79440 AUTHOR ACCEPTED MANUSCRIPT FINAL PUBLICATION INFORMATION The Structure and Composition of a Tropical Dry Forest Landscape After Land Clearance : Azuero Peninsula, Panama The definitive version of the text was subsequently published in Journal of Sustainable Forestry, 30(8), 2011-11-14 Published by Taylor and Francis THE FINAL PUBLISHED VERSION OF THIS ARTICLE IS AVAILABLE ON THE PUBLISHER’S PLATFORM This Author Accepted Manuscript is copyrighted by the World Bank and published by Taylor and Francis. It is posted here by agreement between them. Changes resulting from the publishing process—such as editing, corrections, structural formatting, and other quality control mechanisms—may not be reflected in this version of the text. You may download, copy, and distribute this Author Accepted Manuscript for noncommercial purposes. Your license is limited by the following restrictions: (1) You may use this Author Accepted Manuscript for noncommercial purposes only under a CC BY-NC-ND 3.0 Unported license http://creativecommons.org/licenses/by-nc-nd/3.0/. (2) The integrity of the work and identification of the author, copyright owner, and publisher must be preserved in any copy. (3) You must attribute this Author Accepted Manuscript in the following format: This is an Author Accepted Manuscript of an Article by Griscom, H.P.; Connelly, A.B.; Ashton, M.S.; Wishnie, M.H.; Deago, J. The Structure and Composition of a Tropical Dry Forest Landscape After Land Clearance : Azuero Peninsula, Panama © World Bank, published in the Journal of Sustainable Forestry30(8) 2011-11-14 http:// creativecommons.org/licenses/by-nc-nd/3.0/ © 2013 The World Bank The structure and composition of a tropical dry forest landscape after land clearance; Azuero Peninsula, Panama H.P. GRISCOM 1*, A.B. CONNELLY2, M.S. ASHTON 3, M.H.WISHNIE4, J.DEAGO5 1 James Madison University, Biology Department, MSC 7801, Harrisonburg, Virginia 22807, USA 2 Independent Evaluation Group, The World Bank, Washington, DC 20433, USA 3 Yale School of Forestry and Environmental Studies, 205 Prospect Street, New Haven, CT 06511, USA 4 Equator, LLC, 250 Park Ave S., New York, NY 10003, USA 5 Native Species Reforestation Project (PRORENA), Smithsonian Tropical Research Institute, Apartado 2072 Balboa, Ancon Panama *Corresponding author. email address: griscohp@jmu.edu; tel: 540-908-7294 Abstract Characterization of preexisting flora is an essential preliminary step for successful land rehabilitation projects. This descriptive study was undertaken in a fragmented dry tropical forest region in Panama. Five different habitat types were selected: active pasture, two-year and five-year abandoned pastures, forested riparian zones, and a forest fragment. Species richness, density, basal area, dispersal modes, and phenology of trees as well as their uses were determined. Diversity of pre-existing seed resources as well as natural regeneration was poor after two and five years post cattle removal, suggesting that at an early successional stage, enrichment planting is necessary. Guazuma ulmifolia and Cordia alliodora dominated the pastoral landscape, representing 63% of all inventoried trees in the active pasture. More than half the trees within pastures (76%) had some use, with shade for cattle as the most common reason for leaving trees in the landscape. The largest trees and the greatest diversity were found within the less managed forested riparian zones because of inaccessibility and water conservation. The pastoral landscape is largely shaped by farm management as well as ecological selection process which will in turn affect successional processes. KEYWORDS. Guazuma ulmifolia, pasture, riparian forest, reforestation, regeneration, succession, Tabebuia 2 Acknowledgements: We would like to thank The Achotines Laboratory of the International Tropical Tuna Commission and Manager Vernon Scholey for the use of their research and lodging facilities; Oswaldo and Pedro Batista for permission to conduct experimental research on their land; Dario and Dilsa Ramirez for providing an introduction to their community. Logistical support was provided by PRORENA (Panama Native Species Reforestation Project), Smithsonian Tropical Research Institute, and the Yale School of Forestry and Environmental Studies. The research would not have been possible without the technical support of field assistants, Daniel Mancilla and Belgis Madrid or the generous time and knowledge provided by the farmers of Los Asientos and Oria Arriba. Financial support was provided by the USDA fellowship through the Yale School of Forestry and Environmental Studies and The New York Botanical Garden. 3 INTRODUCTION Dry tropical forests were the most common forest type along the Pacific coast of Central America until cattle ranching transformed the landscape in the early part of the last century (Janzen, 1983; Calvo et al., 2009). In Central America, pastures remained productive for decades (Reiners et al., 1994; Aide et al. 1996; Thomlinson et al., 1996; Holl, 1999; Harvey & Haber, 1999), most notably in dry regions where soil fertility was higher (Murphy & Lugo, 1986). Eventually, high cattle stocking rates, repeated burning, and the removal of native vegetation resulted in severe soil erosion followed by a decline in land productivity (Heckadon, 1984; Janzen, 1988). Cattle ranching is no longer a viable option for many land holders and with the recent increase in land prices along the coast, selling is a desirable option (Ankerson, 2005; Kull et al., 2010). Consequently, there is an opportunity to rehabilitate one of the least known and most threatened tropical ecosystems because many new landowners place a high value on diversity, forest cover, and ecosystem health (Ankerson, 2005; Kull et al., 2010). Most of these landowners are purchasing land for vacation, recreational purposes, or the development of resorts. Natural history descriptions of the flora and fauna within these regions are essential for a baseline understanding of the successional trajectory of old pastures. Old pastures on the Azuero peninsula in Panama may follow several developmental pathways. The land may remain a savanna in a permanent state of arrested forest succession where grass perpetuates the fire disturbance regime (Hopkins, 1983; Olivares and Medina, 1992); land owners may convert the land to teak (Tectona grandis L.f.) 4 plantations (Sloan, 2008), or the land may be rehabilitated into a native forest ecosystem that provides ecological and economic services. Natural regeneration may be the most cost-effective and realistic strategy for rehabilitating pastures to one with productive uses (Lugo, 1988, Aide, 2000; Hooper et al., 2004). Other more intensive strategies, such as reforestation with exotic or native nurse tree plantation systems, although highly effective (Parotta, 1992; Guariguata et al., 1994; Ashton et. al., 1997; Montagnini, 2001; Wishnie et al., 2007) are more costly. This choice depends on economics, past management history and the current distribution, diversity, abundance, autecology, and phenology of relic trees and fragments. The dominant type of dispersal is critical to the colonization of abandoned sites. Many tree species in dry tropical forests (40-70%) have been reported to be dispersed by animals (Janzen, 1988; Castilleja, 1991; Gentry, 1995; Vieira et al., 2006). Animal- dispersed species, especially large-seeded species, may be lost through time in landscapes devoid of forest fragments (Nepstad et al., 1996). With the exception of howler monkeys, which usually stay within riparian zones, large seed-dispersing mammals (e.g. agoutis, tapirs, spider monkeys), are noticeably absent from this study site. Janzen (1988) observed that wind-dispersed species dominated the regenerating cohort in ten year old abandoned pastures suggesting that animal-dispersed species are lost through time in dry tropical systems. Pasture trees that persist in the landscape are generally species that are fire-tolerant and readily resprout (Janzen, 1983; Swaine, 1992; Hooper et al., 2002; Griscom et al., 2009). Characterizing the landscape also involves understanding the rationale for why farmers leave certain trees. The number and diversity of pasture trees is a function of 5 farmer’s perceptions. Isolated trees or trees within riparian zones are selectively removed or conserved based on their perceived beneficial and detrimental characteristics. In Central America, farmers frequently leave trees or plant trees in the form of live fences, isolated trees, and riparian forests (Budowski, 1981; Guevara et al., 1986; Harvey & Haber, 1999; Estrada et al., 2000, Harvey et al.,2005). Isolated trees are left in the landscape to provide shade and fodder, construction material, firewood and fence posts (Budowski, 1981; Guevara et al., 1986; Harvey & Huber, 1999; Barrance et al., 2003; Cordero and Boshier, 2003; Leon and Harvey, 2006; Garen et al., 2009) while trees within riparian zones are conserved to provide shade and cool water for the cattle (Garen et al., 2009). These riparian zones are typically thin, degraded stripes of vegetation, representing the most common forest habitat remaining in a pasture matrix (Graeme & Blake, 2001). Pasture trees also have important ecological values, some of which are recognized by farmers. Trees help prevent soil erosion (Nair, 1991; Bird et al., 1992), they increase water quality by reducing sediment and nutrient run-off (Fennessey & Cronk, 1997), and they serve as seed sources for timber species (Uhl et al., 1981; Guevara & Laborde, 1993). In addition, trees attract animal seed-dispersers into the pastures by functioning as foraging, nesting, perching and/or roosting sites (Guevara et al., 1986; Nepstad et al., 1991; Gerhardt and Hytteborn, 1992; Guevara et al., 1992; Guevara et al., 1993; Estrada et al., 1993; Silva et al., 1996; Aide, 2000; Griscom et. al, 2007). Land-owners must balance the benefits of keeping trees on their land with other concerns, such as tree competition with forage grass (Harvey & Huber, 1999; Barrance et al., 2003) and water flow. 6 Pastures have been inventoried in dry regions in Columbia (Cajas-Giron & Sinclair, 2001) and Honduras (Gordon et al., 2003) but these studies focused on tree uses. Other studies in dry tropical ecosystems have conducted inventories in secondary forests (Taylor, 1963; Holdridge, 1967; Janzen, 1983; Murphy and Lugo, 1986; Sabogal, 1992; Mizrahi et al., 1997; Colon et al., 2006), riparian forests (Taylor, 1963; Elizondo and Jimenez, 1988; Meave and Kellman, 1994; Glander et al, 1996; Pither and Kellman, 2002), chronosequences of successional stages (Ruiz et al., 2005; Lebrija-Trejos et al., 2008; Marin et al., 2009; Powers et al., 2009), or active pastures (Esquivel et al. 2008). Few studies have compared secondary forest, riparian forest, active pasture and recently abandoned pastures in terms of species richness and basal area as well as their usefulness in the landscape. For the following study, we selected five types of habitat (active cattle pasture, riparian forest within an active pasture, secondary forest fragment, and recently abandoned pasture of two ages). The active pasture, riparian forest and forest fragment provided data on the existing capital of the landscape, in terms of a seed source while the two year abandoned pasture and five year abandoned pasture provided data on expected successional pathway trends in terms of species density, basal area and diversity. Active pastures and riparian forest were common habitat types within this landscape whereas forest fragments ( > 20 ha) and abandoned pastures were rare. However, a research laboratory purchased land in the area which consisted of a 70 ha secondary forest and two parcels of active pasture protected from cattle and fire at two different time periods. This particular study site was also the location of a long-term study site on forest restoration (Griscom et al., 2005, 2007, 2009). 7 The aim of our study was two fold: 1) to characterize and compare the floristics of different habitats and successional stages by quantifying the following response variables: species richness, stem count, basal area, and phenology and 2) to determine the rationale for selectively leaving certain species in the landscape by conducting interviews of farmers in the area. This characterization will enable us to better assess the rehabilitation potential of one farm, which can serve as a model for how to proceed with the restoration of other similar pastures in the area. METHODS Site Description This study was conducted in the Los Santos province on the Azuero peninsula, Panama (7º 15’30’’ N, 80º00’15’’ W). The hilly, relatively steep topography was the last land to be cleared in the province. Valuable timber (Cedrela odorata L., Dalbergia retusa Hemsl, and Pachira quinata (Jacq.) Dugand.) was selectively logged in the early part of the 1900’s. In the 1940’s and 1950’s, the land was cleared for cattle ranching. Conversion to pasture accelerated in 1978 following construction of the road between the towns of Pedasí and Tonosí (Heckadon, 1984). Cattle grazing is currently the dominant land use in the region but many large parcels are being sold off to landowners who are interested in alternative management options, such as vacation homes, recreation, and environmental restoration. The study site receives an average of 1700mm of rainfall. This precipitation range falls into the dry tropical forest ecosystem classification although temperature, 8 elevation and evapotranspiration are other important considerations of the Holdridge life zone classification (Holdridge, 1967). The dry season is pronounced with five months out of the year receiving no rain (December through March). Rain begins in late April and ends in late November. The annual average temperature is 25º. The undulating terrain, ranging in elevation from 10 to 100 meters, is a mosaic of pastures planted with African grasses Hyparrhenia rufa (Nees) Stapf and Panicum maximum Jacq., forested riparian zones, isolated trees, and live fences. The following mosaic of habitats were selected because they were part of a long- term restoration project and in close proximity to each other (thus, controlling for more heterogeneity in the landscape): an active pasture, riparian forests within the active pasture, a protected secondary forest, a two year abandoned pasture and a five year abandoned pasture. The active pasture and riparian forest were part of an eighty-five hectare cattle farm that had been ranched for fifty years. The active pasture was grazed by a herd of 50 to 70 Brahman cattle from mid-July until mid-March. The secondary forest and abandoned pasture sites were part of an adjacent 100 hectare property. Prior to1985, the property had been grazed by cattle and burned for approximately 30 years. The forest had never been cleared for pasture but had most likely been grazed by cattle in the understory, burned, and selectively logged. Sampling Design Several different sampling methods were used to inventory tree species (> 5 cm dbh) within this landscape due to constraints of each forest type. Within the 85 ha active pasture, linear riparian zones were inventoried in transects while open active pasture were 9 inventoried as circular plots. Four forested riparian fragments within active pastures were randomly selected from eight riparian forested areas in the active pasture. Riparian fragments were inventoried by randomly selecting two points along the stream (at least 20 meters apart) and establishing two rectangular subplots (10m width by 100 meters length) parallel to the bank of the stream. Eight circular plots of 50 meter radius were randomly selected with the active pasture. Plots were located within different watersheds. Within the 100 ha property, four 100 m2 plots, at least 100 meters apart, were randomly selected within a two-year abandoned pasture, a five-year abandoned pasture and a secondary forest. All plots were replicated within a habitat type but not at the landscape scale because protected forest fragments and recently abandoned pasture were unusual for this area. In each of the three habitat types, individuals greater than 10 cm dbh were identified and measured for basal area. Stem count was also calculated. Three individuals for each of the twenty most common wind and animal- dispersed species in the pasture were randomly selected from 100 individuals and permanently tagged for monitoring phenology. Once a month for two years, the absence or presence of flowers and fruit were recorded. Presence was determined if greater than 20% of the canopy were fruiting or flowering. Phenology was monitored to determine fruit availability and to identify species that provided fruit when little else was available. Semi-structured interviews were carried out with 36 landholders selected at random from a list of 926 farms registered with the Ministry of Agrarian Reform (Summer, 2003). The landowners were selected from the registration list for the corregimientos of Oria Arriba and Los Asientos in the province of Los Santos. All the 10 landowners had secure land rights and were devoted to cattle ranching. Most of the farmers interviewed relied on their farm as their primary, though not always exclusive, source of income. Interviews included a walking tour of the farm. The objective was to determine common uses of trees in farms and rationale for keeping trees in the landscape. Analysis Importance value indices (Curtis and McIntosh, 1951) for each species at all sites were calculated as the sum of relative density, relative frequency and relative basal area. The qualitative similarity between habitats was calculated with Jaccard’s Index (Krebs, 1989) as follows: Cj = a / (a + b + c) where C is the index, “a” represents the number of tree species present in both habitats, “b” represents species present only in the first habitat, and c represents species present only in the second habitat. The presence of a species in a habitat was scored as 1, and its absence was scored as 0. EstimateS (Version 8.0) (Chaos Estimate) was used to statistically estimate species richness in the different habitat types. Stem density and basal area could not be statistically analyzed as plot size varied in different habitat types. However, these variables were standardized as per hectare to allow for comparisons between habitats. RESULTS Dominant Species The most important species varied between habitat types although the Leguminosae family had the most number of species regardless of habitat type. In the secondary forest, Calycophyllum candidissimum (Vahl) DC. (IVI = 15.37) and Tabebuia rosea (Bertol.) 11 DC. (IVI = 13.99) had the greatest importance values whereas in the riparian forest, Guazuma ulmifolia (Lam) (IVI = 16.87) and Hura crepitans L. (IVI = 10.87) had the greatest importance values (Table 1). In the active pasture, Guazuma ulmifolia (IVI = .5) and Cordia alliodora (Ruiz & Pav.) Oken (IVI = .33) had the greatest importance values. Guazuma ulmifolia and C. alliodora represented 63% of the tree species in active open pasture. In the abandoned pasture of two and five years, very few species were represented and Guazuma ulmifolia dominated, representing 92% and 37% of the stems, respectively. By comparison, in the riparian zones, Guazuma ulmifolia represented 22% of the total stems. The only habitat type where G. ulmifolia was rarely encountered was within the secondary forest. (INSERT TABLE 1) Dispersal Modes Within the secondary forest, wind-dispersed species were slightly more common than animal-dispersed species in terms of species number (55%) and individuals (57%), in part due to the dominance of two wind-dispersed species, Calycophyllum candidissimum and Tabebuia rosea. The most common animal-dispersed species in the secondary forest fragment were small understory trees, Genipa americana Linn. and Pouteria campechiana (Kunth) Baehni, both of which have large fleshy fruits dispersed by bats and monkeys. Within the riparian zones, 64% of the individuals (and 52% of the species) were animal-dispersed. In the active pasture, wind and cattle were the dominant seed dispersal modes due to the abundance of Guazuma ulmifolia and Cordia alliodora. Fifty-five percent of the total stems inventoried were dispersed by animals (representing 15 species) and 44% by 12 wind (representing 11 species). In the abandoned pasture of five years, 41% percent of the total stems inventoried were dispersed by animals (representing 4 species) and 58% by wind (representing 7 species). In the abandoned pasture of two years, 92% percent of the total stems inventoried were dispersed by animals (representing 1 species, G. ulmifolia) and 8% by wind (representing 2 species) (Table 1). Stem count and basal area The forest fragment had a greater stem count (900/ha) than any of the other habitat types, including the forested riparian zone (180 stems/ha) (Table 2). However, basal area in the forest fragment (13 m2/ha) was similar to riparian zones (15.3 m2/ha), both of which were greater than the other habitat types (p<0.05). In terms of size classes, within the two year and five year abandoned pasture, 100% of the trees were less than 15 cm dbh (Fig. 1). Seventy-two percent of the trees were less than 15 cm dbh within the secondary forest. Forested riparian zones and active pastures had the largest diameter trees with only 35% and 43% of the trees in the smaller diameter size class (< 15 cm dbh). (INSERT FIGURE 1 HERE) Species Richness Of all habitat types, riparian zones had the greatest estimated number of species (N=38) according to the rarefraction curves of EstimateS (Table 2). A total of 31 species in 16 plant families were identified in the riparian zones. However, species richness varied greatly between forested riparian zones. The most species rich riparian forest had 21 species whereas the most species poor riparian forest had 8 species. The species poor riparian zones were dominated by Guazuma ulmifolia. (INSERT TABLE 2 HERE ) 13 Within the secondary forest a similar number of species were identified (29 species in 15 plant families). Number of species in the forest was estimated to be 30 species, similar to the active pasture. Active pastures were represented by 25 tree species. The active pasture and the riparian zone were the most similar according to Jaccard’s Index (Table 3). Only 12 species in 8 plant families were inventoried in the five year abandoned pasture and three plant species were identified in two year abandoned pasture. Five year pastures were estimated to have 14 species and two year pastures were estimated to have 7 species. (INSERT TABLE 3 HERE) Phenological Data All recorded wind-dispersed species (except for Cedrela odorata L.) flowered during the dry season (December-February) (Fig. 2). The most common month for flowering was February, in the middle of the dry season. Animal-dispersed species were more variable with the majority of flowering beginning in February and lasting through May. However, most animal-dispersed tree species flowered from March through May when wind- dispersed species were in fruit. Animal-dispersed species were then in fruit during the rainy season (May through November), when wind-dispersed species were no longer fruiting. Only four animal-dispersed species had fruit during the dry season (Bursera simaruba (L.) Sarg., Byrsonima crassifolia (L.) Rich. Cecropia peltata (L.), and G. ulmifolia). (INSERT FIGURE 2 HERE) 14 Interviews The majority of species in the pasture inventories (76%) had at least one use recognized by local farmers. Trees were reported to provide shade and supplemental fodder for cattle, and to supply construction material, live fence posts, and firewood. The most common use of pasture trees was to provide shade for cattle (29 species, 64% of the species). Tree shade was reported to be essential for the health and survival of the cattle during the five month dry season. Many trees also had timber value (16 species, 35% of the species). However, timber is usually used directly by landowners for construction, rather than sold into timber markets. Occasionally, farmers retain or actively plant high value timber species (e.g. Cedrela odorata and Cordia alliodora) to diversify their income. With the exception of Cordia alliodora, all of the timber species inventoried in pastures had other uses. Like timber, trees harvested for firewood (4 species, 9% of the species) had at least two additional purposes. Live fence posts were represented by 23% of the species (8 species total) while fodder was associated with 18% of the species (9 species) (Table 1). Edible fruit (Bysonima crassifolia, Hymenaea courbaril L., Manilkara achras (Mill) Fosberg, medicine (Bursera simaruba, Hymanaea courbaril), and fish poison (Hura crepitans L.) accounted for less than 10% of inventoried species. Half of the species had multiple uses (51% of the species). Trees with at least three different uses were the following: Bursera simaruba, Byrsonima crassifolia, Calycophyllum candidissimum (Vahl) DC. Cedrela odorata, Cochlospermum vitifolium Willd (Spreng), Enterolobium cyclocarpum Jacq. Griseb, Genipa americana, Guazuma ulmifolia, and Pachira quinata (Table 1). 15 DISCUSSION Tree diversity within pastures is shaped by management and ecological constraints. These processes in turn determine the ability of abandoned pastures to regenerate. The majority of trees at this study site on the Azuero peninsula had some recorded use and most species had multiple uses. Farmers selectively leave certain tree species in pastures but most species will eventually be harvested for construction after they reach a certain size. Within the active pasture, almost 50% of the trees were less than 15 cm dbh. Two and five years later after cattle had been removed from the pastures (and the land sold), all trees were less than 15 cm dbh, suggesting timber harvesting prior to sale. Enterolobium cyclocarpum (Jacq.) Griseb., a common, large remnant tree in pastures, is increasingly used for construction because the locally preferred timber species (Cedrela odorata) has become scarce. A few species of remnant trees dominated active and abandoned sites. Within active pastures, Guazuma ulmifolia and Cordia alliodora represented 63% of all inventoried trees. Guazuma ulmifolia also had the highest importance value in a secondary forest in Nicaragua (Sabogal, 1992). After pastures had been abandoned for three to five years, G. ulmifolia was one of the few trees remaining as well as successfully regenerating. Guazuma ulmifolia and C. alliodora were both abundant in active and abandoned pastures because they are particularly resilient to fire and grazing. They have no reproductive barriers as G. ulmifolia is dispersed by cattle and is a prolific basal stem sprouter and C. alliodora is dispersed by wind and readily sprouts from roots. These two species are also used by farmers either for livestock shade and fodder (G. 16 ulmifolia) (Janzen, 1982; Sabogal, 1992; Cajas-Giron & Sinclair, 2001) or timber (C. alliodora) (Cajas-Giron & Sinclair 2001). Of the farmers we interviewed, shade for cattle was the most common reason for leaving trees in the landscape. Interestingly, this trend varies by region. In Los Santos, where this study took place, 71% of farmers reported leaving trees for food and shade whereas in the very dry tropical region of Rio Hato in Panama, no farmers left trees for cattle (Garen et al, 2009). Several farmers informed us that in years of drought when pasture grass is scarce, their cattle have almost exclusively survived on the fruit and leaves of G. ulmifolia. Guazuma ulmifolia was partially deciduous, losing 50-75% of its leaves during the dry season. According to farmers, this species does not deplete water resources within riparian zones as fast as other species, such as Hura crepitans. Thus, farmers selectively leave G. ulmifolia within areas close to water, which explains why this species was the most commonly encountered along some streams. Almost all farmers reported leaving trees within riparian zones to protect water sources (Garen et al., 2009). The dominance of Guazuma ulmifolia is expected to decrease as succession proceeds as it is a light-demanding species. In Costa Rica, Kalacska et al. (2004) found that although G. ulmifolia was one of the two most common species in the early successional stage, its importance declined through time. We also recorded very few G. ulmifolia trees within the secondary forest fragment. Instead, Calycophyllum candidissimum had the greatest importance value, similar to inventories conducted in forests of Santa Rosa, Costa Rica (Janzen, 1983). This may be because it was selected 17 for within forest fragments as it has multiple uses (construction, firewood, live fence). The second most common species, Tabebuia rosea, is also a useful timber species. Less managed, relic forests were found within forested riparian zones bordering perennial streams. These areas had an even size class distribution, the largest relic trees (> 45 cm dbh) and the greatest species richness. Riparian forests were often never cleared either because farmers desired shade trees, water protection (Garen et al., 2009) or because they were located at the base of steep, inaccessible slopes. The more accessible forested streams are likely to be more intensively managed. Farmers often girdle tree species (e.g., Hura crepitans) due to perceived competition with cattle for water during the dry season. The Pacific coast of Panama has potential, despite its highly disturbed appearance. From this study, we have achieved a better understanding of the missing or unbalanced arboreal elements of the landscape. Forested fragments, especially along inaccessible riparian zones, are critical to successful restoration projects as these areas are where the greatest diversity of tree species may be found. Recently abandoned pastures were extraordinarily low in tree numbers and species with Guazuma ulmifolia dominating the landscape even after five years post abandonment. These areas will require more intensive restoration efforts to accelerate forest succession. Enrichment planting with other fleshy-fruited tree species would increase diversity of the flora and fauna. Seeds could be collected from mature forests and riparian zones and planted in species poor sites. A few species were especially important because they fruit in the dry season when resources are limiting (e.g., Bursera simaruba and Cecropia peltata). In some cases, 18 such as with the two year abandoned pasture with no diversity of remnant trees, plantations of native species are recommended to increase diversity. 19 REFERENCES Aide, T.M., Zimmerman, J.K.M., Rosari, M., & Marcano, H. (1996). Forest recovery in abandoned cattle pastures along an elevational gradient in northeastern Puerto Rico. Biotropica, 26, 537-548. Aide, T.M. (2000). Forest regeneration in a chronosequence of tropical abandoned pastures: implications for restoration ecology. Restoration Ecology, 8, 328-338. Ankersen, T.T., Regan K.E., & Mack, S.A., (2006). Towards a bioregional approach to tropical forest conservation: Costa Rica’s greater Osa bioregion. Futures, 38, 406-431. Ashton, P.M.S., Gamage, S., Gunatilleke, I.A.U.N., & Gunatilleke, C.V.S. (1997). Restoration of a Sri Lankan rainforest: using Caribbean pine Pinus caribea as a nurse for establishing late-successional tree species. Journal of Applied Ecology, 34, 915-925. Barrance, A.J., Flores, L., Padilla, E., Gordon, J.E., & Schreckenberg, K. (2003). Trees and farming in the dry zone of southern Honduras I: campesino tree husbandry practices. Agroforestry Systems 59, 97-106. Budowski, G. (1981). Agroforestry in Central America. In: Agroforestry: proceedings of a seminar held in CATIE (pp 13–40). Turrialba, Costa Rica. Bird, P.R., Bicknell, D., Bulman, P.A., Burke, S.J.A., Leys, J.F., Parker, J.N., Vandersommen, F.J., & Voller, P. (1992). The role of shelter in Australia for protecting soils, plants, and livestock. Agroforestry Systems, 20, 59-86. 20 Cajas-Giron,Y.S., & Sinclair, S.F. (2001). Characterization of multistrata silvopastoral systems on seasonally dry pastures in the Caribbean Region of Colombia. Agroforestry Systems, 53, 215-225. Calvo-Alvarado A., McLennan, B., Sanchez-Azofeifa A., & Garvin T. (2009). Deforestation and forest restoration in Guanacaste, Costa Rica: Putting conservation policies in context. Forest Ecology and Management, 258, 931– 940. Colon, S.M. (2006). Recovery of a subtropical dry forest after abandonment of different land uses. Biotropica, 38, 354-364. Curtis, J.T., & McIntosh, R.P. (1951). An upland forest continuum in the prairie-forest border region of Wisconsin. Ecology, 32, 476-496. Elizondo, C.L.H., & Jiménez, Q.M. (1988). El Escobio. A forested savannah in Liberia, Guanacaste. Costa Rica. Revista de Biologia Tropical, 36, 175-186. Esquivel, M.J., Harvey, C.a., Finegan, B., Casanoves, F., & Skarpe, C. (2008). Effects of pasture management on the natural regeneration of neotropical trees. Journal of Applied Ecology, 45, 371-380. Estrada, A., Coates-Estrada, R., & Meritt, D. (1993). Bat species richness and abundance in tropical rain forest fragments and in agricultural habitats at Los Tuxlas, Mexico. Ecography, 16, 309-318. Estrada, A., Cammarano, P., & Coates-Estrada, R. (2000). Bird species richness in vegetation fences and in strips of residual rain forest vegetation at Los Tuxlas, Mexico. Biodiversity and Conservation, 9, 1399-1416. 21 Fennessy, M.S., & Cronk, J.K. (1997). The effectiveness and restoration potential of riparian ecotones for the management of nonpoint source pollution, particularly nitrate. Critical reviews in environmental science and technology, 27, 285-317. Gerhardt, K., Hytteborn, H. (1992). Natural dynamics and regeneration methods in tropical dry forests: An introduction. Journal of Vegetation Science, 3, 361–364. Glander, K., Nisbett, E., & Richard, A. (1996). Community structure and species density in tropical dry forest associations at Hacienda la Pacifica in Guanacaste Province, Costa Rica. Brenesia, 45-46, 113-142. Gordon, J.E., Hawthorne, W.D., Sandoval, G., & Barrance, A.J. (2003). Trees and farming in the dry zone of southern Hondurus II: the potential for tree diversity conservation. Agroforestry Systems, 59, 107-117. Graham, C.H., & Blake, J.G. (2001). Influences of patch and landscape level factors on bird assemblages in forest patches and edge in Veracruz, Mexico. Ecological Applications, 11, 1709–1721. Griscom, H.P., Ashton, P.M.S., & Berlyn, G.P. (2005). Seedling survival and growth of three native species in pastures: implications for dry forest rehabilitation. Forest Ecology and Management, 218, 306-318. Griscom, H.P., Kalko, E.K.V., & Ashton, M.S. (2007). Frugivory by Small Vertebrates Within a Deforested, Dry Tropical Region of Central America. Biotropica, 39, 278-282. 22 Griscom, H.P., Griscom, B.W., & Ashton, P.M.S. (2009). Forest regeneration from pasture in the dry tropics of Panama: effects of cattle, exotic grass, and forested riparia. Restoration Ecology, 17, 117–126. Guariguata, M., Rheingans, R., & Montagnini, F. (1994). Early woody invasion under tree plantations in Costa Rica: implications for forest restoration. Society for Economic Restoration, 3, 222-260. Guevara, S., Purata, S., & Van der Maarel, E. (1986). The role of remnant forest trees in tropical secondary succession. Vegetatio, 66, 77-84. Guevara, S., Meave J., Moreno-Casasola,P., & Laborde, J. (1992). Floristic composition and structure of vegetation under isolated trees in neotropical pastures. Journal of Vegetation Science, 3, 655-664. Guevara, S., & Laborde, L. (1993). Monitoring seed dispersal at isolated standing trees in tropical pastures: consequences for local species availability. Vegetatio, 107/108, 319–338. Harvey, C.A., & Haber, W.A. (1999). Remnant trees and the conservation of biodiversity in Costa Rican pastures. Agroforestry Systems, 44, 37-68. Harvey, C.A., Villaneuva, C., Villacis, J., Chacon, M., Munoz, D., Lopez,M., Ibrahim, M., Gomez, R., Taylor, R., Martinez, J., Navas, A., & Saenz, J. (2005). Contribution of live fences to the ecological integrity of agricultural landscapes. Agriculture, Ecosystems, and Environment. 111, 200-230. 23 Heckadon-Moreno, S. (1984). Panama’s Expanding Cattle Front: The Santeño Campesinos and the Colonization of the Forests. Ph. D. dissertation, University of Essex. Holdridge, L.R. (1967) Life zone ecology. Tropical science center, San Jose, Costa Rica. Holl, K. D. (1999). Factors limiting tropical rain forest regeneration in abandoned pasture: seed rain, seed germination, microclimate, and soil. Biotropica, 31, 229– 242. Hooper, E.R., Condit, R., & Legendre, P. (2002). Responses of 20 native tree species to reforestation strategies for abandoned farmland in Panama. Ecological Applications 12, 1626–1641. Hopkins, B. (1983). Successional processes. In: Bourliere, F. (Ed.), Tropical Savannas, Elsevier, New York, pp. 605–616. Janzen, D. (1982). Natural history of Guacimo fruits (Sterculiaceae, Guazuma ulmifolia) with respect to consumption by large mammals. American journal of botany, 69, 1240-1250. Janzen, D. (1983). Costa Rican Natural History. The University of Chicago Press, Chicago. Janzen, D. (1988). Management of habitat fragments in a tropical dry forest: growth. Annuals of the Missouri Botanical Garden, 75, 105-116. Krebs, C.J. (1989). Ecological Methodology, Harper Collins Publishers, New York. Kull, C.A., Ibrahim, C.K., & Meredity, T.C. (2010). Tropical forest transitions and globalization: neo-liberalism, migration, tourism, and international conservation agendas. Society and Natural Resources, 20, 723-737. 24 Lebrija-Trejos, E., Bongers, F., Perez-Garcia, E.A., & Meave, J.A. (2008). Successional change and resilience of a very dry tropical deciduous forest following shifting agriculture. Biotropica, 40, 422-431. Leon, M.C., Harvey, C.A. (2006). Live fences and landscape connectivity in a neotropical agricultural landscape. Agroforestry Systems, 68, 15-26. Lugo, A.E. (1988). The future of the forest: ecosystem rehabilitation in the tropics. Environment, 30, 17-45. Marin, G.C., Mulualem, T., Gonzalez-Rivas, B., Oden, P.C. (2009). A chronosequence analysis of forest recovery on abandoned agricultural fields in Nicaragua. Journal of Forestry Research, 20, 213-222. Meave, J., & Kellman, M. (1994). Maintenance of rain forest diversity in riparian forests of tropical savannas. Implications for species conservation during Pleistocene drought. Journal of Biogeography, 21, 121-135. Mizrahi, A., Prado, J.M.R., Jimenez-Osornio, J. (1997). Composition, structure and management potential of secondary dry tropical vegetation in two abandoned henequen plantations of Yucatan, Mexico. Forest Ecology and Management, 96, 273-282. Montagnini, F. (2001). Strategies for the recovery of degraded ecosystems: Experiences from Latin America. Interciencia, 26, 498-513. Murphy, P.G., Lugo, A.E. (1986). Ecology of tropical dry forest. Annual Review of Ecology and Systematics, 17, 67-88. 25 Nair, P. (1991). State of the art of agroforestry systems. Forest Ecology and Management, 45, 5-29. Nepstad, D., Uhl, C., & Serrao, E. (1991). Recuperation of a degraded Amazonian landscape: forest recovery and agricultural restoration. Ambio, 20, 248-255. Nepstad, D., Uhl C., & da Silva, J.M.C. (1996). A comparative study of tree establishment in abandoned pasture and mature forest in eastern Amazonia. Oikos, 76, 25–39. Olivares, E., & Medina, E. (1992). Water and nutrient relations of woody perennials from tropical dry forests. Journal of Vegetation Science, 3, 383-392. Parrotta, J. (1992). The role of plantation forests in rehabilitating degraded tropical ecosystems. Agriculture, Ecosystems, and Environment, 41, 115–133. Pither, R., & Kellman, M. (2002). Tree species diversity in small, tropical riparian forest fragments in Belize, Central America. Biodiversity and Conservation, 11, 1623- 1636. Powers, J.S., Becknell, J.M., Irving, J., & Perez-Aviles, D. (2009). Diversity and structure of regenerating tropical dry forests in Costa Rica: Geographic patterns and environmental divers. Forest Ecology and Management, 258, 959-970. Reiners, WA., Bouwman, A.F., Parson, W.F.J., & Keller, M. (1994). Tropical rain-forest conversion to pasture-changes in vegetation and soil properties. Ecological Applications, 4, 363-377. Ruiz, J., Fandino M.C., & Chazdon et al, R.L. (2005). Vegetation structure, composition, and species richness across a 56-year chronosequence of dry tropical forest on Providencia island, Colombia. Biotropica, 37, 520-530. 26 Sabogal, C. (1992). Regeneration of tropical dry forests in Central America, with examples from Nicaragua. Journal of Vegetation Science, 3, 407-417. Silva, J.M.C., Uhl, C., & Gregory, M. (1996). Plant succession, landscape management, and the ecology of frugivorous birds in abandoned Amazonian pastures. Conservation Biology, 10, 491-503. Sloan, S. (2008). Reforestation amidst deforestation: Simultaneity and succession. Global Environmental Change, 18, 425– 441 Swaine, M.D. (1992). Characteristics of dry forest in West Africa and the influence of fire. Journal of Vegetation Science, 3, 365-374 Taylor, B.W., 1963. An outline of the vegetation of Nicaragua. Journal of Ecology, 51, 27-54. Thomlinson, J.R., Serrano, M.I., Lopez, T.D., Aide, T.M., & Zimmerman, J.K. (1996). Land-use dynamics in a post-agricultural Puerto Rican landscape (1936-1988). Biotropica, 28, 525-536. Uhl C., Clark K., Clark, H., & Murphy, P. (1981). Early plant succession after cutting and burning in the upper Rio Negro region of the Amazon Basin. Journal of Ecology, 69, 631-649. Uhl C., Buschbacher, R., & Serrao, E. (1988). Abandoned pastures in Eastern Amazonia. Patterns of Plant Succession. Journal of Ecology, 76, 663-681. Vieira, D.L.M., Scariot, A. (2006). Principles of natural regeneration of tropical dry forests for restoration. Restoration Ecology, 14, 11-20. Wishnie, M.H., Dent, D.H., Mariscal, E., Deago, J., Cedeno, N., Ibarra, D., Condit, R., & Ashton, P.M.S. (1997). Initial performance and reforestation potential of 24 27 tropical tree species planted across a precipitation gradient in the Republic of Panama. Forest Ecology and Management, 243, 39-49. 28 Table 1. Plant species importance values by inventory location. Importance values were calculated as the sum of relative density, relative frequency and relative basal area. Tree uses are denoted as the following: SH = shade, FO = fodder, TI = timber, FR = fruit, FI = firewood, LF = live fence, M = medicine. Inventories were conducted in secondary forest (SF), forested riparian fragment (RZ), active pasture (AP), five-year abandoned pasture (5-AP), and two-year abandoned pasture (2-AP). Plant Species Dispersal Use Plant Family SF RZ AP 5-AP 2-AP Acacia collinsii Animals Leguminosae 1.279 0.03 Albizia adinocephala Wind Leguminosae 0.83 0.83 Albizia guachapele Wind SH, FO Leguminosae 2.24 Allophyllus psilospermus Animals Sapindaceae 0.616 Annona purpurea Animals SH, FO Annonaceae 1.75 0.03 Apeiba tibourbou Wind SH Tiliaceae 0.629 3.392 Astronium graveolens Wind TI, SH Anacardiaceae 3.194 3.569 0.03 Bauhinia sp. Wind Leguminosae 0.794 0.417 Bursera simaruba Animals LF, FI, M Burseraceae 2.714 3.062 0.06 0.338 Byrsonima crassifolia Animals SH, FI, FR Malphigiaceae 0.06 Calycophyllum TI, SH, LF, candidissimum Wind FI Rubiaceae 15.37 2.952 0.03 Cecropia peltata Animals FO Cecropiaceae 5.73 0.07 Cedrela odorata Wind TI, SH, LF Meliaceae 1.731 1.94 0.09 0.339 Ceiba pentandra Wind SH Bombaceae 0.07 Chlorophora tinctoria Animals Moraceae 3.858 Chomelia spinosa Animals Rubiaceae 0.966 3.004 Coccoloba laseri Animals SH Polygonaceae 0.415 0.1 Cochlospermum TI, SH, LF, vitifolium Wind FO Cochlospermataceae 3.565 2.16 5.438 0.418 Cojoba rufescens Animals Leguminosae 0.363 Cordia Wind TI Boraginaceae 2.109 4.427 0.33 0.339 29 alliodora Cordia panamensis Animals Boraginaceae 0.782 2.33 0.677 Dalbergia retusa Wind TI, SH Leguminosae 3.139 0.91 1.693 Diphysa robinioides Wind TI, SH Fabaceae 1.27 0.17 Enterolobium cyclocarpum Animals TI, SH, FO Fabaceae 8.911 0.08 Erythroxyllum sp. Animals Erythroxylaceae 0.355 Eugenia coloradoensis Animals SH, LF Mytaceae 0.83 0.03 Genipa americana Animals SH, LF, FO Rubiaceae 7.8 2.24 Guazuma ulmifolia Animals SH, FO, FI Sterculiaceae 0.64 16.87 0.5 9.48 11.71 Hura crepitans Ballistic SH Euphorbiaceae 10.87 0.07 Licania arborea Animals SH Chysobalanceae 0.04 Lonchocarpus felipei Wind TI, SH Leguminosae 1.97 0.03 Lonchocarpus velutinus Wind TI, SH Leguminosae 0.344 Luehea speciosa Wind SH, FO Tiliaceae 1.356 Machaerium microphyllum Wind Leguminosae 1.016 Manilkara achras Animals TI, SH Sapotaceae 1.41 Pachira quinata Wind TI,SH, LF Bombaceae 0.88 Platymiscium pinnatum Wind TI, SH Leguminosae 0.03 Pouteria campechiana Animals TI Sapotaceae 5.857 1.98 0.05 Sapium glandulosum Animals SH Euphorbiaceae 1.04 3.602 0.06 Sciadodendron excelsum Animals SH, LF Araliaceae 5.102 0.07 Spondias mombin Animals SH, FO Anacardiaceae 0.77 5.161 0.09 Sterculia apetala Animals SH Sterculiaceae 0.04 Swartzia simplex Animals Leguminosae 0.83 Tabebuia guayacan wind TI Bignoniaceae 0.379 Tabebuia rosea Wind TI, SH Bignoniaceae 13.99 2.33 0.06 30 Table 2. Stem count, basal area, diversity, and dispersal mode by habitat type. One standard error of the mean is given in parenthesis. Species richness was estimated by EstimateS Version 8.0. Dispersal mode percentages were calculated from the total number of stems found in each habitat type. Stem Count Basal Area Species Abiotic Biotic Forest Type (1 ha) (m2/ha) Richness Disp. Disp. Secondary Forest 900 (117) 17.7 (7) 30 0.57 0.42 Riparian Forest 180 (74.9) 14.3 (5.4) 38 0.35 0.64 5 Year Abandoned 494 (286) 1.88 (1.36) 14 0.58 0.41 2 Year Abandoned 231 (177) 1.00 (.87) 7 0.08 0.92 Active Pasture 12.0 (8.3) 1.26 (1.06) 32 0.44 0.55 31 Table 3. Tree species similarity between habitats using Jaccard’s Index. Values range from 0 to 1 where 1 represents 100% overlap and 0 represents no overlap of species in two habitats. 2-Year 5-Year Active Riparian Abandoned Abandoned Pasture Zone 2-Year Abandoned 5-Year Abandoned 0.2 Active Pasture 0.04 0.12 Riparian Zone 0.07 0.22 0.51 Secondary Forest 0.16 0.33 0.31 0.43 32 Figure 1. Size class distributions of trees >5 cm dbh within an active, eighty-five hectare cattle pasture, forested riparian fragments, 5-year abandoned pasture, 2-year abandoned pasture, and a secondary forest. 33 Dry Season Wet Season Dec. Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. Wind Dispersal Calycophyllum candidissimum X X Diphysa robinioides Cordia alliodora Cochlospermum vitifolium Bombacopsis quinata Ceiba pentandra Tabebuia ochracea Tabebuia rosea Astronium graveolens Cedrela odorata Animal Dispersal Brysonima crassifolia Sciadodendron excelsum Pouteria campechiana Bursera simaruba Spondias mombin Enterolobium cyclocarpum Guazuma ulmifolia Genipa americana Sapium glandulosum Cecropia peltata Figure 2. Flower and fruiting patterns of the ten most common wind-dispersed and the ten most common animal-dispersed tree species throughout the year within an active pasture in Panama. Black-filled boxes = presence of flowers, Gray-filled boxes = presence of fruits. Gray-black checkered boxes = presence of fruits and flowers. 34