ENVIRONMENT DEPARTMENT PAPER NO. 96 Environmental Economics Series THE WORLD BANK ENVIRONMENT DEPARTMENT Paying for Biodiversity Conservation Services in Agricultural Landscapes Stefano Pagiola Paola Agostini José Gobbi Cees de Haan Muhammad Ibrahim Enrique Murgueitio Elías Ramírez Mauricio Rosales Juan Pablo Ruíz May 2004 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the Environment Department of the World Bank by calling 202-473-3641. The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Manufactured in the United States of America First printing May 2004 Contents ACKNOWLEDGMENTS V ABBREVIATIONS VII EXECUTIVE SUMMARY IX Chapter 1 Introduction 1 Chapter 2 Silvopastoral Practices 3 On-site benefits 4 Biodiversity benefits 5 Other benefits 5 Chapter 3 Barriers to Adoption 7 Chapter 4 Payments for Environmental Services 11 Chapter 5 From Theory to Practice 13 What is being paid for? 13 How should payments be made? 16 Avoiding perverse incentives 19 Chapter 6 Monitoring Results 21 Changes in land use 21 Impact of land use change on environmental services 21 Impact of the project on participating households 22 Chapter 7 Conclusions 23 Assessing the success of the project 23 Assessing the success of the approach 24 Replicating the approach 24 Environmental Economics Series iii Paying for Biodiversity Conservation Services in Agricultural Landscapes APPENDIXES A -- Land Use Change in Central America and Colombia, 1990­2000 27 B -- Degraded Pastures and Silvopastoral Practices 29 REFERENCES 33 BOX 1 World Bank Support for PES 12 FIGURES 1 Typical time profile of benefits of silvopastoral systems 8 2 Effects of PES on the profitability of silvopastoral systems 19 TABLES 1 Changes in pasture land and forest area in Colombia, Costa Rica, and Nicaragua 3 2 Average land use in farms in Quindío, Colombia 7 3 Initial investment costs for selected silvopastoral practices 8 4 Environmental service indices used in the RISEMP 15 5 Example of payment computation 18 iv Environment Department Papers Acknowledgments The authors are members of the team that those of the World Bank, CATIE, CIPAV, prepared and is implementing the RISEMP Nitlapan, or LEAD-FAO. We would like to project. Stefano Pagiola, Paola Agostini, Cees thank Ken Chomitz, John Kellenberg, Agustin de Haan, and Juan Pablo Ruíz are at the World Arcenas, and Benjamin Kiersch for helpful Bank; José Gobbi and Muhammad Ibrahim at comments and suggestions. An earlier draft of CATIE (Costa Rica); Enrique Murgueitio at this paper was presented at the Third BioEcon CIPAV (Colombia); Elías Ramírez at Nitlapan Workshop, Contract Mechanisms for (Nicaragua); and Mauricio Rosales at LEAD- Biodiversity Conservation, Montpellier, France, FAO. May 22­24, 2003; comments from participants at that workshop are also gratefully The opinions expressed in this paper are the acknowledged. authors' own and do not necessarily reflect Environmental Economics Series v Abbreviations ABC American Bird Conservancy CAPE Cape Action Plan for the Environment CATIE Centro Agronómico Tropical de Investigación y Enseñanza (Center for Teaching and Research on Tropical Agronomy) CIPAV Centro para la Investigación en Sistemas Sostenibles de Producción Agropecuaria (Center for Research on Sustainable Agricultural Production Systems) CRP Conservation Reserve Program EBI Environmental Benefits Index FAO Food and Agriculture Organisation of the United Nations GEF Global Environment Facility LEAD Livestock, Environment and Development Initiative NGO Non-governmental organization PES Payments for environmental services PSA Pago por Servicios Ambientales (Payment for Environmental Services, Costa Rica) PSAH Pago por Servicios Ambientales Hidrológicos (Payment for Hydrological Environmental Services, Mexico) RISEMP Regional Integrated Silvopastoral Ecosystem Management Project WBI World Bank Institute Environmental Economics Series vii Executive Summary Adoption of improved silvopastoral practices Designing the mechanism required addressing in degraded pasture areas is thought to provide issues such as (1) measuring the actual amount valuable local and global environmental of environmental services being provided, so benefits, including biodiversity conservation, that appropriate payments can be made; (2) However, these practices are insufficiently providing payments in a way that resulted in attractive to individual land users for them to the desired change in land use; and (3) adopt them spontaneously, particularly due to avoiding the creation of perverse incentives their high initial costs. This paper describes the (for example, for land users to cut down contract mechanism developed for the existing trees so as to qualify for additional Regional Integrated Silvopastoral Ecosystem payments for tree planting). Two variants of Management Project, which is being the proposed payment mechanism are being implemented with financing from the Global tested, with participating land users assigned Environment Facility (GEF). The project is randomly to one or the other. The project also testing the use of the payment-for-service includes extensive monitoring of the mechanism to encourage the adoption of effectiveness of each mechanism in stimulating silvopastoral practices in three countries of adoption of the proposed measures and of the Central and South America: Colombia, Costa resulting impact on environmental services and Rica, and Nicaragua. The project has created a on household welfare. These features, together mechanism that pays land users for the global with the three-country approach, will provide environmental services they are generating, so in the coming years a very rich dataset for that the additional income stream makes the testing the use of contract mechanisms for proposed practices privately profitable. biodiversity conservation. Environmental Economics Series ix 1 Introduction As natural habitats have come to be requires complex certification schemes and is increasingly restricted and degraded, not always feasible. increasing attention has been paid to conserving biodiversity in agricultural A further approach, which has received landscapes. This can be both an end in itself, increasing attention in recent years, is to driven by the realization that agricultural provide direct payments for the provision of landscapes can have high levels of biodiversity, biodiversity services (Pagiola and Platais, and a means of complementing conservation in forthcoming; Pagiola and others, 2002; Landell- protected areas (Pagiola and others, 1997; Daily Mills and Porras, 2002; Ferraro, 2001: Ferraro and others, 2001). Classical approaches to and Kiss, 2002). This approach internalizes conservation, attempting to preserve pristine what had been an externality, ensuring that it is habitats within protected areas, are necessary taken into consideration in decisionmaking. but insufficient in the face of growing pressure This is the approach taken by the Regional on land. Integrated Silvopastoral Ecosystem Efforts to enhance biodiversity in agricultural Management Project (RISEMP), which is being landscapes need to consider the incentives implemented with financing from the Global faced by individual land users, who decide Environment Facility (GEF). The project is what practices to use on their land, generally piloting the use of payments for environmental without considering what biodiversity benefits services as a means of generating biodiversity different land use practices may have. When conservation and carbon sequestration services biodiversity-friendly agricultural practices are in watersheds at three sites in Colombia, Costa the most profitable, there is a happy Rica, and Nicaragua. convergence of private and social interests. This This paper examines the contract mechanisms is the case of jungle rubber in Indonesia, for developed for the RISEMP. It begins by example (Thiollay, 1995; Tomich and others, describing the specific context in which the 1998). But biodiversity-friendly agricultural project is being implemented, that of degraded practices are not necessarily the most profitable pastoral areas in Central and South America. It from the perspective of individual land users. then describes the potential for silvopastoral In some cases, the profitability of biodiversity- practices to address this problem, which would friendly practices can be boosted by inducing provide both local and global benefits. But the consumers to pay a premium for their outputs, on-site benefits of silvopastoral practices alone as in the case of shade-grown coffee (Pagiola are insufficient to justify their adoption by and Ruthenberg, 2002). But this approach Environmental Economics Series 1 Paying for Biodiversity Conservation Services in Agricultural Landscapes farmers. Paying land users who adopt these technical characteristics of the practices being practices for the biodiversity and carbon promoted, the specific biodiversity and carbon sequestration services they generate can tip the sequestration being sought, and the economics balance towards adoption. The RISEMP is of silvopastoral practices from the land users' piloting an effort to do so. The factors which perspective. As this is a novel approach, the led to the design of the contract used in the RISEMP includes extensive monitoring efforts. RISEMP are described next. These include the 2 Environment Department Papers 2 Silvopastoral Practices Cattle production has long been associated with encourages this process, by granting titles to deforestation in Latin America (Barbier and land that is deemed to be `improved' (that is, others, 1994; Binswanger, 1991; Browder, 1985; cleared and used for agriculture). Downing and others, 1992; Kaimowitz, 1996; Kaimowitz and others, 2004; Mahar, 1988; Table 1 summarizes changes in pasture and Mertens and others, 2002; Myers, 1981; Repetto forest area in Colombia, Costa Rica, and and Gillis, 1988; Schneider, 1994), and as such Nicaragua. Forest cover has been in retreat has been an important cause of the loss of throughout the region. The area under annual natural habitat and biodiversity in the region. In crops has fallen in many countries (Nicaragua is most countries, the prevailing policy framework an exception, as the end of unrest in the early encouraged deforestation for timber extraction 1990s allowed a considerable expansion of and conversion of forest areas to pastures and agricultural land into areas that had been crops, which were encouraged by subsidized unsafe). There has been some expansion of credit, guaranteed prices, and other incentives. permanent crops, although this trend has The extent of these policy distortions has been reversed in recent years, due to low coffee substantially reduced in recent years (Faminow, prices. Permanent pasture, on the other hand, 1998) but pressure from poor landholders and-in has expanded steadily in all countries for which some areas-large scale ranches continues to data are available, although at different rates. result in large-scale deforestation in many areas. Appendix A provides more detailed data on In many countries, the legal framework land use changes in Central America and Table 1. Changes in pasture land and forest area in Colombia, Costa Rica, and Nicaragua Colombia Costa Rica Nicaragua Area, 2000 Change Area, 2000 Change Area, 1995 Change (`000 ha) 1990-2000 (`000 ha) 1990-2000 (`000 ha) 1990-1995 (%) (%) (%) Annual crops 2,818 ­14.7 225 ­13.5 2,457 25.2 Permanent crops 1,766 6.2 281 12.2 291 14.3 Permanent pasture 40,925 2.1 2,339 0.4 4,820 a .. Natural forest area 49,650 ­3.6 1,966 ­7.5 3,278 b ­26.4c Notes: a. data from 1990; b. data from 2000; c. 1990­2000 change. Source: World Bank World Development Indicators database. Environmental Economics Series 3 Paying for Biodiversity Conservation Services in Agricultural Landscapes Colombia in the last decade. These data show widely used in some countries of tropical that these patterns were common throughout America, provides an inexpensive the region. alternative for fencing and supplements livestock diets. In addition to the environmental problems caused by the initial loss of forest, traditional · Livestock grazing in forest plantations. In approaches to pasture are often unsustainable. this system, grazing is used to control the After an initial period of high yields, soil invasion by native and exotic grasses, thus fertility is depleted and grass cover diminishes, reducing the management costs of the resulting in soil erosion, contamination of water plantations. supplies, air pollution, further loss of biodiversity, and degradation of landscapes. Appendix B illustrates some of these systems in Lower income for producers results in the RISEMP project sites, as well as the continuing poverty and in pressure to clear degraded pastures they are meant to replace. additional areas. On-site benefits Silvopastoral systems, which combine trees with pasture, offer an alternative to prevalent cattle Silvopastoral systems can provide a range of on- production systems in Latin America. They site benefits (Dagang and Nair, 2003). The provide a deeply rooting, perennial vegetation introduction of trees in pasture areas can which is persistently growing and has a dense improve pasture productivity. Silvopastoral but uneven canopy. These systems can be systems tend to increase nutrient re-cycling grouped in four major categories (Murgueitio, across a deep portion of the soil profile 1999): occupied by the root systems of a wide variety of plants associated with silvopastoral systems. · Systems in which high densities of trees and Depending on the species of trees being used shrubs are planted in pastures, providing and on local climate characteristics, trees extract shade and diet supplements while water and nutrients from soil horizons protecting the soil from packing and inaccessible to grasses, and deposit the nutrients erosion. on the ground with the natural fall of foliage, twigs, and fruits. The biomass and amount of · Cut and carry systems, which replace nutrients released by pruning the trees of the grazing in open pasturelands with stables in agroforestry systems varies depending on the which livestock is fed with the foliage of kind of management in use. As much as 18 different trees and shrubs specifically tons/ha of dry matter can be deposited on the planted in areas formerly used for other ground annually, and the amount of nitrogen agricultural practices. Cut and carry flowing through the system can reach values of systems have been particularly successful in up to 380 kg/ha/year (Alpizar and others, Central America and in Colombia 1983). In addition, the trees can provide direct (Benavides, 1994). benefits in the form of products such as fruit, · Use of fast-growing trees and shrubs for fuelwood, fodder, and timber. From the farmers' fencing and wind screens. This system, perspective, the benefits of silvopastoral systems derive from (a) additional production 4 Environment Department Papers Silvopastoral Practices from the tree component; (b) maintaining and/ silvopastoral systems can serve as biological or improving pasture productivity; (c) corridors, helping to connect remaining diversification of production; and (d) habitats. At the regional level, silvopastoral contribution to the overall farming system (for systems may play an important role in the example, by providing fodder or income at a implementation of the Mesoamerican Biological time when other sources do not) (Current and Corridor, given the vast area of pasturelands in others, 1995). The shade provided by trees may Central America and Colombia. It is expected also enhance livestock productivity, especially that these corridors would provide adequate milk production. habitat for wildlife while facilitating seed dispersal and the regeneration of the native vegetation (Saunders and Hobbs, 1991). Biodiversity benefits The increased complexity of silvopastoral Other benefits systems relative to traditional pastures means they often bring important biodiversity benefits Silvopastoral systems are capable of fixing (Dagang and Nair, 2003). These take two main significant amounts of carbon in the soil under forms. First, they tend to support much higher the improved pastures and in the standing tree species diversity than traditional pastures. biomass (Fisher and others, 1994). Research in Second, they help connect protected areas. Colombia (Ramirez, 1997), Panama, and Costa Rica (CATIE, 1999; Pfaff and others, 20000) has Silvopastoral systems have been shown to play shown that soils under silvopastoral systems a major role in the survival of wildlife species have higher carbon content. Additional carbon by providing scarce resources and refuge; to is sequestered by the trees found in such have a higher propagation rate of native forest systems. Moreover, grass-based pastures tend to plants under these scattered trees; and to sequester most of the carbon in the deeper part provide shade for grazing animals, and shelter of the soil profile (between 40 and 100 cm for wild birds (Harvey and Haber, 1999). Food depth), thus making it less prone to oxidation, availability for wild birds is high in and hence loss (Fisher and others, 1994; silvopastoral systems, and the complex Beinroth and others, 1996). structure of the vegetation provides a more adequate nesting substrate and better protection Silvopastoral systems are also likely to affect against predators than other agroecosystems. water services, though the specific impact is Silvopastures and other agroforestry systems likely to be site specific. Infiltration generally also harbor a larger and more complex increases with the presence of trees, reducing assemblage of invertebrates than monoculture superficial runoff with its attendant soil erosion. pastures (Dennis and others, 1996). By Improved livestock management can help providing alternative sources of fuelwood and reduce compaction, thus further reducing other wood products, silvopastoral systems can surface runoff. The presence of trees also leads also help reduce pressure on remaining natural to increased evapotranspiration, however, thus habitats. tending to decrease water yield (Bosch and Hewlett, 1982; Bruijnzeel, 1990). In agricultural landscapes characterized by the fragmentation of the natural habitats, Environmental Economics Series 5 Paying for Biodiversity Conservation Services in Agricultural Landscapes In hilly areas, trees have an additional species is important. Trees of different root protective role in the ecosystem, that of depths are required for effective soil anchorage, preventing landslides (Bruijnzeel, 1990). Not in particular during torrential rain events only is the presence of trees essential for soil accompanying tropical storms. protection on slopes, but also the variety of 6 Environment Department Papers 3 Barriers to Adoption Despite their many benefits, silvopastoral An important constraint to the adoption of systems have only been adopted to a limited silvopastoral practices is their limited extent (Dagang and Nair, 2003). National-level profitability from the perspective of individual data typically do not distinguish land uses with land users. Establishing silvopastures can entail a sufficient level of detail, but Table 2 shows high initial costs, as shown in Table 3. Increasing average land use in the Quindío, Colombia, the livestock herd to take advantage of the RISEMP project site. Pasture with no or low tree increased fodder production entails additional density dominates all other land uses (Mejía, costs. In addition, there are opportunity costs 2004). Overall tree cover is low, although there is resulting from the time lags before the systems a significant amount of forest remnants, most of become productive (particularly important in which is riparian forest. Permanent crops, systems with substantial tree components). mostly coffee, account for about 10 percent of the area. Coffee was once the dominant land use Figure 1 illustrates the typical time profile of in this area, but it has been replaced by pasture returns to adoption of silvopastoral practices. in the last decade due to low coffee prices. The example shown is of a 20 ha farm in Fodder banks are practically non-existent: only 7 Nicaragua raising livestock for milk and meat. in 110 farms surveyed had any, with an average About 15 ha are used for unimproved pasture, of less than 1 ha each. with 2 ha devoted to cultivating basic grains for Table 2. Average land use in farms in Quindío, Colombia Hectares Percentage Annual crops 0.9 2.6 Permanent and semi-permanent crops 3.7 10.8 Pasture with no trees 21.9 64.1 Pasture with low tree density 0.6 1.8 Pasture with high tree density 0.0 0.1 Fodder banks 0.0 0.1 Plantations, riparian forest, and forest remnants 7.0 20.5 Total 34.2 100.0 Source: Mejía, 2004. Environmental Economics Series 7 Paying for Biodiversity Conservation Services in Agricultural Landscapes Table 3. Initial investment costs for selected silvopastoral practices Matiguás-Río Quindío, Esparza, Blanco, Colombia Costa Rica Nicaragua Improved pasture (US$/ha) 375 250 } 265 Planting 100 trees in improved pasture (US$/ha) 55 50 Planting 1,000 leuceana trees (US$/ha) 1,000 Protein bank (US$/ha) 960 660 475 Live fencing (US$/km) 700 610 390 Source: Gobbi, 2002. the household's own consumption and 3 ha in production would increase from just under brush. As shown in the figure, current land uses 2,000 liters annually to over 4,000 liters. Once generate a net farm income of about US$800 a the silvopastoral systems have been established, year. The proposed investment involves net farm income would rise to about US$1,200 a switching 3 ha of pasture to improved pasture year-a 50 percent increase. In the first years, with low tree density, and establishing a 0.75 ha however, farm income would be substantially fodder bank. This would allow an increase in lower because of the up-front investment costs the livestock herd from 14 to 15, but more and the time lag before the trees grow importantly it would result in a substantial sufficiently to provide benefits. Only in the fifth increase in the productivity of the herd due to year following the initial investment would the greater availability of higher quality fodder farm returns rise above those of the current land and the shade provided by the trees. Milk use practices. As a result, these investments are Figure 1. Typical time profile of benefits of silvopastoral systems 1,250 1,000 )$S 750 (U e Current practices m 500 coni Silvopastoral practices mrafte 250 N 0 No PES -250 NPV (US$): 440 IRR (%): 12 -500 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Year Note: 20ha farm in Matiguás, Nicaragua. 8 Environment Department Papers Barriers to Adoption financially marginal: in this case, the rate of knowledge of these on-site benefits further return to adoption of silvopastoral practices is reduces the perceived benefits to land users. less than 12 percent, and the net present value is only US$440 (over 50 years, at a 10 percent Even if silvopastoral practices are financially discount rate). viable, the high initial investment costs required pose problems for credit-constrained land users. The low rates of return to the adoption of In the Quindío project site, only 25 percent of silvopastoral systems are typical. Estimates households had access to credit in the past five prepared for the RISEMP show rates of return of years (Mejía, 2004). Access to credit is higher in between 4 and 14 percent, depending on the the Matiguás-Río Blanco area in Nicaragua, country and type of farm (Gobbi, 2002). Other thanks to presence of several NGOs that offer studies found similar results; White and others credit. About 50-75 percent of households in this (2001), for example, found rates of return to area report having used credit in the past five adoption of improved pasture in Esparza, Costa years (Ramírez and others, 2004). However, Rica, of 9 to 12 percent. These estimates, of credit is often only available for specific course, only consider the on-site benefits of purposes and with collateral requirements that silvopastoral practices. The biodiversity are difficult for farmers to meet. conservation and carbon sequestration benefits are not considered in the farmers' The long-term nature of investments in most decisionmaking. silvopastoral practices means that tenure security is an important factor in their adoption This problem is compounded by a lack of (Deininger and others, 2003, Meinzen-Dick and awareness by farmers of some of the on-site others, 2002). Tenure is not a constraint in the benefits offered by silvopastoral systems, such three study sites, however. In the Costa Rica and as reduced dependency on chemical fertilizers Colombia project sites, all farmers have formal and pesticides, savings in water for irrigation, ownership of the land (though they may not all soil protection and enhanced fertility, and the have titles). In the Nicaragua project site, most potential for additional incomes from ranchers occupy public land, but long-term harvesting fruit, fuelwood, and timber. Limited occupancy gives them secure tenure. Environmental Economics Series 9 Payments for 4 Environmental Services From the land users' perspective, the supported by a World Bank loan and GEF grant biodiversity conservation and carbon under the Ecomarkets Project (World Bank, sequestration benefits are externalities. As such, 2000) (Box 1). The town of Heredia has they do not take them into consideration in established an `environmentally adjusted water making their land use decisions, thus reducing tariff', the proceeds of which are used to pay the likelihood that they will adopt practices that landholders to maintain and reforest watershed generate such benefits, including silvopastoral areas (Castro, 2001; Cordero, 2003). In a systems. Recognition of this problem and of the separate initiative, hydropower producer La failure of past approaches to dealing with it has Manguera SA is paying the Monteverde led to efforts to develop systems in which land Conservation League to maintain under forest users are paid for the environmental services cover the watershed from which its plant draws they generate, thus aligning their incentives its water (Rojas and Aylward, 2002). In with those of society as a whole. The simple Colombia, irrigation water user groups and logic of Payments for Environmental Services municipalities in the Cauca valley are paying to (PES) is that compensating land users for the conserve the watersheds that supply them with environmental services a given land use water (Echevarría, 2002b). In 2003, Mexico provides makes them more likely to choose that created the Payment for Hydrological land use rather than another. Environmental Services program (Pago por Servicios Ambientales Hidrológicos, PSAH), There has been considerable experimentation which pays for the conservation of forests in with PES and other market-based approaches in hydrologically critical watersheds using recent years (Pagiola and Platais, forthcoming; revenue from water charges (Bulas, 2004). In Pagiola and others, 2002; Landell-Mills and southern Mexico, the Scolel Té project is paying Porras, 2002). Latin America has been a farmers to provide carbon sequestration particularly fertile ground for such services (Tipper, 2002). In Ecuador, the city of experimentation (Pagiola and Platais, 2001). Quito has created a water fund with Costa Rica has developed an elaborate, nation- contributions from the water utility and the wide PES program, the Pago por Servicios electric power company to pay for conservation Ambientales (PSA) (FONAFIFO, 2000; Pagiola, in the protected areas from which it draws its 2002). Under the 1997 Forestry Law, land users water (Echevarría, 2002a). can receive payments for specified land uses, including new plantations and conservation of The bulk of PES programs to date have focused natural forests. The PSA program is now being on water services, reflecting both the urgency of Environmental Economics Series 11 Paying for Biodiversity Conservation Services in Agricultural Landscapes addressing water issues in many developing Conservation International has also used the countries and the relative ease with which the approach (which it calls `conservation beneficiaries of water services can be identified concessions' or `conservation incentive (Pagiola and Platais, forthcoming). The agreements') in several cases, including Guyana approach has been used for biodiversity and Peru (Hardner and Rice, 2002; Rice, 2003). benefits in a few cases, mostly with GEF Many of these efforts have focused on relatively support as in the case of the Costa Rica untouched areas, however, rather than on Ecomarkets Project. Environmental NGO agricultural areas. Box 1 World Bank Support for PES The World Bank is working with several countries to develop PES programs, through loans, technical assis- tance, and capacity building. In addition to the RISEMP project, current operational activities involving PES include: ! Costa Rica. The Ecomarkets Project, which supports the country's PES program, includes a US$32.6 million loan from the World Bank to help the government ensure current levels of environmental service contracts and a US$8 million grant from the GEF to assist the program's conservation of biodiversity (World Bank, 2000). ! Guatemala. The Western Altiplano Natural Resources Management Project includes a component aimed at testing and piloting PES mechanisms at the local level and supporting the development of the required national policy framework and instruments (World Bank, 2003a). ! Venezuela. A GEF-financed project focusing on Canaima National Park is under preparation, including a mechanism to channel watershed conservation payments made by hydropower producer CVG-EDELCA. ! Mexico. The World Bank provided technical support to the government's efforts to establish the Payment for Hydrological Environmental Services program. ! Dominican Republic, Ecuador, and El Salvador. Pilot PES programs are under preparation in these coun- tries. El Salvador is at the most advanced stage. ! South Africa. The Cape Action Plan for the Environment (CAPE), under preparation, aims to use a PES approach as one of the tools to encourage conservation in the Cape Floristic Region. ! BioCarbon Fund. The newly created BioCarbon Fund is examining the potential for buying carbon seques- tration services generated by land use change. For example, one proposal would pay for carbon sequestered by improving shade-grown coffee systems in the Mexican uplands. In addition, the World Bank's training arm, the World Bank Institute (WBI), has provided training on PES targeted to technical personnel in ministries, conservation agencies, and nongovernmental organizations (NGOs) involved in implementing PES programs. It should be noted that the World Bank did not originate the PES concept. The World Bank has played an important role in launching such projects primarily because its borrowing countries have requested its assis- tance in doing so. By virtue of its role in assisting many countries, it has been able to cross-fertilize efforts in individual countries with the lessons learned in others (Pagiola and Platais 2003). 12 Environment Department Papers 5 FromTheory to Practice Although the PES approach is intuitively Development Initiative (LEAD), hosted by the appealing, putting it into practice is far from Food and Agriculture Organisation (FAO). It is simple (Pagiola and Platais, 2003). The financed by a US$4.5 million GEF grant, with theoretical elegance of a blackboard concept the World Bank acting as implementing agency. needs to be translated into actual In each country, field activities are being implementation arrangements on the ground. undertaken by local non-governmental The remainder of this paper describes the organizations (NGOs): the Centre for Research approach adopted to do so by the RISEMP on Sustainable Agricultural Production Systems project. (Centro para la Investigación en Sistemas Sostenibles de Producción Agropecuaria, The RISEMP, which began implementation in CIPAV) in Colombia, the Center for Teaching July 2002, is seeking to pilot the use of the and Research on Tropical Agronomy (Centro payment for environmental services approach to Agronómico Tropical de Investigación y encourage the adoption of silvopastoral Enseñanza, CATIE) in Costa Rica, and Nitlapan practices in degraded pastures areas in Central in Nicaragua, with CATIE coordinating the and South America (World Bank, 2002). The work. The American Bird Conservancy (ABC) is project is being implemented in three providing technical assistance for the microwatersheds: Quindío, in Colombia; development of a common and consistent Esparza, in Costa Rica; and Matiguás-Río methodology for the monitoring of biodiversity Blanco, in Nicaragua. Participating land users at the three project sites. enter into contracts under which they receive a payment for the environmental services that they generate. They receive annual payments What is being paid for? over a two- or four-year period, based on the Contracting for land users to provide increment in environmental services provided biodiversity benefits is all very well in theory, relative to the baseline situation for that but in practice it is clearly unrealistic to ask particular farm. Through this mechanism, the them to deliver biodiversity. A way is needed to project aims to establish silvopastoral systems communicate what is desired to potential on 3,500 ha, thus enhancing the environmental participants in ways that they can understand. benefits generated in watersheds covering about The typical solution has been to offer to pay not 12,000 ha. for biodiversity itself, but for land uses that are hospitable to biodiversity (Pagiola and others, The project was prepared with support of the 2002). multi-donor Livestock, Environment and Environmental Economics Series 13 Paying for Biodiversity Conservation Services in Agricultural Landscapes But land use alone can be a relatively blunt carbon sequestration indices are presented in instrument. In Costa Rica's PSA program, for Table 4. example, most contracts call for conservation of existing forest, and pay all participants the same The biodiversity conservation index was scaled amount (FONAFIFO, 2000; Pagiola, 2002). with the most biodiversity-poor land use While this approach has the virtue of simplicity, (annual crops) set at 0.0 and the most it fails to recognize the very different levels of biodiversity-rich land use (primary forest) set at services that different land uses can provide. 1.0. Within this spectrum, the points given to The biodiversity-friendliness of agricultural each specific land use were set by a panel of practices is not a binary, yes/no proposition. On experts, taking into consideration factors such the contrary, there is a spectrum of effects, as the number of species (of plants, birds, small ranging from relatively inhospitable systems mammals, and insects), their spatial such as monocultures with heavy agrochemical arrangement, stratification, plot size, and fruit use to relatively hospitable systems such as production. Higher scores were given to land organic coffee grown under a diverse shade uses that have greater potential to maintain the canopy of native species. Location also matters: original biodiversity of the region. Note that the biodiversity-friendly practices in proximity to index estimates the environmental benefits of all protected areas, for example, might be more land uses, and not only silvopastoral practices. valuable by helping to buffer and protect them. This approach can take into consideration the Failing to take these differences into account different impact that different land uses are risks either under-paying for desirable land likely to have on biodiversity. There are, of uses, or over-paying for relatively less desirable course, limitations. The biodiversity impact ones (Pagiola and Platais, forthcoming). depends not only on the characteristics of the The solution adopted in the RISEMP was to land use, but also on its location, its extent, and prepare a list of land uses and associate each its relationship to other land uses. At the pilot with a point system upon which payments are scale of the RISEMP, issues of location are not based. This approach is similar to that of the significant, as all three pilot areas were Environmental Benefits Index (EBI) used in US specifically chosen for their proximity to Conservation Reserve Program (CRP) (NCEE, protected areas or to corridors between them. 2001). Separate indices were developed for the All three project areas were selected in part biodiversity conservation and carbon because of their location in ecologically- sequestration benefits of each land use. These sensitive areas. The Quindío project site is in two indices were then aggregated to form an one of the most severely degraded regions of environmental service index to be employed as Colombia, with few, mostly unconnected the basis for calculating payments to remnants of natural habitats. Restoring a degree participants. A similar index for water benefits of habitat heterogeneity and connectivity would was not included, partly because of the lack of increase the chances of survival of species data needed to develop it, and partly because requiring large home ranges in an area improved water flows would be national considered as a priority for bird conservation. benefits, and thus are not eligible for GEF The Esparza area in Costa Rica is in the vicinity funding. The biodiversity conservation and of conservation areas such as La Fortuna, the 14 Environment Department Papers From Theory to Practice Table 4. Environmental service indices used in the RISEMP (Points per hectare, unless otherwise specified) Carbon Environmental Biodiversity sequestration service Land use index index index Annual crops (annual, grains, and tubers) 0.0 0.0 0.0 Degraded pasture 0.0 0.0 0.0 Natural pasture without trees 0.1 0.1 0.2 Improved pasture without trees 0.4 0.1 0.5 Semi-permanent crops (plantain, sun coffee) 0.3 0.2 0.5 Natural pasture with low tree density (< 30/ha) 0.3 0.3 0.6 Natural pasture with recently-planted trees (> 200/ha) 0.3 0.3 0.6 Improved pasture with recently-planted trees (> 200/ha) 0.3 0.4 0.7 Monoculture fruit crops 0.3 0.4 0.7 Fodder bank 0.3 0.5 0.8 Improved pasture with low tree density (< 30/ha) 0.3 0.6 0.9 Fodder bank with woody species 0.4 0.5 0.9 Natural pasture with high tree density (> 30/ha) 0.5 0.5 1.0 Diversified fruit crops 0.6 0.5 1.1 Diversified fodder bank 0.6 0.6 1.2 Monoculture timber plantation 0.4 0.8 1.2 Shade-grown coffee 0.6 0.7 1.3 Improved pasture with high tree density (> 30/ha) 0.6 0.7 1.3 Bamboo (guadua) forest 0.5 0.8 1.3 Diversified timber plantation 0.7 0.7 1.4 Scrub habitats (tacotales) 0.6 0.8 1.4 Riparian forest 0.8 0.7 1.5 Intensive silvopastoral system (>5,000 trees/ha) 0.6 1.0 1.6 Disturbed secondary forest (> 10 m2 basal area) 0.8 0.9 1.7 Secondary forest (> 10 m2 basal area) 0.9 1.0 1.9 Primary forest 1.0 1.0 2.0 New live fence or established live fence with (per km) 0.3 0.3 0.6 frequent pruning Multi-story live fence or wind break (per km) 0.6 0.5 1.1 Note: The environmental service index is the sum of the biodiversity and carbon sequestration indices. Monteverde Reserve complex, and the Alberto Reserve, and is very close to one of the priority Brenes Biological Reserve. More biodiversity- areas for bird conservation in the country. If this friendly land use practices would help the approach were to be scaled up and applied on a chances of survival of several species occurring broader scale, location effects could be in these protected areas. The Matiguás-Río incorporated either by varying the points for Blanco watershed in Nicaragua is part of the activities in different locations or by varying the buffer zone of the Cerro Musún Natural payment per incremental point. Issues of scale Environmental Economics Series 15 Paying for Biodiversity Conservation Services in Agricultural Landscapes and contiguity are harder to address. Some Note that under RISEMP, biodiversity and biodiversity benefits may be obtained only after carbon sequestration benefits are given equal appropriate land uses cover a minimum area, or weight in calculating payments. The two indices if the areas covered are contiguous rather than could easily be de-coupled, however, with scattered. To an extent, these effects might be separate payment levels for each kind of addressed by adding bonus points if the area environmental service. Alternatively, different covered by a given land use passes a threshold. weighting schemes could be used to give Such an approach could quickly result in an proportionally more weight to one or the other, excessively complex point system, however. depending on the interest of those making the Another approach would be to set a minimum payments. participation threshold for the PES program to take effect; this approach was followed by New This index approach was tested with potential York City, for example (A. Appleton, pers. participants, and is proving quite intelligible to comm.) them in practice. Dissemination materials such as posters and handbooks have been prepared A similar procedure was used to establish the showing precisely what the payments would be carbon sequestration index, with different land for specific land uses. uses given points according to their capacity to sequester stable carbon in the soil and in hard How should payments be made? wood through the years. Recent studies indicate that secondary forest can fix an average of 10 The second challenge in developing an tonnes of carbon per year in wood and in the appropriate contract is the need to understand soil. As secondary forest has a value of 1.0 in the the economics of the farming system, so that the index, 0.1 points correspond to an estimated appropriate amount and form of payment can sequestration of 1 tonne of carbon. Data from be determined. Payments for environmental studies conducted by CATIE were used to services will have the desired effect only if they calibrate the carbon sequestration index. reach the land users in ways that influence their decisions on how to use the land. As data were insufficient to derive country- specific indices, the same index is being used in Analysis of the time path of benefits generated all three countries. Data from the monitoring by silvopastoral systems showed that they are efforts will be used to improve the indices, and unattractive to land users primarily because of it is expected that these will differ from country their substantial initial investment, and because to country. of the time lag between investment and returns, as shown in Figure 1 above. This leads to the Should downstream water users be willing to hypothesis that a relatively small payment pay for hydrological services, the approach provided in the early period of adoption would could also be extended by adding an index be sufficient to `tip the balance' between current denoting the contribution of each land use type and silvopastoral systems. This effect works by to the desired water services, though increasing the net present value of investments developing such an index would certainly prove in silvopastoral practices, but also by reducing difficult. the initial period in which adoption of these 16 Environment Department Papers From Theory to Practice systems imposes net costs on land users. By the offered insufficient to justify the investment.) In time payments end, the silvopastoral practices Mexico, a specific study was commissioned of themselves are ready to begin generating the opportunity cost of land (Jaramillo, 2003) to income for land users. The payments also provide a basis for payments levels under the alleviate the liquidity problems faced by many PSAH; no study was made of the magnitude of land users and help them finance the required benefits. Zelek and Shively (2003) propose a investments. scheme to pay the opportunity costs of Philipino farmers who adopt practices that sequester Based on this analysis, it was decided to carbon. Paying the opportunity cost of adopting provide a relatively small, up-front payment to the desired practice also accords well with participating land users. This payment is of GEF's policy of paying for the incremental costs US$75 per incremental point, per year over a of generating global environmental benefits. four-year period, up to a maximum of US$4,500 per farm (US$6,000 in Colombia, where input In terms of payments for carbon emissions prices are higher). Both of these aspects deserve reductions, the US$75/point/year payment further discussion. level is equivalent to paying US$7.5 per tonne of carbon sequestered, or US$2 per tonne of CO2 In principle, the amount should be no less than equivalent. This compares favorably to current the land users' opportunity cost (or they will world prices for carbon emissions reduction of not participate), and no more than the value of US$3­5 per tonne of CO2 equivalent (World the benefit provided (or it would not be Bank, 2003b)--although these payments worthwhile to provide the service). In practice, typically require a higher degree of assurance of the actual value of the benefit provided is the permanence of the emissions reduction and extremely difficult to estimate, and particularly a more intensive monitoring regime than the so for benefits such as biodiversity RISEMP offers. No similar comparison is conservation. In contrast, the farmers' possible for payments for biodiversity opportunity cost can usually be estimated conservation. relatively easily. For this reason, as well as to limit the budgetary requirements of the In general, emerging guidelines for payments payment, payment levels are usually set at for environmental services indicate that slightly more than the opportunity cost of the payments should be on-going rather than finite main alternative land uses. All the existing (Pagiola and Platais, forthcoming). In Costa systems of payments for environmental services Rica's PSA program, for example, payments for implicitly or explicitly use this approach. Costa forest conservation contracts are for 5 years, but Rica's PSA program, for example, currently they are renewable indefinitely by mutual pays US$45/ha/year for forest conservation. agreement. The logic for this is simple: if This payment has proven to be quite attractive, environmental services are to be generated over with far more applications for this contract than a long period of time (presumably, indefinitely), the program has been able to finance. (In then payments for these services should also be contrast, a payment of US$538/ha over 5 years made over a similarly long period. Ending for reforestation has proven to be less popular, payments sooner creates the risk that land users as many landowners consider the payment will revert to their previous land use practices. Environmental Economics Series 17 Paying for Biodiversity Conservation Services in Agricultural Landscapes This is a risk that has been observed time and alternative contract will receive a similar time again in projects that attempted to change amount over a two-year period. Farmers were land use practices, such as soil conservation or assigned randomly to one or the other contract. reforestation projects (Lutz and others, 1994). This risk was thought to be relatively low in this Table 5 illustrates the application of this contract instance, as the silvopastoral practices, once for the 20 ha farm in Matiguás, Nicaragua, used established, are privately more profitable (see in the previous example. In the baseline year, Figure 1). Moreover, the payments represent the farm has 2.5 ha under annual crops, 14.5 ha only a small portion of the necessary investment under natural pasture without trees, and 3 ha costs, thus making it unlikely that land users under brush (tacotal). Motivated by the project, would adopt practices they intend to abandon it converts 3 ha of its pasture to higher tree solely to receive the payments. In an effort to densities: 1 ha a year for the first three years. It determine the long-term sustainability of the also plants an 0.5 ha fodder bank, and fences off mechanism, a sub-group of participants is being the scrub areas so that secondary forest can given a slightly modified contract, in which the regenerate. Finally, it plants trees along 1.5 km payments are frontloaded: rather than receiving of its fence lines. Using the environmental them over a four-year period, farmers with this service index in Table 4, the resulting scores can Table 5. Example of payment computation Years from contract signing 0 1 2 3 4 Land use Crops (annual, grains, tubers) (ha) 2.5 2.5 2.5 2.5 2.5 Natural pasture with low tree density (ha) 14.5 13.0 12.0 11.0 11.0 Natural pasture with high tree density (ha) 0.0 1.0 2.0 3.0 3.0 Fodder bank (ha) 0.0 0.5 0.5 0.5 0.5 Scrub habitat (ha) 3.0 0.0 0.0 0.0 0.0 Secondary forest (ha) 0.0 3.0 3.0 3.0 3.0 Total area (ha) 20.0 20.0 20.0 20.0 20.0 Wire fences with trees (km) 0.0 0.5 1.0 1.5 1.5 Environmental service score (points) Crops (annual, grains, tubers) 0.0 0.0 0.0 0.0 0.0 Natural pasture with low tree density 8.7 7.8 7.2 6.6 6.6 Natural pasture with high tree density 0.0 0.9 1.8 2.7 2.7 Fodder bank (monocrop) 0.0 0.5 0.5 0.5 0.5 Scrub habitat 3.9 0.0 0.0 0.0 0.0 Secondary forest 0.0 5.7 5.7 5.7 5.7 Wire fences with trees 0.0 0.5 1.0 1.5 1.5 Total points for the farm 12.6 15.4 16.2 17.0 17.0 Baseline points 12.6 Incremental points 2.8 3.6 4.4 4.4 Income from environmental services (US$) 126 210 270 330 330 Note: 20ha farm in Matiguás, Nicaragua. 18 Environment Department Papers From Theory to Practice Figure 2. Effects of PES on the profitability of silvopastoral systems 1,250 1,000 )$S 750 (U e Current practices m 500 coni Silvopastoral practices mrafte 250 Silvopastoral practices with PES 0 N No PES With PES -250 NPV (US$): 440 1,450 IRR (%): 12 19 -500 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Year Note: 20ha farm in Matiguás, Nicaragua. be calculated for the baseline and for each approach entailed a substantial risk of creating subsequent year. These scores are then used to perverse incentives. "Bueno, corto todo," was a compute the payments due to the farmer, common reaction by land users when told they including the initial baseline payment for would not be compensated for pre-existing existing services (see below) and the main trees: "fine, I'll cut them all." It might have been payment for incremental services provided possible to avoid this risk among project under the project. Figure 2 shows the impact of participants by imposing contractual restrictions these payments on the time profile of benefits to on such actions, though this would certainly adopting silvopastoral practices, and the have required an increased monitoring effort, resulting impact on the profitability of the and thus increased costs. But there was also a investment. What had been a marginally viable broader risk that non-participants in investment now becomes more attractive. surrounding areas would postpone adopting silvopastoral practices that they might have been tempted to adopt, so that they might wait Avoiding perverse incentives for a project to come and compensate them for The initial plan involved paying land users only doing so. As a result, the initial plan was for incremental improvements in land use modified to allow for a payment to be made for practices. The extent to which land users had pre-existing environmental services. A one-time already adopted practices that conserved payment of US$10/point will be made for the biodiversity or sequestered carbon prior to the baseline points, up to a maximum of US$500 per project was to have been reflected in their farm. This payment has the further benefit of baseline environmental service index, and only helping to alleviate financing constraints to increments to this index were to be implementing silvopastoral practices. compensated. It soon became clear that this Environmental Economics Series 19 Paying for Biodiversity Conservation Services in Agricultural Landscapes As part of the effort to avoid perverse damaging activities being simply displaced, so incentives, the contract also specifies that that there is little net benefit. The RISEMP burning in pastures is banned (except in areas minimizes this problem by computing the devoted to food security, where burning is points on which payments are made over the allowed in the first two years), and that the entire farm and basing payments on the net contract will be terminated if the participants points over the entire farm. If land users cut cut down primary or secondary forest in their down trees in one plot even as they plant them farms. in another, the negative points earned from the adverse change will offset those gained from the A related problem is that of minimizing positive change. `leakage'-that is, avoiding environmentally- 20 Environment Department Papers 6 Monitoring Results Monitoring is always important, as it allows the group will be selected so as to have similar effectiveness of the project to be determined. It characteristics as the households in the target can also allow mid-course corrections to be group, but will not receive any payments for made if they should prove necessary. The need environmental services. The target group itself for monitoring is particularly high in pilot will be partitioned into two groups, so that the projects, which are intended to serve as guides impact of technical assistance provided by the for future projects. project can also be tested. The main part of the group will receive both the payments and technical assistance, while a subgroup of 30 Changes in land use households will only receive the payments for For the purpose of monitoring compliance with environmental services (to the extent that they the contract and computing payments owed, adopt the recommended practices), without the observation of participants would be sufficient. technical assistance. Each household in each of However, the RISEMP has the broader objective these groups will be monitored every other year. of piloting the application of PES systems to Their land use will be monitored, and a socio- silvopastoral practices. This requires not only economic survey will be conducted. monitoring the behavior of participants, but determining that the project has been Impact of land use change on instrumental in affecting this behavior. environmental services Monitoring the degree to which the project is encouraging participants to undertake the To verify that the silvopastoral systems desired changes in land use entails monitoring promoted under the project actually generate the changes in land use of the participants the expected environmental benefits, themselves, and of a control group (so that the biodiversity and carbon sequestration will be impact of the project itself can be distinguished monitored in all land use types in the three pilot from other trends that might affect land use). areas. For biodiversity, counts of bird species will be the main indicator of biodiversity used, To address these issues, monitoring will be but they will be complemented by studies of undertaken on three groups. In addition to the butterflies, ants, and mollusks. Factors such as target group of participants (80 households in endemicity and rarity in the species observed Colombia, 100 in Costa Rica, and 100 in will be taken into consideration. Water quality Nicaragua), a control group of 30 households will be monitored only in the Colombian site, as will be monitored in each country. This control funds did not permit a more general assessment Environmental Economics Series 21 Paying for Biodiversity Conservation Services in Agricultural Landscapes of the contribution of these systems to improved Impact of the project on participating water quality. households In addition to the project's impact on the global The changes will be compared to baseline environment, it is also important to understand measurements made at the start of the project. A its impact on household welfare: does welfare study of the Matiguás-Río Blanco sites in increase, and if so how much and in what way, Nicaragua, for example, found 131 bird species and are there differences in how welfare (Pérez and others, 2004). The key test for the increases across income groups? Payments for project will be whether biodiversity increases environmental services have been hypothesized significantly compared to the baseline. as having the potential for improving the The results of the monitoring will also be used welfare of the poor in target areas (Pagiola and to revise and refine the biodiversity and carbon others, 2003), but there has been little empirical sequestration indices. These revised indices work on this topic to date. Data collected could be used to determine payments under any through the socioeconomic survey will help to future project. They will not, however, affect address these questions. It will allow low- payments under the current project, which will income households to be identified, and will be made based on an ex ante estimates of the provide a variety of measures of household global environmental benefits of each land use, welfare, including income-related measures as expressed in the environmental service index. (total income and income variability) and other indicators (such as health status). 22 Environment Department Papers 7 Conclusions The RISEMP project demonstrates that direct offered, the greater adoption is likely to be. contracts for biodiversity conservation in However, higher payments also militate against agricultural landscapes are not just a theoretical the cost-effectiveness of the approach. curiosity. Such contracts are possible, if the link Moreover, a higher payment per point also between land use and biodiversity is known. increases the risk that participants will adopt Specific details depend on the economics of the otherwise un-profitable practices only system being promoted. temporarily, so as to receive the payment, with the intention of abandoning them later. The It will be some time before the effectiveness of payment level was initially set at US$50 per the mechanisms discussed in this paper can be environmental service index point, but was determined. The intensive monitoring being raised to US$75 after field staff reported that undertaken will allow a very detailed analysis participants considered US$50/point of this effectiveness, including consideration of insufficiently attractive to justify widespread numerous exogenous factors that might affect it. adoption of silvopastoral practices. Given the This project will thus allow both an overall novelty of the approach, there is also a potential conclusion on the effectiveness of the approach credibility problem of the project's promise to and provide data for its refinement. Already, pay for environmental services. The baseline however, a number of key questions can be payments probably played an important role in identified, some specific to the particular this sense, over and above their benefits in approach used in the RISEMP, some that apply avoiding perverse incentives. more broadly to PES approaches in general, and some that bear to the potential for replicating Will changes in land use be sustainable? The the approach on a wider scale. RISEMP project is based on the hypothesis that silvopastoral practices, once adopted, are more profitable to land users than current practices. If Assessing the success of the project this hypothesis is correct, then adoption should Will the desired land use changes be induced? The be sustainable with no further assistance. To test extent to which the desired land use changes are this hypothesis some participants are being paid induced depends in part on the payment level over a two-year rather than a four-year period. being sufficient to `tip the balance' between current and improved practices. The Will improvements in biodiversity conservation be appropriate level of the payment was a subject significant? Baseline studies show that there is of intense debate. The higher the payment some pre-existing biodiversity in each of the Environmental Economics Series 23 Paying for Biodiversity Conservation Services in Agricultural Landscapes project areas. The challenge for the project will outright requires paying for the entire present be to demonstrate whether it has improved this value of the future flow of benefits up-front. In biodiversity significantly. The project will contrast, a PES approach makes payments over monitor changes in biodiversity levels closely. time. PES will also attract lower-opportunity cost lands, while a protected area approach Will the poor benefit? If the farm models prepared usually privileges conservation benefits and for the project (illustrated in Figures 1 and 2) are thus may include higher-opportunity cost land. correct, the project has the potential of The PES approach is likely to be particularly increasing the net income of participating advantageous if, as may be the case in the households quite substantially. Whether this RISEMP, a short-term payment is sufficient to will occur is being closely monitored. result in sustainable adoption of the desired Constraints that might prevent poorer land uses. Finally, buying land outright may households from participating, and hence from simply not be politically feasible, or may entail obtaining these income increases, are also being undesirable social consequences because of the examined closely. need to relocate the landowners. In the case of the Costa Rica Ecomarkets Project, the PES Assessing the success of the approach approach was found to be much more cost- effective than establishing a protected area of How cost-effective is the PES approach in terms of the same size (World Bank, 2000). biodiversity conservation? Both the cost and the effectiveness of this approach remain to be PES may be the cheaper way to conserve a determined. The two are to some extent given area, but the level of conservation is likely inversely related. Payments will be high if land to be lower. If the land were bought outright users adopt practices with high point values, and placed in a protected area, it could be and low if they adopt practices with low point managed optimally from the conservation values. The transaction costs of implementing perspective. Under a PES approach, land use is the project must also be considered. Some have determined by the combination of conservation argued that an incremental conservation dollar benefits (as reflected in the payment) and land would be most effectively spent on other user preferences; in many cases, this will lead to approaches, such as protected areas. Certainly a compromise result. Moreover, some of the protected areas are likely to have much lower reasons that make PES cheaper may also lead to transaction costs than the approach discussed lower conservation benefits: lower opportunity here. But their implementation costs may well cost land is not necessarily the most desirable be higher. Establishing a protected area would from a conservation perspective, for example. require buying the land from its current owners. That is, it would require compensating them for Replicating the approach the loss of the entire flow of benefits it might generate in its most profitable use. The PES How can transaction costs be reduced? The approach only requires compensating them for transaction costs involved in implementing a the difference between the net benefits they PES approach are a key determinant of its cost- obtain under the conservation use and the most effectiveness, its sustainability, and its profitable use. Moreover, buying the land replicability. They also play a critical role in the 24 Environment Department Papers Conclusions extent to which poorer land users can result in sustainable adoption of the desired participate (Pagiola and others, 2003). Because land uses. Rather, longer-term, probably of its pilot nature, the RISEMP has relatively indefinite payments will be needed. In turn, this high costs for detailed monitoring and other means that sustainable long-term financing activities that would not necessarily be needed sources will be required. Even if only short-term in a scaled-up project. The environmental payments are sufficient, substantial additional service index used in the RISEMP allows a very funding flows will be required if the approach is fine-tuned targeting of payments to expected to be extended beyond the pilot areas. For water benefits, but it also imposes relatively high services, potential sources for such financing monitoring costs. There is a need to find proxy can readily be identified-although capturing indicators that are highly correlated with them can be difficult. Moreover, funding biodiversity conservation but are easy and streams for water services can in principle be cheap to monitor, ideally using remote sensing. very sustainable, as they are tied to services that If forest cover provides an adequate proxy, for will continue to be used indefinitely (Pagiola example, it would be relatively cheap to and Platais, forthcoming). All available observe. The current environmental services financing sources for biodiversity conservation, index cannot be monitored solely from remote however, including the GEF, tend to focus on sensing, as it includes elements of the type and relatively short-term projects. Placing funds into quality of vegetative cover. A crucial question a trust fund so as to generate a stream of future which needs to be explored is that of the revenues is one option (GEF, 1999), but it tradeoff between the precision of the index and requires substantial up-front financing. Because the transaction costs involved in implementing of the greater ease of generating long-term it. payment streams for water services, basing payments on water service provision may How can the approach be made sustainable? Most of appear to be an attractive option. The the discussion in this paper would be broadly municipality of Matiguás, for example, is applicable to PES approaches that address interested in using this approach to protect its benefits other than biodiversity. Where water supply. This approach should certainly be biodiversity services differ, however, is in the exploited as much as possible, but two long-term sustainability of payments. Emerging constraints need to be borne in mind: first, lessons indicate that payments under PES water services are very site-specific, and so programs usually have to be made on a long- many areas would not be eligible for payments. term basis if the desired services are to be The project areas near Matiguás, for example, generated sustainably (Pagiola and Platais, are downstream of the water intakes for the 2003, forthcoming). The specific practices municipal water supply system and so would promoted by the RISEMP project may not not be included in a water service-based PES require a long-term payment, but this is likely to program. Second, the most desirable activities be the exception rather than the rule. In the from the perspective of generating water more general case in which the systems with the services are not necessarily the same as those highest external benefits are not the most that generate the biodiversity and carbon profitable to land users, the short-term payment sequestration services sought in the RISEMP approach used in the RISEMP is unlikely to project. Basing payments on water services, Environmental Economics Series 25 Paying for Biodiversity Conservation Services in Agricultural Landscapes therefore, will require additional scientific generate all the desired biodiversity and carbon research to improve the understanding of how sequestration benefits, and the need for separate land use affects water services. In general, financing for this purpose will remain. therefore, water-based payments will not 26 Environment Department Papers Appendix A -- Land Use Change in Central America and Colombia, 1990­2000 Area (`000 ha) Change in area (%) 1990 1995 2000 1990-1995 1995-2000 1990-2000 Annual crops Colombia 3,305.0 2,399.0 2,818.0 -27.4 17.5 -14.7 Costa Rica 260.0 225.0 225.0 -13.5 0.0 -13.5 El Salvador 550.0 582.0 560.0 5.8 -3.8 1.8 Guatemala 1,300.0 1,355.0 1,360.0 4.2 0.4 4.6 Honduras 1,462.0 1,600.0 1,068.0 9.4 -33.3 -26.9 Nicaragua 1,963.0 2,457.0 .. 25.2 .. .. Panamá 499.0 500.0 500.0 0.2 0.0 0.2 Permanent crops Colombia 1,661.9 2,077.4 1,765.8 25.0 -15.0 6.2 Costa Rica 250.2 291.0 280.8 16.3 -3.5 12.2 El Salvador 259.0 273.5 250.7 5.6 -8.3 -3.2 Guatemala 487.9 553.0 542.2 13.3 -2.0 11.1 Honduras 358.0 346.9 358.0 -3.1 3.2 0.0 Nicaragua 254.9 291.4 .. 14.3 .. .. Panamá 156.3 .. .. .. .. .. Permanent pasture Colombia 40,093.8 40,093.8 40,924.8 0.0 2.1 2.1 Costa Rica 2,328.3 2,338.5 2,338.5 0.4 0.0 0.4 El Salvador 640.2 750.1 793.6 17.2 5.8 23.9 Guatemala 2,504.7 2,602.3 2,602.3 3.9 0.0 3.9 Honduras 1,499.3 1,532.9 1,510.5 2.2 -1.5 0.7 Nicaragua 4,819.6 .. .. .. .. .. Panamá 1,473.7 .. .. .. .. .. Forest area Colombia 51,519.5 49,649.9 -3.6 Costa Rica 2,124.1 1,965.8 -7.5 El Salvador 192.7 120.2 -37.6 Guatemala 3,383.0 2,851.7 -15.7 Honduras 5,974.9 5,381.9 -9.9 Nicaragua 4,455.4 3,277.8 -26.4 Panamá 3,394.0 2,873.0 -15.4 Notes: Annual crops category includes temporary pasture. .. indicates data not available. Source: World Bank World Development Indicators database. Environmental Economics Series 27 Appendix B -- Degraded Pastures and Silvopastoral Practices Photo 1 -- Silvopastoral systems, in which pastures include significant tree cover, tend to be more productive and to provide a much better habitat for biodiversity. Here, cattle graze in pastures with high tree cover in Esparza, Costa Rica. All images in Appendix B are by Stefano Pagiola Environmental Economics Series 29 Paying for Biodiversity Conservatioon Services in Agricultural Landscapes Photo 2 -- Degraded pasture in Matiguás, Nicaragua Photo 3 -- Improved pastures with high tree cover in Matiguás, Nicaragua 30 Environment Department Papers Appendix B -- Degraded Pastures and Silvopastoral Practices Photo 4 -- Degraded pastures in Quindío, Colombia Photo 5 -- Fodder banks in Quindío, Colombia Environmental Economics Series 31 References Alpizar, L., H.W. Fassbender, and H. Browder, J. 1985. Subsidies, Deforestation, and the Heuveldop. 1983. "Estudio de Sistemas Forest Sector of the Brazilian Amazon. Agroforestales en el Experimento Central Washington: World Resources Institute. del Catie." Turrialba: CATIE (processed). Bulas, J.M. 2004. "Implementing Cost Recovery Barbier, E.B., J.C. Burgess, J. Bishop, and B. for Environmental Services in Mexico." Aylward. 1994. The Economics of the Tropical Paper presented at World Bank Water Week, Timber Trade. London: Earthscan. Washington, 24-26 February 2004. Beinroth, F.H., M.A. Vázquez, V.A. Snyder, P.F. Castro, E. 2001. "Costarrican Experience in the Reich, and L.R. Pérez Alegría. 1996. "Factors Charge for Hydro Environmental Services Controlling Carbon Sequestration in of the Biodiversity to Finance Conservation Tropical Soils." Washington: US Department and Recuperation of Hillside Ecosystems." of Agriculture, Natural Resources Paper presented at the International Conservation Service. Workshop on Market Creation for Biodiversity Products and Services, OECD, Benavides, J. E. (ed.). 1994. Árboles y Arbustos Paris, 25-26 January 2001 (processed). Forrajeros en América Central. Serie Técnica. Informe Técnico No.236. Turrialba: CATIE. Centro Agronómico Tropical de Investigación y Enseñanza (CATIE). 1999. "Capacidad y Binswanger, H. 1991. "Brazilian Policies that Riesgos de Actividades Forestales en el Encourage Deforestation in the Amazon." Almacenamiento de Carbono y World Development, 19, pp.821-829. Conservación de Biodiversidad en Fincas Privadas del Área Central de Costa Rica." Bosch, J.M., and J.D. Hewlett. 1982. "A Review Paper presented at the IV Semana Científica of Catchment Experiments to Determine the del CATIE, Turrialba, Costa Rica. 6-9 April, Effect of Vegetation Changes on Water Yield 1999 (processed). and Evapotranspiration." Journal of Hydrology, 55, pp.3-23. Cordero, D. 2003 "Tarifa de Agua Basada en Costo de Protección de Cuencas, en Bruijnzeel, L.A. 1990. Hydrology of Moist Tropical Heredia." Heredia: Empresa de Servicios Forests and Effects of Conversion: A State of Públicos de Heredia (ESPH) (processed). Knowledge Review. Paris: UNESCO International Hydrological Programme. Current, D., E. Lutz, and S.J. Scherr. 1995. "The Costs and Benefits of Agroforestry to Environmental Economics Series 33 Paying for Biodiversity Conservation Services in Agricultural Landscapes Farmers." World Bank Research Observer, Downstream Cooperation in the Protection 10:2, pp.151-180. of Rural Watersheds." Land-Water Linkages in Rural Watersheds Case Study Series. Dagang, A.B.K., and P.K.R. Nair. 2003. Rome: FAO (processed). "Silvopastoral Research and Adoption in Central America: Recent Findings and Faminow, M.D. 1998. Cattle, Deforestation, and Recommendations for Future Directions." Development in the Amazon: An Economic, Agroforestry Systems, 59, pp.149-155. Agronomic and Environmental Perspective. Oxford: Oxford University Press. Daily, G.C., P.R. Ehrlich, and G.A. Sánchez- Azofeifa. 2001. "Countryside Biogeography: Ferraro, P.J. 2001. "Global Habitat Protection: Utilization of Human Dominated Habitats Limitations of Development Interventions by the Avifauna of Southern Costa Rica." and a Role for Conservation Performance Ecological Applications, 11, pp.1-13. Payments." Conservation Biology, 15:4, pp.1- 12. Deininger, K., S. Jin, B. Adenew, S. Gebre- Selassie, and B. Nega. 2003. "Tenure Ferraro, P.J., and A. Kiss. 2002. "Direct Security and Land-Related Investment: Payments for Biodiversity Conservation." Evidence from Ethiopia." Washington: Science, 298, pp.1718-1719. World Bank (processed). Fisher, M.J., I.M. Rao. M.A. Ayarza, C.E. Dennis, P., L. Shellard, and R. Agnew. 1996. Lascano, J.I. Sanz, R.J. Thomas, and R.R. "Shifts in Arthropod Species Assemblages Vera. 1994. "Carbon Storage by Introduced in Relation to Silvopastoral Establishment in Deep-rooted Grasses in the South American Upland Pastures." Agroforestry Forum, 7:3, Savannas." Nature, 371, pp.236-238. pp.14-21. Fondo Nacional de Financiamiento Forestal Downing, T.E., H.A. Pearson, and C. Garcia- (FONAFIFO). 2000. El Desarrollo del Sistema Downing (eds). 1992. Development or de Pago de Servicios Ambientales en Costa Rica. Destruction: The Conversion of Tropical Forest San José: FONAFIFO. to Pasture in Latin America. Boulder: Westview Press. Global Environment Facility (GEF). 1999. "Experience With Conservation Trust Echevarría, M. 2002a. "Financing Watershed Funds." GEF Evaluation Summary Report Conservation: The FONAG Water Fund in No.1-99. Washington: GEF. Quito, Ecuador." In S. Pagiola, J. Bishop, and N. Landell-Mills (eds.), Selling Forest Gobbi, J. 2002. "Enfoques Silvopastoriles Environmental Services: Market-based Integrados Para el Manejo de Ecosistemas Mechanisms for Conservation and en Colombia, Costa Rica y Nicaragua: Development. London: Earthscan. Análisis Económico-Financiero Ex-Ante de la Inversión en los SSP Propuestos Para Echevarría, M. 2002b. "Water User Associations Cada País." Turrialba: CATIE (processed). in the Cauca Valley: A Voluntary Mechanism to Promote Upstream- 34 Environment Department Papers References Hardner, J., and R. Rice. 2002. "Rethinking Meinzen-Dick, R., A. Knox, F. Place, and B. Green Consumerism." Scientific American, Swallow (eds). 2002. Innovation in Natural May, pp.88-95. Resource Management: The Role of Property Rights and Collective Action in Developing Harvey, C., and W. Haber, 1999. "Remnant Trees Countries. Baltimore: Johns Hopkins and the Conservation of Biodiversity in University Press for IFPRI. Costa Rican Pastures." Agroforestry Systems, 44, pp.37-68. Mertens, B., R. Poccard-Chapuis, M.-G. Piketty, A.-E. Lacques, and A. Venturieri. 2002. Jaramillo, L. 2003. "Estimations of the "Crossing Spatial Analyses and Livestock Opportunity Costs of Forested Land in Economics to Understand Deforestation Mexico." México: Instituto Nacional de Processes in the Brazilian Amazon: the Case Ecología (processed). of Sao Felix do Xingu in South Para." Agricultural Economics, 27, pp.269-294. Kaimowitz, D. 1996. Livestock and Deforestation in Central America in the 1980s and 1990s: A Murgueitio, R.E., 1999. "Sistemas Agroforestales Policy Perspective. Bogor: CIFOR. para la Producción Ganadera en Colombia." Paper presented at the Seminario Kaimowitz, D., B. Mertens, S. Wunder, and P. Intensificación de la ganadería en Pacheco. 2004. Hamburger Connection Fuels Centroamérica: Beneficios Económicos y Amazon Destruction: Cattle Ranching and Ambientales, Turrialba, Costa Rica, May 24- Deforestation in Brazil's Amazon. Bogor: 26. CIFOR. Myers, N. 1981. "The Hamburger Connection: Landell-Mills, N., and I. Porras. 2002. Silver How Central America's Forests Became Bullet or Fools' Gold? A Global Review of North America's Hamburgers." Ambio, 10:1, Markets for Forest Environmental Services and pp.3-8. Their Impact on the Poor. London: IIED. National Center for Environmental Economics Lutz, E., S. Pagiola, and C. Reiche. 1994. "Cost- (NCEE). 2001. The United States Experience Benefit Analysis of Soil Conservation: The with Economic Incentives for Protecting the Farmers' Viewpoint." The World Bank Environment. Report No.EPA-240-R-01-001. Research Observer, 9:2, pp.273-295. Washington; United States Environmental Mahar, D. 1988. "Government Policies and Protection Agency. Deforestation in Brazil's Amazon Region." Pagiola, S. 2002. "Paying for Water Services in Environment Department Working Paper Central America: Learning from Costa No.7. Washington: World Bank. Rica." In S. Pagiola, J. Bishop, and N. Mejía, C. 2004. "Línea de Base de Colombia." In Landell-Mills (eds.), Selling Forest J.A. Gobbi (ed), "Enfoques Silvopastoriles Environmental Services: Market-based Integrados para el Manejo de Ecosistemas: Mechanisms for Conservation and Reporte sobre la Línea de Base." Internal Development. London: Earthscan. Paper No.2. Turrialba: CATIE. Environmental Economics Series 35 Paying for Biodiversity Conservation Services in Agricultural Landscapes Pagiola, S., J. Kellenberg, L. Vidaeus, and J. Pagiola, S., and I.-M. Ruthenberg. 2002. "Selling Srivastava. 1997. Mainstreaming Biodiversity Biodiversity in a Coffee Cup: Shade-grown in Agricultural Development: Toward Good Coffee and Conservation in Mesoamerica." Practice. Environment Paper No.15. In S. Pagiola, J. Bishop, and N. Landell-Mills Washington: World Bank. (eds.), Selling Forest Environmental Services: Market-based Mechanisms for Conservation and Pagiola, S., N. Landell-Mills, and J. Bishop. Development. London: Earthscan. 2002. "Making Market-based Mechanisms Work for Forests and People." In S. Pagiola, Pérez, A.M., I. Siria, M. Sotelo, I. Arana, F. J. Bishop, and N. Landell-Mills (eds.), Selling Ramírez, E. Ramírez, and M. Ibrahim. 2004. Forest Environmental Services: Market-based "Resultados Preliminares del Monitoreo de Mechanisms for Conservation and la Biodiversidad en Sistemas Silvopastoriles Development. London: Earthscan. de Matiguás y Río Blanco, Dpto de Matagalpa, Nicaragua." Gaia, No.5, pp.1-26. Pagiola, S., and G. Platais. 2001. "Selling Biodiversity in Central America." Paper Pfaff, A., S. Kerr, F. Hughes, S. Liu, G. Sanchez, presented at the International Workshop on D. Schimel, J. Tosi, and V. Watson. 2000. Market Creation for Biodiversity Products "The Kyoto Protocol and Payments for and Services, Paris, 25-26 January 2001 Tropical Forest: An Interdisciplinary (processed). Method for Estimating Carbon-offset Supply and Increasing the Feasibility of a Pagiola, S., and G. Platais 2003. "Implementing Carbon Market under the CDM." Ecological Systems of Payments for Environmental Economics, 35:2, pp.203-221. Services: Initial Lessons of Experience." Paper presented at the Workshop on Ramírez, G.H. 1997. "Evaluación de Dos Ecosystem Services in the Tropics: Sistemas Silvopastoriles Integrados por Challenges to Marketing Forest Function, Cynodon Plectostachyus, Leucaena Spring 2003 Meeting of the International Leucocephala y Prosopis Juliflora." Society of Tropical Foresters, Yale Proceedings V Seminario-Taller University, April 5-6, 2003. Internacional Sistemas Sostenibles de Producción Agropecuaria y Primer Pagiola, S., and G. Platais. Forthcoming. Seminario Internacional Palmas en Sistemas Payments for Environmental Services: From de Producción Agropecuaria para el Theory to Practice. Washington: World Bank. Trópico, Cali, July 31-August 3, 1997. Cali: Fundación CIPAV. Pagiola, S., G. Platais, and A. Arcenas. 2003. "Ensuring the Poor Benefit from Systems of Ramírez, E., AY. Cárdenas, and O. Dávila. 2004. Payments for Environmental Services." "Línea de Base de Nicaragua." In J.A. Gobbi Paper presented at the Workshop on (ed), "Enfoques Silvopastoriles Integrados Reconciling Rural Poverty Reduction and para el Manejo de Ecosistemas: Reporte Resource Conservation: Identifying sobre la Línea de Base." Internal Paper Relationships and Remedies, Cornell No.2. Turrialba: CATIE. University, Ithaca, NY, May 2-3, 2003. 36 Environment Department Papers References Repetto, R., and M. Gillis (eds.). 1988. Murdiyarso, F. Stolle, and A.M. Fagi (eds). Government Policies and the Misuse of Forest 1998. Alternatives to Slash-and-Burn in Resources. Cambridge: Cambridge Indonesia: Summary Report and Synthesis of University Press. Phase II. Bogor: ICRAF. Rice, R. 2003. "Conservation Concessions - White, D., F. Holmann, S. Fuijsaka, K. Reategui, Concept Description." Washington: and C. Lascano. 2001. "Will Intensifying Conservation International (processed). Pasture Management in Latin America Protect Forests? - Or is it the Other Way Rojas, M., and B. Aylward. 2002. "The Case of Round?" In A. Angelsen and D. Kaimowitz La Esperanza: A Small, Private, (eds), Agricultural Technologies and Tropical Hydropower Producer and a Conservation Deforestation. Oxford: CAB International. NGO in Costa Rica." Land-Water Linkages in Rural Watersheds Case Study Series. World Bank. 2000. "Ecomarkets Project: Project Rome: FAO. Appraisal Document." Report No.20434-CR. Washington: World Bank. Saunders, D.A., and R.J. Hobbs. 1991. "The Role Of Corridors in Conservation: What do we World Bank. 2002. "Colombia, Costa Rica, and Know and Where do we go?" In D.A Nicaragua Regional Integrated Saunders and R.J. Hobbs (eds.), The Role of Silvopastoral Approaches to Ecosystem Corridors. Surrey: Beaty & Sons. Management Project: Project Appraisal Document." Report No.21869-LAC. Schneider, R. 1994. "Government and the Washington: World Bank. Economy on the Amazon Frontier." LAC Regional Studies Program Report No.34. World Bank. 2003a. "Guatemala Western Washington: World Bank. Altiplano Natural Resources Management Project: Project Appraisal Document." Thiollay, J.-M. 1995. "The Role of Traditional Report No.25660-GUA. Washington: World Agroforests in the Conservation of Rain Bank. Forest Bird Biodiversity in Sumatra." Conservation Biology, 9, pp.335-353. World Bank, 2003b. "State and Trends of the Carbon Market 2003." Washington: World Tipper, R. 2002. "Helping Indigenous Farmers Bank. Participate in the International Market for Carbon Services: The Case of Scolel Té." In Zelek, C.A., and G.E. Shively. 2003. "Measuring S. Pagiola, J. Bishop, and N. Landell-Mills the Opportunity Cost of Carbon (eds.), Selling Forest Environmental Services: Sequestration in Tropical Agriculture." Land Market-based Mechanisms for Conservation and Economics, 79:3, pp.342-354. Development. London: Earthscan. Tomich, T.P., M. Van Noordwijk, A. Budidarsono, A. Gillison, T. Kusumanto, D. Environmental Economics Series 37