L. George, H. A. L. Knight, and S. A. Temple. "A thing is right when it tends to preserve the integrity, stabil.ty, and beauty of the biotic community. It is not when it tends otherwise'. Aldo Leopold, Sand County Almanac (1949) Aldo Leopold, who is credited with developing In addition to the taxonomic variation, there is the discipline of wildlife management, and who great variation in spatial and temporal scales. The originated the concept of the land ethic, most spatial dimensions shape the objectives of man- certainly had sustainability in mind when he agement. Certain publics of Colorado want to see penned these words. Landscapes that would al- wolves and grizzly bears reintroduced into Colo- low people and wildlife to coexist in harmony rado mountains; others say that these areas are were one of Professor Leganism" (Meine 1988). A not large enough to support viable populations. half-century after he wrote these words, land When we talk about sustainable wildlife manage- managers are revisiting this arena and discover- ment, are we talking of time scales from 500 to ing the difficulties associated with sustainable 1,000 years or evolutionary time? The time scale wildlife management. will have profound impacts on the management When defining sustainability, we need to iden- approaches implemented. If we are not interested tify what we are going to manage, for how long in evolutionary time, then we do not have to we are going to manage it, and for whom are we worryaboutallelicvariation,adecision thatcould going to manage it. When discussing sustainable radically alter the approach followed. wildlife, some people think abou t sustaining eco- When viewing temperate wildlife in terms of systems, while others think about sustaining managing for sustainability, there are three cat- populations of specific species. Historically, we egories into which most species will fall (see could manage for maximum sustained yield of a table 24-1). First, sometemperatewildlifespecies particular species, almost always a game species are specifically managed for sport or commercial (Leopold 1933). Today, in certain very special- harvest (big game, fur bearers). Indeed, it was ized systems such as agriculture, this can still be interest in these species that gave rise to the birth done. This option, however, is becoming less of wildlife management (Leopold 1933), and wild- viable when managing natural or semi-natural lifebiologistshavehadgreatsuccessesinmanag- landscapes. Far too many publics are now inter- ing them for sustained-yield harvests. These spe- ested in far too many diverse species or groups of cies are protected under a diverse array of state species. Since today there is seldom unanimous and federal statutes and treaties. agreement on what we are managing for, perhaps Second, some species are experiencing popu- all we can aspire to is the vague goal of managing lation declines. By their very designation, these for biological diversity. species are presently unsustainable. Concern for Defining and Measuring Sustainability: The Biogeophysical Foundations Table 24-1: A Conceptual Model Illustrating the Categories of Wildlife, the Various Approaches in Managing These Groups, and the Components of the Management Approach Element Single species Multiple species Categories Game, endangered, pest Mammals, birds, reptiles, amphibians, insects, plants, and so forth Approach Manipulate populations Ecosystem processes, landscape mosaic Components Birth, death, dispersal Fire, soil, tree gap, disease (ecosystem processes); area, isolation, juxtaposition (landscape mosaic) themhasbeenresponsiblefordevelopmentof the clining wildlife species is to recover the Endangered Species Act and its goal of recover- population's viability and thenceforth to sustain ing species from impending extinction (see Rohif it indefinitely. Hence, when viewing species 1991). Wildlife biologists have only recently fo- threatened withextinction, wearenotinterested, cused attention on species facing extinction, and by and large, in recovering their populations so the list of species with declining populations is they can subsequently be harvested. The goal of growing far faster than any organized effort to managing wildlife communities is to "maintain deal with it. For example, the United States gov- orpreserve thestatusof thesespeciesasmembers ernment has placed more than 650 species on the of the wildlife community of an area. Status is list of endangered or threatened species, but an- usually evaluated by the relative abundance of other 600 just as vulnerable have yet to be in- thespecieswithinthecommunity"(Templel986). cluded (Gibbons 1992). The third category includes species that are Harvested wildlife neither game species nor endangered species. Concern for these species has generated interest Wildif anag e haduitsons The in the maintenance of biological diversity and its overexploitation of wildlife populations. The de- resulting metadiscipline of conservation biology clines of many temperate bird and mammal spe- (Brussard 1985; Soule and Wilcox 1980; Temple cies as a result of indiscriminate slaughter neces- and others 1988). Other than the U.S. Department sitated the creation of state and federal agencies of Agriculture'rsNationalForest ManagementmAct with specific mandates for recovering overex- whichleure'sN ational forest s And, ploited populations of game species and produc- gasad foquires viablegpop ationsl fofrnati vr ing sustainable harvests for sportsmen. The suc- grasslands for viable populations of native verte- cse fteeaece r mn h ra brates, the vast majority of wildlife species pres- cesses of these agencies are among the great ently do not have adequate legislation to ensure conservation accomplishments of the twentieth their sustainability. It has been proposed that an century and were based on the theory of sus- EaeredtEcosy. Acthis enecessar to en tained-yield harvesting (U.S. Department of Endangered Ecosystem Act IS necessary to en- thIneir18) sure the integrity and sustainability of these di- the Interior 1987). , ~~~~~~~~~The logistic growth curve is at the heart of verse species (Noss and Harris 1986). managing a wildlife population for sustained- yield harvesting. Once a population has reached its carrying capacity, it is not necessarily produc- Managing for sustainable wildlife tive because birth is balanced by mortality. The population can produce a sustainable harvest if A classification system that places wildlife into some of the individuals are removed. If the yield one of two single-species categories (game or is increased until it equals the number being endangered) and all other species into wildlife recruited to the population by birth, the popula- communities will clearly necessitate different tion is being harvested at its maximum sustain- goals for sustainable populations and communi- able level, and it will stabilize at about half its ties. The goal of managing harvested wildlife is to density at carrying capacity. producea sustainable surplus that canbe cropped Although exploitation certainly reduces abun- indefinitely, whereas the goal of managing de- dance, below certain levels of exploitation, popu- 390 Sustainable Management of Temperate Wildlife: A Conceptual Model lations can be sustained indefinitely. This in- focusing on birth and death rates, populations of volves the concept of density dependence. By declining species that are spatially isolated may reducing an animal's population, the impacts of require measuring these demographic param- predation, competition, and disease are also re- eters as well as other variables (dispersal rates duced. This results in higher survival rates and and size, isolation, and shape of habitat frag- productivity of the exploited population, causing ments containing the population). increased growth rates and producing more Whereas wildlife managers are relatively con- harvestable individuals. A number of population fident in their abilities to maintain sustainable variables are necessary to monitor and predict populations of many game species, they are just sustainability forharvested species. This includes now learning the correct approaches and require- knowing birth rates and survival rates, each un- ments to restore declining animal populations to der various environmental conditions, and some long-term viability. The approach used is to ma- estimate of population size. nipulate the species so that the population grows Thequestionofwhetherhumanexploitationis beyond the point of inflection on the logistic compensatory (for example, hunting replaces growth curve and reaches carrying capacity. For other kinds of mortality) is of fundamental im- species threatened with extinction, the goal is to portance in managing garne species. That hunt- have the population close to carrying capacity so ing is a compensatory form of mortality, rather astominimizethespeciesexperiencingtheharm- than additive, is the unspoken assumption un- ful effects of low population size (loss of het- derlying the existence of game seasons and bag erozygosity, skewed sex ratios). This approach is limits on North American wildlife. Wildlife especiallyrelevanttoendangeredspeciesbecause, managers assume there is a harvestable surplus unlike harvested species, species experiencing of animals that are going to die from some cause declines in population are often naturally rare (such as predation or starvation); hunting is an and occur in low densities. acceptable alternative to these other forms of A problem associated with managing for de- mortality. If hunting were additive to these natu- clining species is how much change in a popula- ral forms of mortality, then it would serve as a tion must occur before one becomes alarmed and population depressant. seeks a specific cause. Given the fact that popula- Exploitation can be compensated for only up tion parameters normally vary from year to year, to a point. At some level of harvest, hunting whatsortofmagnitudeofchangeshouldoccurto mortality adds to the total natural causes of mortal- show alarm? The duration of a trend in a popula- ity of a population. For populations that produce tion parameter might be more important than the offspring that would be added to the subsequent magnitude per se of the change. The geographic breeding stock in the absence of exploitation, extent of a change in population parameters may mortality from hunting is additive and will result determine its significance. A deviation in one in a population reduction. local population may not be cause for concern, whereas a change that occurs over a largeportion of Endangered wildlife the species' distribution might be cause for action. When managing for endangered species, it is The perspective most relevant to today's concern essential to identify both the proximate and ulti- for managing single-species populations focuses mate causes responsible for declines in popula- on nonharvested species that are experiencing tion. The proximate factor-or factors-is what declinesinpopulation.Althoughthereareanum- causes the decline and usually relates to the ber of sweeping explanations for population de- species' demographics; the species is either expe- clines, important causes include chemical con- riencing a decrease in natality or an increase in taminants, overexploitation, and loss and frag- mortality or both. The ultimate factor is the rea- mentation of habitat (Soule 1991). Habitat frag- son the species is declining and is almost always mentation disrupts dispersal between spatially a severe change in environment (habitat frag- isolated populations, which can result in declines mentation, environmental toxin). in heterozygosity and allelic diversity, decreased Obviously, both proximate and ultimate fac- viability of populations, and alterations in com- tors need to be addressed when attempting to munity assemblages with resultant disruptions recover a species' population; however, the first in important interspecific interactions. Whereas priority is to deal with the proximate factor and populations of harvested species may require then to address the ultimate factor. If curtailing 391 Defining and Measuring Sustainability: The Biogeophysical Foundations the proximate factor can prevent a species from r-selected). In addition, they often have relatively going extinct in the wild, restoring the species broadecologicalrequirementsandhaveincreased back to the wild once the ultimate factor has been fitness near ecological edges. In contrast, species corrected may not be necessary. experiencing declines in population are often in- There are six general categories by which ei- capable of responding to rapid changes in their ther birth or death rates of an endangered species environment and show a lower reproductive ca- can be manipulated to cause an increase in popu- pacity. Instead, they defer breeding to an older lation. Three of these categories focus on increas- age, have small litter or clutch sizes, and require ing natality, while the remaining three concen- a high degree of parental care. These species are trate on decreasing mortality (see table 24-2). often long-lived, leading to selection for large body size and repeated reproduction during the Dichotomy between harvested wildlife lifetime of the individual. In addition, these spe- and endangered wildlife cies maybe habitat specialists, having a relatively narrow range of ecological conditions where they A dichotomy between harvested wildlife and do best, or show an affinity for a type of habitat endangered wildlife relates to their evolutionary that is being destroyed or fragmented. strategies. The reactions of a species to exploita- These distinctions, of course, are relative; a tion cannot be forecast without knowing the cat- species may be r-selected relative to another but egory of animal concerned. Some specieslive ina K-selected relative to a third. For example, a relatively stable environment and have not javelina is r-selected relative to a moose but is K- evolved a mechanism for responding rapidly to selected relative to a mouse. This is an important changes, while others live in relatively fluctuat- point, because r-selected species are more likely ing habitats and are better able to adjust to rapid cted because the have changes. In the one case, there has been selection to wtsadeavy cop ec t h toward population stability, and in the other, greaterrecuperative powers. toward rapid reproduction, that is, the concept of able candidates for harvesting, yet the most ex- r- and K-selection (MacArthur and Wilson 1967: treme of all, the elephant, has often been recom- Pianka 1970). mended as such by conservationists anxious to Game species, by and large, have great repro- show that the elephant could "pay its way." Al- ductive potential in that they tend to breed at an ternatively,r-selected speciesmayalsodeclineas early age, have large clutch or litter sizes, and a result of overharvesting. Waterfowl have per- show reduced amounts of parental care (that is, haps the richest history of sport hunting in North Table 24-2. Six Approaches to Reversing Population Declines of Species Threatened with Extinction Approach Tactic Example Reduce mortality by stopping Enact legislation Endangered Species Act overexploitation Reduce mortality by controlling Control exotic species Feral pigs on Isla Santiago, Galapagos, predators, competitors, diseases, Ecuador (Goblentz, and Baber 1987) or parasites Reduce mortality by supplementing Provide supplemental food White-tailed sea eagles in Sweden limiting factors (Helander 1978) Increase recruitment by increasing Double-clutching in birds Peregrine falcon in eastern United number of offspring per breeding States (Barclay and Cade 1983) Increase recruitment by increasing Provide additional nesting Puerto Rican parrots (Snyder 1978) number of breeding attempts sites Captive breeding and Captive breeding and Guam rail (Haig, Ballou, and reintroduction hacking in birds Derrickson 1990) 392 Sustainable Management of Temperate Wildlife: A Conceptual Model America yet are experiencing widespread de- diseases caused by insects, and fire. This ap- clines. Most hunted duck species show all the proachassumesthatifecosystemsfunctionprop- characteristics of r-selected species in that they erly, then the naturally occurring biological di- are short-lived, have large clutch sizes, breed versity is intact. A problem with this approach is early, and require a minimal amount of parental that ecosystem function can become more impor- care. Loss and fragmentation of habitats are tant than species composition. Accordingly, the viewed as the greatest threats facing waterfowl: actual identity of a species is not as important as between the 1780s and the 1980s,a 53 percent loss its functional role in the community (that is, as in the original amount of wetlands has occurred decomposer or consumer). This approach deval- within the United States (Dahl 1990). ues the identity of species. An exotic species, if it performs the same function as a native species Wildlife commnun ities (that is, as pollinator) is no worse than the equiva- lent native species (see Lugo 1992). Wildlife communities, although they comprise A third approach in managing for wildlife the vast majority of temperate wildlife, have re- communities is to manipulate habitat and land- ceived theleast attention from wildlife managers. scapes in such a way as to influence groups of This may be duet to the absence of any economic species collectively in the desired direction. For importance attributed to these species, although example, in order to maintain songbird commu- nature viewers would no doubt dispute thatistate- nities of mature forests, land managers would ment. It may also be due to the seemingly impos- want to minimize landscape and habitat charac- sible task of simultaneously managing so many teristics associated with increasing edge and iso- species. Whereas wildlife biologists are comfort- lation, and decreasing size, of the stand. able with the idea of managing single species, Landscapeandstand characteristicsthatinflu- they were, until recently, largely unfamiliar with en dlife stand incterstat ize, approaches to managing assemblages of species stand shape, connectivity of stands, amount and (Noss and Harris 1986). Indeed, the attempt of distributionofvegetativesuccessionalstages,and land managers to identify indicator species, key- distributionoand densityaofvarioushabitatpatch stone species, umbrella species, or guild-indica- These landscapes have to be viewed at different tor species demonstrates their reluctance to leave spatial scales, which necessitates management the single-species approach behind. with two basic components (Saunders, Hobbs, When managing for wildlife communities, at and Margulesl99). First,thepatchesthemselves least three approaches may be employed: (1) the have to be managed so as to maintain or simulate species approach, (2) the ecosystem approach, internal dynamics of natural systems. These in- and (3) the landscape approach. The species ap- ternaldynamicsincludesuchdiversenaturaldis- proach, as its name implies, focuses on one spe- turbances as tree gap formation, fire, epidemics cies that may be affecting a collection of species. of insects, flow of energy, and cycling of nutri- This approach is similar to the single-species ents.Second, managementehas to focusyon factors approach in that it concentrates on demographic that are external to the patches but that influence variables, such as birth and death rates, and eco- internal patch dynamics. These include pollu- logical concepts, such as predation, competition, tion, urbanization, spread of exotics, and alter- diseases, and parasites. For example, this ap- ations of water regimes. For larger fragmnents, proach might attempt to increase productivity of emphasis should be placed on the internal dy- a community of songbirds by controlling cow- namics; for smaller fragments, emphasis should bird populations to reduce nest parasitism, A nmc;frsalrfamns nthsssol bird opultion to educ nes parsitim. A concentrate on the external influences. External problem with this approach is that it focuses on factr as d o rcan in cite), species and, unless it is firmly embedded in man- however, can be important regardless of the agement at the landscape level, exhausts the re- remnant size. sources of any land management organization. Since most impacts on habitat fragments origi- The second approach focuses on ecosystem nate from the surrounding landscape, there is processes and draws on ecosystem science for its clearly a need to depart from traditional notions inspiration. Land managers taking this approach of management and to look instead toward inte- are most concerned with ecological processes grated landscape management (Pickett, Parker, such as the formation of forest gaps, outbreaks of and Fiedler 1992; Saunders, Hobbs, and Margules 393 Defining and Measuring Sustainability: The Biogeophysical Foundations 1991). Traditional management of natural re- Dichotomy and conflict between single- sources stopped at the reserve boundary; fluxes species and wildlife community management of water, pollutants, and organisms do not. Plac- ing the conservation reserves firmly within the An important distinction between managing a context of the surrounding landscape and at- single species and managing a wildlife commu- tempting to develop complementary manage- nity is that in the first case, the manager focuses ment strategies seem to be the only way to ensure on population variables such as birth, death, and thelong-termviabilityoftemperatewildlifecom- dispersal, while in the second, the manager ad- munities. Problems associated with a landscape dresses ecosystem processes and landscape fea- approach include difficulty of acquiring enough tures in such a way that the wildlife community of protected areas to be able to manage effectively at interest shows collective population responses in an adequate spatial and temporal scale. In addi- the desired direction (such as increased birth rates). tion, parks and reserves are traditionally set aside A dilemma of considerable import when man- for reasons other than their landscape configura- aging wildlife one species at a time is that the tions relative to biological diversity. Reserves are single-speciesapproach inevitablyresultsinman- usually protected for their recreational or aes- agement that contradicts other management. Put thetic purposes, not because of their productivity as a question, if sustainable management of one or unique vegetation. A sequence of steps that species or resource alters the population dynam- could be modified by land managers faced with ics of another species, are we actually sustainable the dilemma of managing landscapes for biologi- managers? Wildlife management tends to sim- cal diversity is presented in box 24-1. This strat- plify wildlife communities by managing for a egy incorporates aspects of the ecosystem and handful of harvested species. When managing landscape approaches. for sustained yield of ungulates, what happens if the process entails altering populations of preda- tors, competitors, diseases, or parasites or entails changing landscape patterns, which alters edge/ Box 24-1: Steps for Managing habitat interior ratios or connectivity? Following Temperate Wildlife Communities this, if the management practices are successful, what happens after the population increases and 1. S .et clear obectives (maximizing diversity we begin to see alterations in ecological commu- of native wildlifp, minimizing the number of nities? For example, browsing by white-tailed species that fall below minimum populations). deer can profoundly affect the abundance and Step2. Associate wildlifecommunities with spe- population structure of several woody and her- cific habitat configurations, using environmen- baceous plant species. Because deer wander talsuitability(vegetativestructure,geomorphol- widely, the effects of high densities of deer can ogy, primary productivity, climate) as a surro- greatlymodify the vegetationof anarea(Alverson, gate for animal demography. Waller, and Solheim 1988). By reducing preda- Step 3. Determine the minimum subsets of habi- tors of white-tailed deer, we may be altering the tat fragments required to represent the diversity naturally occurring vegetation of regions (Ander- of a given area; know the distribution of species son and Loucks 1979). and tvpes of ecosystems. This has, to a degree, occurred already. Wild- Step 4. Assess the potential sensitivity of groups life managers have a long historyof manipulating of species to changes in landscape; this may be habitats to increase edge, because this generally based on the life history strategies of the species results in higher species richness and greater or on the amount and distribution of suitable populations of certain game species (cottontail habitats.ppltoso etl gm pce ctotl habitats. rabbit ruffed grouse; Leopold 1933). This in- Step 5. Manage the system to meet the predeter- crease in amount of edge habitat, along with the mined objectives; management will have to be crease in habitat alongenatiothe ongoing, and a monitoring system should be pervasive increase in habitat fragmentation by devised to measure responses of habitats and other types of human activities (such as road species. building), has been partly responsible for the decline in some North American songbird popu- Saunders, Hobbs, and Margultes 1991. lations (Robbins, Dawson, and Dowell 1989; Terborg 1989). 394 Sustainable Management of Temperate Wildlife: A Conceptual Model Instead of accentuating thedifferencesbetween Models for endangered species, historically, single-speciesandmultiple-speciesmanagement, were very much like those for game species in however, what is needed is a synthetic approach that they were deterministic and yielded useful rather than the existing dualism. For example, the insights on the relative importance of topics such Hawaiian islands have an inordinate number of as birth and death rates (Grier 1988). In the ensu- endangered species. We need to focus on particu- ing years, alternative approaches have been de- lar aspects of the life history of each of these veloped. For example, the U.S. Fish and Wildlife species in order to recover them while at the same Service developed habitat suitability index mod- time takinganecosystemand landscapeapproach els, which assess the sensitivity of wildlife to to ensure the integrity of the life support systems. habitat perturbations (Schroeder 1986; see Van One way of describing an ecosystem is to mea- Horn and Wiens 1991 for a critical review). In sure its productivity; another is to count the num- addition, population viability analysis and ber of species it supports. metapopulation models are being used for spe- cies experiencing declines in population (Gilpin and Hanski 1991; Soule 1987). Population viabil- Use of models in managing ity analysis has been of immense importance for for sustainable wildlife managing small populations that are particularly sensitive to stochastic phenomena such as loss of Models are a collection of symbols that achieve a genetic diversity and catastrophic events (see purposeful representation of realityby providing Hedrick and Miller 1992). The metapopulation an abstraction that can be manipulated by chang- concept is equally app.icable to harvested species ing the symbols or their relationships to each and is the first attempt to integrate landscape other (Jeffers 1982). Management decisions in- ecology formally into the dynamics of a species' volve very complex living systems for which land population (Murphy and Noon 1992). managers frequently do not have the data neces- The final frontier for wildlife modelers will sarytomakedecisions.Thusmodelspermitprob- focus on wildlife communities. Although only a lem solving that would be too large and too few attempts have been made to model species expensive for experimental approaches. Although communities, the modelsavailable have provided models can reduce the uncertainty of the conse- extremely important insights into how landscape quences of management decisions, decisionmakers configurationscan influence wildlifeassemblages are all too eager to embrace poorly tested, poorly (Temple 1986; Temple and Cary 1988). analyzed models. Wildlife managers have used modeling for a variety of purposes, but its chief function has Datasetsusedformonitoringandpredicting been to examine how populations change in size. sustainability of temperate wildlife This reflects back to the life table approach in population estimation (Caughley 1977). By using There are a number of well-established monitor- natality, mortality, and population age structure, ing programs for terrestrial wildlife in North wildlifebiologistscanlookat thegrowthrateand America. These projects generally cover large determine how environmental variables may af- spatial scales and are organized by federal, state, fect population size. Given a populationestimate, or private organizations. They transcend the for example, managers can then determine what boundaries of protected areas and deal with both effects different rates of harvest may have on the single-species and multiple-species assemblages, population and what different environmental although with fewexceptions, they focusonbirds factors might influence it. and mammals.Theseprogramsrangefrommerely Single-species models can be convenientlybro- noting the presence orabsenceof a species within ken into those that focus on game species and ana to collengdtta t maypeued to those that do not. By and large, models for game anaeat oplltion data t aybuise to species are deterministic and have been used for detect population trends. Data acquisition is of- a wide variety of questions ranging from the ten designed to serve the specificobjectivesof the influence of habitat, balance of energy, and im- agency or organization involved. Accordingly, portance of demographic changes (Gaudette and many of the data bases are widely scattered, are Stauffer 1988; Hobbs 1989; McCullough and oth- often incompatible, and can be inaccessible to ers 1990; Medin and Anderson 1979). potential users. 395 Defining and Measuring Sustainability: The Biogeophysical Foundations In spite of limitations of some of the sampling sists of fifty three-minute stops 0.8 kilometer methods and spatial and temporal scale of the apart and is run one morning each year, at the data sets, the information available is consider- height of the breeding season, starting at a half ableand could beused to monitor continent-wide hour before sunrise (see Robbins, Bystrak, and trends for certain taxa. For instance, variations in Geissler 1986). bird populations, including information on re- * Christmas Bird Count (National Audubon Society), productive success and mortality, could bemoni- whose goal is to determine continent-wide tored by using trends and density of winter bird distribution and abundance patterns of most populationsfromChristmasbirdcountsandwin- birds wintering in North America. First con- ter bird population studies, monitoring nesting ducted in 1900 with twenty-seven people, the populationsand densities from thebreedingbird survey counted twenty-six localities, includ- surveysandbreedingbirdcensuses,determining ing two in Canada. In 1985-86, 38,346 people reproductive success from the Nest Record Card parficipated in 1,504 countsintheUnited States, Program, and determining mortality and dis- Canada,MiddleandSouthAmerica,Bermuda, persal from the Bird-Banding Laboratory. Al- and the West Indies. Each count covers a circle though these data sets are tremendously valu- 24 kilometers in diameter, and at least eight able, they have seldom been integrated into na- hours must be spent counting at each site. The fionwideappraisalsofbird status (but seeRobbins, survey takes place within a two-week period Bystrak, and Geissler 1986; Root 1988). Presently, around Christmas (see Root 1988). these data sets cannot be used to determine the North American Nest Record Program, whose cause of changes in population.*NotAmrcnesRcrdPga,whe Some existing monitoinngpprog i goal is to record data on avian breeding biol- th mefollowing: monitoring programs include ogy such as the nesting season, clutch size, the following: incubation period, nesting period, and nesting * The State Natural Heritage Program (Nature success; 300,000 nest records have been com- Conservancy), whose goal is to provide com- pleted for 555 species, of which more than prehensive information on both species and 150,000 have been computerized. This program ecosystem diversity that can be used for ac- would allow the reproductive success of North quiring, designating, and managing protected American birds to be monitored on an annual areas. Natural Heritage Data Centers exist in basis, but this hasyet to be done forany species. every state of the United States as well as in * Breeding Bird Census and Winter Bird Population Latin America and elsewhere. The heritage Study(CornellLaboratoryofOrnithology),whose program inventories are continually updated goal is to estimate species density of nesting or through a system for gathering and ranking winteringbirds withinparticular typesof habi- information. They begin with broad informa- tat throughout the United States. Study plots tion searches and are often supplemented with of a fixed size and location are surveyed using detailedfieldsurveys.Thedataarestoredinboth the spot-mapping method (see Droege and manual and computer files (see Jenkins 1988). Sauer 1989; Robbins 1981). * Breeding Bird Survey (U.S. Fish and Wildlife * Bird Atlas Project, whose goal is to provide an Service and Canadian Wildlife Service), whose annual sample of changing abundance of bird goal is to estimate population trends of birds species, primarily todeterminethepresenceor that nest in North America north of Mexico absence of a species. The survey is usually and that migrate across international bound- based on blocksof land that cover 1 of latitude aries. The survey began in 1965 and provides and 'of longitude (see Laughlin, Kibgbe, and information, both locally by ecological or po- litical regions and on a continental scale, on Eagles 1982). short-term changes in populations that can be * Upland game bird call counts (U.S. Fish and correlated with specific weather incidents, re- Wildlife Service), whose goal is to discern coveryperiodsfollowingcatastrophicdeclines, population trends of upland game birds, normal year-to-year variations, long-term particularly mourning doves and woodcock. population trends, and invasions of exotics. A Some states also conduct call counts for bob- sampling scheme is based on degree blocks of whites, ring-necked pheasants, and other 1 latitude and I longitude. Each route con- quail species (see Eng 1986a). 396 Sustainable Management of Temperate Wildlife: A Conceptual Model * Waterfowl surveys (U.S. Fish and Wildlife Service, Measurements of sustainable wildlife commu- Canadian Wildlife Service, and state wildlife nities may fall within eithera speciesapproach or agencies), whose goal is to assess population an ecosystem/landscape approach. The use of trends in waterfowl. Annual trend data are particular species as surrogates to monitor the gathered in May and June along aerial east- health of wildlifecommunities found earlypopu- west transects distributed throughout breed- larity (Code of Federal Regulations 1985; ing areas from the prairies to the boreal forest Severinghaus 1981; U.S. Department of the Inte- and into the tundra. Segments of transects are rior 1980a, 1980b); however, it has recently re- covered intensiveiy by ground crews within ceived considerable criticism (Landres, Verner, twenty-four hours of the air counts, and a andThomasl988;Mannan,Morrison,andMeslow visibility index is applied to the aerial counts. 1984; Szaro 1986; remple and Weins 1989; Verner These surveys are conducted over about 80,500 1984). There are two inherent assumptions when kilometers (about 2 percent of the breeding using indicator species as representative of rela- habitat). Brood counts are conducted to assess tions between wildlife and habitat. First is the success of annual waterfowl breeding and are assumption that theindicator speciesisan appro- done from the air and from the ground. Fi- priateagentforalargersuiteof speciesof interest nally, an annual winter inventory isconducted and that a change in the indicator species' popu- to determine the size and distribution of the mid- lation is reflective of widespread changes in habi- winter waterfowl population (see Eng 1986b). tat and environment. Second is the supposition * Ungulates (state wildlife and conservation agencies), that a change in the population of the indicator whose goal is to assess population trends from species can be used to predict the environmental year to year in order to set hunting seasons and variables responsible for these changes (Van bag limits. Horne and Wiens 1991). These assumptions fail * National Wetlands Inventory Project (U.S. Fish on both conceptual and empirical grounds and Wildlife Service), whose goal is to collect (Landres, Verner, and Thomas 1988). information on the location, extent, status, and Alternatives do exist, however, for taking a trends of U.S. wetlands. Maps are available for species approach to monitor viable wildlife com- 51 percent of the lower forty-eight states, 14 munities. For instance, a useful index might be percentofAlaska, andallofHawaii.Anational ratios that reflect the number of species going assessment of the status and recent trends in extinct over the number of species extant, the U.S. wetlands has been completed. The results number of species being listed or delisted as are published in the Fish and Wildlife Service endangered over the number of species extant, or report, Wetlands of the United States: Current the total number of native species divided by the Status and Recent Trends (U.S. Departmnent of total number of species (both native and exotic) the Interior 1992). found in the area of concern (Anderson 1991). A landscape approach could include monitoring changes in patterns of land use (percentage of forest being converted to farmland), changes in Measures of sustainability patch characteristics and landscape mosaics (dis- tribution of patch sizes, connectivity, shape, and Indicatorsof sustainable wildlife populationsand so forth), or ratios reflecting theamount of an area communities are necessary to evaluate progress of interest that is in a natural or disturbed condi- toward meeting management goals. For single- tion. Ecosystem approaches could be measure- species populations, measurements that are nec- ments of human resources necessary to maintain essary include changes in population sizes or functioning ecosystems or incidences of ecosys- densities, birth and death rates, and dispersal tem processes such as formation of tree gaps, patterns. Specifically, for harvested species, man- fires, or outbreaks of disease. agers need to monitor changes in populations and harvest rates, as well as changes in age and sex ratios. For endangered species, indicators of The challenge sustainability include reaching the minimum vi- able population estimates or being taken off the How does a society complete the transition from list of endangered species. It is essential to monitor a wildlife management paradigm that focuses on population status after these events have occurred. single-species management to a paradigm in 397 Defining and Measuring Sustainability: The Biogeophysical Foundations which land stewards strive to preserve the integ- Barclay,J. H.,and T. J. Cade. 1983. "Restorationof rity, stability, and beauty of the biotic commu- the Peregrine Falcon in the Eastern United nity? This is clearly more than just a scientific States." Biological Conservation 1, pp. 3-40. dilemma, although natural resource managers Brussard, P. F. 1985. "The Current Status of Con- areexperimentingwith new approachesand tech- servation Biology." Bulletin of the Ecological nologies. This challenge embraces the diverse Society of America 66:1, pp. 9-11. and complex fabrics and colors of our society as a Caughley, G. 1977. Analysis of Vertebrate Popula- whole. Diverse publics find it easy to support tions. New York: John Wiley and Sons. efforts to save whooping cranes or grizzly bears, forhereonecaneasilyseetheresultsofconcerted Coblentz, B. E., and D. W. Baber. 1987. "Biology efforts. For example, so many individuals of spe- and Control of Feral Pigs on Isla Santiago, cies x were alive in 1980, and, with the concerted Galapagos, Ecuador." Journal of Applied Ecol- efforts of agencies and environmental groups, ogy 24, pp. 404-18. this number had doubled by 1990. Dahl, T. E. 1990. "Wetlands Losses in the United The publics are not nearly as facile, however, States 1780's to 1980's." U.S. Departmentof the with ecosystem and landscape approaches. Size, Interior, Fish and Wildlife Service, Washing- shape,and connectivityof habitat patchesor natu- ton, D.C. ral fire regimes or forest gap fon-nation are nei- Droege, S., and J. R. Sauer. 1989. "North Ameri- ther easily understood nor easily appreciated by can Breeding Bird Survey Annual Summary alargelyecologicallysemi-literatecitizenry.Other 1988." U.S. Fish and Wildlife Service Biological than the overwhelmingly important goal of in- Report 89:13, pp. 1-16. creasing citizen ecological literacy, I suggest that Eng, R. L. 1986a. "Upland Game Birds." In A. Y. scientists need to stress the values of ecosystem Cooperrider, R. J. Boyd, and H. R. Stuart, eds., functions to society as a whole. When discussing Inventory and Monitoring of Wildlife Habitat, pp. the wealth of benefits accrued to society as a 407-28. Denver, Col.: U.S. Department of the result of watershed management, flood control, Interior, Bureau of Land Management Service soil formation, decrease in soil erosion, theproper Center. balanceof atmosphericgases,or the full spectrum of biological diversity, the scientific community .1986b."Waterfowl."InA.Y.Cooperrider, needs to emphasize that these values are only R. J. Boyd, and H. R. Stuart, eds., Inventory and fully realized when ecosystems are healthy. Once Monitoringof WildlifeHabitat,pp.371-86.Denver, a society understands that ecosystem processes Colo.: U.S. Departmentof thelnterior,Bureau of are more basic to an ecosystem than elements Land Management Service Center. (read species) and that the processes create and Gaudette, M. T., and D. F. Stauffer. 1988. "Assess- ensure the maintenance of the elements, then ingHabitatofWhite-tailedDeerinSouthwest- perhaps we will truly begin to manage for sus- ern Virginia." Wildlife Society Bulletin 16, pp. tainable wildlife communities. 284-90. Gibbons, A. 1992. "Mission Impossible: Saving All Endangered Species." Science 256, pp. 1386. References Gilpin, M., and 1. Hanski, eds. 1991. "Metapo- pulation Dynamics: Empirical and Theoretical Alverson, W. S., D. M.Wailer, and S.lI.Solheim. Investigations." Biological Journalof the Linnean 1988. "Forests Too Deer: Edge Effects in Invety42 pp.1-336. Northern Wisconsin." Conservation Biology 2, Society 42, pp. 1-336. pp. 348-58. Grier, J. W. 1988. "Modeling Approaches to Bald Eagle Population Dynamics." Wildlife Society Anderson, J. E. 1991. "A Conceptual Framework Bulletin 8, pp. 316W22. for Evaluating and Quantifying Naturalness." Conservation Biology 5, pp. 347-52. Hansen, A. J., S. L. Garman, B. Marks, and D. L. Urban. 1993. "An Approach for Managing Anderson, R. C., and 0. L. Loucks. 1979. "White- VertebrateDiversityacrossMultiple-useL.and- tail Deer (Odocoileus virginianus) Influence on scapes.r"Ecological Applications 3, pp. 481-96. the Structure and Composition of Tsuga Hedrick P.W. andP.S.Miller.1992."Conserva- canadensis Forests." Journal of Applied Ecology tion Gec TniqueS and Funda ental 16 pp 855-61. tion Genetics:Techniquesand Fundamentals." 16E pp. 8557-61. Ecological Applications 2, pp. 30-46. 398 Sustainable Management of Temperate Wildlife: A Conceptual Model Helander, B 1978. "Feeding White-tailed Sea Noss, R. F., and L. D. Harris. 1986. "Nodes, Net- Eagles." In S. A. Temple, ed., Endangered Birds: works, and MLUMs: Preserving Diversity at All Management Techniques for Preserving Threat- Scales." Environmental Management 10, pp. 299- ened Species, pp. 149-59. Madison, Wisc.: Uni- 309. versity of Wisconsin Press. Pianka, E. R. 1970. "On r- and K-selection." Ameri- Hobbs, N. T. 1989. "Linking Energy Balance to can Naturalist 104, pp. 592-97. Survival in Mule Deer: Development and Test Pickett, S. T. A., V. T. Parker, and P. L. Fiedler. of a Simulation Model." Wildlife Monographs 1992. "TheNewParadigminEcology:Implica- 101, pp. 1-39. tions for Conservation Biology above the Spe- Jeffers,J.N.R. 1982.Modelling. London: Chapman cies Level." In P. L. Fiedler and S. K. Jain, eds., and Hall. Conservation Biology, pp. 66-88. New York: Jenkins,R. E.,Jr. 1988. "Information Management Chapman and Hall. for the Conservation of Biodiversity." In E. 0. Robbins, C. S. 1981. "Reappraisal of theWinter Bird Wilson, ed., Biodiversity, pp. 231-39. Washing- Population Study Technique." In C. J. Ralph and ton, D.C.: National Academy Press. J. M. Scott,eds., Studies inAvian Biology, vol.6,pp. Landres, P. B., J. Verner, and J. W. Thomas. 1988. 62-57. Cooper Ornithological Society. "Ecological Uses of Vertebrate Indicator Spe- Robbins, C. S., D. Bystrak, and P. H. Geissler. cies: A Critique." Conservation Biology 2, pp. 1986. Breeding Bird Survey: Its FirstFifteen Years, 316-28. 1965-1979. Research Publication 157. Wash- Laughlin, S. D., D. F. Kibgbe, and P. F. J. Eagles. ington, D.C.: U.S. Department of the Interior. 1982. "Atlasing the Distribution of the Breed- Robbins, C. S., D. K. Dawson, and B. A. Dowell. ing Birds of North America." American Birds 1989. "Habitat Area Requirementsof Breeding 35, pp. 6-19. Forest Birds of the Middle Atlantic States." Leopold, A. 1933. Game Management. New York: Wildlife Monograph 103. Charles Scribner's Sons. Rohl f, D. J. 1991. "Six Biological Reasons Why the . 1949. A Sand County Almanac. New Endangered Species Act Doesn't Work-and York: Oxford University Press. What to Do about It." Biological Conservation 5, Lugo, A. E. 1992. "More on Exotic Species." Con- pp. 273-82. servation Biology 6, p. 6. Root, T. 1988. Atlas of Wintering North American MacArthur, R. H., and E. O. Wilson. 1967. The Birds: An Analysis of Christmas Bird Count Data. Theory of Island Biogeography. Princeton, N.J.: Chicago, Ill.: University of Chicago Press. Princeton University Press. Saunders, D. A., R. J. Hobbs, and C. R. Margules. Mannan, R. W., M. L. Morrison, and E. C. Meslow. 1991. "Biological Consequences of Ecosystem 1984. "The Use of Guilds in Forest Bird Manage- Fragmentation: A Review." Conservation Biol- ment." Wildlife Society Bulletin 12, pp. 426-30. ogy 5, pp. 18-32. McCullough, D. R., D. S. Pine, D. L. Whitmore, Schroeder, R. 1986. "Habitat Suitability Index T. M Masfild,andR. . Dcke. 190. Models: Wildlife Species Richness in "Linked Sex Harvest Strategy for Big Game Shelterbelts." FWS/OBS-82/10.28. U.S. Fish "Linkd Se Harest tratey fo BigGame and Wildlife Service, Washington, D.C. Management with a Test Case on Black-tailed ' Deer." Wildlife Monographs 112, pp. 1-41. Severinghaus, W. D. 1981. "Guild Theory Devel- Medin, D. E., and A. E. Anderson. 1979. "Model- opment as a Mechanism for Assessing Envi- ing theDynarnicsof a Colorado Mule DeerPopu- ronmental Impact." Environmental Management lation." Wildlife Monographs 68, pp. 1-77. 5, pp. 187-90. Meine, C. 1988. Aldo Leopold: His Life and Work. Snyder, N. F. R. 1978. "Puerto Rican Parrots and Madison,Wisc.: University of Wisconsin Press. Nest-site Scarcity." In S. A. Temple, ed., Endan- Madson, Wise.: University of Wisconsin Press. gered Birds: Management Techniques for Preserv- Murphy, D. D., and B. R. Noon. 1992. "Integrat- ing Threatened Species, pp. 47-53. Madison, ing Scientific Methods with Habitat Conser- Wisc.: University of Wisconsin Press. vation Planning: Reserve Design for North- Soule, M. E., ed. 1987. Viable Populations for Con- ern Spotted Owls." Ecological Applications 2, servation. Cambridge, England: Cambridge pp. 3-17. University Press. 399 Defining and Measuring Sustainability: The Biogeophysical Foundations .1991. "Conservation: Tactics for a Con- Terborg, J. W. 1989. Where HaveAll the Birds Gone? stant Crisis." Science 253, pp. 744-50. Princeton, N.J.: Princeton University Press. Soule, M. E., and B. A. Wilcox, ed s. 1980. Conser- U.S. Department of the Interior. 1980a. "Habitat vation Biology: An Evolutionary-Ecological Per- Evaluation Procedures (HEP)." Ecological Ser- spective. Sunderland, Mass.: Sinauer Associ- vices Manual 102. U.S. Department of the Inte- ates. rior, Fish and Wildlife Service, Division of Szaro,R. C. 1986. "Guild Management: An Evalu- Ecological Services, Washington, D.C. ation of Avian Guilds as a Predictive Tool." . 1980b. "Standards for the Development Environmental Management 10, pp. 681-88. of Habitat Suitability Index Models." Ecological Temple, S. A. 1986. "Predicting Impacts of Habi- Services Manual 103. U.S. Department of the tat Fragmentation on Forest Birds: A Compari- Interior, Fish and Wildlife Service, Division of son of Two Models." In J. Verner, M. L. Ecological Services, Washington, D.C. Morrison, and C. J. Ralph, eds., Wildlife 2000: .1987. Restoring America's Wildlife: 1937- Modeling Habitat Relationships of Terrestrial Ver- 1987. Washington, D.C.: U.S. Fish and Wildlife tebrates, pp. 301-04. Madison, Wisc.: Univer- Service. sity of Wisconsin Press. . 1992. Wetlands of the United States: Cur- Temple, S. A., E. G. Bolen, M. E. Soule, P. F. rentStatusandRecentTrends.Washington,D.C.: Brussard, H. Salwasser, and J. G. Teer. 1988. U.S. Fish and Wildlife Service. "What's So New about Conservation Biol- U.S. Government. 1985. Code of Federal Regula- ogy?" Transactions of the North American Wild- tions. 36 CFR, chapter II 219, pp. 19-64. Wash- life and Natural Resources Conference 53, pp. ington, D.C.: Government Printing Office. 609-12. Van Home, B., and J. A. Wiens. 1991. "Forest Bird Temple, S. A., and J. R. Cary. 1988. "Modeling Habitat Suitability Models and the Develop- Dynamics of Habitat-interiorBird Populations ment of General Habitat Models." Fish Wild- in Fragmented Landscapes." Conservation Biol- life Research 8. U.S. Fish and Wildlife Service, ogy 2, pp. 34047. Washington, D.C. Temple,S. A.,andJ. A. Wiens. 1989. "Bird Popula- Verner, J. 1984. "The Guild Concept Applied to tions and Environmental Changes: Can Birds Be Management of Bird Populations." Environ- Bio-indicators?" American Birds 43, pp. 260-70. mental Management 8, pp. 1-14. 400 '/ ~--5v Sustainability of Wildlife and Natural Areas Kent H. Redford and John G. Robinson The term sustainability can be defined rather from temperate wildlife in two major ways. First, easily. Itisa characteristic of some entity thatlasts the tropicsarea long way from most international through time. The problem with the concept of decisionmaking centers and therefore seem easier sustainability, asused in conservation and devel- to deal with than temperate areas whose prob- opment arenas, concerns the identity of that en- lems stare decisionmakers in the face. It is easier tity. What do we want to sustain? When this to promulgate facile solutions in areas rarely vis- question is posed, there is a general agreement itedandpoorlyunderstood.Second,mrorepeople that itissomeaspectofthebiogeosphere.Thereis are living in areas termed natural in the tropics also agreement that the reason we want to sustain than in temperate areas; in all cases, they rely on this aspect is to support human activities. But wildlife,toagreaterorlesserextent.Thepresence thereafter, the common ground is difficult to find. of these people in tropical areas, including indig- Do we want to sustain the environment, natural enous, traditional, and colonist populations, cre- capital, or ecosystems? For which human groups atesa suite of problems, not generally regarded as do we want to sustain it: local communities, na- relevant to temperate natural areas. tional states? At what level will these human The second term to examnine is wildlife. In activities be supported: at present levels or do we temperate areas, particularly thosein Europe and aspire to improve the quality of human life? And North America, there is a single-species tradition for how long will we sustainably use that re- of wildlife rnanagement, which has largely de- source? Answers to these questions reflect differ- fined the approaches taken to wildlife and its use. ent understandings of what parts of the Thistraditionhasgivenhistoricallydefinedwild- biogeosphereareimportanttohumansurvival, life as those species of birds and mammals that different values that people place on preserv- hunters pursue for sport. This limited definition ing different parts of the biogeosphere, and hasbeguntobemodified,suchthatinmanyparts different expectations and aspirations for human of the world, wildlife is now being defined as all development. wild plant and animal species. Any discussion of sustainability must there- The third term to examine is natural areas, fore define its universe of interest very precisely. which refers to areas in which natural rather than In this chapter, we address the question of the anthropogenic processes dominate. It is a cat- sustainability of tropical wildlife and natural areas. egory that includes landscapes and implicitly, Before moving on to a discussion of their therefore, processes of both a biotic as well as an sustainability,weneedtoexaminethesetermscare- abiotic nature. The degree to which an area is fully and attempt to provide precise definitions. natural is the subject of much debate. Data from First, we must consider the fact that the wild- botany, archaeology, and anthropology collected life and natural areas under consideration are in many parts of the world are showing that located in the tropics. Tropical wildlife differs anthropogenic effects are ubiquitous in tropical Defining and Measuring Sustainability: The Biogeophysical Foundations areasandthatthevirginhabitatsosoughtafterby and certainly the one to which most frequent ecologists may not exist. Despite this fact, most reference is made. observers, on visiting an area, would agree on Biodiversity, then, refers to the varietyand vari- whether it is a natural area or not. This is to say ability among living organisms, the ecological that certain areas provide a gestalt of naturalness complexes in which they naturally occur, and the thatcorrespondsto thelackof obvious, large-scale, ways in which they interact with each other and human intervention. with the geosphere. Biological diversity can be measured at different levels (genes, species, higher taxonomic levels, communities and biotic pro- Biodiversity cesses, and ecosystems and ecosystem processes) and at different scales (temporal and spatial). Combining thebroader definition of wildlife-all Biological diversity, at its different levels, can be wild plant and animal species-with the general measured in number or relative frequency. definition of natural areas-landscapes includ- Genetic diversity refers to the variability within ing biotic and abiotic processes in which human a species, as measured by the variation in genes actions are relatively insignificant-results in a (chemical units of hereditary information that termthat,inthischapter,wedefineasbiodiversity. can be passed from one generation to another) In this context, the term serves asa useful replace- within a particular species, population, variety, ment for wildlife and natural areas. subspecies, or breed. All genetic diversity ulti- Biodiversity, one of the watchwords of the mately arises at the molecular level, based on the 1990s, is usually either not defined or defined in propertiesofnucleicacids.Thereisnosingleway vague and highly variable ways (Noss 1990). to measure genetic diversity, but it can be as- Despite the imprecise ways in which the term is sessed by DNA and protein polymorphism as used, policymakers have taken what was an aca- well as by detection of polymorphism in quanti- demic working concept (biodiversity) and put tativemorphologicaltraits(Bawaandothersl991). hundreds of millions of dollars into funded pro- Humans have affected the genetic diversity of grams to conserve it (Redford and Sanderson plants and animals directly through the process 1992). Largelyinresponse to this flushof funding, of domestication, as well as indirectly through many interestgroups have taken up the bannerof hunting, gathering, habitat alteration, and har- biodiversity conservation, enlarging the defini- vesting (see, for example, Ledig 1992; Ryman and tion to the point that thecurrent global biodiversity others 1981). Very few people have investigated strategy suggests that biodiversity be defined to the ubiquitous effects of human activity on this include human cultural diversity, manifested in component of biodiversity. diversity in languages, religious beliefs, land Species diversity refers to the variety of living management practices, art, and so forth (IUCN species on earth and is measured at the local, 1991). This definition allows Manhattan or Sao regional, or global scale. It can be measured in a Paulo to be considered on equal footing with the number of different ways, which differentially Great Barrier Reef of Australia when it comes to weight presence versus frequency of different levels of biodiversity and makes impossible any species at a given locality. The species is the unit coherent discussion of biodiversity conservation. that biologists most commonly use to categorize In this chapter, we propose restricting the the variation of life. It is also the unit best under- meaning of biodiversity, deliberately excluding stood by lay people. As a result of this, and the human cultural diversity as well as the diversity pioneering efforts of taxonomists, much of the represented in domesticated plants and animals. attention on biodiversity has been focused at the We have, in addition, divided the concept into species level. There is, however, considerable various components and incorporated processes, disagreement on how to define a species and how a frequently neglected component of biodiversity to measure the diversity of species. (Franklin 1988; Noss 1990). In doing this, we Diversityofhighertaxonomiclevels(genera,fami- provide a definition that closely approximates lies, and so forth) refers to the variety of organ- our charge, which is to examine wildlife and isms within a given region at a taxonomic level natural areas. We provide a formal definition, higher than the species level. It is clear that the based on the Office of Technology Assistance patterns of diversity manifested at the species definition (OTA 1987), generally regarded as level are by no means always the same as those being the most comprehensive of all definitions, demonstrated at the generic level and higher. 402 Sustainability of Wildlife and Natural Areas When the objective is to preserve the greatest (di Castri and Younes 1990; Ehrlich and Mooney genetic variation, species from different higher 1983; Walker 1989). taxa should be selected; that is, the community Biodiversity can be measured in a variety of that contains the most species may not contain the spatial and temporal scales. The time scale can greatest amount of unique genetic information vary from a few hours to decades or centuries. (Mares 1992; Platnick 1991). This component of Biodiversity occursat all spatial scales,from local biodiversity appears to be particularly important through regional to global, and the forces respon- when comparing marine with terrestrial systems sible for observed patterns of biodiversity may in that the patterns of genetic diversity within vary according to such scales (see Auerbach and taxonomic groups may be different in these two Shmida 1987). It is therefore important to specify systems (Thorne-Miller and Catena 1991). both the temporal and spatial scalebeingconsid- Communitiesandbiotic process diversityrefersto ered as well as the level at which biodiversity is the variety at the level of a group of organisms being discussed. belonging to a number of different species that In addition to specifying temporal and spatial co-occur in the same habitat or area and interact scaleswhendefiningbiodiversityatagivenlocal- through trophic and spatial relationships. In- ity, it is also important to specify whether or not cluded in this level are communities of organ- the component in question is being measured in isms, defined in a given time frame. This qualifi- terms of presence/absence or relative frequency. cation is importantbecause more and more ecolo- The difference between these two is crucial, par- gistsareappreciatingthedynamicnatureofcom- ticularly when assessing the effects of human munitycomposition(Hunter,Jacobson,andWebb actions. For example, at a species level, large 1988). Biotic processes include processes such as game mammals and birds at a given site may all pollination, predation, and mutualism. be present but in relative frequencies greatly af- Ecosysterns-leveldiversityrefersto thevarietyof fected by hunting (Redford 1992). It is a much communities of organisms and their physical more difficult task to conserve the different com- environment interacting as an ecological unit. ponentsofbiodiversityintheirrelativefrequency The ecosystem level can be divided into ecosys- than simply in terms of their presence or absence. tem types and ecosystem processes. Ecosystem types are bounded communities interacting with the abiotic environment, such as gallery forest. The sustainability of tropical biodiversity One of the critical differences between this level and the community level is the inclusion of eco- In our discussion of tropical diversity, it is clear system processes, such as fire and nutrient cy- thatbiodiversityhasdifferentcomponents.These cling. Ecosystem processes can be classified on are not vague, and they can be precisely defined. the basis of (1) functional attributes, that is, the Now let us consider the sustainability of this capacity of the ecosystem to capture, store, and diversity. Sustainability, as a concept, presup- transfer energy, nutrients, and water; or (2) struc- poses that this diversity will be used by people, tural attributes, relating to abundance and distri- but that the use-and the biodiversity-will not bution of species of various sizes and shapes, be lost in the process. These interdependent re- such as species as structural types; or (3) func- quirements are evident in all definitions that in- tional types, referring to the abundance and dis- clude sustainability, virtually all of which in- tributionof specieswithsuch functional attributes volve the concept of development. The World as the capacity to fix nitrogen and behave as Commission on Environment and Development predators, pollinators, and so forth (Anderson (otherwise known as the Brundtland Commis- and others 1991). sion) in 1987 defined sustainable development as Biodiversityconservationat the ecosystem level development that "seeks to meet the needs and seeks to preserve the basic trophic structure of an aspirations of the present without compromising ecosystem and the patterns of energy flow and the ability to meet those of the future" (WCED nutrient cycling resulting from that structure 1987, p. 8; our italics). In Caring for the Earth, (McNaugh ton 1989). Conservation ofbiodiversity sustainable development is defined as "improv- at this level is in large part conservation of prop- ing the quality of human life while living within erties and processes, not of species or assem- the carrying capacity of supporting ecosystems blages of species, because of the substitutability (IUCN 1991, p. 10, our emphasis). The concept of and redundancy of species within an ecosystem sustainable development is widely accepted as 403 Defining and Measuring Sustainability: The Biogeophysical Foundations bridging the need to conserve natural systems peccaries, and you must address the question and the need to allow human beings the use of of minimum ecologically operational popula- these systems. Many people would agree with tion sizes of the species. the statement that the important questions facing * However, if you are interested in sustainability the world community no longer have to do with at the level of ecosystem processes, then you the relationship between development and the probably do not care about white-lipped pec- environment but instead must now focus on how caries per se but would settle for sustaining sustainabledevelopmentcanbeachieved(Lelel991). other species that interact with the rest of the However, when biodiversity is considered as biological community and the abiotic environ- an equal partner in sustainable development ment in a manner similar to the white-lipped schemes, it is rarely possible to address adequately peccary. the dual requirements of use and conservation. * And, at the level of landscapes or natural areas, Most authors who discuss sustainable develop- questions of sustainability can be addressed ment schemes use only vague, undefined terms q y v to te nohuma, ornatual, using Munasinghe and McNeely's adaptation when referring to the nonhuman, or natural, of Perring's definition: "Sustainability from a component: biophysical perspective is linked to the idea * Natural capital, defined variously as "a stock that the dynamic processes of the natural that yields a flow of useful goods and services" environment can become unstable as a result (Daly 1991, p. 21) or "the soil and atmospheric of stresses imposed by human activity. structure, plant and animal biomass, etc. that, Sustainability in this scenario refers to the taken together, forms the basis of all ecosys- maintenanceofsystemstability,whichimplies tems" (Costanza 1991 p. 76) limiting the stress to sustainable levels on those * "Essential ecological processes and life sup- ecosystems that are central to the stability of port systems" (IUCN 1980) the global system" (chapter 2 of this volume). * The "Earth'svitalityand diversity" (IUCN 1991). Of course, at this level, even more than at the previous one, a concern about peccaries has These definitions obfuscate a fundamental con- little relevance. tradiction between use and conservation: that is, . any use tends to reduce the biodiversity of a Onyheii.osb ,spcfp cseyh Whenuseconsideringduc the biotype of development, the objectives of that devel- system (Robinson 1993). When considering the opment, and the levels of biodiversity targeted natural world at the broad level of the ecosys- for conservation will it be possible to assess a tem-as In these definitions-it appears to be pomassibe tobotheuse aefindtconserve. Whpens conpriori the costs of human action on tropical wild- possible to both use andIconserve. When consid- life and natural areas. That there will be costs is ering biodiversity, as defined here, it is obvious clear. Proponents of sustainable development that there are real tradeoffs to consider.cla.Ponetofstmbedvlpet assure us that there are cost-free solutions to Because use tends to decrease biodiversity, a consdertio of sutiabt mus spcf wha spring us from the trap of environmental degra- considegree io of losusi tainable.y Tist ispea f aitr dation as a price for human development. They dere of losi'cetbe Ti sarirr are wrong, and they are usually wrong because decision based on the specific components of rey wrong nantheoy areusl wo bc asre biodiversity that one wishes to conserve and use, hydfn aueol ntrso hti e thodiverndsi thet huane grous foconserv whih the, quired for human survival. As is clear from our iheneeds beinge utain theoualy fof hif toe definition of biodiversity, cost-free solutions are system virtually impossible. If one only conserves those which these people aspire, and the time over p whc sutiablt is deie. For. example, parts of the biosphere that are essenfial for human which sa ilife, many elements of the natural tropical land- * If you are interested in the sustainability of a scape will be lost. population of the white-lipped peccary (Tayassu Virtually every activity, be it sustainable agri- pecari), you should not sanction use that would culture, natural forest management, use of land imperil the demographic sustainability of that by indigenous people, or hunting, has been said population. to be related to the conservation of biodiversity. * If, however, you are interested in sustaining In short, virtually nothing is said not to conserve biological communities and biotic processes, biodiversity. Yet, it is clear to all intelligent ob- then you should not sanction use that would servers that at all levels, biodiversity conserva- threaten the ecological role of white-lipped tion has been dealt with in a monolithic fashion: 404 Sustainability of Wildlife and Natural Areas a given activity was said to either conserve Bawa, K., B. Schaal, 0. T. Solbrig, S. Stearns, A. biodiversity, or destroy it, with no intermediate Templeton, and G. Vida. 1991. "Biodiversity possibilities allowed. This in turn is due to a lack from the Gene to the Species." In 0. T. Solbrig, of precise definitions combined with a desire to ed., From Genes to Ecosystems: A Research Agenda proffer cost-free solutions. for Biodiversity, pp. 15-36. Cambridge, Mass.: The detailed definition of biodiversity given IUBS. above divides biodiversity into its various com- Costanza, R. 1991. "The Ecological Effects of ponents. This division allows an assessment of Sustainability: Investing in Natural Capital." the effects of different types of land use on the In R. Goodland, H. Daly, and S. El Serafy, eds., different components of biodiversity and there- "Environmentally Sustainable Economic De- fore a priori acknowledgment of what compo- velopment: Building on Brundtland," pp. 72- nents of biodiversitycanbeconserved under that 79. Environment Working Paper 46. World system. This in turn allows establishment of crite- Bank, Washington, D.C. ria to select the version of system of land use that Daly, H. 1991, "From Empty-world to Full-world most effectively conserves these biodiversity com- Economics: Recognizing an Historical Turn- ponents. It also allows an a priori acknowledg- meto htbodvrit opnet ilntb ing Point in Economic Development." In R. ment of what biodiversity components will not be Goodland, H. Daly, and S. El Serafy, eds., conserved in areas subjected to that type of land Goodrnd H. Dally and. El SErafyi Ded use. This recognition of the specific costs- Eviromentally Ssanable Eom De- biodiversitycomponents-ofagiventypeof habi- 2E nvn Buil ing oapdr pp. 18- tat alteration is vital, for it allows the establish- 26. Environment Working Paper 46. World ment of alternate areas where those biodiversity Bank, Washington, D.C. components can be maintained. di Castri, F., and T. Younes. 1990. "Ecosystem Usingbiodiversitycost-benefit calculus allows Function of Biological Diversity." Summary the creation of landscapes in which different sys- report of an IUBS/SCOPE workshop, June 29- tems of land use can be combined in such a 30,1989. Biology International, special issue 22. fashion as to maintain every component of Ehrlich, P. R., and H. A. Mooney. 1983. "Extinc- biodiversity under some type of land use, while tion, Substitution, and Ecosystem Services." maximizing sustainable human development at BioScience 33:4, pp. 248-54. the landscape level. Under such a scheme, rehabili- Franklin, J. F. 1988. "Structural and Functional tated land could be coherently integrated with Diversity in Temperate Forests." In E. 0. Wil- tree plantations, extractive reserves, and national son, ed., Biodiversity, pp. 166-75. Washington, parks, each system of land use conserving a dif- D.C.: National Academy Press. ferent component of biodiversity. It is vital to point out that national parks with minimal hu- Hutr 1988. rPaleoecology and the Coarse-filter man activity-the natural areas-are keystones Apoc 18 "Pae niog adteoarsiter in this scheme, for they are the oniy areas to con- CosrvatontBioog Biological Diversity." serve all components of biodiversity. Wildlife, in contrast, can be managed in a variety of land use IUCN (International Union for the Conservation settings, depending on the specific objectives. of Nature). 1980. World Conservation Strategy. With the United Nations Environment Pro- gram and the World Wildlife Fund. Gland, References Switzerland. .1991. Caringfor the Earth: A Strategyfor Andersen, R., E. Fuentes, M. Gadgil, T. Lovejoy, Sustainable Development. With the United Na- H. Mooney, D. Ojima, and B. Woodmansee. tions Environment Program and the World 1991. "Biodiversity from Communities to Eco- Wildlife Fund. Gland, Switzerland. systems." In 0. T. Solbrig, ed., From Genes to Ledig, F. T. 1992. "Human Impacts on Genetic Ecosystems: A Research Agenda for Biodiversity, Diversity in Forest Ecosystems." Oikos 63, pp. pp. 73-82. Cambridge, Mass.: IUBS. 87-108. Auerbach,M.,andA.Shmida.1987."SpatialScale Lele, S. M. 1991. "Sustainable Development: A and the Determinants of Plant Species Rich- Critical Review." World Development 19, pp. ness." TREE 2:8, pp. 238-42. 607-21. 405 Defining and Measuring Sustainability: The Biogeophysical Foundations Mares, M. A. 1992. "Neotropical Mammals and Robinson, J. G. 1993. "The Limits of Caring: Sus- the Myth of Amazonian Biodiversity." Science tainable Living and the Loss of Biodiversity." 255, pp. 976-79. Conservation Biology 7, pp. 22-28. McNaughton, S. J. 1989. "Ecosystems and Ryman, N., R. Baccus, C. Reuterwall, and M. H. Conservation in the Twenty-first Century." Smith. 1981. "Effective Population Size, Gen- In D. Western and M. C. Pearl, eds., Conserva- eration Interval, and Potential for Loss of Ge- tion for the Twenty-first Century, pp. 109-20. neticVariabilityin GameSpeciesunder Differ- New York: Oxford University Press. ent Hunting Regimes." Oikos 36, pp. 257-66. Noss, R. F. 1990. "Indicators for Monitoring Thorne-Miller, B., and J. Catena. 1991. The Living Biodiversity: A Hierarchical Approach." Con- Ocean: Understanding and Protecting Marine servation Biology 4:4, pp. 355-64. Biodiversity. Washington D.C.: Island Press. OTA (Office of Technology Assessment). 1987. Walker, G. 1989. "Diversity and Stability in "Technologies to Maintain Biological Diver- Ecosystem Conservation." In D. Western sity." U.S. Government Printing Office, Wash- and M. C. Pearl, eds., Conservation for the ington, D.C. Twenty-first Century, pp. 121-30. New York: Platnick, N. 1. 1991. "Patterns of Biodiversity: Oxford University Press. Tropical vsTemperate." Journal of Natural His- (WCED) World Commission on Environmentand tory 25, pp. 1083-88. Development. 1987. Our Common Future. New Redford, K. H. 1992. "The Empty Forest." York: Oxford University Press. BioScience 42, pp. 412-22. WRI, IUCN, and UNEP (World Resources Insti- Redford, K. H., and S. E. Sanderson. 1992. "The tute, International Union for the Conservation Brief Barren Marriage of Biodiversity and of Nature, and United Nations Environment Sustainability?" Bulletin of the Ecological Society Program). 1992. Global Biodiversity Strategy. of America 73:1, pp. 36-39. New York. 406 Tropical Water Resource Management: The Biophysical Basis Jeffrey Edward Richey, Eneas Salati, Reynaldo Luiz Victoria, and Luiz Antonio Martinelli We would like to acknowledge the support of the U.S. National Science Foundation and the NASA Earth Observing System. The tropics of the world-those regions between continental scales; these cycles are of fundamen- the tropics of Cancer and Capricorn-include tal importance not only to the maintenance of environments ranging from the wettest on Earth natural systems but to any human occupation. to the driest. They include the great tropical rain Then we have the problem of how to incorpo- forests of South and Central America, Africa, and rate such knowledge into programs for develop- southeast Asia and the Pacific rim. They include ment and resource management. Actually doing the deserts of Northern Africa and the cerrado of this requires a series of state-of-the-art advances Brazil. They include the most dense populations in science and in communication. In fact, our on Earth and the most sparse. understanding of the hydrological cycle at the Among the most publicized links between glo- relevant regional and continental scales is sur- bal change and the tropics are the emissions of prisingly rudimentary. The overall scientific and greenhouse gases, particularly the carbon diox- policy communities should recognize that there ide associated with burning vegetation, and the are new technologies and new ways of doing potential loss of biodiversity. At a more immedi- scientific business, each of which may be done by ate management level, the tropics pose a different communities that rarely communicate with each set of problems for water resources than do tem- other. There is the difficult problem of bridging perate zones. In temperate regions, management the gap between those who purportedly know is a relatively clear-cut, if politically difficult, how the system functions (the environmental problem of the supply, allocation, and quality of scientist), the maker or implementer (agency) of watercomplicatedbycurrentand futurevagaries policy, not to mention the end user. In the case of in climate. Throughout the tropics, problems of projects in the developing nations, the develop- water quality due to both natural disease vectors mentbanks may play a key role,yet their capabil- and pollution may be so severe as to imperil ity to assimilate information about the physical human health, and lack of flood control threatens basis of the problem may need to be enhanced. millions. Loss of forest on the high-relief topogra- Yet scientists often take the position that their job phy of Asia and Central America leads to massive is only to make information, usually just "their" erosion and loss of fertility. information, available, not to be active in its dis- These resource topics are symptomatic of the semination. broader issues of the natural hydrological and In this chapter, we focus on how the biophysi- biogeochemical cycles in the diverse tropical ba- cal basis for the sustainability of water resources sins. One of the most significant challenges for for a representative area of the tropics, the Ama- defining the basis for sustainability is to deter- zon basin of northern South America, might be mine how these cycles function on regional to defined. In terms of developing a biophysical Defining and Measuring Sustainability: The Biogeophysical Foundations model for tropical water resources, the basin is A definition of the biophysical basis * Representative. The Amazon consists of a vari- of sustainability of the Amazon ety of bioclimatic zones (rain forest, savannah, and cerrado) that could be considered repre- Objective 1: To develop a biophysical definition of sentative of other tropical zones for purposes sustainabilityforbothpracticalapplicationandusein of anextended model and represents a seriesof fonrulating a full definition of sustainable develop- hydrological and chemical regimes that are ment in all its ramifications. typical of the world's rivers. typical ofthe worlds rivers.Definition and structure of sustaqinability * Quantitativelysignificant. The Amazon isa large D s area, accounting for a significant portion of the The sheer physical size and logistics of problems humid tropics and providing 20 percent of the posed by tropical basins present challenges that world's river discharge to the oceans. The con- must be resolved in order to develop a reasonable densationalenergyreleasedbyconvectivepre- definition of a biophysical basis. At these scales, cipitation within the basin has been shown to river basins are natural integrators of surficial be of sufficient magnitude to affect global pat- processes. The water and dissolved and particu- terns of climate. late materials observed in the main channel of the * Qualitatively significant. The Amazon is Amazon and other large floodplain rivers are the primarily undisturbed. It thus provides a products of processes occurring in the drainage unique natural laboratory as to how large- basin across widely varying temporal and spatial scale systems function in a natural state. This is scales. An understandingof how these substances a very important attribute for theoretical as areroutedfromprecipitationthroughtheirdrain- well as practical purposes. age systems to the oceans would yield important information on the processes controllingregional- We will show that present knowledge, though scale hydrological and biogeochemical cycles. For sparse, reveals the key role of the forest in main- example, the carbon measured in the main chan- taining the dynamic equilibrium of the Amazo- nel isa mixtureof carbonoriginating fromsources nian ecosystem. In summary, the forest controls thousands of kilometers away in upland regions, the dynamics of the basin, the balance of energy, as well as carbon introduced continuously (spi- theyield of sediment, thebalanceof nutrients, the raled) from theadjacent floodplain. Organiic mat- diversity of species, the quality of surface water, terof both sources has been subjected to transport the quality of soil, and the stock of soil in the and reactive processes within the channel. Of biosphere. Potential consequences of deforesta- particular importance for biogeochemistry are tion include modifications in the basin's convec- the storage of water in various parts of the drain- tive rainfall regime and downstream changes in age system for periods of weeks and the transfer the river's flow and transport of nutrients and of this water between the various physiographic sediments. In the Amazon, as elsewhere, natural reservoirs. oscillations in the hydrological cycle and the pro- Given these realities, the critical question that cesses influencing those oscillations must be dis- concerns the biophysical sustainability of the tinguishedbefore possibleanthropogenic impacts Amazon is the extent to which the land use of a can be truly attributed. The lack of knowledge of particular region of the basin can be altered with- the basic functioning mechanisms of the out affecting the overall regime of rainfall and Amazonia ecosystem and of the most suitable runoff or of production and decomposition of methods for achieving a sustainable develop- organic matter. Further, the sustainability ques- mentof theregion is the main reason that many of tion can be defined on the basis of a setof interac- the agricultural and cattle ranching projects fail. tive questions dealing with hydrological and bio- Specific measurement, modeling, and remote- geochemical cycles integrated across landscapes sensing programs can be identified that would in the context of the phytogeography and physi- dramatically improve the base of knowledge and cal structure of the basin (see figure 26-1). management for the basin. The geological structure of the basin, includ- The synthesis represented in this chapter is ing soil types, topography, and drainage net- based primarily on Martinelli and others (forth- works establishes the overall physical matrix in coming), Victoria and others (1991), Salati (1987), which the biotic world can come to life. Within and Salati and others (1989, 1991). this overall framework, the relevant question for 408 Tropical Water Resource Management: The Biophysical Basis Figure 26-1: Biophysical Synthesis To what extent can the land use of a particular region be altered? Hydrologic cycle * Regional hydrologic modeling * Data acquisition: rain, discharge, energy * Site water and energy exchange with atmosphere Phytogeography Biogeochemistry * Vegetation assemblages Biological diversity * River chemical signals * Remotely sensed attributes . Relation of physical to species world * Production/oxidation * Succession modeling * Dynamics modeling Physical structure * Soil property distributions * Topographic and drainage networks the hydrology of the basin is how the climatic and existence of communitiesdepends on subtle varia- surface features of the basin determine the tem- tions in the moisture and temperature regime. An poral distribution of runoff and the spatial pat- important question is how the distributions of tern of moisture storage. vegetation communities (phytogeography) are It is convenient to think of basin-scale bio- influencedbyhydrologic,geologic,andultimately geochemical and watercycles as a combined prob- human factors. 1emin theregionalbalanceand subsequentdown- The hypothesis can be posed that each vegeta- stream routing of water. The balance of water at tion community has characteristic soil properties any site and time can be described by and predictable seasonal patterns of moisture, R = P - ET ± SM (26-1) net radiation, and evapotranspiration that can be P is precipitation, R is the effective runoff, established by (point) measurements at specific where T is evapot ation, and SM ischange sites and extended by inference to similar com- ET is actual evapotranspiration, and SM is change munities. The first issue is to identify characteris- in soil moisture and storage of groundwater. To tic assemblages of vegetation over large areas, provide overall constraints on fluxes of water and ultimately with regard to the physical differences energy using data that are realistic to obtain, the between areas. The second issue is to bring these plan starts at the regional scale, with river dis- communities to life; that is, to discover what their charge (runoff). Given the heterogeneous nature successional patterns are and then how they are of precipitation and collectors of precipitation, differentially fixingand transformingcarbonand river discharge is a robust integrator of the long- related elements. term hydrologic properties of a drainage basin. The key to these questions is first to character- Evapotranspiration can then be constrained an- ize assemblages of species and then to identify nually as the difference between precipitation thosepropertiesthatcanbeextrapolatedtolarger and runoff and examined in more spatial and regions. A key output would be the definition of temporal detail via regional applications by en- functional groups of vegetation, where a func- ergy calculations. Ifthedistributionsofprecipita- tional group is defined by attributes that are tion, runoff, and evapotranspiration can be calcu- unifying (similar structure, function, and implicit lated with sufficient precision, insight into the taxonomy) and that can be determined (prefer- regional distributions of soil moisture is possible. ablyby satellite). Then thesegroups canbecoupled Phytogeography is one of the main links among to landscape issues of short-term community suc- the physical structure of the basin, hydrology, cession: Why and how do existing functional and the distribution of plant communities. As groups grade from one type to another? How and shown below, the hydrologic cycle depends inti- at what rate does natural (short-term) succession mately on the vegetation. Conversely, the very operate dynamically (for example, filling gaps, 409 Defining and Measuring Sustainability: The Biogeophysical Foundations blowing down, dying)? The physical attributes of able. Hydrologic and ecological modelers have the functional groups include biomass (nutrients, traditionally focused most of theirattention at the C) and structure (canopyarchitecture, LAI), which very small to moderate spatial scales of square thus provides the bridge to the biogeochemistry. meters to several hectares. The integrated effect The next major problem is to couple the hydro- of small-scale cycling may or may not influence logic cycle and knowledge of the phytogeogra- cycling at larger temporal or spatial scales. The phy with the biogeochemistry of the Amazon question of how spatially to average the hydro- basin. For example, how does the structure and logic parameters of mesoscale areas when their biological diversity of the Amazon ecosystem component parameters are spatially variable and control thecycling of water, carbon, and nutrients poorly characterized at a smaller scale is one that undernaturalconditionsandunderdifferentcon- has drawn the attention of many investigators. ditions of land use? Heterogeneity of the environment with regard to The hard question of linkage is in how the mechanisms that produce runoff, measurement biogeochemical dynamics and community struc- and logistic realities, and differences in response ture might respond to changes in the state of the between smaller and larger catchments make it system that occur abruptly (direct clearing or difficult to extrapolate from one site to larger areas. physical intervention) or more subtly (such as Traditionally, large-scale hydrologic models changes in ambient environment: moisture, tem- have inferred the nature of hydrologic systems on perature, carbon dioxide) either as propagating the basis of input-output data; without consider- edge effects or regional effects. ing microscale physics in their derivation. Most With the above perspectives and information, of them are "black box" in nature and some are it should then be possible to address the problem highly conceptual, but all have lumped, physi- of how information on the biophysical perspec- cally meaningless parameters.These modelsneed tive of the Amazon ecosystem can be used to sufficiently long meteorologic and hydrologic address issues of biodiversity. records for their calibrations; they have not used such data as basin geomorphology, soils, and A toolbox vegetation; their parameters are not measurable in the field or by remote sensing and are difficult The degree of geographic variation of land sur- to interpret. Furthermore, lumped parameter face properties in this continental-scale drainage models do not provide space-time distributions basin and the variety of possible measurements of water within thebasin and thereforeare of little and scientific and resource interests require a use for the study of biogeochemical processes. toolbox of models, both heuristic and mathemati- However, by representing a large-scale basin as a cal, and of field measurements to force some homogeneousone-dimensional system with uni- discipline and cooperation on activities. A range form climate forcing, current physically based, of such modelsis necessary to handle processcsat one-dimensional models have also shown great different scales, and some attention needs to be uncertainties in their predictions. These uncer- placed on consistency between the physics and tainties are primarily caused by spatial variabili- chemistry represented at various scales. Within ties in thebasin's physical characteristics and, for the guidance provided by such models, field pro- a large-scale basin, a lack of spatial uniformity in grams must provide the pertinent data, if not the climate forcing, such as fractional coverage of needed check on reality. rainstorms. Underchangingclimateand land use conditions, the uncertainties in the model pre- MODELING dictions increase. Therefore, it is essential for Immediately, any such model must deal with the large-scale models to incorporate these spatial issue of scale; that is, we have to transfer our variabilities. understanding from very small scales (where it is Driven by the need to improve the representa- possible to do field research), to regional scales (a tion of the land surface processes, general climate river basin), and ultimately to the Amazon as a models have been developed recently that take whole and its interactions with the global atmo- into account morphological and physiological sphere. The first problem for a large drainage characteristics of the vegetation as well as physi- basin such as the Amazon is to establish the cal characteristics of the soil. These include the hierarchy of time and space at which the pro- Simple Biosphere Model (SiB; Sellers and others cesses that control the fluxes of interest are oper- 1986). Given the properties of vegetation and 410 Tropical Water Resource Management: The Biophysical Basis soils and soil moisture, plus wind speed, air tem- cally at a point using micrometerologic data are perature and humidity at a reference level above available but have not been useful or practical for the canopy, visible and near-infrared incoming application at the basin scale. Most models of solar radiation, and precipitation, these models rainfall and runoff treat evaporation as a function can be used to calculate the fluxes of vapor more of potential evaporation, which in turn is based accurately than models using simple formula- on a bulk transfer coefficient and a reduction tions of the balance of energy. For application to factor. These methods of calculating evaporation the Amazon, SiB-type models have to be cali- depend on knowledgeof soil moisture, whichcan brated with measurements of air temperature vary greatly over relatively small distances. A and humidity above the canopy, precipitation mesoscale representation of this process must, and downward fluxes of solar and near-infrared therefore, incorporate information about the spa- radiation, and wind speed for typical ecosystems tial variability-and mean value-of soil mois- (terra firme forests, igap6 forest, campina forest, ture. Regional potential evapotranspiration can floodplain vegetation, and vegetation of areas be calculated using the climatological data (inso- under use, such as grass). lation, humidity, temperature, wind speed) avail- able from the meteorological network using the HYDROLOGIC NETWORK Monteith (1973) method adapted to a tropical As the backbone of large-scale hydrologic analy- forest (Villa Nova, Salati, and Matsui 1976). Based ses, the Departamento Nacional de Aguas e on detailed eddy correlation methods, Energia Eletrica (DNAEE) maintains a gaging Shuttlesworth (1988) has modified the empirical network of 600 precipitation, river stage, and (bulk) formulation fitted for the micrometeoro- meteorological (solar radiance, surface tempera- logical data and shown that this model yields ture, humidity, wind) stations throughout the results within 15 to 20 percent of the actual fluxes basin. Daily records of at least a ten-year duration and describes the seasonal cycle. Given the vari- areavailable for moststations, and longerrecords ability in the base climatological data, it is not exist for specific stations. These data constitute always possible to calculate evapotranspiration the primary data set available for the water re- even on a monthly basis for all years. Where data sources of the Amazon. Specifically, precipita- are limited, it may be necessary to calculate the tion patterns can be determined by analyzing the monthly or annual means and to derive the distri- data from rain gage networks. Because of the bution of these averages from theoretical consid- well-known sparseness of precipitation data in erations. Remote sensing of the radiances corre- these networks, it is necessary to improve the sponding to these energy terms may be the best methods for obtaining such information from way to do this. Within the likely evolution of the satellitedata. Satellite observations can be related capability for detecting these terms over remote to the rain gage network by using proxy precipi- basins, however, the challenge is large. For ex- tation records derived from measurements of ample, at the present time it is not possible to outgoinglongwaveradiationobtainedfromGOES measuresoilmoistureunderdensetropicalforest or AVHRR. The proposed Tropical Rainfall Mis- canopies using remote sensing. sion (TRMM) could provide valuable data on the distribution of rainfall. RESEARCI I INSTITUTIONS OF THE AMAZON River stage records, the data from which dis- AND PROCESS STUDIES charge is calculated, are among the most com- The complement to the network type of informa- plete and the most accurate data available for tion are the detailed, generally more process- remote basins. The DNAEE maintains stage sta- oriented studies carried out by individual scien- tions at seven sites along the Amazon mainstem, tists and research teams, usually in affiliation at one to five sites along major tributaries (Rio witharescarchinstitutioninvolvedintheregion. Madeira, Japura, and so forth), and more intermit- In the Amazon, these include the following. tently along subtributaries. The density is greatest The Instituto Nacional de Pesquisas da in the most populated regions and in the areas with Amazonia (INPA, Manaus) has a broad mandate the greatest potential to generate hydropower. in areas related to hydrology and ecology. The Evapotranspiration isa much more stable pro- Museu Paraense Emilio Goeldi (MPEG, Belem) cess than precipitation (it is lcss variable over complements INPA, with strengthsinanthropol- time,andvariationsoccuroveralongerperiodof ogy and history. The Instituto de Pesquisas time). Methods for determining evaporation lo- Espaciais (INPE, Sao JosedosCampos) is the lead 411 Defining and Measuring Sustainability: The Biogeophysical Foundations organization in Brazil for remote sensing of the drainage network within which the mainstem Amazon, including a LANDSAT receiving sta- and its extensive floodplain (vdrzea) receive in- tion.Itisalsothebaseformanyoftheclimatologi- puts from a series of tributaries of different sizes cal studies in the Amazon, has impressive com- (see figure 26-2). It is characterized topographi- puter facilities, and is participating in field cam- cally by a great plain at altitudes lower than 200 paigns in the Amazon. The Centro de Energia meters. This plain is more than 3,400 kilometers Nuclear na Agricultura (CENA, Piracicaba) is a long from east to west and 2,000 kilometers wide research branch of the University of Sao Paulo from north to south. The great plain is bounded at specializing in the application of sophisticated the north by the Guyanan Shield (Guyana pla- stable and radioactive isotope and analytical teau), composed of ancient pre-Cambrian rocks. chemical techniques to environmental problems. In this plateau, with altitudes averaging from 600 With support from the International Atomic En- to 700 meters, the highest elevation in the Brazil- ergy Agency, it has pioneered the application of Fog Peak (Pico da Neblina) is 3,014 meters high. these techniques to the large-scale problems of To the south, the plain is bounded by the Brazil- hydrology in the Amazon. The Universityof Para ianplateau,alsocomposedof pre-Cambrianrocks, (Belem,) has a good geoscience department and with average heights of 700 meters; to the west, it has conducted studies in climate issues. The is bounded by the Andean mountain range, of Empresa Brasileira de Pesquisa Agropequaria tertiary origin, dividing the slopes of the Atlantic (EMBRAPA, Brasilia) maintains a number of field side from those of the Pacific. The Andean moun- sites in the Amazon for studies of soil fertility. tain range forms a semicircle, opening toward the Several large international projects are work- east, and has altitudes above 4,000 meters. Since ing in collaboration with these institutions. its emergence, the Andean mountain range has CAMREX,amultidisciplinarygroupfromCENA, been the main source of sediments for the Ama- INPA, and the University of Washington have zon plain. Today, about 13.5 tons per second of been working together since 1982 on the bio- materialiserodedfromtheAndes.Itisimportant geochemistryand hydrology of the Amazon river to stress the small surface gradient along the main system. EOSRAM, the group of CAMREX, to- channel of the Amazon River; the vertical drop gether with INPE is working on a large-scale between Iquitos in Peru and the estuary 2,375 analysis of the Amazon basin through the NASA kilometers downstream is only 107 meters. Yet, Earth Observing System mission. ABRACOS is a although from a geomorphologic viewpoint the joint Anglo-Brazilian study on the effects of great plain exists, it is in fact divided by innumer- changes in land use on the micrometeorology of able tributaries and streams that have cut deep specific sites. The Max Plank Institute (Germany) furrows into the soil and created a complex micro- has maintained a long-term interest in the Ama- structure of hills, gullies, and plains, with local zon, particularly in the area of limnology. slopesoftenexceeding45.1 nevaluatingtheselands ORSTOM (France) has had particular interests in for agriculture or cattle ranching, this fact is very fisheries, hydrology, and soil chemistry. importait, since erosion can rapidly eliminate the fertile soil stratum after the forest cover is removed. The Amazon mainstem has a total length of The Amazon hydrologic cycle: 6,771 kilometers, running from the Andes to the The biophysical basis of the intact system Atlantic Ocean. Formed by the Ucayali and Maranion in the Andes, it is first called the Rio Objective 2: To assess the state of the science of these Solimocs in Brazilian territory, changing to the measurements. Amazon after the junction with the Rio Negro (for clarity, we refer to all sections as the Amazon Physical structure River). Of the north-draining tributaries, the Ica The physical structure of the Amazon has two and Japuri rivers have Andean origins but are mostly lowland drainages. The Rio Negro drains components: the geomorphology and t.rainage primarilythecaatingaforestontheGuyanaShield, networkand thedistributioi and fertility of the soil. although its major tributary, the Rio Branco, drains a drier savannah region. Of the south-draining GEOMORPHOLOGY ANE, DRAINAGE NErWORK tributaries, theJutaf,Jurua,and Purusriversdrain The Amazon isaclassic riverbasin, with a central the sediments of the sub-Andean trough and of plain bordered by highlanids and a terrestrial the central plain, while the Rio Madeira begins in 412 Tropical Water Resource Management: The Biophysical Basis Figure 26-2: The Amazon Drainage Basin 50N 700 650 60°W 550 500 5 N 7 65 I 5 O0 Note: DNAEE gaging stations include Sao Paulo de Olivenca (SPO). Santo AntBnio do Ica (SAI), Itapeua (Ita), Manacapurul (Man), and Obidos (Obi). Major tributaries are indicated by name. the Bolivian Andes and passes across the Brazil- sources of water for the varzea is important for ian Shield and the central Amazon plain. The analyzing the nutrient cycling of the region and tributaries of the lower course of the Amazon, the for estimating the extent of biogenic gas fluxes; Trombetas and Uatuma rivers, are shield-drain- floodplains are an important source of methane ing rivers that have large "mouth bays," where to the troposphere. sediments are deposited. The main channel also receives input from smaller, ungaged tributaries SOIL DISTRIBUTIONS AND FERTIUT'Y and unchanneled varzea areas. The majority of the soils are chemically poor, with In the reach between Sao Paulo de Olivensa kaolinite and iron and aluminum oxides donu- and Obidos, floodwaters and direct precipitation nating the clay mineral fraction. About 80 percent regularly inundate about 40,000 square kilome- of the soils show low levels of exchangeable basic ters of varzea through an extensive network of cations. The high rainfall regime is the main rea- drainage channels (paranzas) and overbank flow son for the leaching processes and consequent during the 7 to '10 meter rise and fall of the river loss of exchangeable bases and silica. The nutri- over the course of a year. Approximately 10,000 ents lost with the intense leaching cannot be re- square kilometers are covered by thousands of placed either by the poor geological substrate or permanent lakes that range in size from less than by nutrients derived from the decomposition of a hectare to more than 600 square kilometers and organic matter that could be retained in the soil. are typically 6 to 8 meters deep at high water. As In contrast, the physical properties of the Ama- the river falls, land is exposed again, and the lakes zon soils are generally good. They are well drained, become isolated from the main channel, with due to their very stable sand and silt granular depths decreasing to 1 to 2 meters. Determining structure, originating from the cementing action the relative distribution of mainstem versus local of iron oxides, aluminum, and organic matter. 413 Defining and Measuing Sustainability: The Biogeophysical Foundations As a consequence, due to their low natural Another type of forest is the inundated forest, fertility, the majority of the Amazon soils are not which occurs in places periodically inundated by suitable for agriculture. Phosphorus is one of the rivers. The two most important types are varzea most problematic nutrients in the Amazon. Al- and igap6 forests. The first are found in the flood- though generally present at normal concentra- plains of white-water rivers, which are character- tions for soil samples, it is rarely available, due to ized by a rich soil. The second are found at the its adsorption to the oxides, hydroxides, and ka- margins of black-water rivers, consequently with olinite minerals. Another common problem is poorersoilsthanthoseof thefloodplainsof white- aluminum toxicity due to the characteristic acid- water rivers. ity of many tropical soils. Deficiencies like that A second general type of vegetation, occupy- would normallybecorrectablebyliming; tropical ing a smaller area than the forest formations, is soils,however, haveahighbuffercapacity, which the savannah. There are also two main types of makes liming or any other amending difficult savannah: terra firme and inundated. The and expensive. savannahs on terra firme are generally open grass- lands, with or without woody vegetation. Inun- Phytogeography dated savannah occurs mainly in the lower Ama- Classification of types of vegetation for the whole zon, between the Negro and Xingu rivers. The ofassification basin ty s ofrvegressedto r litte sice typical vegetation of these areas are grasses, of the Amazon basin has progressed little since shrubs, small vines, and several floating species. based on SLAR images. In addition, the Amazon Finally, there is the type of vegetation that grows hase along hLRimages. of botaddicalcltiong. M h Aaof over pure leached white sand, generally classi- has a long history of botanical collecting. Much of fied as campina, campinarana, or chavascal. what is known about Amazon flora is derived As reported by Nelson (1992), Gentry (1986) from approximately 290,000 herbarium speoc- has shown a strong correlation between annual mens. Unfortunately, by far the majority of col- rainfall and diversity of woody plants in lowland lected specimens are from "collecting islands," neotropical forests: lowland dry forestsgenerally such as the Manaus area, and little is known about have about 50 species greater than or equal to 2.5 the regional distribution of the documented flora. centimeters DBH per 0.1-hectare plot, moist for- The vegetation in the Amazon basin has been estsabout lOO to l50species,wet forestsabout200 described asdifferent formations. The forest for- species, and pluvial forests about 250 species. mation with the largest area is the terra firme Based on thiscorrelation,diversityofedaphically foqureskismte of foest haszan areaof3.8million similar sites should increase progressively from square kilometers in the Brazilian Amazon. The thedrytransversezone(dryforest)throughManaus terra firme forest can be divided in different for- (moist forest) to west of Iquitos (pluvial forest). mations. The most common is the dense forest, Land use may be considered as a special aspect with the greatest biomass, occurring mainly in of pteogay because, tpical thema places where no major factors limit its growth. eof agricultral orinustrialdlopmen mi Contrasting with the dense forest is the open to change radically the type of vegetation. Al- forest, with lower trees and biomass and higher though the regrowth differs from the original concentrations of shrub and liana species due to thon, t fowth ners om the orger greater penetration of light. According to Pires vegetation, it forms a new subset of the larger array of phytogeographic classes. Most of the and Prance (1985), the lower biomass of this type carin of pthefogretin theA . Mon bs tha of forest is caused by a low water table, poor cern ftefrs nteAao ai a ofdforestagis causd by easlow water tab rle,ioo been to establish pasture and to a lesser extent drainage and long dry seasons with low relative crops, timber, andcharcoal fuel;clearingforwhat- humid ity. The same authors divide the open for- ever reason typically is followed within a decade opes into threet witypelms: oend forest without.p by regrowth of shrubs and trees having low spe- open forest withpl and gina forest. cies diversity and biomass in comparison with The third group within the genera tralstera firme the original forest. Cleared areas and areas of group is the dry forest, which is a transition forest regrowth differ from the forest in important hy- foundtat the bordergbetween Amazoniazand cen- drologic parameters such as temperature, net tral Brazil. This region is characterized by long radiation, soil moisture, and leaf area index. In dry seasons with low relative humidity. Finally, .th isturd,h araterainbemea * ~~~the disturbed areas, these parameters canbe mea- there are the montane forests, generally occurring sured on the ground and the measurements ex- at higher altitudes, at the border of the basin. trapolated regionally by remote sensing. 414 Tropical Water Resource Management: The Biophysical Basis Hydrologic cycle November and the lowest, 301C, occurs in July. Overall, Septemberand Novemberare the months The first-order calculation of the water budget with the highest temperature, coincident with the (equation 26-1)yieldsthefollowing. The Amazon minimum precipitation. This pattern can be ex- basin encompasses an area of 6.4 million sq2uare plained by the portioning of the solar energy. kilometers, with an average precipitation of2,200 During the rainy months, a larger part of the millimeters a year. These figures represent a flux energy is used as latent heat, promoting evapora- of 14.1 trillion cubic meters a year of water into tion, while in the dry season a higher proportion the basin. The ultimate discharge of water from is used as sensible heat, increasing the air tem- trillion cubic meters a year. Therefore, approxi- perature. This isothermy results from the great mately 60 percent of the yearly precipitation within quantity of water vapor in the atmosphere. the basin is returned to the atmosphere, where it WATER VAPOR FLUX may again become precipitation. These averaged Water vapor of the Amazon region originates calculations can be refined, as follows. primarily in the Atlantic Ocean and enters the region with the trade winds, which blow year SOLAR ENERGY AND TEMPERATURE round from the east. Fluxes decrease from east The amount of solar energy reaching the upper to west across the basin. Precipitable water atmosphere in the Amazon remains practically vapor in the region averages 35 millimeters or constant the year round. For instance, in the city higher with a seasonal variation of 10 millime- of Manaus, situated in the central Amazon, the higherew,theaaverageawater or store- solar input varies from a maximum of 885 calories in the atmosphere above the Amazon basin is of per square centimeter a day in January to a mini- the order of 0.2 trillion tons. The greenhouse mum of 767 calories per square centimeter a day absorption of outgoing longwave radiation by in June. Solar radiation reaching the Earth varies this significant mass of water vapor largely primarily as a function of cloud cover. Data are accounts for the remarkable isothermal behav- available on the extent of variations in solar en- ior observed in the region (low fluctuation of ergyatselectedsitesinthecitiesofBelem,Manaus, surface temperature between day and night). andRioBranco.Theyearly(insolationratio-m/n Comparison of the seasonal cycle of the ratio-in the areas mentioned is below 50 percent basinwide, vertically integrated divergence of and varies during the year. The solar energy atmospheric moisture and the Amazon reaching the upper canopy of the forest is around atmow mOis and te fron 425 calories per square centimeter a day (Ribeiro streamflow at Obidos (500 kilometers from the and others 1982; Villa Nova, Salati, and Matsui lags the seasonal divergence of atmospheric 1976). Villa Nova calculated that 210 calories per moisture in the basin by approximately three squarecentimeteradayareusedinevapotranspi- months. Therefore, three months can be taken ration processes and that 215 calories per square as a first-order estimate of the time that water centimeter a day are consumed in heating the air resides in the Amazon hydrological system and are diverted into other processes. These data (Marques, Salati, and Santos 1980). At the south- indicate that evapotranspiration and water bal- eMr boundary of Amazonia, the direction of ance are of great importance for the energy bal- water vapor fluxes is from north to south for ance of the region. almost the entire year. This shows that water An important characteristic of the region's cli- vapor from Amazonia can influence the con- mateisthesmallvariationinthemonthlyaverage centration of water vapor in the atmosphere temperature, especially in the central strip below above the Brazilian Highlands. an altitude of 200 meters (see figure 26-3). For instance, in the city of Belem, the highest monthly PRECIPITATION average temperature, 26.9C, occurs in Novem- Precipitation is more variable than temperature ber and the lowest, 24 C, in March. In Manaus, the (figure 26-3). Fluctuations in the intertropical con- highest average monthly temperature, 27.9CC, vergence zone induce wet and dry seasons alter- occurs in September and the lowest, 25.8 C, be- nating between the northern and southern sides tween February and April, with a variation of of the basin. A pronounced difference in wet and only 2.1-C. In the city of Iquitos, the highest dry seasons between the northern and southern average monthly temperature, 32 C, occurs in sides is caused by the slow seasonal migration of 415 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 26-3: Precipitation and Temperature 500 '. . . 26.5 400 . . ... . 28.0 g <*, N . S . _ 400 *D 26.2 0 27.5 300 0 25.9 0 .27.0 o (35 *. 200 0D 26.5 a m 200 / 25.6 ,2 26.0 E 0~~~~~~~I- ~~~~~~~~100 '~- o- 25. 25.5 oF. . I I , ,. .25.0 0C 25.0 J F M A MJ J A S O N D J F M A M J A S O N D Months the continental convective bands over tropical transpiration (Salati 1986; Salati and Marques South America. The most important synoptic- 1984). Several independently used methods have scale (rain-producing) systems in the Amazon indicated the importance of the Amazon forest in may be sea breezelines that sometimes propagate this process. inland all the way to the Andes. The secondary The first studies concerning the meteorology maximum of precipitation in the southern Ama- and hydrology of the Amazon measured funda- zon can likely be accounted for by the interaction mental climatological parameters like tempera- of cold fronts with convective precipitation. Near ture, precipitation, solar radiation, and winds. the ocean, rainfall rates may reach up to 3,250 With respect to the water balance in the region, millimeters a year, and the same pattem is ob- studies were conducted at two scales. At the local served in the northwestem Amazon region. The scale, studies involved measurements made in highest precipitation in the basin is observed in small basins of a few square kilometers, while at the Andes, with values up to 7,000 millimeters a theregional scale,studiesinvolved measurements year; the minimum of 1,750 millimeters a year made over thousands of square kilometers. The occurs in the central part of the basin. In Belem first three studies encompassed large areas of the and Manaus, the rainy season goes roughly from basin. Molion (1979) used the climatonomic December to June. It is interesting to note that the method; Villa Nova, Salati, and Matsui (1976) distribution of rainfall is different above and be- used an adaptation of theclassicPenmanmethod; low the equator. There is a lag of six months in the and Marques, Salati, and Santos (1980) used the maximum level of precipitation north and south aerological method. The others used small water- of the equator. Salati and Marques (1984) esti- sheds.Forinstance,RibeiroandVillaNova(1979), mated an average of 2,300 millimeters a year for Leopoldo and others (1982) and Shuttlesworth the basin for the period 1972-75, while M. N. G. and others (1984) all used the Reserva Ducke Ribeiro (personal communication) estimated an forest, a 1.3-square-kilometer ecological station average of 2,100 millimetersa yearfor Manaus for near Manaus. Distinct methods were used for the the period 1911-85. estimations: Ribeiro and Villa Nova (1979) used both the Penman and the Thomthwaite methods, EVAPOTRANSPIRATION while Leopoldo and others (1982) and Jordan and As indicated by themassbalance in thebasin, one Heuneldop (1981) used water balance, and fi- of the most prominent features of the forest in the nally Shuttlesworth and others (1984) used eddy Amazon is its capacity to recycle a considerable correlation measurements over a period of ap- amount of water through the process of evapo- proximately one year. 416 Tropical Water Resource Management: The Biophysical Basis Theoverallresultswerevariablebutconverged 1991). The same authors suggest that there is a to a common conclusion. For instance, the studies substantial input of water from the evaporation of carried out on the Reserva Ducke Forest revealed open water surfaces, mainly during the dry season. a difference of 1 millimeter a day between the Evaporative process from open bodies of water smaller (3.7 millimeters a day; Shuttlesworth and may contribute up to 40 percent of the total flux of others 1984) and the larger (4.6 millimeters a day; recycled water. Possible sources of free water avail- Leopoldo and others 1982) values. The relative able for evaporation are rivers, lakes, and water contribution to rainfall ranged from 48 to 81 per- deposited on the surface of leaves of the canopy. In cent. Despite this variability, it is clear that a large fact, the canopy intercepts 10 to 20 percent of water, amount of water, sufficient to contribute at least asafunctionoftheintensityanddurationofrainfall 48 percent of the rainfall, returned to the atmo- (Franken and others 1982). Thus trees may play an sphere.through evapotranspiration. additional role in the Amazon water cycle, where A second independent type of analysis is pos- the water stored on their leaves after precipitation sible, using the water stable isotopes (180 and occursprovidesanimportantsourceofwatertothe deuterium) as tracers of the precipitation/evapo- atmosphere. Victoria and others (1991) suggest that transpiration sequence (for an explanation of iso- the interception process maybe larger than thought tope terminology, see Dall'Ollio 1976; Salati and beforeandconsequentlymaybeanimportantsource others 1979). Evapotranspiration is an isotopi- of water evaporation to the atmosphere. cally nonfractionating process, thus returning to The most important fact is that the water vapor the atmosphere water of isotopic composition flux originating in the Atlantic Ocean is not of similar to its source (rainwater), which in turn is sufficient magnitude to explain the rainfall and isotopically heavier than the atmospheric water the vapor outflux in the basin. As a direct conse- vapor.Therefore,asamassofairmovesinland,it quence, it is necessary to assume that receives an input of isotopically heavier water evapotranspired water recirculates in the basin. supplied by the forest evapotranspiration. This technology was used first in the Amazon by DISCHARGE Dall'Ollio (1976), who divided the Amazon basin The most striking features of Amazon River dis- into different sectors and analyzed the evolution charge are its magnitude and its highly damped of the water vapor through the behavior of oxy- hydrograph (see figure 264). Although differ- genanddeuteriumisotopesineach.Amodelwas ences in discharge of 7-10 meters are common developed based on the assumption that, as a along the main stem, there is only a twofold to mass of air moves inland and loses water through threefold difference between low and high dis- precipitation, rainfall becomes progressively de- charge. Sao Paulo de Olivenca has average mini- pletedinheavyisotopes.Thepatternofthisdeple- mum and maximum discharges of 20,000 and tion, assuming steady state conditions in the at- 60,000 cubic meters a second, Manacapuru aver- mosphere, can be modeled by a Raleigh-type ages 70,000 and 130,000cubic meters a second, equation.Forthe Amazon however, the observed and Obidos averages 100,000 and 220,000 cubic depletion in samples of rainfall water was much meters a second, respectively. The total Amazon smaller than expected based on the amount of input to the Atlantic includes the Tapajos, Xingu, precipi ta tion and the Raleigh law (Dall'Ollio 1976; and Tocantins rivers, for a mean annual input of Salatiandothersl979).Thediscrepancywascred- about 200,000 to 220,000 cubic meters a second. ited to the recycling of water through the evapo- The damped hydrograph of the main stem re- transpiration of the forest. As a result, the actual flects in part the offset input from tributaries. The depletion in the isotopic composition of the rain- peak flows from the northern and southern tribu- fall is smaller than the values predicted by the tariesare typically three monthsoutof phaseasa Raleigh equation. result of the seasonal differences in precipitation. Dall'Ollio's model was developed with a lim- Average tributary discharges range from about ited set of data, consisting of monthly rainwater 3,000cubicmetersasecondfortheJutaiandJurua samples of fifteen stations during the years 1972- rivers to about 30,000 cubic meters a second for 73. The extension of this data base with continu- the Negro and Madeira rivers. ing sampling did not change the trends and con- For each reach of the river, inputs from the last firmed the importance of forest evapotranspira- reach upstream and from the large, gaged tribu- tionintheAmazonwatercycle(Victoriaandothers taries constitute major inputs. In addition, we 417 Defining and Measuring Sustainability: The Biogeophysical Foundations Figure 26-4: Discharge along Ordinate and Tributary Gaging Stations, 1972-84 (housands of square meters) Sao Paulo do Olivenca 6 Manacapu72 80 84 072 76 80 84 .1 ! S11 ~~~~~~~~72 76 80 84 70 - O ' ,,, ., I,, .,, .,,5 - 125 7 ° 72 76 a0 82 472 76 80 8 140~~~~~~~~~~~~~~~~~~~~ 72 76 80 84~~~~~~~~~~~~~2 76 so 8 50~~~~~~~~~~~~~~~~~~~~2 72 76 80 84 0,~~~~~~~~~7 7 8 8 241 72 76 80 8084 140 . 72 76 0 84A~~AAA72 76 80 8 70 - 72~16 76vv 8vv 84 ' 80~~~~~~~~~~~~~~~~~~~~ 72 76 80 84 ~ ~ 7 7 8 8 N35e -acltdfo NE tg eod o anse SnoAtnod ~ n tpu r o hw) Yer In0ae 418~~~~~~~~~~~6 Tropical Water Resource Management: The Biophysical Basis mustassesstheotherpotentialsources(andsinks) ods, the discharge history must be considered. of water. These (ungaged) flows range from 3,000 The only long-term discharge record available for cubic meters a second during the dry season to the Amazon is a record of the stage of the Rio 7,000 cubic meters a second during the wet season Negro at Manaus, covering the period 1903 to the in the upriver and downriver sections; midriver present. That is, the Manaus record represents a flows are about half of these values. These esti- ninety-year integration of runoff and, ultimately, mates of flow for individual paranas and ungaged climatic conditions over 3 million square kilome- tributaries from thecalculationsof area precipita- ters of the Andean and western Amazon water- tion compare reasonably well to direct measure- shed. Thesedata canbe used tocalculateadischarge ments of discharge on those rivers. Overall, ex- time series for Manacapuru (see figure 26-5). change was greatest during early falling to mid- The mean discharge at Manacapuru for the falling water in the upriver and downriver reaches, period 1903-85 was 94,600 cubic meters a second. with a net flow from the floodplain to the main Minimum discharge varied between 48,000 and stem of about 20,000 cubic meters a second. Net 84,000 and maximum discharge between 100,000 exchanges were generally lower in the midriver and 140,000 cubic meters a second. Variability of reach, where the area of the floodplain is rela- the Amazon hydrograph is obviously dominated tively small. Therefore, water derived from local by the annual cycle. To reveal the nonseasonal drainage through paranas and small tributaries variability of the Amazon hydrograph more constitutes a significant component of the water clearly, the long-term mean annual cycle was budget of the main stem. These flows correspond removed, producing a deseasonal ized to about 30 percent of the flow at Itapeua and hydrograph (the lower part of figure 26-5). Over cumulatively to about 25 percent of the flow at the period 1903-26, there were pronounced oscil- Obidos. lations about the mean, with differences between minimum and maximum deseasonalized dis- Long-term variability in the climate! charge of 30,000 to 40,000 cubic meters a second. discharge record The minimum anomaly on record, 45,000 cubic meters in 1926, has been attributed to a period of The data from the DNAEE records represent a extensive drought and fires. From 1927 to 1962, short period of time. In order to determine whether the oscillationsexhibited a comparable frequency, the data from the detailed discharge records start- with reduced amplitude of 10,000 to 20,000 cubic ing in 1972 are representative of longer time peri- meters a second. Near the end of a secular trend Figure 26-5: Long-Term Discharge Record for Manacapuru and Deseasonalized Q'hydrograph, 1903-85 150 (a) 100- 50 *(b) 1900 1920 1940 1960 1980 Years Note: Arrows indicate occurence of ENSO events. Source: After Richey and others 1989. 419 Defining and Measuring Sustainability: The Biogeophysical Foundations of increasing discharge, which began around 1963, This analysis demonstrates the long-term link the maximum anomalies on record, 30,000 cubic between atmospheric circulations outside the meters a second in 1973 and again in 1976, were basin and discharge within the basin, but it does obtained. Thereafter, discharge returned to its not shed light on the actual physical mechanisms long-term mean value. Power spectrum analysis involved. It has been suggested that the descend- of the deseasonalized hydrograph reveals a pro- ing branch of zonal circulation over the equato- nounced spectral peak at 2.4 years. The tendency rial Pacific could be shifted eastward to over for regular oscillations on the two- to three-year Amazonia during the Southern Oscillation nega- time scale is evident in the deseasonalized time tive phase, suppressing convection and hence series itself. Similarly, recurrence intervals are precipitation, while the ascending motions asso- dominated by the two- to three-year flows. The ciated with theSouthemOscillationpositivephase deseasonalized hydrograph exhibits no signifi- would be strengthened, promoting increased pre- cant linear trend over the period of record. cipitation over Amazonia and northeastern Bra- These data indicate that the period of 1972 to zil. However, only a portion of the variance in the the present is indeed representative of the histori- discharge regime is linked to the ENSO phenom- cal pattern, and results obtained from these data enon. Relations between runoff and precipitation could be applied to the longer-term record. This are not straightforward, due in part to the conclusion, however, leads to the provocative carryover storage (basin memory effects) typical problem of determining the factors influencing of large catchments. The influences of local cli- the interannual variability that is observed. The mate, such as the boundary layer convergence oscillations of river discharge predate significant mechanisms and the steady progression of indi- human influences in the Amazon basin and re- vidual fronts and air mass boundaries character- flect both extrabasin and local factors. istic of the region, contribute to variability in Climate records and general circulation model discharge. calculations suggest that interannual variations These patterns of interannual variability indi- in the precipitation regime and hence discharge cate that considerable caution must be exercised of the Amazon may be linked to changes in the in determining anthropogenic impacts, particu- general circulation of the atmosphere over the larlywiththeuseofshort-termrecordsoverlarge tropical PacificOceanassociated with theElNifio- areas,. Conversely, it would be difficult to iden- Southern Oscillation (ENSO) phenomenon. To tify a unique deforestation effect in the highly test this hypothesis, the river discharge anoma- damped discharge regime of the Amazon River lies can be compared to atmospheric pressure mainstem.Thelikelihoodof linkagesbetween the anomalies at Darwin, Australia, a widely used Amazon basin and large-scale atmospheric circu- index of the ENSO. Qualitatively, the months of lations reinforces the importance of determining maximum pressure anomalies (Southern Oscilla- the factors controlling the hydrology of the basin tion negative phase) corresponding to ENSO in the face of extensive changes in land use. A warm events preceded the negative flow anoma- rigorous analysis of the regional hydrology of the lies in most cases. Major ENSO events, for ex- basin using field measurements and remote sens- ample, in 1925-26 and 1982-83, are reflected in ing integrated through realistic, physically based pronounced low discharges. The converse effect, modeling must be considered as the long-range high discharge associated with the Southern Os- goal for assessing and managing sustainable de- cillation positive phase, is also apparent. The velopment in the basin. unusually cold waters of the eastern Pacific in 1989 were accompanied by very high discharge in The implications for changes in land use the Amazon. These observations can be corrobo- rated statistically through cross-spectral analysis Patterns of land use in the Amazon basin are between the deseasonalized Amazon and Dar- changing. As summarized by Nelson (1992), the win records and between monthly mean sea sur- total area deforested was about 400,000 square face temperature in the eastern and central equa- kilometers, as of August 1989, which corresponds torial Pacific, a key oceanic indicator of ENSO, to 8.1 percent of the area and 10.2 percent of the and Amazon discharge for the period 1946-85. forests. For the 11.6-year period 1978-89 the rate We conclude from these results that the variabil- was about 21,000 square kilometers a year (0.53 ity in the Amazon hydrograph on two- to three- percent of the naturally forested area a year). year time scales is coupled with the ENSO cycle. Since the peak of 1987, rates have decreased, due 420 Tropical Water Resource Management: The Biophysical Basis to a number of possible factors, including weak- temperature was attributed to the decrease in ening of the threat of agrarian reform by 1988, roughness length. In Shukla, Nobre, and Sellers, heavy rainfall in 1989, drastic reduction in the theCOLLAGeneralClimateModel,coupled with money supply for investments in 1990, and wide- the Simple Biosphere Model (SiB) of Sellers and spread popular sentiment opposed to deforesta- others (1986), was used with a horizontal resolu- tion in Brazil and abroad, spurring more effective tion of 2.8 longitude by 2.8 latitude face rough- law enforcement. In 1991, deforestation wasabout ness length. The annual mean budget of surface half the 1989 rate. energy for Amazonia in the two simulations shows Although speculationisconsiderable,few data that absorbed solar radiation at the surface is are available on the effects of disturbances on the reduced in the deforestation case (186 watts per hydrologic cycle of the Amazon. Several ap- square meter) relative to the control case (204 proaches have been taken to speculating on pos- watts per square meter) because of the higher sible consequences. albedo for grassland compared to forest (21.6 and 12.5 percent, respectively). That plus the larger INTUITION AND BACK-OF-THE-ENVELOPE outgoing longwave radiation from the surface CALCULATIONS due to the higher surface temperature in the Salati and others (1979) postulated that a major deforested case mean that the amount of net alteration in vegetative cover of the region would radiative energy available at the surface for par- lead to changes in the climate at the microscale tition into latent and sensible heat flux is smaller and mesoscale. Changes would be felt particu- in the deforested case than in the control case (146 larly through variations in the albedo, in the and 172 watts per square meter, respectively). rainwater residence time, increase in runoff, and Also, as remarked in Shukla, Nobre, and Sellers, decrease inevapotranspiration. Increases in maxi- less precipitation is intercepted and evaporated mumtemperaturesanddailythermalamplitudes, again because the surface roughness and the due to a decrease in precipitation, would also be canopy water-holding capacity of the pasture are expected. Such alterations would be felt not only relatively small. Furthermore, the transpiration in the Amazon region itself but also in other rates are reduced due to the reduced moisture- nearby regions, especially the Brazilian central holding capacity of the soils under pasture. plateau. Bearing suchevidencein mind, it isclear An interesting result is that the reduction in that a major alteration in vegetative cover of the calculated annual precipitation was larger than region would lead to changes in the climate at the the reduction in evapotranspiration (642 and 496 microscale and mesoscale. millimeters, respectively), which suggests that changes in the circulation of atmosphere may act LARGE-SCALE MODEL SIMULATIONS to reduce further the convergence of moisture Morerecently,simulationmodelshavebeenused flux in the region, a result that could not have to speculate as to what the consequences of turn- been anticipated without the use of a dynamic ing the Amazon into pasture might be. Two of the model of the atmosphere, as noted in Shukla, most recent general climate model simulations of Nobre, and Sellers (1990). This, in turn, implies tropical deforestation were conducted by the that runoff also decreased for the deforested case, UnitedKingdomMeteorologicalOffice(Leanand since the decrease in precipitation was larger Warrilow 1989) and by Shukla, Nobre, and Sellers than the decrease in evapotranspiration. (1990). In Lean and Warrilow, the model's hori- Taken together, the results of these studies zontal resolution was 3.75 longitude by 2.5 'lati- seem to suggest that the regional climate is sig- tude, and all the model's vegetation north of 30' nificantly sensitive to the removal of tropical South in South America was replaced by grass. forest. In general, the somewhat short period of Although the total area in which the model's integration in these studies precludes drawing vegetation changed was almost twice that used in conclusions about the significance of changes in DHS and in Shukla, Nobre, and Sellers, their global climate or even changes in regions adja- results were similar to those in DHS: surface cent to Amazonia. temperature increased by 2.5C and evapotrans- The conversion of tropical forested areas into piration decreased for the pasture scenario com- pastures and other types of short vegetationcauses pared to the forest one. Addi tionally, it was found changes in the microclimate of the disturbed ar- that simulated precipitation was reduced over eas; if the size of the perturbed area is sufficiently Amazonia. As in DHS, the increase on surface large, even the regional climate may be altered. 421 Defining and Measuring Sustainability: The Biogeophysical Foundations Depending on their scale, these alterations may effect on the biological, chemical, and physical cause changes in climate at the global level and processes in the top soil. Plants, animals, and affect regions distant from the tropical forests. microorganisms living in that layer experience temperature, humidity deficit, and water stresses LOCAL CHANGES IN CLIMATE not present in the remarkably constant microcli- Changes also occur in albedo and in energy and mate of the forest floor. water balances. There is a tendency toward less water infiltration, more runoff during rainy peri- REGIONAL CHANGES IN CLIMATE ods, and less runoff during prolonged dry peri- The summation of local changes in climate over a ods. An important conclusion of the micrometeo- sufficiently large quasi-contiguous area (say, rological studies conducted at Ducke Reserve, larger than 1 million square kilometers) mnight near Manaus in central Amazonia (summarized change the transport of water vapor and the wa- in Shuttlesworth 1988), is that the annual flux of ter balance at a regional level with consequent latent heat into the atmosphere is close to its changes in the energy balance. Climatic alter- potential value, that is, 20 percent smaller than ations and the scale at which they occur depend the potential evapotranspiration during the dry on the geographic location and its geomorphol- season and about 10 percent above the evapo- ogy. For instance, even small changes in the transpiration rate during the rainy season. low-level wind regime on mountainous areas Shuttlesworth suggests that there might be a re- such as the Andean cordillera can cause a large duction of between 10 and 20 percent in the change in the temporal and geographic distribu- evapotranspiration for pastures as compared to tion of rainfall. It is not possible yet to predict that for the rain forest, mostlybecause the albedo accurately, by means of model simulations of of grass is higher than the albedo of tropical climate, regional changes in climate associated forests (thus, the amount of available energy is with the observed patterns of deforestation. An smaller, other things being equal). That reduc- important reason for such a limitation is that tion, in turn, might cause rainfall to decrease by when current climate models are integrated in a 10 percent, he suggested. Yet, this hypothetical control model, such as attempting to mimic the scenario takesintoaccountonlychangesinevapo- observed climate, they commonly fail to repre- transpiration due to changes in the availability of sent important aspects of the regional climate. radiative energy. Important changes also occur One problem is, of course, resolution. It is ex- due to the decrease in surface roughness at the pected that only when model resolution becomes soil level. Loss of organic matter in the top soil of the order of 100 kilometers (current climate and fauna in the soil, compaction due to agricul- model resolution is typically between 200 and 500 tural practices and overgrazing, and soil erosion kilometers) will the models probably capture the may cause large changes in the physical and finer details of the regional climate. Yet, the re- chemnical characteristicsof the predominantlyclay sults of recent climate model simulations of Ama- soils of the Amazonian terra firme forest. Those zonian deforestation, reviewed in the previous changeslikelycombine to reduceinfiltration rates section, suggest the following changes at the re- drastically, increase surface runoff during rainy gional level to be likely following extensive de- periods, and decrease soil moisture in the shal- forestation of tropical forests: increase in surface lower rooting zone of the grass vegetation prima- and soil temperature, increase in the diurnal fluc- rily during the dry season. Decreased availability tuation of temperature and specific humidity of soil moisture also reduces evapotranspiration. deficit, and reduction in evapotranspiration and Comparative measurements of the diurnal PBL moisture. In two of the three studies (Lean cycle of canopy and subsurface temperature at and Warrilow 1989; Shukla, Nobre, and Sellers cleared and forested sites in Ibadan, Nigeria, and 1990), yearly average precipitation and runoff in Surinam showed a large increase of soil and air decreased for Amazonia as a whole for pasture temperatures (more than 5'C and 3 C, respec- vegetation compared with forest. The annual re- tively) for the cleared areas compared to the for- duction in rainfall in these two simulations was ested ones. Notbeing in the shade of a tall canopy, larger than the correspondingreduction inevapo- the diurnal fluctuation of ground temperature transpiration, thus explaining the reduction in and humidity deficit was much larger for the runoff. It is likely, however, that runoff will in- cleared sites in these two studies as well. Those crease following rainy periods, that is, runoff changes in soil microclimate have a profound (and river streamflow) will be higher after defor- 422 Tropical Water Resource Management: The Biophysical Basis estation during the rainy season and will de- of maximum rainfall over tropical South America crease during the dry season. from central Amazonia in January to February to Central America in June to July. GLOBAL CHANGES Tropical forest areas also have a characteristic Tropicalforestscontributeinmanywaystomain- energy balance that contributes to the transport tain the present dynamic and chemical equilib- of energy as latent heat (water vapor) from the rium of the atmosphere. Forests represent a car- equatorial regions to those of greater latitude. bon reservoir, both through their areal and root This is particularly conspicuous in central Brazil, systems as well as through organic matter in the southern Bolivia, Paraguay, and northern Argen- soil. Estimates indicate that the Amazon region tina, where, due to the generally southward possessesareserveofcarbonataboutl00gigatons. low-level circulation, most of the water vapor Therefore, conversion of forests into pastures will present comes from Amazonia. Therefore, changes release carbon dioxide from the biosphere into in atmospheric moisture in Amazonia due to the atmosphere, likely enhancing the greenhouse deforestation might affect the precipitation of the warming. adjacent regions to the south. The burning associated with the process of convertingforestsinto pasturesalso releasesgreat Hydroelectric impoundments quantities of particles and compound gases into the atmosphere. These particles cause changes in Given the vastness of the Amazon hydrologic theatmosphere,especiallyinitschemicalcompo- cycle, the desire to hansoelfraction of the sition and energy balance. To understand and energy in the form of hydroelectric power is predict any possible large-scale change in climate obvious. In 1987 the Brazilian Ministry of Energy due to tropical deforestation, it is crucial to know presen ted the Plan 2010, with the main objective theextenttowhichthepatternsof rainfall change of transforming the Amazon into an energy-ex- when rain forests are converted into grasslands. port region by that year. That plan was the center It is well known that the tropical regions function of debates about its potential environmental im- as atmospheric sources of heat through the re- pactsandeventherealenergyneedsofthecountry. lease of latent heat by the condensation in convec- The first and by far the largest dam is the leaseofloudsaThenheatbytseonreleaseddrienslatvec-se Tucurui damon the Rio Tocantins, 300 kilometers tive clouds. The heat so released drives large-scale south of Bel6m, built primarily to supply power tropical circulations (of the Hadley-Walker type) to the electrometallurgical industry of the Carajs with ascending motion over the tropical regions, to the e astergical indus h as mostly over Amazonia, tropical Africa, and the region of the eastern Amazon. Tucurui has a Indonesia-western Pacific region, and descend- nominal capacity of 8,000 megawatts, with about ing motion over the dry subtropics, primarily half of that currently on line. Much less efficient over the subtropical oceans. It is conceivable that are the other dams of the Amazon. The Balbina a significant reduction in rainfall over Amazonia power plant, built at the Rio Uatuma, north of (say, greater than the 20 percent reduction sug- Manaus, generates only 110 megawatts from an gested by the model simulations described in inundated area 2,360 square kilometers, similar Lean and Warrilow 1989 and Shukla, Nobre, and to the Tucuruf. Balbina thus generates only 0.1 Sellers 1990) might have an effect in these tropical megawatt per flooded square kilometer against circulations. However, it is unclear what these 3.3 megawatts per flooded square kilometer in changes would be and how they would manifest Tucurui, and its estimated construction cost of themselves in terms of climatic changes in the US$3,000 per kilowatt is about four times higher tropics but away from the perturbed areas and in than the cost of building Tucuruf. The Samuel the extra-tropics. Regarding the extra-tropics, dam, constructed 50 kilometers south of Porto Paegle (1987) suggests a possible link between Velho, Rondonia, is another example of the prob- tropical convection and quasi-stationary features lems associated with hydropower dams in the of the large-scale circulation over North America. Amazon. The flatness of the area is also a problem He suggests that the westward shift of the sub- at Samuel. To keep water from spilling into adja- tropical jet stream from the east coast of North cent drainage basins, a huge dike, of almost 60 America in boreal winter to the west coast in kilometers in extension, had to be constructed. spring and a concomitant westward shift of the The nominal capacity of Samuel is 217 mega- North American longwave trough maybelinked to watts, but as in Balbina, availability of enough the seasonal, northwestward migration of the area water will probably be a problem. The estimated 423 Defining and Measuring Sustainability: The Biogeophysical Foundations size of the inundation area is 540 square kilome- could be a useful means of trackingor monitoring ters,generatingthereforeO.4megawattpersquare a situation, where a full-blown field campaign kilometer, a bit more than Balbina, but still much would be impossibly expensive to maintain, and less thanTucurui.The cost of constructing Samuel the results too complicated to explain. As such, it was US$5,000 perkilowatt,almost 1.5 timehigher could provide a useful bridge between the com- than Balbina. plexities of the physical world and the complexi- These dams have had controversial effects on ties that managers must confront and sell. How the environment and indigenous peoples. An inclusive and universal should such a interplay between changes in land use and hy- sustainability index be? There is an elegance to droelectric potential is shown by the following the concept that it should be universally appli- example. The sedimentological studies done for cable, across all systems and all scales. Such a the construction of the Samuel dam did not take sustainability index would have to be something into account the possible changes in land use and like an increase in carbon dioxide. Although per- theireffectonthesedimentloadoftheRioJamari. haps indicative of the planet's health, such a Studies that are presently being carried out using sustainability index is by definition so broad that the 20Pb technique to calculate the sedimentation it carries little new information. Rather, it might rate in a lake near the dam are showing increases make sense to explore more regionally based in the rates that might be closely correlated to indexes that summarize the biophysical informa- deforestation or tin mining activities in the basin tion of that region in a context as rich in informa- (Forsberg and others 1989). These rates, if contin- tion as possible. ued, will result in a drastically shortened life for In this spirit, we explore here a sustainability the dam. index based on the water cycling characteristics of the Amazon. Ultimately, the integrity of the Amazon ecosystem depends on a dynamic equi- Future measurement protocols librium between the vegetation and water cycle, for maintaining sustainability such that about half of the rainfall can be derived from evapotranspiration. To reflect this, a simple Objective 3: To assess proposals and suggest research sustainability index might be the ratio of evapo- for indicators andlor indexes of sustainability that can transpiration to precipitation, where a deviation be collected easily and frequently used to monitor the from 0.5 induced by a change in land use would susfainability of large regions composed of a mosaic of indicate a loss of sustainability. In practice, this ecosystem types, up to continent size. would be difficult to monitor, and deviations from 0.5 could occur for other reasons. The in- Indexes for monitoring and predicting verseratio,runoff toprecipitation,wouldbeeasier sustainability to monitorbut would suffer from the samelackof precision. Indicators of sustainability encompass simple A more robust sustainability index would be measures such as a sustainability index and com- the behavior of runoff to precipitation relative to plex measures such as modeling, prediction, and the PO content of precipitation by region (re- management. Once determined, a coreset of mea- member the previous discussion of stable iso- surements must be collected now and into the topes; see figure 26-6). An intact system would be future. characterized by a runoff to precipitation ratio of about 0.5 and a al8O of about -6 /,. In a system SIMPLE INDICATORS: A SUSTAINABILITY INDEX where the recycling of water is affected, runoff A sustainability index that somehow summa- would increase (and evapotranspiration would rizes the health or integrity of ecological and decrease)relativetoprecipitation,whiletherelative human systems in a manner comparable to how amount of 18O would decrease; for example, the the gross national product summarizes economic a'sO value of rain would become more negative. health is an attractive management target. To what use would such an index be put, and what MORE COMPLEX INDICATORS: MODELING, information should it contain? Many resource- PREDICTION, AND MANAGEMENT based sectors have their targets-for example, A more complex analysis of sustainability that catch or crop yield-but what are the analogs for would allow prediction of management options natural systems? A robust sustainability index should be based on regional modeling. To simu- 424 Tropical Water Resource Management: The Biophysical Basis Figure 26-6: Ratio of Runoff to Precipitation (1) Model for distributed rainstorms. The basin Relative to ;180, by Region area modeled here is larger than the area of rain- storm events that generate runoff and recharge soil moisture. Physically based models for storm runoff and water balance dynamics require spa- tially and temporally realistic inputs from rain- .10 storms. Precipitation patterns have been observed to have a well-defined hierarchic structure. Ac- cording to their areal extend, storms can be clas- "O /s y . sified as synoptic area (larger than 10,000 square e 0.5 kilometers), large mesoscale area (1,000 to 10,000 S _ square kilometers), or small-scale area (100 to 1,000 square kilometers). Within the large mesos- cale area, either inside or outside a small mesos- -4 we 4cale area, are areas made of several convective de) O,yonro-18 oPnclDlatl cells. Each cell ranges from 10 to 30 square kilo- meters in area. The life spans of these different scales of rainstorms are, respectively, days for synoptic storms, several hours for large mesos- cale storms, and a few hours for small mesoscale late the hydrologic and biogeochemical conse- storms and convective cells. In this hierarchic quences of changes in land use, an Amazon tribu- structure, rainfall intensity increases as the scale tary-scale (10,000 to 100,000 square kilometers) decreases. Rainstorms of the small mesoscale or model forbasin runoff and waterbalancedynam- cell clusters are considered significant to basin ics coupled with the DNAEE data base is neces- hydrology and are used as basic rainstorm input sary. For the purpose of routing both water and for the model. its chemical load, the model must provide de- <2) Hillslope runoff production model. Precipita- tailed information about dominant hydrologic tion must thenbe partitioned through a model for processesarrayed in the horizontal, including the the production of hillslope runoff and soil mois- residence time of water within each flow path- ture dynamics. Storm runoff on hillslopes may be way. The model must also recognize the con- produced by three mechanisms: infiltration ex- straints of this logistically difficult region. cess or Horton runoff, saturation excess runoff, For example, our research group is formulat- and subsurface stormflow. Horton overland ing a tributary-scale model (RAM) for storm run- flow occurs when the rainfall intensity exceeds off and water balance dynamics for the Amazon the infiltration capacity of soil. Saturation ex- River basin (Zhang, Richey, and Dunne 1990). cess runoff occurs when the soil becomes satu- The model is based on the realistic parameteriza- rated to the surface from below. This mecha- tion of microscale physics concerning the pro- nism results in the partial areas of runoff pro- duction of runoff and the dynamics of soil mois- duction located adjacent to streams. The partial ture on hillslopes and flow routing through ac- areas may expand or contract during and be- tual drainage networks. The model allows ex- tween storms. Subsurface stormflow occurs plicit incorporation of key vegetation, soil, and when there is significant downslope flow be- geomorphologicfeaturesofthebasinandaccepts low the hillslope surface and soil saturation realistic space-time rainstorms. The model is de- from below does not reach the surface. The signed to make maximum and more direct use of response of any particular basin may be domi- the information about land surface and climate nated by a single mechanism or may involve forcing, which have been and are to be obtained more mechanisms that occur at the same time in from the field and by current and anticipated different parts of the basin. Hydrologic models remote-sensingtechnology.Themodelintegrates for large areas should consider these different several physically based submodels concerning runoff production mechanisms. In the heavily the primary components of the land phase of the vegetated Amazon River basin, it is also neces- hydrologic cycles. The four submodels include sary to take into account the effects of vegeta- the following: tion on the production of runoff. 425 Defining and Measuring Sustainability: The Biogeophysical Foundations Given these processes, RAM will need to con- flow within a drainagebasin, the assumption of a vert the rainstorms into hillslope runoff and soil spatially uniform velocity of travel has proved to moisturestorageintimeandspace.Themicroscale be a reasonable approach. processes will be parameterized and averaged within the large-scale model. Separate submodels MEASUREMENT SCHEMES must be developed for the interception and stor- Acoresetofmeasurementsmustbemadeinto the age of moisture by land surface vegetation, soil future. Precipitation, discharge, temperature, infiltration and infiltration excess runoff, soil humidity, and wind fields must be measured as moisture dynamics, subsurface stormflow, and well as possible over as wide an area as possible. saturation excess runoff. The DNAEE network is fundamental to this ef- (3)Modelforevapotranspiration. Amodelwillbe fort. It should be maintained and enhanced (in- used to calculate moisture flux due to evapo- cluding automatic weather stations). To tie to transpiration during periods between storms or biogeochemical cycles, measurements of atmo- on "unwetted" lands between patches of storms. spheric gases indicative of ecosystem function The model will determine the antecedent condi- (carbon dioxide, methane) should be monitored, tionofsoilmoisturefortheproductionofhillslope preferably at selected "towers" coupled with runoff. Evapotranspiration from land surface soil measurements of atmospheric boundary layer. and layers of vegetation will be calculated using River chemistry at the mouths of key tributaries, several of the detailed biophysics models being with more in areas subject to a change in land use, developed elsewhere (the Simple Biosphere should be monitored. The actual design of such Model, SiB) as nodes. These models are designed sampling programs is controlled primarily by the as a subgrid model of general climate models for pragmatic realities of what is economically and calculating the transfer of surface energy, mass, logistically feasible. and momentum between the atmosphere and the vegetated land surface. They are physically con- sistent with the hillslope submodel for runoff Management options for achieving production and soil moisture dynamics. The sustainability of production of goods hillslope model will provide realistic initial and and services and of ecosystem integrity boundary soil moisture conditions for SiB-type models, while SiB-type models will provide de- The problem of how to preserve the Amazon is tailed informationon moisture flux fromlayersof extremely complex. Use and preservation of the soil and vegetation to atmosphere and anteced- Amazon's resources have important social and ent conditions of soil moisture for the hillslope political aspects. Pressure on the Amazon rain model for runoff production and soil moisture forestcontinues toincrease,asthehumanpopula- recharge. tion grows and the needs of the people outstrip the (4) Geomorphologically based channel routing capacity of the forest to sustain them. For example, model. Once runoff to the stream network has the recent decline of the Zona Franca (free trade been generated, the model will integrate the pa- zone) in Manaus may soon force 100,000 people out rameterized and averaged microscale processes of work, many of whom will then turn to the interior of hillslopes and route the hillslope hydrographs if their needs cannot otherwise be met. to the outlet of the basin. The model will be Theapplicationofscienceand technologycom- developed using a generalized basin geometric bined with resource management must play a and stream network structure, to facilitate the key role in addressing the sustainability of the integration and routing that occurs in a large Amazon. Specific technical issues include prob drainage basin. In the model, structure and ge- lemssuchasrecoveringdegradedareasandiden- ometry of the basin network are quantified with tifying natural products from plants or animals a width and an area function. Hydrologic re- that can be exploited sustainably. Conservation sponse of a basin is closely related to the general- goalsmust include thepreservationof biodiversity ized drainage network by the width function as well as maintenance of dynamic processes of rather than by Horton or Strahler laws. With the the underlying hydrologic and biogeochemical generalized structure of the basin network, the cycles essential to the structure and function of flow is routed to the basin outlet by assuming a the basin. How can scientific information (data, spatially constant velocity of travel. For routing models)beusedinpoliticaldecisionmaking?How 426 Tropical Water Resource Management: The Biophysical Basis can demographic and economic information re- Martinelli, Luiz A., Reynaldo L. Victoria, Jeffrey ally be used for modifying current patterns of E. Richey, J. Mortatti, and A. H. Devol. Forth- land use and developing the elusive target of self- coming. "The Amazon Basin: Natural Cycles sustaining development? and Land Use Change." In summary, sustainabilityof the tropicsshould Molion, L. C. 1979. "A Climatonomic Study of the mean not only biodiversity but, in particular, the Energy and Moisture Fluxes of the Amazon biophysical processes of the underlying hydro- Basin with Considerations of Deforestation Ef- logic and biogeochemical cycles essential to the fects." Ph.D. diss., University of Wisconsin, structure and function of these basins. Studies of Madison. these cycles in their natural state are important . L . r o because they give mankind the time and the tools Monteith, J. L. 1973. Prnciples of Environmental needed to learn how the necessities of inhabitants Physics. New York: Elsevier. can be reconciled with the necessary preservation, Nelson, B. W. 1992. "Classification of Amazon in a truly self-sustained development. Such studies Vegetation by Satellite Remote Sensing. Uma provide a basic data set, against which possible Estrategia Latino-Americana para a effects of changes in land use can be tested. Amazonia." Fundacao Memorial de America Latina, Sao Paulo, Brazil. Paegle, H. 1987. "Interactions between Convec- References tive and Large-scale Motions over Amazonia." In R. Dickinson, ed., The Dall'Ollio, A. 1976. "A composicao isot6pica das Geophysiology of Amazonia, pp. 347-87. New precipitaq6es do Brasil: Modelos isot&rmicos e York: John Wiley. a influencia da evapotranspiracao na bacia Pires, J. M., and G. T. Prance. 1985. "The Vegeta- Amazonica." Master's thesis, Universidade de tion Types of the Brazilian Amazon." In G. T. S3o Paulo, Piracicaba. Prance and T. E. Lovejoy, eds., Key Environ- Forsberg, B. R., J. M. Godoy, R. L. Victoria, and L. ments, Amazonia, pp. 109-45. New York: A. Martinelli. 1989. "Development and Ero- Pergamon Press. sion in the Brazilian Amazon: A Geochrono- Ribeiro, M. N. G., E. Salati, N. A. Villa Nova, and logical Case Study." Geolournal 19, pp. 402-05. C. G. B. Demetrio. 1982. "Radiaqao solar Franken, W., P.R. Leopoldo, E. Matsui,and M. N. G. disponivel em Manaus (AM) e sua duracao Ribeiro. 1982. "Interceptaqaodasprecipitagoes com o brilho solar." Acta Amazonica 12, pp. em florestas Amaz6nica de Terra-Firme." Acta 339-46. AmazBnica 12, pp. 15-22. Ribeiro, M. N. G, and N. A. Villa Nova. 1979. Gentry, A. H. 1986. "An Overview of Neotropical "Estudos climatol6gicos da Reserva Florestal Phytogeographic Patterns with an Emphasis Ducke, Manaus, AM. III. Evapotranspiraqao." on Amazonia." In M. Dantas, ed., Proceedings, Acta Amaz6nica 9, pp. 305-10. Vol. 2, Flora and Forest, pp. 19-36. First sympo- Richey, Jeffrey E., J. B. Adams, and Reynaldo L. sium on the humid tropics. EMBRAPA-CPATU Victoria. 1990. "Synoptic-scale Hydrological Documentos 36. Belem. and Biogeochemical Cycles in the Amazon Jordan, C. F., and J. Heuneldop. 1981. "TheWater River Basin: A Modeling and Remote-sens- Budget of an Amazonian Rain Forest." Acta ing Perspective." In H. Mooney and R. Amaz6nica 11, pp. 87-92. Hobbs, eds., Remote Sensing of Biosphere Lean, J., and D. Warrilow. 1989. Nature 342, pp. Functioning, pp. 249-68. Berlin: Springer- 411-13. Verlag. Leopoldo, P. R., W. Franken, E. Matsui, and E. Richey, Jeffrey E., J. 1. Hedges, A. H. Devol, P. D. Salati. 1982. "Estimativa da evapotranspiracao Quay, Reynaldo L. Victoria, Luiz A. Martinelli, de floresta Amaz6nica de Terra-Firme." Acta and B. R. Forsberg. 1990. "Biogeochemistry of Amaz6nica 12, pp. 23-28. Carbon in the AmazonRiver." LimnologyOcean- Marques, J., Eneas Salati, and J. M. Santos. 1980. ography 35, pp. 352-71. "CAlculo da evapotranspiraco real na bacia Richey, Jeffrey E., L. A. Mertes, Reynaldo L. Amazonica atraves do m6todo aerol6gico." Victoria, B. R. Forsberg, T. Dunne, E. Oliveira, Acta Amazonica 10, pp. 357-61. and A. Tancredi. 1989. "Sources and Routing 427 Defining and Measuring Sustainability: The Biogeophysical Foundations of the Amazon River Floodwave." Global Bio- Salati, Eneas, Reynaldo L. Victoria, Jeffrey E. geochemical Cycles 3, pp. 191-04. Richey, and Luiz A. Martinelli. 1991. "Forests: Richey, Jeffrey E., C. Nobre, and C. Deser. 1989. Their Role in Global Change, with Special Ref- "Amazon River Discharge: 1903 to 1985." Sci- erence to the Brazilian Amazon." Proceedings of ence 246, pp. 101-03. the Second World Climate Conference. Cambridge, Richey, Jeffrey E., and Reynaldo L. Victoria. 1994. Mass.: Cambridge University Press. "C, N, and P-Export Dynamics in the Amazon Sellers, P. J., Y. Mintz, Y. C. Sud, and A. Dalcher. River." In R. Wollast, ed., Interactions of C, N, P, 1986. "A Simple Biosphere Model (SiB) for Use and S Biogeochemical Cycles. Berlin: Springer- within General Circulation Models." Journal of Verlag. Atmospheric Science 43, pp. 505-31. Salati, Eneas. 1986. "The Climatology and Hy- Shukla, J., C. Nobre, and P. Sellers. 1990. "Ama- drology of Amazonia." In G. T. Prance and T. zon Deforestation and Climate Change." Sci- E. Lovejoy, eds., Amazonia. Oxford, England: ence 247, pp. 1322-25. Pergamon Press. Shuttlesworth, W. J. 1988. Quarterly Journal of the . 1987. "The Forest and Its Hydrologi- Royal Meteorological Society. cal Cycle." In R. Dickinson, ed., The Shuttlesworth,W.J.,J.H.C.Gash,C.Lloyd,C.J. Geophysiology of Amazonia, pp. 347-87. New Moore, J. Roberts, A. de Marques Filho, G. York: John Wiley. Fisch, V. de P. Silva Filho, M. N. G. Ribeiro, L. Salati, Eneas, A. Dall'Ollio,J. Gat, and E. Matsui. C. B. Molion, L. D. de A Sa, C. A. Nobre, 0. M. R. 1979. "Recycling of Water in the Amazon Cabral, S. R. Patel, and J. C. de Moraes. 1984. Basin: An Isotope Study." Water Resources "EddyCorrelationMeasurementsofEnergyPar- Research 15, pp. 1250-58. tition for Amazonian Forest." QuarterlyJournalof Salati, Eneas, and J. Marques. 1984. "Climatol- the Royal Meteorological Society 110, pp. 1143-62. ogy of the Amazon Region." In H. Sioli, ed., Victoria, Reynaldo L., Luiz A. Martinelli, J. The Amazon Limnology and Landscape Ecology Mortatti, and Jeffrey E. Richey. 1991. "Mecha- of aMightyTropicalRiverandltsBasin,pp.87- nisms of Water Recycling in the Amazon Ba- 126. Dordrecht: Dr. W. Junk Publishers. sin: Isotopic Insights." Ambio 20, pp. 384-87. Salati, Eneas, Reynaldo L. Victoria, Luiz A. Villa Nova, N., Eneas Salati, and E. Matsui. 1976. Martinelli, and Jeffrey E. Richey. 1989. "De- "Estimativa de evapotranspiracao na bacia forestation and Its Role in Possible Changes Amazonica." Acta Amaz6nica 6, pp. 215-28. in the Brazilian Amazon." In R. DeFries and Zhang, W., Jeffrey E. Richey, and T. Dunne. 1990. T. F. Malone, eds., Global Change and Our "Modeling Runoff and Water Balance Common Future: Papersfrom a Forum, pp. 159- Dynamics for Large Scale Tropical Drainage 71. Washington, D.C.: National Academy Basins." AGU Chapman Conference, Lake Press. Chelan, June. 428 Tropical Water Resource Management: The Biophysical Basis struction of dams in the Amazon. Large dams Comments such as Tucurui, Balbina, and Samuel undoubt- edly produce a lot of changes in the water cycle, Jose G. Tundisi the quality of water, and the biota (Tundisi, Matsumura-Tundisi, and Calijuri 1992; Tundisi, The chapter deals with the fundamental question forthcoming). of sustainability in the largest basin on Earth: the Finally, the authors discusse the design of a Amazon. Some basic questions and problems are sustainability index and provide a basic idea for addressed, mainly related to the physical effects such an index based on thebehaviorof runoff and of the Amazon on the communities and the con- precipitation relative to 1'0 content by region. trol of biogeochemical cycles by the large and The search for such indexes is vital for the future diverse biomass of plants and animals. management of the system. The discussion of Particularly for the long-term climatological several types of models to be applied is also and hydrologic cycles, the records are described pertinent because such models help in the devel- and discussed. Of special interest is the analysis opment of predictive capabilities. of long-term variability, the interannual varia- Thegivensuggestionsforfuturemeasurements tions in the precipitation regime, the discharge of are interesting and pertinentbut, probably, should water, and the conclusion that extrabasin atmo- be more emphatic. The construction of a perma- spheric circulation interferes with discharge of nent data bank is the key for the sustainability the Amazon. The description of physical struc- issue. A more complete analysis and detailed tureand typeanddistributionofsoilprovidesthe discussion of the sampling schemes and of the background for application of the sustainability overall problem would fill an existinggap. Equally concept. important is the discussion of the management The patterns of land use are presented; the options for the Amazon basin and the various possible changes that would occur in the water alternatives. This was only slightly mentioned by cycle with the change in vegetative cover are the authors since this is not the main scope of the discussed. However, there seems to be a scarcity chapter, but management and management al- of data about this problem. This is a key issue: the ternativesareundoubtedlyone of themajorques- extent to which a major change in the vegetative tions to be considered in the Amazon region. cover of the Amazon would produce changes at Twoimportantconclusionscanbedrawnwhen the micro, meso, and macroscale in the climate, analyzing this chapter. (a) First, a sustainability What would be the implications forglobal changes index for the whole region would be impossible, in the planet? The sections on large-scale simula- due to its size, diversity, and local and regional tions and the local or regional changes in climate biogeophysical characteristics. (b) A much stron- deal with this problem and try to give answers ger data base is clearlyneeded.The bottleneck to based on tendencies. The possible changes in the any proposition about sustainability is clearly the atmosphere produced by forest burning are dis- scarcity of continuous data in sub-basins and cussed,particularly the effect of chemical compo- small watersheds. Any efficient decision about sition and the energy balance on tropical circula- the occupation and management of the Amazon tions. More information is clearly needed before basin has to be based on a reliable, continuous, any realistic scenario can be made. and strong data bank. The hydroelectric power plant in Amazonia produces many changes because it is a source of methane and other gases to the atmosphere. It Other remarks also has a considerable impact on the quality and regional cycle of water, mainly on the Balbina and There is no doubt that more research is needed on Samuel reservoirs. Pattermns of land use changed the hydrologic cycle, not only on a regional and extensam ely reserv Ptternsaofnlandas ue haned. continental scale in the tropics but also in selected representative local watersheds. The quantifica- Satellite imagery indicates that these dams tion of this cycle is an important problem, some- have a short life cycle due to the increase of what neglected in tropical regions. For example, suspended material carried to the reservoir after the key role of forests in the water cycle needs to intensive deforestation. It remains to be deter- be known for selected watersheds, especially for mined what would be the ideal pattern for con- different types of vegetation. 429 Defining and Measuring Sustainability: The Biogeophysical Foundations The integration of natural processes by the scale, the physiological responses of the vegeta- river basins is an important concept in the defini- tion are also important. The physiology of the tionofsustainability.Thequestionposedregard- vegetation in the Amazon region is one of the ing land use and interference with the cycle of important links among the physical processes, water and of carbon and organic matter is funda- the biogeochemical dynamics, and the commu- mental. One particular problem in this cycle is the nity structure. storage and retention time of water in several subsystems (river, floodplain, lakes, wetlands). The Amazonian region is very dynamic. One References main forcing function is the fluctuation in the level of waterthat sets up seasonal patternsand to Tundisi, Jose G. 1994. "Tropical South America: some extent regulates the chemical and biological Present and Perspectives." In R. Margalef, ed., response of the lotic and lentic systems. The bio- Limnology Now: A Paradigm of Planetary Prob- logical diversity is dependent not only on the lems. yearly, physical changes such as fluctuations in Tundisi, Jose, T. Matsumura-Tundisi, and M. C. the level of water but also on periodic shifts of the Calijuri. 1992. "Limnology and Management river patterns and drainage, which produce of Reservoirs in Brazil." In M. Straskraba,Jose patches of isolated forests and oxbow lakes, remi- G. Tundisi, and A. Duncan, eds., A Comparative niscent of river meanders. Reservoir Limnology and Water Quality Manage- The definition of the main functional groups of ment. Dordrech: Kluwer. vegetation is an important output, but on a finer 430 Limitations in Measuring Sustainability Richard A. Carpenter Sustainability, as in sustainable development, I important to correct these expectations, particu- take to mean the continuation of the potential for larly in international development assistance production from, and use of, mnanaged ecosys- agencies, where sustainable development is now tems. Demand for food, fiber, and other natural a catchphrase. These agencies should support products is increasing so that production and use research and monitoring to improve measure- of ecosystems everywhere are at a high level. ments of sustainability if their own projects are to Management practices to get the most out of these be examples of sustainable development. renewable natural resources (soil, water, plants, This chapter presents evidence of the limita- and animals) may overwhelm their resilience and tions in measuring sustainability from a wide self-maintenance capabilities. And yet, forgoing varietyof managedecosystemsaround the world. too great a margin of production in order to In sum, the following excerpts and abstracts con- diminish risk of degradation may worsen the lot vincingly document the severe limitations and of poor, needy people. inadequacy of biophysical measurements and It is urgently necessary to be able to measure indicators of sustainability, on which important and predict the sustainability of alternative strat- decisions and econometric analysis mustbebased. egies in agriculture, forestry, fisheries, and other managed ecosystems. Year-to-year data on har- vests are obviously the most common and avail- Relevance to policy and management able measure. But yields may be maintained, and degradation obscured, by increasing inputs such Asian Development Bank. 1990. Economic Policies as fertilizer, technology, and improved varieties for Sustainable Development. Manila: Asian Devel- of plants or animals. When decreasing yield is the opment Bank. first or only sign of unsustainability, then some The Asian Development Bank recently spon- damage to the fundamental potential of the eco- sored studies in seven countries to determnine system may already have occurred and may be how sustainable development might be imple- difficult to repair. mented in practical programs. Natural variations are often quite large com- Inthepreparationofthisreportnotaskproved pared with mean values of measured parameters. pared with mean values of meaSUTed paramers. more daunting than the assembly of reliable This means that the ratio of signal to noise is statistical indicators of recent trends and the lowthatitisdifficulttodetectwhethersomechange current state of the environment. in an ecosystem is due to humnan intervention or not. These limitations in measuring sustainability * Aside from the most glaring cases where offi- are not well understood by many officials and cials and researchers are aware of what is policymakers in economic development, who happening and can describe conditions in gen- assume a degree of certainty and understanding eral terms most eloquently, the lack of quanti- that is not justified by the state of the science. It is tative environmental information comparable Defining and Measuring Sustainability: The Biogeophysical Foundations withthestatisticsavailableregardingeconon-dc action more certain. . . . Problems such as parameters is a major obstacle to integrated enormous loss of topsoil, prospects of global economic-cum-environmental planning. warming, and storage of hazardous waste * Statisticalinformationontheenvironmentisscarce, materials, will very likely have such severe often inaccurate, seldom comparablefrom country consequences that by the time the information to country, and rarely available in a time series base is adequate for the construction of a ro- covering a sufficient number of years to indicate bust predictivemodel,itmaybetoolate to take trends in a reliable way. Thus, descriptions re- corrective action. main anecdotal and lack the hard edgeof quan- tification which is necessary for analysis and Carpenter, S. R. 1990. "Large-Scale Perturbations: policy formulation.... Opportunities for Innovation?" Ecology 71, pp. * Without question one of the most important 2038i43. findings of this study is that while reliabledata It is often impossible to replicate ecosystem on conventional economiic parameters are plen- experiments or otherwise to establish a normal tiful, statistical information on the condition of range of variation. the environment is scarce and poorly orga- * The variability of community and ecosystem nized [emphasis added] . variates may be so great that experiments with only two replicates cannot detect perturbation Brussard, P. F. 1991. "The Role of Ecology in effects unless they are very large. . . . The Biological Conservation." Ecological Applications intrinsic variability of ecosystems may be so 1, pp. 6-12. large that rather powerful manipulations * It is usually taken as a sine qua non that sen- would be needed to detect responses even if sible management begins with a solid, funda- experiments could be replicated. mental understanding of a species' ecological Carter, G. C., and B. I. Diamondstone. 1990. relations and natural history. Unfortunately Directions or Internationall Con m atible Envi- we are woefully short on this information for rons for newaYork: Comphere Pub- mos spcis o cosevatonconcern ............... Fi- ronmental Data. New York: Chemisphere Pub- most species of conservation cishiegrCrporation nally, the solid, underlying science necessary lMshing Corporation. for developing appropriate conceptual mod- Monitoring programs are poorly coordinated, els for biodiversity management is generally according to this report of a recent workshop on lacking; even worse, what is known is rarely environmental data measurement and use: applied. * This results in significant variations in sam- pling methods, measurement standards, qual- Cairns,J. 1991. "TheNeed for Integrated Environ- ity control, and procedures to evaluate the mental Systems Management." In J.Cairnsand T. quality of the documentation of the data, and V. Crawford, eds., Integrated Environmental Man- often makes it impossible to assess the quality agement. Chelsea, Mich.: Lewis Publishers, Inc. of the measurements. It is often impossible to Proactive or preventive measures are desir- integrate independently collected data sets for able to maintain the environment in good condi- problem solving because of the great difficul- tion. Changes in management practice, however, ties in reconciling one data set with another. are difficult to justify. Often, data sets may turn out to havelittleor no * The uncertainty of the outcome is often unac- value, especially for third-party use in inter- ceptably high. Predicting the precise environ- disciplinary problems. mental benefits .. . is a very riskybusiness. This * Quality control and data compatibility are re- is one of the reasons there isa reluctance to take quired forall measurements,but theserequire- any action. It is presumed that [making big mentsareespeciallydifficulttosatisfyinbiota, management changes] would be very costly due to the intrinsic nonuniformity of intraspe- and the biological benefits not entirely clear. cific characteristics, the uncontrollable differ- The same thing is true of efforts to restore the ences in test environments, and numerous dif- tropical rain forests. In both situations, the ficulties related to the sampling process. problem is that the science of ecology still does . In trying to estimate total exposure, the largest not have the robust predictive models needed uncertainties stem from the lack of good mea- to make the outcome of a particular course of surement systematics for biota, poorly con- 432 Limitations in Measuring Sustainability trolled procedures for the analysis of biologi- Liverman, D. M., M. E. Hanson, B. J. Brown, and cal tissues, as well as a dearth of relevant R. W. Merideth, Jr. 1988. "Global Sustainability: models.... Toward Measurement." Environmental Manage- To date, no single internationally adopted set ment 12, pp. 133-43. of standards exists for statistical concepts such Several organizations compile and publish as "bias," "precision," and "limit of detection." extensive data bases on environmental and social Work must proceed towards agreement on measures, often as appendixes to documents that these measures so that numerical data bases expandonorinterpretthedata: the World Bank's can be based on a single set of definitions. World Development Report; the World Resources Institute's series; the Organization for Economic Cocklin, C. R. 1989. "Methodological Problems in Cooperation and Development's State of the Envi- Evaluating Sustainability." Environmental Con- ronment; the Worldwatch Institute's State of the servation 16, pp. 343-51. World; and the Population Reference Bureau's New Zealand has embodied sustainability in World Population Data Sheet. its laws, according to Cocklin, but the evaluation * Although theserecent contributions imply that of particular initiatives for sustainable develop- progress is being made in the development ment has not been adequately addressed. and critical analysis of sustainability indica- * The problems are magnified in the context of tors, in many cases the existing or proposed sustainability. Most simply, how do we in fact indicators are not the most sensitive or useful measure whether a system is sustainable or measures. not? In principle and in certain applied cases, it can be a relatively easy task to establish the Magnuson, J. J. 1990. "Long-term Ecological Re- sustainability of individual resources or sys- search and the Invisible Present. Uncovering the tems. Harvesting models for renewable re- Processes Hidden Because They Occur Slowly or sources, such as fisheries, forests,and aquifers, Because Effects Lag Years Behind Causes." bear testimony to this. But if we adopt the BioScience 40, pp. 495-01. holistic interpretation of sustainability, it will * As with observational studies, field experi- almost certainly prove impossible to define ments also can be susceptible to serious misin- any single measure of sustainability at a gen- terpretation if they are not conducted in the eral-system level. context of long-term ecological research. For example, only with 132 years of records ConsultativeGrouponlnternational Agricultural has the general warming trend been detected in Research (CGIAR). 1990. Sustainable Agricultural the southern ocean oscillation index and corre- Production: Implications for International Agricul- lated with ice cover. tural Research. FAO Research and Technology Paper 4. Rome: Food and Agriculture Organiza- Sawhill, John C. 1991. "Into the Future." Nature tion of the United Nations. Conservancy (November-December), p. 9. The Technical Advisory Committee of CG IAR The Nature Conservancy acquires ecosystems, has characterized the problems of sustainability often of large scale, in order to protect them from in agricultural systems. anthropogenic impacts. But, is effective conser- * At the level of the production system, mea- vation on this scale practical or even possible? surements can be made on a continuing basis * In short we need to learn enough about how during the course of routine experimentation. ecosystems function to improve land] preserve The question arises, however, of the extent to design and to intervene successfully in man- which the wider aspects of sustainability might agement. Similarly, we need to develop new be quantified, using a judicious combination ways of measuring success. No one's tried of theoretical considerations, experimentation, conservation on this scale before, and the meth- and modelling. TAC does not consider that ods for monitoring progress simply don't exist. Centres should be solely responsible for devel- opingmethodologies for thispurpose, but they United Nations Economic Commission for Latin should collaborate with institutions and orga- America and the Caribbean. 1991. Sustainable nizations that specialize in the assessment of Development: Changing Production Patterns, Social agriculturalchangeand environmental impact. Equity, and the Environment. Santiago, Chile. 433 Defining and Measuring Sustainability: The Biogeophysical Foundations Understanding, and even monitoring, of how * Research programs exist to develop specific natural systems function is so inadequate that sustainable natural resources (e.g., sustainable environmental impacts go unrecognized in forestry or sustainable agriculture). However, decisionmaking. current research efforts are inadequate for dealing There can be no policy of sustainable develop- with sustainable systems that involve multiple re- ment without a more detailed knowledge of sources, multiple ecosystems, and large spatial the acceptable limits of exploitation of ecosys- scales. Moreover, much of the current research tems. The establishment of these standards is focuses on commodity-based managed sys- essential to development and is even more tems, with little attention paid to the important in those cases where direct inter- sustainability of natural ecosystems whose vention is selected. However, one of the main goods and services currently lack a market problems in the region is the lack of sufficient value. information to establishadequate environmen- * [Current] efforts are not presently united in a tal standards. Incentives for scientific and tech- comprehensive research framework. Such a nological research in all relevant fields are framework is needed because ecological pro- needed if this situation is to be improved.... cesses link natural and managed populations There is substantial evidence that marine and to ecosystemsand because common ecological coastal resources are being mismanaged and principles underlie effective management abused. All too often, the decisions taken in strategies. respect of these resources are dictated by nar- row, short-term interests. Such decisions are Lovejoy, T. E. "Diverse Considerations." In E. 0. usually taken without having a full scientific Wilson, ed., Biodiversity,pp.421-27. Washington, knowledge of the potential long-term adverse D.C.: National Academy Press. effects, or without even caring about those [There are major] limitations deriving from effects.*[hr r aollmttosdnlgfo our relatively shallow knowledge of flora and fauna . . . we do not even know the extent of Health of ecosystems biological diversity on our planet to the near- est order of magnitude.... biologists can say relatively little about which species occur Cairns, J., and B. R. Niederlehner. 1989. "Adapta- where, which are in dangerof extinction, where tion and Resistance of Ecosystems to Stress: A protected areas should be established, and Major Knowledge Gap in Understanding An- where heavy environmental modification for t hr opogenic Perturbations." Speclations in Sci- development is permissible .... If we want to ence and Technology 12, pp. 23-30. perpetuate the dream that we are in charge of * The capacity of ecosystems to adapt to anthro- our destiny and that of our planet, it can only pogenic stress is presently poorly understood. be [by] maintaining biological diversity-not Unfortunately, therearefew places in theworld by destroying it. where human activities do not have a major influence on natural systems. The capacity of Mooney, H. A. 1991. "Emergence of the Study of natural systems to generalize an adaptation Global Ecology: Is Terrestrial Ecology an Impedi- from one stress to new stresses is virtually ment to Progress?" EcologicalApplications 1,pp.2-5. unknown. ... Terrestrial ecology lacks fundamental knowl- * Mechanisms of adaptation at the community edge and even a plan to acquire such knowledge. level and the influence of the type of stress, * Whyisitthatweareperceivedtolackacoherent either common in evolutionary history, such body of knowledge, and effort, that will readily asorganicenrichment, or unprecedented, such interface with the current informnation on the as synthetic pesticides, are only partially un- physicaldriversandresponderstoglobalchange? derstood, and their reversibility is not assured once the stress is removed. O'Neill, R. V., C. Hunsaker, and D. Levine. 1990. "Monitoring Challenges and Innovative Ideas." Ecological Society of America. 1991. "The Sus- U.S. Environmental Protection Agency Interna- tainable Biosphere Initiative: An Ecological Re- tional Symposium on Ecological Indicators, Oc- search Agenda." Ecology 72, pp. 371-412. tober 15-19,1990, Ft. Lauderdale, Florida. 434 Limitations in Measuring Sustainability Ecosystems are complex, and it is difficult to yields or biomass yields per hectare, and net predetermine what aspects of system structure value added to production. of dynamics will respond to a specific insult. It is equally difficult to interpret whether a response Henderson, C. 1987. "Famines, Droughts, and the is a stabilizing compensatory mechanism or a 'Norm' in Arid Western Rajasthan: Problems of real loss of capacity to maintain the ecosystem.. Modeling Environmental Variability." Research . . The problems are compounded in broad moni- in Economic Anthropology 9, pp. 251-80. toring programs designed to assess ecosystem * Despite its enormous impact on such popula- "health" at regional and continental scales. It is tions, the problem of environmental variabil- challenging in the extreme to monitor ecosystem ity has received little sustained attention in response, at any scale, to past insults as well asan anthropology... This situation reflects the fact unknown future array of impacts. that research is often carried out within a single year's time. Most descriptions of production outline the events of an "average" or "normal" Agriculture year, under the assumption that the knowledge of a single annual cycle is sufficient to under- Carpenter, R. A., and D. E. Harper. 1989. "To- stand the relationshipsbetween local producers' wards a Science of Sustainable Upland Manage- activities and their environmental parameters. ment in Developing Countries." Environmental * A key heuristic device assumed in many eco- Management 13, pp. 43-54. logical models is that of cyclical resonance: * A . . . hypothesis is that statistically reliable resources vary around some mean or a state data can be obtained from experiments in up- that is characterized as "normal." . . . Quite land situations, although natural variations of simply, there is no such thing as a cyclical soils,weather,andvegetationaregreat .... The pattern that can be identified by examining objective of the work is to provide credible differences in annual rainfall totals. quantitative information to help policy and decision makers and resident farmers to plan Nickum, J. E. 1989. "Volatile Waters: Is China's and implement improved practices based on Irrigation in Decline?" East-West Center Work- ecological principles . . . The signal-to-noise ing Paper. East-West Center, Honolulu, Hawaii. ratio in these field experiments is low, and the More often than not messages from China's detection of changes due to human interven- irrigation front this past decade have been dire tion in soil erosion, nutrient movement, and reportsofinadequatesupport,poormorale,and plant productivity is difficult. lostbattles.IsChina'sirrigationsectorintrouble, Consultative Group onlInternational Agricultural beset by hostile disbursers and myopically indif- Research (CGIAR). 1990. "Report of the Commit- ferent farmers? It is not at all clear that it is. tee on Sustainable Agriculture. Consultative * What is clear is that the aggregate body counts Group Meeting May 21-25, 1990." The Hague, of irrigated [land] must obscure more than Netherlands. they inform about a very complex, dynamic, * Absolute and universal measurement of and diverse set of conditions. sustainability is difficult and probably liesout- Rapport, D. J. 1990. The Use of Indicators to Assess side CGIARC responsibilities. No single mdi- the State of Health of Ecosystems: An Historical Over- cator is likely to incorporate many normative view. Notebook of the Intemational Symposium judgments that are difficult to quantify-such on Ecological Indicators. Environmental Man- as the reversibility of degradation, the critical agement AssessmentProgram, U.S. Environmen- threshold of decline, or the level of diversity tal Protection Agency, Washington, D.C. necessary to protect the future genetic base of agriculture. * Historical references to degradation of agricul- tural lands and forests can be found in the * Severalquantifiableindicatorstakenovertime writings of Plato and even earlier. Yet many can provide data along crucial dimensions that centuries later, we are far from consensus as to help to indicate the sustainability of most agri- the identity of a minimal but sufficient set of cultural production systems. These include, indicators by which to measure changes in the especially, soil organic matter, soil acidity, crop state of nature. 435 Defining and Measuring Sustainability: The Biogeophysical Foundations Rerkasem, K., and A. T. Rambo. 1988. tages of tropical grassland with respect to soil Agroecosystem Research for Rural Development. erosion and streamflow levels .... What do we Chiang Mail, Thailand: Multiple Cropping Cen- really know about the hydrological properties tre, Faculty of Agriculture, Chiang Mai Univer- of tropical grasslands? More specifically, how sity and SUAN. do these properties vary with different levels Sustainability is an emergent property of of burning and grazing, and how do these agroecosystems and is not identical to compare with properties of cultivated or for- sustainability as meant by ecologists in natural ested land? ecosystems. * Unlike natural ecosystems, agricultural eco- Binns, T. 1990. "Is Desertification a Myth?" Geog- systems are purposive; they are managed by raphy 15, pp. 106-13. people to achieve socially defined objectives * Over the years the definition of desertification and their emergent properties are defined in has moved from the "expansion of desert-like terms of their relationship to meeting these conditions" to "a process of sustained land objectives... . Sustainability is not a measure- (soilandvegetation)degradationinarid,semi- ment of the ability of the agroecosystem to arid, and dry sub-humid areas, caused at least persist over time on its own but instead refers partly by man," and reducing productive po- to its ability to persist with an acceptable level tential toanextent whichcan neitherbereadily of human inputs such as labor, fertilizer, or reversed by removing the cause nor easily pesticides. reclaimed without substantial investment.... A further problem revealed by studies of Walters, C. J., and C. S. Holling. 1990. "Large- drought, degradation, and desertification is a scale Management Experiments and Learning by lack of reliable statistics over any considerable Doing." Ecology 71, pp. 2060-68. length of time ... data needed to classify land * In no place can we claim to predict with cer- are available for very few areas and for very tainty either the ecological effects of the activi- few years, and in most of Africa little is known ties, or the efficacy of most measures aimed at about rangecondition, crop yield, or the extent regulating or enhancing them. of soil erosion. * Every major change in harvesting rates and Blaikie, P. 1989. "Environment and Access to management policies is in fact a perturbation Resources in Africa." Africa 59, pp. 18-40. experiment with highly uncertain outcome, no matter how skillful the management agency is . In the area of understanding and evaluating in marshallingevidenceand arguments in sup- environmental degradation in Africa, the fol- port of the change. lowing causes of uncertainty emerge. First, there is the problem of data-its scantiness, unreliability, irrelevance, and ambiguity. Sta- Soil tistics are seldom in the right form, are hard to come by, and even harder to believe let alone Andrus, C. 1986. "Soil Erosion and Streamflow interpret. from Tropical Grasslands: How Much Do We Really Know?" East-West Center Working Paper. Faeth, P., R. Repetto, K. Kroll, Q. Dai, and G. East-West Center, Honolulu, Hawaii. Helmers. 1991. "Paying the Farm Bill: U.S. Agri- cultural Policy and the Transition to Sustainable * Land clearing and fire have transformed large Agriculture." Washington, D.C.: World Resources areas of tropical forest in Asia and the Pacific Institute. into grassland. .. . Government agencies in a *In the field, erosion-induced productivity number of Asian and Pacific countries view*.. changes are almost impossible to isolate and grasslands as undesirable and have adopted measure accurately.... Because there is no measures to halt expansion of grasslands and satisfactory methodology for separating the convert portions of the area back to trees . .. interacting effects of many factors on crop [Clonflicts have spawned a rash of statements yields, soil productivity declines due to soil concerning the relative merits or disadvan- erosion can be easily masked. 436 Limitations in Measuring Sustainability Forestry knowledge of the internal functioning of the old-growth ecosystem. Bruijnzeel, L. A. 1990. "Summary and Conclu- sions." In Hydrology of Moist Tropical Forests and Leopold, L. B. 1990. Ethos, Equity, and the Water Effects of Conversion: A State of Knowledge Review, Resource. Washington, D.C.: U.S. Forest Service. pp. 175-84. Netherlands: IHP Comnittee. * The Forest Service advertises its dedication to * Of the two main components of ET [evapo- multiple use. But its research gives no assur- transpiration], rainfall interception has fre- ancethatitspolicyofclear-cuttingfollowedby quently been overestimated because of inad- monoculturewillresultinsustainability .... To equate sampling designs. The second major my knowledge no Forest Service research is component(transpiration)isoftenonlyknown aimed at evaluating over the long-term the indirectly and unreliably.... The information effect of changing a multi-storied, mixed-spe presented in this report leads to the observa- cies stand to an even-aged, single species for- tion that the adverse environmental condi- est. . . . There has been no attempt to make tions so often observed following "defores- measurements to test the validity of this as- tation" in the humid tropics are not so much sumption. the result of deforestation per se but rather of poor land use practices after clearing of the McNitt,B.1991."DiscussingFreeTradeandTropi- forest. cal Forests." Conservation Exchange 9, pp. 5-8. Dover, M., and L. M. Talbot. 1987. "To Feed the * The problem is that no one knows what sus- Earth: Agro-Ecology for Sustainable,Development." tainable management of the tropical forest is. WoldRsorcsIntiue,Wah n.C. No one knows how to do it. Sustainable timber World Resources Institute, Washington, D.C. extraction from these forests is still a concept, * If diversity causes stability, the most species- not a reality (Guillermo Castilleja, resource rich communities such as tropical rain for- specialist with the National Wildlife ests and coral reefs-should be able to with- Federation's International Department). stand the greatest disruption at human hands. In fact, these communities are among the most Metz, J. J. 1991. "Vegetation Assessment and Re- fragile. search Methods for Community Forestry in Nepal." East-West CenterWorking Paper 27. East- Harris, L. D. 1984. The Fragmented Forest: Island West Center, Honolulu, Hawaii. Biogeography Theory and the Preservation of Biotic * A ... reason to abandon the use of sustainable Diversity. Chicago, Ill.: University of Chicago yield calculations is that the data on which Press. such estimates can be based do not exist. Esti- * The actual requirements of individual species, mates of biomass and productivity are based populations, and communities have seldom on exhaustive, destructive samplings of repre- been known, nor has the available information sentative examples of major forest types. always been employed in site selection and * Not only are the number of such studies com- planning for nature reserves. The use of lands pleted in Nepal an inadequate data base, but surrounding nature preserves has typically this work is based on and hence only valid for been inimical to conservation, since it has usu- forests with undamaged trees. ally involved heavy use of pesticides, indus- trial development, and the presence of human Talbot, L. M. 1990. "A Proposal for the World settlements in which fire, hunting, and fire- Bank's Policy and Strategy for Tropical Moist wood gathering feature as elements of the local Forests in Africa." Draft for discussion. World economy. Resources Institute, Washington, D.C. * Thus, although a great deal of scientific * Increasingly, however, scientists are question- information was available, it was not in a form ing whether sustainability of commercial log- readily usable for comprehensive planning, ging in natural tropical moist forests has ever nor was it clear that the scheduling of timber been demonstrated, and whether it is, indeed, operations in hundreds of districts and ever possible in other than plantation type thousands of old-growth tracts would be situations.... [Environmentalists] point to the greatly or immediately improved by increased unquestionable loss of tropical forests follow- 437 Defining and Measuring Sustainability: The Biogeophysical Foundations ing lumbering operations and they say that the tional fisheries, will be reached well before the claim of "sustainability" is a smoke screen to end of the century. However, this progressive cover destruction of irreplaceable forests for increase conceals great variability of natural financial gain. resources and many problems. * If environmental assessments are to be valid, the chemical measurements on which they are Water based must be reliable and adequate for their intended use. This is specially important where Birkeland,C. 1990. "Caribbean and Pacific Coastal the information is to be used for decision mak- Marine Systems: Similarities and Differences." ing and the enforcement of regulations or for Nature and Resources 26, pp. 3-12. legal purposes. Productivity can be high in areas of low nutri- * It has recently become clear that many chemni- ent input because nutrients are perpetually cal measurements in the sea made more than recycled at several levels: microbiologically, ten years ago are dubious, makingitdifficult to physiologically (between the animal and the establish time trends.... There is a growing plant tissues of symbionts), ecologically (in the awareness of the need to validate measure- detrital foodweb and in other interactions be- ments and to be confident of their reliability tween trophic levels), hydrodynamically (by and adequacy for environmental assessment. eddies, gyres, high residence time of water in lagoons, nd otherenclosedcoastal eatures) * There is a parallel and equally important need lagoons, and other enclosed coastal features), fr silaquitcorlofblgcldt. and through retention of nutrients from the foresmi, la qualitycconrolsofebiolgicalvdata water column to the reef surface by active 'dwverse asbinherentlymsreenariemoh current producing suspension-feeders such as diverse and inherently more variable than sponges .p. . chemical measurements, it is more difficult to Asponutrient input increases on coral reefs up to distinguish errors due to sampling from those * As nutrient input increases on coral reefs up to due to analytical procedures. Some biological a point, the growth of benthic vegetation and methodsarewell standardised,butotherslare] rates of primary productivity increase. This rather idiosyncratic. can sometimes lead to overgrowth and even mortality of corals. Pollution of shallow ma- * In addition, the inherent variability of popula- rine habitats by eutrophication (nutrient in- tions and communities leads to a diversity of put) leads to dominance by benthic algae. Long- techniques and many biological data are "snap- term users of the Great Barrier Reef have noted shots" of dynamic and variable parameters an increase in algal cover and a decrease in rather than determinations aimed at absolute, coral cover attributed to increased urbaniza- or even relative, values of stable characteris- tion, industrialization, and agricultural devel- tics. This makes comparison between different opment in relatively heavily populated areas data sets difficult, unlesscomparable methods on the Queensland coast. The mechanism by have been used and statistical limits to the which this operates may be increased nutrient estimates obtained. input by sewage or by terrestrial runoff from * At a stage where it is necessary to look for baring the soil and exposing the coast to erosion. relationshipsbetweenenvironmentalfactorsand biological responses, thislackofsystematicqual- Group of Experts on the Scientific Aspects of ity control for the biological components of an Marine Pollution (GESAMP). 1990. "The State of investigation isinconsistent with thecritical scru- the Marine Environment." Regional Seas Reports tiny given to chemical observations, and detracts and Studies 115. United Nations Environment fromthevalueofthederivedrelationships.Ma- Program, Nairobi. jor efforts are needed to improve this situation. * The most recently reported [global fisheries * The collection of biological material for analy- catch], for 1987, achieved a new record of 92.7 sis of contaminant content is rarely under- million tonnes, and preliminary figures for taken with sufficient safeguards to ensure that 1988 indicate a further increase to 94 million the samples are representative of the popula- tonnes. It is nowexpected that the figure of 100 tion. ... The accurate identification of species million tonnes, which many believed to be the is still a significant problem in some biological maximum sustainable global yield of conven- samples, and there are contentious attributions 438 Limitations in Measuring Sustainability and difficulties .. . [Liow levels of contamina- on Global Warming and Sustainable Develop- tion could build up insidiously in the sea with ment, International Center for Living Aquatic subtle effects causing damage to wide areas in Resources Management, Bangkok, June 10-12, the long term. 1990. International Center for Living Aquatic * Becauseofthedifficultyofrecognizingchanges Resources Management, Manila, Philippines. of this kind against the background of natural * Accurate catch data are probably severely un- variability, they can be studied only indirectly derestimated as recent data from the Philip- through a combination of experimental ap- pines suggest that even if accurate data are proaches (both in the laboratory and in obtained on marketable catch (which is sel- mesocosms), field surveys, and modelling of dom the case) there isstillanother67percentof process dynamics. the catch that is consumed or processed lo- cally. Furthermore, many different species are Hofer, T. 1990. Deforestation: Changing Discharge involved in reef fisheries, about sixty speciesin and Increasing Floods: Myth or Reality? Berne, Swit- the Philippines. Accurate data on their capture zerland: University of Berne. and details on their interactions are not avail- * Climatological data are easily available. Hy- able.Theunderlyingproductivesystemincoral drological infornation is restricted, not only reef communities is also very complex. It is for foreigners but also for Indian scientists. expected that global climate change will have This data situation makes research in India adverse effects on coral reef systems through verydifficult.... [Tlhereisnoevidencethatthe sea level rise, sea temperature increase, and flooded area in the Gangetic Plain has been increased ultraviolet radiation. ICLARM is increasing over the last decades. . . The deci- working with Philippine researchers to model sive statement of the Chenab investigation is coral reef fisheries systems, but we concluded representative for all the four analyzed river that there was little scope for research on the systems.... It is not possible to identify signifi- effect of global climate change on coral reefs. cant changes in the discharge characteristics The data base is weak, and there is no strong caused by eventual man-induced ecological environmental signal thatlinkscoral reefsand degradation of the watershed ... . Either the climatechange.Furthermore,thecurrentthreat trends are statistically not significant or they of siltation from terrestrial agricultureand for- are parallel for precipitation and discharge. estry mismanagement and habitat destruction by harmful fishing methods (cyanide, pesti- Karr, J. R. 1991. "Biological Integrity: A Long- cides, and dynamite) may destroy the coral neglected Aspect of Water Resource Manage- areas well before the effects of global climate ment." Ecological Applications 1, pp. 66-84. change are apparent. Current estimates sug- * Although perception of biological degrada- gest that 75 percent of the Philippine coral reef areas have been destroyed or severely de- tion stimulated current state and federal legis- grad Cal reefsaref th sevtreats lation on the quality of water resources, that graded Coral reefs are facing the same threats biological focuswas lost in the search for easily as tropical forests. measured physical and chemical surrogates. * Predictions on the consequences of global cli- The "fishable and swimmable" goal for the mate change for developing countries should Water Pollution Control Act of 1972 .. . and its only be undertaken where there are good data charge to "restore and maintain" biotic integ- bases and strong signals between climate rity illustrate that law's biological underpin- change and the production system. In most ning. Further, the need for operational defini- cases the first step to research the consequences tions of terms like "biological integrity" and of climate change on fisheries will be to de- "unreasonable degradation" and for ecologi- velop suitable data on local production. There cally sound tools to measure divergence from maybe a need to setupa network of long-term societal goals have increased interest in bio- benchmark sitestoallow foranexaminationof logical monitoring. changes in fish populations and in the aquatic environment over time. MacKay, Kenneth Tod. 1991. "Global Warming, Fisheries, and Policy for Sustainable Develop- Morrison, R. J. 1991. Assessment and Control of ment." Presented at the International Conference Marine Pollution in the South Pacific Islands. Pro- 439 Defining and Measuring Sustainability: The Biogeophysical Foundations ceedings of the Pacific Science Congress, Hono- * [Despite] three decades of intense activity of lulu, Hawaii, May 27-June 1, 1991. Fiji: Univer- protection and restoration, monitoring, public sity of the South Pacific Suva. health, measuring microbial indicators, . . . The knowledge gaps that have been identified most environmental monitoring programs fail include the following: to provide the information needed to under- a. A lack of long-term data to facilitate the stand the condition of the marine environment of temporal and spatial trends. or to assess the effects of human activity on it. v arc .gton Monitoring is generally not well coupled with There are one or two notable exceptions which eerhporm n eindt mrv are discussed urther below.research programs and designed to improve are discussedgefurther below, our of toxic the appropriateness of routine measurements b. A knowledge of the behaviour of toxic and allow interpretation of the implications of contaminants (e.g., pesticides) in critical envi- monitoring results. ronments is unavailable. The marine environment is complex and vari- c. There is no basis for estimating the health able, and it is often difficult to detect, identify, impacts of the microbiological contamination and measure anthropogenic impacts clearly. caused by uncontrolled sewage discharges. These factors, coupled with limitations to sci- d. Few baselines exist for the development entific knowledge, emphasize the need for re- of programmes studying ecosystem changes alistic expectations. as pollution indicators. * [R]isk-free decision making is not possible. e. Apart from certain fish species, informa- When well developed, applied, and used, en- tion on the extent of use of marine resources is vironmental monitoring can help quantify the lacking. magnitude of uncertainty, thereby reducing but not eliminating uncertainty in decision National Research Council. 1990. Managing making. Troubled Waters: The Role of Marine Environmental Monitoring. Washington, D.C.: National Acad- Regier, H. A. 1992. "Indicators of Ecosystem In- emy Press. tegrity." In D. Mackenzie, ed., Proceedings of the * [Miore than 133 million is spent annually on International Symposium on Ecological Indicators, Ft. monitoring programs in the U.S. to acquire Lauderdale, Florida, September 1990, pp.183-200. Bark- information for environmental management ing, England: Elsevier Applied Science Publishers. decisions and ultimately to ensure protection ... * Our new indicators will begin to have practical The general perception is that the costs of consequences early in the next century since a monitoring programs, as currently conducted, data series is not likely to be invested with much often exceed their utility and benefit. credibility until it is at least ten years long. 440 Distributors of World Bank Publications ARGENTINA The Middle EastObserver KOREA, REPURLIC OF SOUTrH AFRICA, BOTSWANA Carlos Hirsch, SRL 41, Sherif Street Pan Korea Book Corporation Forseaglk WIw Galert Guemes Cairo P0D. BOX 101, Kwangwhanuzn Oxford University Prenss Florida 165, 4th Floor-Ofc. 453/465 Seoull Southemn Africa 1333 Buenos Aime FINLAND P.O. Box 1141 Akuteeminen KIClrakauppa Korea Stock Book Centre Cape Town NM1 Oficinsa del Llbmo Internaclonal P.O. Box 128 PJD. Bosx 34 Albentl40 SF-00101 HeLsInkJIO Yeoeido Fonumiscriplim wardm 1082 Buenos Aime Seoul Intemnational Suabsaiption Service FRANCE P.O. Box 41095 AUSTRALIA, PAPUA NEW GUINEA. World Bank Publications MALAYSIA Crawlgsl FI1JI, SOLOMON ISLANDS, 66, avenue d'IEn%a University of Malay. Cooperativ Johannesburg 2024 VANUATU, AND WESTERN SAMOA 75116 Paris Booliahop, Limited D.A. Informiation Services P.O.wB 1127, jalan Pantai Baru SPAIN 648 Whitehorse Road GERMANY 59700 Kuala Luanipur Murdi-Prensa Libros, S.A. Mrtcham 3132 UNO-VerLag CasteIo 37 Victoria Poppelsdorfer Allee 55 MUXICO 28001 Madrid 531 15 Bonn ENMTSC AUSTRIA Apartado Peasl 224160 UJbreria Internacdonal AEDOS Gerold and Co. GHANA 14060 Thlpan, Mexico D.iE Consell de Cent 391 Graben 31 Greenwich Mag.snrd Books 0800 Barcelona A-tOll Wien Rivera Beach Hotle NETHERLANDS PO Boa 01 198 De LLndeboomn/Inr-Publlkatlea SRI LANKA AND THE MALDIVYES BANGLADESH Osw-Accas P.O.Bo 202 Lake HouseBookshop Micro Industries Development 7lh0AEHaakabergen P.O. Box 244 Assistance Society MIDAS) GREECE 100. SirChfttampalamA. HouseS5, Road 16 PapasotiriouS.A. NEW ZEALAND GardinerMawatha Dhanmrondr R/Area 35, Stoumrara Str. EB90DNZUld. Colootbo 2 Dhaka 1209 106882 Athena Private Mail Bag 99914 New Market SWEDEN BELGIUM HONG KONG, MACAO Auckland FritresFackboksforetaget Jean De Lannoy Asia 2(000 Ltd. Regeringsgatan 12, Bo. 10356 Av. du Roi 202 46-48Wyndham Street NIGERIA S-l 0847 Stockholm 1060 Btnssels Winning Centre University Preslm ited 7th Floor Three Crowns Buildinig Jericho Wennergren-Willlsnns AB BRAZIL Central Hong Kong Private Mail Bag 5095 P.O0. Box 1305 Publicacoes Tecnicas Intemnaclonais LAda. Ibadan S-1 71 25 Solna Rua PeixotoGCoride, 209 HUNGARY 01409 Sao Paulo, SP Foundation for Market Economy NORWAY SWITZERLAND Domnbovari UJt 17-19 Narvesen Informration Center Librairie Payst CANADA H-Ill?7 Budapest Book Departmrnez Case poetalee 3212 Le Diffuscur P0D.5Bo 6125 Etterstad CR 1002 LausanLne l51 ABoul. de Mortagne INDIA N.0602Oslo 6 Boucherville, Qu"bec Allied Plublishers Private Ltd. Van DLernman Editions Techniques.- ADECO 14B 5176 751 Mount Road PAKISTAN P.O. Box 465 Mad ras- 6100002 Mirza Book Agency CHl121 IGeneval16 Re nou f Publishi ng Co. 65, Shahrah-e-Quald-e-Azamn 1294 Algoma Road INDONESIA P.O. Box No. 729 TANZANIA Ottawa, Ontario Pt. Indira LimitdW LahoreS4LOO Oxford University Press Kl B3W8 ja la nBorobud u r20 Maictaba Street P.O.Box 1181 PERU P.O.5Box5299 CHINA Jakarta 10320 Edltorial Desarrollo SA Dares Salaam China Financial & Economic Apantaldo 3824 Publishing House IRAN LIitrs I THAILAND 8. Da Fo Si Dong Jie ICowkab Publishers Central Department Store Beibng P.O. Box 19575-511 PHILIPPINES 306 Silom Road Tehran Internationtal Book Center Bangkok COLOMIBIA Suite 1703, Cltyland 10 Infoenlace Ltda. IRELAND Contdom-inium Tower 1 TRINIDAD & TOBAGO Apartado Aerno 34270 Government Supplies Agency Ayala Aventue, I-.V. dela Systeffatics Studies Unit Bogota D.E. 4-5 Hiarcou n Road Costs Extension #9 Watt Siredt Dublin 2 Makat, Metro Manila Curepe COTE DyIVOIRE Trinidad, West Indies Centre dErdition ritde Diffusion ISRAEL POLAND Atfricaines (CODA) Ycmzrot Literature Ltd. Internationial Publishing Service UGANDA 04 BY.1 541 P.O. Boa 56055 UIl. Ptekna 31/37 Custro Ltd. Abid)an 04 Plateau Tel Aviv 61560 00-677 Warszawa lst Floor, Room 4, Geogiadis Chamnbers P.O. Box 9997 CYPRUS R.O.Y.lInterrnational PORTUGAL Plot (69) Kampala Center of Applied Research P.O.8. 13056 Livsriar Portugai CyprusCollege Tel Aviv 61130 Rus Do Camse 70-74 UNITED KINGDOMI 6, Diogenes Street, Engomli 1200 Lisbon Mlcroinfo Ltd. P.O. Bon 20D6 ITALY P0..Box 3 Ni c os,ia Licosa CommLssionaria SantsonLi SPA SAUDI ARLABIA, QATAR Aiton, Hampshire CU34 2WG Via Duca Di Calabria, I /I jarir Book Store angland CZECH REPUBLIC Casella Postale 552 P.O. Box 3196 National Informnation Center S01 25 Firenze Riyadh 11471 ZAMBIA P.O. Bon 668 University of Zamnbia Bookshop CS- Il137 Pragu e I JAMAICA SLOVAK REI'UBUC Great EastRoad Campus Ian Randle Publishers Ltd. Slovart G.T.G ltd. P.O. Box 32379 DENMARK 206 Old Hope Road Kruplnska4i Lanka Sarnf-ndsLitteratuar Klngton 6 PA).Box iS2 Rse noerns A Ili II 852 99 BratislavaS5 ZIMBABWE DK-1970 Frederiksberg C JAPAN Lonigman Zimbabwe (Put.) Ltd. Eastern Book Serice SINGAPORE,TAIWAN, Tourle Road, Ardbennle DOMINICAN REPUBLIC Hongo 3-horme, Bunkyo-ku 113 MYANM4AR,BRUNEI P.O. Box STI125 rd,tora Taller, C. por A. Tokyo Cower Asia Padife Pte Ltd. Southserto Res,iauraci6n e lsaibe la Cai6lica 309 Golden Wheel Building H-arare Aparrado de Comreos 219011A KENYA 41, Kallang Pudding, #04-ID Santo Domiungo Africa Book Servce (E.A.) Ltd. Singapore 1334 Quaran House, Mfangano Street EGYPT, A RAB REPLTBLIC OF P.O. Box 45245 Al Ahram Naimobi Al Calaa Street Calm The World Bank Headquarters European Office Tokyo Office U 1818 H Street, N.W. 66, avenue d'lena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome Chiyoda-ku, Tokyo 100, Japan Telephone: (202) 477-1234 Telephone: (1) 40.69.30.00 Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.66 Telephone: (3) 3214-5001 Telex: Mc[64145woRLDBANK Telex: 640651 Facsimile: (3) 3214-3657 Mci 248423 WORLDBANK Telex: 26838 Cable Address: NTBAFRAD WASHINGTONDC in conjunction with: 4 I~ The United Nations The United Nations University L University Headquarters UNU Office in North America UNU Office in Europe 53-70 Jingumae, 5-chome (UNUONA) c/o UNESCO Shibuya-ku, Tokyo 150 Room DC2-1462 Rooms 7B 111-113 Japan United Nations 1, Rue Miollis New York, N.Y. 10017 75732 Paris Cedex 15 Telephone: (81) (3) 3499-2811 France Facsimile: (81) (3) 3499-2828 Telephone: (212) 963-6387 Internet: unu@igc.apc.org Facsimile: (212) 371-9454 Telephone: (33) 45 68 30 07/10 Internet: unuona@igc.apc.org Facsimile: (33) (1) 40 65 91 86 1_1LU_ __ 9 80821 347 Cover design by Soleil Associates ISBN 0-8213-3134-5