xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00204
El papel de los bosques tropicales primarios en las comunidades de aves en un paisaje agrcola mixto
The role of primary forest on tropical avian communities in a mixed agricultural landscape.
Digitized by MVI
The conversion of tropical forest to agricultural systems is both widespread and increasing. As a result, much of tropical biodiversity may depend on altered landscapes for their long-term survival. The purpose of this study was to compare the avifauna of dairy cattle pastures and forested coffee plantations to assess their capacity to maintain avian biodiversity in San Luis, Costa Rica. Bird assemblages were examined for three sites of each landscape type along a distance gradient (100m, 250m, and 500m) from primary forest. Bird diversity decreased significantly with distance from forest in both landscapes (between 100m and 250m; pasture: t = 9.58, df = 210.55, P < 0.05; coffee: t = 8.04, df = 353.28, P < 0.05). At 250m and 500m, coffee sites were significantly more diverse than pasture sites (t = 2.30, df = 290.08, P < 0.05, and t = 2.57, df = 141.77, P < 0.05, respectively). Relative abundance of forest-habitat species decreased in both landscapes, but the reduction was only significant for pastures (between 100m and 250m; pasture: X2 = 13.807, df = 1, P < 0.05; coffee: X2 = 0.647, df = 1, P > 0.05). Moreover, pasture sites at 250m and 500m had significantly higher percentages of habitat-generalists and open-habitat species than corresponding coffee sites (250m: X2 = 14.313, df = 1. P < 0.05; 500m: X2 = 4.609, df = 1, P < 0.05). These findings suggest that forested coffee farms may serve to better maintain native biodiversity than pastureland after habitat transformation. Nevertheless, distance from forest is the key factor in the maintenance of bird communities in human-dominated landscapes, as short distances from primary forest (~250m) are responsible for large decreases in avian diversity. Incorporation of conservation areas in the expanding array of agricultural landscapes is necessary to best maintain native diversity in converted landscapes.
La conversin de los bosques tropicales en los sistemas agrcolas est creciendo y se est extendiendo; por eso, la mayora de la diversidad tropical tendr que fiarse de los hbitats alterados para sobrevivir a largo plazo. El propsito de esta investigacin fue comparar la diversidad de aves en los potreros lecheros y los cafetales forestales para examinar su capacidad de mantener la biodiversidad en San Luis, Costa Rica. Las comunidades de aves fueron examinadas por hbitat en tres sitios con respecto a distancia (100m, 250m, 500m) del bosque primario. La diversidad disminuy significativamente con la distancia en ambos hbitats (entre los 100m y los 250m; potrero: t = 9.58, df = 210.55, P < 0.05; cafetal: t = 8.04, df = 353.28, P < 0.05). A los 250m y a los 500m, los cafetales tuvieron ms diversidad que los potreros (t = 2.30, df = 290.08, P < 0.05, y t = 2.57, df = 141.77, P < 0.05, respectivamente). La abundancia de las especies forestales disminuy con la distancia del bosque en ambos ambientes, aunque la reduccin solamente fue significativa en los potreros (entre los 100m y los 250m; potrero: X2 = 13.807, df = 1, P < 0.05; cafetal: X2 = 0.647, df = 1, P > 0.05). Adems, los potreros de los 250m y los 500m tuvieron porcentajes significativamente ms altos de especies generalistas y de ambientes abiertos que los cafetales (los 250m: X2 = 14.313, df = 1. P < 0.05; los 500m: X2 = 4.609, df = 1, P < 0.05). Estos resultados indicaron que los cafetales forestales son mejores para mantener la diversidad original despus de la conversin de los hbitats naturales. Sin embargo, la distancia del bosque es el factor ms importante en mantener las comunidades aviares tropicales en los ambientes impactados por la gente. Adems, las distancias ms cortas (250m) pueden causar reducciones graves en la diversidad de aves. Por eso, se necesitan incluir reas conservadas en ambientes agrcolas para mantener la diversidad de aves tropicales.
Text in English.
Land use, Rural
Costa Rica--Puntarenas--Monteverde Zone--San Luis
Uso del suelo, rural
Costa Rica--Puntarenas--Zona de Monteverde--San Luis
Tropical Ecology Fall 2005.
Impact of agriculture on birds
Ecologa Tropical Otoo 2005
Impacto de la agricultura en las aves
t Monteverde Institute : Tropical Ecology
The role of primary forest on tropical avian communities in a mixed agricultural landscape Matt Warner Department of Soil, Water, and Climate, University of Minnesota ABSTRACT The conversion of tropical forest to agricultural systems is both widespread and increasing. As a result, much of tropical biodiversity may depend on altered landscapes for their long term survival. The purpose of this study was to compare the avifauna of dairy cattle pastures and forested coffee plantations to assess their capaci ty to maintain avian biodiversity in San Luis, Costa Rica. Bird assemblages were examined for three sites of each landscape type along a distance gradient (100m, 250m, and 500m) from primary forest. Bird diversity decreased significantly with distance fr om forest in both landscapes (between 100m and 250m; pasture: t = 9.58, df = 210.55, P < 0.05; coffee: t = 8.04, df = 353.28, P < 0.05). At 250m and 500m, coffee sites were significantly more diverse than pasture sites (t = 2.30, df = 290.08, P < 0.05, an d t = 2.57, df = 141.77, P < 0.05, respectively). Relative abundance of forest habitat species decreased in both landscapes, but the reduction was only significant for pastures (between 100m and 250m; pasture: X 2 = 13.807, df = 1, P < 0.05; coffee: X 2 = 0.647, df = 1, P > 0.05). Moreover, pasture sites at 250m and 500m had significantly higher percentages of habitat generalists and open habitat species than corresponding coffee sites (250m: X 2 = 14.313, df = 1. P < 0.05; 500m: X 2 = 4.609, df = 1, P < 0.05). These findings suggest that forested coffee farms may serve to better maintain native biodiversity than pastureland after habitat transformation. Nevertheless, distance from forest is the key factor in the maintenance of bird communities in huma n dominated landscapes, as short distances from primary forest (~250m) are responsible for large decreases in avian diversity. Incorporation of conservation areas in the expanding array of agricultural landscapes is necessary to best maintain native divers ity in converted landscapes. RESUMEN La conversiÃ³n de los bosques tropicales en sistemas agrÃcolas estÃ¡ creciendo y se estÃ¡ extendiendo; por eso, la mayoria de la diversidad tropical tendrÃ¡ que fiarse de hÃ¡bitats alterados para sobrevivir a largo plazo. El propÃ³sito de esta investigaciÃ³n fue comparar la diversidad de pÃ¡jaros en los potreros lecheros y los cafetales forestales para examinar su capacidad de mantener la biodiversidad en San Luis, Costa Rica. Las comunidades de pÃ¡jaros fueron examinadas en tres sitios por hÃ¡bitat con respecto a distancia (100m, 250m, 500m) del bosque primario. La diversidad disminuyÃ³ significativamente con la distancia en ambos hÃ¡bitats (entre los 100m y los 250m; potrero: t = 9.58, df = 210.55, P < 0.05; cafetal: t = 8.04 , df = 353.28, P < 0.05). A los 250m y a los 500m, los cafetales tuvieron mÃ¡s diversidad que los potreros (t = 2.30, df = 290.08, P < 0.05, y t = 2.57, df = 141.77, P < 0.05, respectivamente). La abundancia de las especies forestales disminuyÃ³ con la dis tancia del bosque en ambos ambientes, aunque la reducciÃ³n solamente fue significativa en los potreros (entre los 100m y los 250m; potrero: X 2 = 13.807, df = 1, P < 0.05; cafetal: X 2 = 0.647, df = 1, P > 0.05). AdemÃ¡s, los potreros de los 250m y los 50 0m tuvieron porcentajes significativamentes mÃ¡s altos de especies generalistas y de ambientes abiertos que los cafetales (los 250m: X 2 = 14.313, df = 1. P < 0.05; los 500m: X 2 = 4.609, df = 1, P < 0.05). Estos resultados indicaron que los cafetales for estales son mejores para mantener la diversidad original despuÃ©s de la conversiÃ³n de los hÃ¡bitats naturales. Sin embargo, la distancia del bosque es el factor mÃ¡s importante en mantener las comunidades aviares tropicales en los ambientes impactados por la gente. AdemÃ¡s, las distancias mÃ¡s cortas (250m) pueden causar reducciones graves en la diversidad de aves. Por eso, se necesitan incluir Ã¡reas conservadas en ambientes agrÃcolas para mantener la diversidad de pÃ¡jaros tropicales.
INTRODUCTION Across the globe, humans are rapidly converting tropical landscapes into a wide array of agricultural systems (Vitousek et al. 1997). According to the 2005 FAO Global Forest Resources Assessment, 13 million ha of forest are lost worldwide each year, primarily for c onversion into agricultural landscapes, and Central America has the highest rate of tropical forest transformation at a rate of 1.3% (285,000ha) per year. Moreover, it is now evident that current conservation areas alone do not contain the spatial extent, security, or stability required to ensure the safety of threatened species and habitats across the tropics (SoulÃ© and Sanjayan 1998, Terborgh 1999, Musters et al. 2000, Luck and Daily 2003). In addition, few large tracts of natural habitat, beyond those already in designated protected areas, are likely to be preserved (Halffter 1994, Ehrlich et al. 1995, Vitousek et al. 1997, Daily et al. 1998). As a result, understanding the impacts of converted landscapes on native biodiversity will be critical for fut ure conservation efforts. Whereas the vast majority of research to date has focused on biodiversity losses in response to a given habitat change, it is becoming increasingly important to examine the potential conservation value of converted landscapes (Da ily 1997). Tropical forest transformation has profound impacts on native biodiversity. Studies across a wide array of taxa, such as butterflies, terrestrial arthropods, and birds, have examined the importance of size of remaining forest fragments on the retention of native diversity, while also examining differences in diversity between fragment interior and surrounding transformed areas (e.g. Daily and Ehrlich 1995, Djupstrom 1997, Hughes et al. 2002, Sekercioglu 2002). In short, diversity declines sha rply with decreasing fragment size and only a few forest species can flourish outside fragments in human altered landscapes. Instead, they are often replaced by habitat generalists and open usually found in high abundance, flourishing and distributed over large geographical ranges, unlike forest habitat species that are often endemic or restricted to pristine or remnant forests. Therefore, local species diversity decreases outside of forests but may be partially maintained through weedy species additions. However, regional diversity declines sharply due to the loss of forest species in converted landscapes. By far, most biodiversity studies in human altered landscapes have centered on bird s (e.g. Carveth 1997, Estrada et al. 1997, Wunderle 1998, Daily et. al 2001, Browne 2002, Luck and Daily 2003). Birds follow the general patterns of decreasing diversity with fragment size and replacement of forest species by weedy species in transformed habitats. Various studies have also examined the role of distance from the edge of primary forest in bird community maintenance (Guevara and Laborde 1993, Dunn 2000, Luck and Daily 2003). Increasing distance from forest results in large decreases in both forest species richness and abundance across a wide variety of converted habitats. A large number of recent avian studies have examined changes in species composition for particular agricultural landscapes following land transformation. For example, temp erate studies have suggested that fewer forest species remain in pastureland than in agricultural cropland after forest conversion (Tucker 1997, Peterson et al. 1998). Tropical agricultural systems, however, have been understudied, leaving a poor understan ding of biodiversity in specific landscape types (Estrada et al. 1997). Most tropical studies of this nature have focused on bird assemblages in coffee plantations,
with results indicating that shade grown coffee plantations may serve as important refuges for bird diversity (Perfecto et al. 1996, Greenberg et al. 1997, Wunderle 1998, Browne 2002). The ability of cattle pastures, a prevailing element of converted tropical landscapes, to maintain bird communities remains largely unknown, as most studies have simply clumped pastureland with various other farming activities (such as small scale banana, bean, sugar cane, taro, and yucca plots) to form a broad mixed agriculture category (e.g. Estrada et al. 1993, Duncan and Chapman 1999, Daily et al. 2001, Luck a nd Daily 2003). Hence, there is need to examine avian communities in this characteristically common component of tropical countryside landscapes. This study centers on the use of countryside habitats by avian communities in the San Luis Valley of north ce ntral Costa Rica. Contrary to other regions of the greater Monteverde area, bird communities of San Luis are understudied and poorly known (Fogden 1993). While covered by premontane wet forest (Holdridge 1967 75 years ago, the study area has since been co nverted to support a mixture of small scale agricultural activities, while maintaining a bordering ridge of continuous primary forest ( ~ 10, 500ha; Burlingame 2000). This study compares the ability of the two dominant agricultural landscapes, dairy cattle p astures and forested coffee plantations, to maintain bird diversity while also assessing the importance of habitat distance from forest. Specifically, the goals of this investigation are to: 1) compare the species richness, composition, and diversity of bi rds communities with respect to distance from adjacent primary forest in the two aforementioned agricultural land types; 2) compare avifaunal richness, composition, and diversity between the two land use practices; and 3) quantify the potential conservatio n value of these agro ecosystems. Materials and Methods This study was conducted in the agricultural landscape of San Luis Arriba, Costa Rica. Six 0.5 ha study sites (three pasture, three forested coffee) were selected with respect to a distance gradie nt from an adjacent ridge of continuous primary forest (100m, 250m, and 500m). Hence, there were two sites, one from each landscape type, at each of the aforementioned distances (Diagram 1). The six sites were uniform within land use practice (see Study S ites ) and encompassed a narrow altitudinal band (1130m 1190m). Each site was surveyed twice from October 27 November 14, 2005, for a total of 12 observation periods, with censuses beginning at 5:30 AM and ending at 9:30 AM . Birds were identified visual ly using binoculars and A Guide to the Birds of Costa Rica (Stiles and Skutch 1989). Aural identification was used for birds that were seen but could not be clearly identified visually. Both species richness and abundance values were recorded for each si te, and data were totaled over the two survey periods. Birds were also categorized into one of four habitat preferences (edge, forest, open, generalist) according to species specific information from A Guide to the Birds of Costa Rica (Stiles and Skutch 1 989). Soaring birds, such as raptors, swifts, and swallows, were not included in this study.
Study Sites The three pasture study plots were located in distinct cattle pastures, separated by both forested riparian strips (2 10m wide) and barbwire fe nces. Each pasture (3 5 ha) was in active cattle rotation with approximately 15 head of cattle per pasture. In addition to forested riparian strips, each pasture contained planted hedgerows and freestanding trees. Within each 0.5 ha survey plot there we re six to eight freestanding trees, and effort was made to include no tree that would be visited disproportionately, for example prolifically fruiting trees. The three coffee sites were located in three different forested coffee farms. All 0.5 ha site s were homogeneous in that they contained a mixture of shade grown coffee, banana and citrus trees, and a thin surrounding border of forest vegetation (2 5m wide). Common shade trees included mature Cecropia obtusifolia (Cecropiaceae), Billia colombiana (H ippocastanaceae), and Ficus pertusa (Moraceae). Additionally, random patches of trees and hedges dotted the landscape between farms. Pesticide use was nonexistent. RESULTS Abundance A total of 878 birds from 80 species was identified in the six pastur e and coffee sites. Abundance values differed both within and between habitat types, although there was no clear relationship to either distance from forest or landscape type (Figure 1). Pasture abundance values showed little deviation. The pasture site closest to the forest (P1) had the most individuals with 163, followed by the farthest site (P3) with 162, and the middle site (P2) with 149, for a total of 474 individuals. There were no significant differences in DIAGRAM 1. Visual representation of pasture sites (P1, P2, P3) and forested coffee sites (C1, C2, C3) with distance from primary forest in San Luis, Costa Rica. One site fro m each land use category was situated at each distance (100m, 250m, and 500m) for a total of six study sites. Elevation ranged from 1130m 1190m.
any pasture combination (P1 P2 P3: X 2 = 0.77, df = 2, P > 0.05; P1 P2: X 2 = 0.63, df = 1, P > 0.05; P1 P3: X 2 = 0.00, df = 1, P > 0.05; P2 P3: X 2 = 0.54, df = 1, P > 0.05). Coffee plots, on the other hand, displayed greater variance. The middle coffee site (C2) had the highest abundance value with 198 individuals, followed by the closest site (C1) with 156. However, only 50 individuals were identified in the farthest site (C3), giving a total of 404 individuals in coffee farms. Contrary to pastures, abundance differed significantly in all co ffee comparisons (C1 C2 C3: X 2 = 86.40, df = 2, P < 0.05; C1 C2: X 2 = 4.98, df = 1, P < 0.05; C1 C3: X 2 = 54.54, df = 1, P < 0.05; C2 C3: X 2 = 88.32, df = 1, P < 0.05). When comparing across land use types, significant differences in abundance existed betw een medium and far sites, but not between sites nearest the forest (P2 C2: X 2 = 6.92, df = 1, P < 0.05; P3 C3: X 2 = 59.17, df = 1 P < 0.05; P1 C1: X 2 = 0.15, df = 1, P > 0.05). FIGURE 1. Avian abundance of pastures and forest ed coffee farms with respect to distance from forest. Abundance remained relatively stable for pasture sites (P1 = 163, P2 = 149, P3 = 162), while coffee sites showed much larger fluctuations (C1 = 154, C2 = 198, C3 = 50). Species Richness Species rich ness declined with distance from forest for both pasture and coffee landscapes (Figure 2). For pastures, P1 had the greatest number of species with 43, while P2 and P3 each contained 23. The difference in species richness between all three pastures was s ignificant (X 2 = 8.99, df = 2, P < 0.05), as were differences between P1 and P2 (X 2 = 6.06, df = 1, P < 0.05) and P1 and P3 (X 2 = 6.06, df = 1, P < 0.05), a difference of 20 species (46.51%) in both cases. Species richness did not differ between sites P2 a nd P3, as both had equal richness (X 2 = 0.00, df = 1, P > 0.05). For coffee plots, species
richness was greatest in C1 (41) followed by C2 (27) and C3 (21). As with pasture sites, significant differences were found between all three sites (X 2 = 7.10, df = 2, P < 0.05) and between sites C1 and C3 (X 2 = 6.45, df = 1, P < 0.05). However, declines in species richness from sites C1 to C2 and sites C2 to C3 were not significant (C1 C2: X 2 = 2.88, df = 1, P > 0.05; C2 C3: X 2 = 0.75, df = 1, P > 0.05). Moreove r, there were no significant differences in species richness between landscape types (P1 C1: X 2 = 0.05, df = 1, P > 0.05; P2 C2: X 2 = 0.32, df = 1, P > 0.05; P3 C3: X 2 = 0.09, df = 1, P > 0.05). equal sample sizes on species richness (Figure 3). Values derived from this index are strictly comparative and provide only a relative evaluation of richness. For both landscapes, use of this index resulted in approximate evenness of species richness in medium and far distance sites, while near sites remained most species rich. FIGURE 2. Species richness values over a distance gradient from primary forest. While no differences existed between land use types, richness was sig nificantly greater in near sites than far sites for both pasture and coffee (P1 P3: X 2 = 6.06, df = 1, P < 0.05; C1 C3: X 2 = 6.45, df = 1, P < 0.05).
FIGURE 3. Smarg comparative richness values, reflecting differences in sample size between sites, with respect to distance from primary forest. Richness is greatest in near sites and essentially the same in medium distance and far sites. Evenness Site evenness varied within and between landscape categories (Figure 4). Evenness values give a representation of how evenly species in a community are distributed by abundance, with 1 being perfectly even and 0 completely uneven. For both landscapes, evenness was highest for sites nearest the forest and lowest at medium distance sites . For pasture sites, P1 was the most even (E = 0.90) followed by P3 (E = 0.80) and P2 (E = 0.78). For coffee sites, C1 and C3 shared the same evenness value (E = 0.92) while C2 was the least even (E = 0.82).
FIGURE 4. Evennes s values for species composition along a distance gradient from primary forest. Evenness was highest for sites nearest the forest and showed a drastic decrease within 250m, indicating that short distances have large effects on avian community evenness. E venness increased from 250m to 500m in both landscapes, though the magnitude of this movement was much greater in coffee farms, possibly due to the homogenization of dominant open and generalist species at greater distances. Diversity Avian diversity di ffered significantly both within and between landscapes using a modified t For both pastures and coffee, the sites nearest the forest were significantly more diverse than medium distance and far sites (P1 P2: t = 9.58, df = 210.55, P < 0.05; P2 P3: t = 9.90, df = 239.07, P < 0.05; and C1 C2: t = 8.04, df = 353.28, P < 0.05; C1 C3: t = 5.85, df = 93.81, P < 0.05). While there was no significant difference in bird diversity between P1 and C1 (t = 0.10, df = 234.79, P > 0.05), significant differences were found between landscapes at both 250m and 500m from primary forest. C2 was more diverse than P2 (t = 2.30, df = 290.08, P < 0.05), and C3 more diverse than P3 (t = 2.57, df = 141.7 7, P < 0.05). For both land use types, far sites had higher diversity than middle sites, although these results were not significant (C2 C3: t = 0.79, df = 113.21, P > 0.05; P2 P3: t = 0.28, df = 306.07, P > 0.05).
r the six pasture and coffee sites with respect to distance from primary forest. In both land types, sites nearest the forest were significantly more diverse than medium distance and far sites, and diversity did not differ significantly between medium and far distances. Additionally, diversity between coffee and pasture sites did not differ significantly for near sites. However, at 250m and 500m coffee sites were significantly more diverse than pasture plots suggesting that land use context is important at distances greater than 100m. Species Composition Similarity Index, which determines species similarity without incorporating abundance differences, ntitative Index, which determines similarity based upon relative abundance of co occurring species (Southwood 1966; Figure 6). In both cases, a value is assigned from 0 1, with 0 indicating no similarity and 1 indicating complete homogeny between sites. Similarity was determined for sites equidistant from forest within all pasture sites and within all coffee sites. Results indicate intermediate or intermediate low differences for shared species between sites. The highest similarity was found using found
FIGURE 6. Sorenson Index (Cs; top) and Sorenson Quantitative Index (Cn; bottom) values of similarity within and between land use types. P1, P2, and P3 represent the three pasture sites at 100m, 250m, and 500m from primary forest respectively. Likewise, C1, C2, and C3 represent the coffee sites along the same gradient. There is intermediate to intermediate low similarity in all comparisons, with the highest similarity occur ring between sites P2 and P3 (Cs = 0.696). Sites P3 and C3 were the most dissimilar (Cn = 0.189). Forest Species The effect of distance from forest on the composition of forest habitat species was compared in both pasture and coffee sites. Although per centages of forest species varied between sites (Figure 7), there were no significant differences within or between landscape types (P1 P2 P3: X 2 = 5.232, df = 2, P > 0.05; P1 P2: X 2 = 3.792, df = 1, P > 0.05; P1 P3: X 2 = 1.194, df = 1, P > 0.05; P2 P3: X 2 = 2.221, df = 1, P > 0.05; C1 C2 C3: X 2 = 1.394, df = 2, P > 0.05; C1 C2: X 2 = 0.020, df = 1, P > 0.05; C2 C3: X 2 = 1.118, df = 1, P > 0.05; P1 C1: X 2 = 0.560, df =1, P > 0.05; P2 C2: X 2 = 2.031, df = 1, P > 0.05; P3 C3: X 2 = 0.310, df = 1, P > 0.05). Nevertheless, significant differences were found for percent abundance of forest species (Figure 8). In pastures, abundance differed significantly between all three sites (X 2 = 24.725, df = 2, P < 0.05) and decreased significantly from the site nearest the forest to both medium distance and far sites (P1 P2: X 2 = 13.807, df = 1, P < 0.05; P1 P3: X 2 = 17.774, df = 1, P < 0.05). However, there was no significant difference between medium distance and far pasture sites (X 2 = 0.182, df = 1, P > 0.05). Abundance also decreased with distance for coffee sites, although these differences were not significant (C1 C2 C3: X 2 = 1.445, df = 2, P > 0.05; C1 C2: X 2 = 0.647, df = 1, P > 0.05; C1 C3: X 2 = 1.237, df = 1, P > 0.05; C2 C3: X 2 = 0.383, df = 1, P > 0.05). When comparing abundance between land types, coffee sites supported more forest species at both medium and far distances (C2 P2: X 2 = 17.178, df = 1, P < 0.05; C3 P3: X 2 = 8.412, df = 1, P < 0.05). However, near coffee an d pasture sites did not differ significantly, despite greater forest species abundance in coffee (X 2 = 1.255, df = 1, P > 0.05).
FIGURE 7: Percentage of forest habitat species for pasture and forested coffee plots along a dista nce gradient from primary forest. Differences within and between landscapes were not significant. FIGURE 8. Percentage of forest habitat birds by abundance for pasture and forested coffee sites along a distance gradient from primary forest. Abundance declined with distance from forest in both land use types, although differences were only significant for pasture sites. Coffee sites supported more forest species than pastures at both medium and far sites. However, near coffee and pasture sites did not differ significantly, despite greater forest species abundance in the coffee site.
Weedy Species The percentage of weedy species (generalist and open habitat species) in pasture and coffee sites is shown in Figure 9. For coffe e sites, weedy species increased with distance from forest (58.54%, 66.67%, 76.19% for sites C1, C2, and C3, respectively). However, differences between sites were not significant (C1 C2 C3: X 2 = 1.945, df = 2, P > 0.05; C1 C2: X 2 = 0.456, df = 1, P > 0.05; C1 C3: X 2 = 1.891, df = 1, P > 0.05; C2 C3: X 2 = 0.519, df = 1, P > 0.05). Pastures sites were similar to coffee sites in that percent weedy species was lowest nearest the forest (55.81%). But, the relationship differed in that a larger percen tage of weedy species was observed at 250m (82.61%) than at 500m (69.51%). The only significant difference in percentage of weedy species for pastures existed between sites P1 and P2 (P1 P2: X 2 = 4.739, df = 1, P < 0.05; P1 P2 P3: X 2 = 4.963, df = 2, P > 0.05; P1 P3: X 2 = 1.187, df = 1, P > 0.05; P2 P3: X 2 = 1.075, df = 1, P > 0.05). Additionally, percentage of weedy species did not differ significantly between corresponding pasture and coffee sites (P1 C1: X 2 = 0.064, df = 1, P > 0.05; P2 C2: X 2 = 1.641, df = 1, P > 0.05; P3 C3: X 2 = 0.243, df = 1, P > 0.05). FIGURE 10. Percentage of weedy (generalist and open habitat) bird species in pastures and coffee farms with respect to distance from forest. This percentage increases with distance from forest in both land types. However, the only significant difference is between near and medium distance pasture sites (X 2 = 4.739, df = 1, P < 0.05). Percent weedy species by abundance was also compared between sites (Figur e 10). Weedy species abundance increased with distance from forest in coffee sites, although no differences between sites were significant (C1 C2 C3: X 2 = 3.693, df = 2, P > 0.05; C1 C2: X 2 = 2.902, df = 1, P > 0.05; C1 C3: X 2 = 1.960, df = 1, P > 0.05; C2 C3: X 2 = 0.107. df = 1, P > 0.05). Pasture sites showed a similar pattern of increasing weedy species abundance with distance, though site P2 had a slightly higher percentage than P3. This difference was not significant (X 2 = 0.220, df = 1, P > 0.05). Unlike coffee
sites, percent weedy species by abundance differed significantly between all three pasture sites (P1 P2 P3: X 2 = 18.915, df = 2, P < 0.05) and between near sites and both medium and far sites (P1 P2: X 2 = 13.688, df = 1, P < 0. 05; P1 P3: X 2 = 11.299, df = 1, P < 0.05). When comparing between land use types, pasture sites had higher weedy species abundance than corresponding coffee sites. As with species richness, diversity, and percent forest species by abundance, these diff erences were only significant for medium and far sites (P1 C1: X 2 = 2.55, df = 1, P > 0.05; P2 C2: X 2 = 14.313, df = 1. P < 0.05; P3 C3: X 2 = 4.609, df = 1, P < 0.05). FIGURE 10: Percentage of weedy species by abundance for pastures and forested coffee farms along a distance gradient from primary forest. Weedy species increase by percent abundance with distance from forest. Pasture sites at 250m and 500m have significantly higher percent abundances of weedy species tha n corresponding coffee sites (P2 C2: X 2 = 14.313, df = 1. P < 0.05; P3 C3: X 2 = 4.609, df = 1, P < 0.05). Dominance Bird species dominance was determined using the three most abundant species in each site. In general, dominance within landscape typ e was greatest at medium and far sites and lowest nearest the forest. For sites C1, C2 and C3, the three most abundant species per site comprised 22.42%, 41.93%, and 38.00% of birds, respectively. Corresponding pasture values were higher indicating greate r dominance in pastureland than in forested coffee farms. Pastures also displayed the same trend of increasing dominance with distance, as the three most abundant species per site comprised 26.99%, 54.36%, and 52.47%, correspondingly. While dominant spec ies varied between sites both within and between land use types, several species were especially dominant: Yellow throated Euphonia ( Euphonia hirundinacea ; generalist species), Social Flycatcher ( Myiozetetes similes ; open habitat species), and Blue gray Ta nager ( Thraupis episcopus ; open habitat
species). Yellow throated Euphonia was one of the top three most abundant species in all sites except site C2, in which it was the fourth most abundant species (P1 = 6.13%, C1 = 6.40%, P2 = 5.37%, C2 = 16.67%, P3 = 9.26%, C3 = 12.00%). Social Flycatcher was a top three most abundant species in four of six sites (C1 = 7.69%, C2 = 16.67%, P2 = 24.83%, P3 = 16.67%), and Blue gray Tanager was a most abundant species in three sites (C2 = 13.64%, P2 = 22.15%, P3 = 26.54 %). Abundances of these three species for the six sites are shown in Figure 11. For pasture sites, Social Flycatchers and Blue gray Tanagers greatly increased in abundance from near sites to medium distance and far sites (Social Flycatchers: P1 = 4.29%, P 2 = 22.15%, P3 = 26.54%; Blue gray Tanagers: P1 = 3.68%, P2 = 24.83%, P3 = 16.67%). In addition, 120 of the 878 birds identified across all sites were Blue gray Tanagers, making it the most abundant species of the study (13.67%). Another noteworthy speci Wilsonia pusilla , forest species ), which differed greatly between pasture and coffee sites. It was highly abundant in all coffee sites (C1: 8.33%, C2: 9.09%, C3: 14.00%), but not in pastures (P1: 1.23%, P2: 0.00%, P3: 0.62%). FIGURE 11. Abundance of three dominant bird species over a distance gradient, in pastures (P) and forested coffee farms (C). With the exception of Yellow throated Euphonias, increases of dominant species were greater for pasture than coffee sites . Abundance decreases from 250m to 500m are explained by increasing dominance of other species.
DISCUSSION This study found that bird communities respond to land transformation in a variety of ways. Avian richness and diversity is highest near remainin g forest habitat and declines rapidly with distance from forest edge, findings that are supported by previous tropical avian studies (Hughes et al. 2002, Luck and Daily 2003). At least some forest birds will use transformed habitat, provided that it is not too far from intact forest. Diversity drops more steeply with distance from forest edge for pastures than forested coffee farms, indicating that land use type greatly affects the resilience of birds after habitat conversion. Even though species richness was similar between the two habitats, pasture is a less attractive choice for forest birds. At both 250m and 500m, forested coffee farms supported higher diversity and maintained a higher percentage of forest habitat species by abundance than pastureland . Conversely, pastures supported a greater abundance of weedy species than coffee plots, indicating that generalist and open habitat species are replacing forest species faster in pastures than forested coffee farms. Despite these findings, both habitats were inhabited primarily by weedy species even at 100m from the forest. Hence, intact forest habitat is the key in maintaining high diversity of tropical forest species. Bird dominance by a few species was characteristic of both landscapes. Dominance , determined by net abundance of the three most abundant species in any given site, was greatest in medium and far sites and lowest nearest the forest for both pasture and forested coffee farms, further implicating the importance of proximity to forest in the maintenance of avian diversity. Moreover, dominance by a few species was more prevalent in pastures than coffee sites, paralleling trends of lower diversity and higher weedy species abundance in pastures. Additionally, certain dominant birds may serve as indicator species for avian communities in these converted habitats. Both pasture and coffee farms contained notably high abundances of three particular weedy species: Blue gray Tanager, Yellow throated Euphonia, and Social Flycatcher. In general, ab undance of these species increased with distance from forest for both pasture and coffee sites. However, increases were larger for pasture sites than coffee sites. These findings further emphasize that coffee farms may be better at supporting diversity fo llowing land transformation, but domination by weedy species is inherent to both habitats. However, bird composition within and between landscape types was only intermediately similar, indicating that large differences in both species richness and relative abundance are the result of both landscape context and distance from forest. forest species was highly abundant in all coffee sites yet virtually absent in pastures. He nce, this species might serve as a key in the assessment of other transformed landscapes, as its presence in a converted habitat may imply higher conservation value than pastureland. Unfortunately, most tropical forests will not remain untouched in our in creasingly human dominated world (Vitousek et al. 1997, Palmer et al. 2004), and because transformation into agricultural land is inevitable, it is pertinent to determine how converted habitats can best maintain native biodiversity. This study shows that in order to maintain high avian diversity in tropical montane habitats, cattle p astures and coffee
becoming more and more removed from source populations (Luck and Daily 2003), and as a result, bird diversity is likely to decline to a subset of wee dy species with isolation from forest habitat. Though forested coffee habitats maintain higher diversity than pasturelands, forest removal would almost certainly reduce the capacity of coffee farms to support native diversity to little better than pasture s in the long run. Ultimately, avian biodiversity will be severely reduced on a regional scale if large forest fragments are not preserved. For converted landscapes, proximity to remnant forest is essential to maintaining long term diversity. Within t he growing matrix of transformed agricultural landscapes, distinct land use practices differ in their ability to support bird communities. This investigation provided a mere snapshot of tropical bird assemblages in two distinct land use practices, cattle p astures and forested coffee plantations. The two habitats clearly differ in their ability to support biodiversity, implying that future conservation successes in converted habitats will require distinguishing between specific agricultural land use practice s. This has been largely overlooked in previous avian studies that have simply lumped a wide variety of unique practices as mixed agriculture habitat (e.g. Estrada et al. 1993, Duncan and Chapman 1999, Daily et al. 2001, Luck and Daily 2003). In order to best support natural communities in an increasingly unnatural world, converted landscapes must incorporate conservation areas (such as large, protected forest fragments and forested riparian areas) within agriculture landscapes. Through this effort, diff erent agricultural practices may, indeed, act as valuable habitat for forest, as well as weedy, species. ACKNOWLEDGEMENTS I am extremely grateful for the help of Alan Masters, who helped me persevere after two failed project ideas and provided persisten t and invaluable words of wisdom and encouragement. I also thank la familia LeitÃ³n Arce de San Luis Arriba who kept my stomach full on those dark, wet mornings, with special gratitude to mi hermano tico, Randy, and my faithful steeds, Brownie and Spike, w ho kept me company on many a morning. I am especially grateful to my three partners in crime in San Luis who pushed through many long walks up La Trocha and never failed to fill up the old gas tank with delicious helado. In addition, thanks to the Bottom Feeders, who would be nowhere without the fearless leadership of Porkfish, and Team Zissou. Lastly, thank you Mom and Pops for all your incredible support in every adventure. LITERATURE CITED Browne, A.S. 2002. Bird diversity in a coffee plantation and neighboring secondary forest. CIEE Fall 2002. 101 108. Burlingame, L.J. 2000. Conservation in the MonteverdeZone. Pages 351 388 in Nadkarni. N.M. and N.T. Wheelwright. Monteverde: Ecology and conservation of a tropical cloud forest. Oxford University Press , New York, USA. Carveth, K.A. 1997. Avian diversity in forest fragments. CIEE Fall 1997. 144 156. Daily, G.C. 1997. Countryside biogeography and the provision of ecosystem services. in P. Raven. Nature and human society: The quest for a sustainable world. National Research Council, National Academy Press, Washington D.C., USA. Daily, G.C., and P.R. Ehrlich. 1995. Preservation of biodiversity in small rainforest patches: rapid evaluation using butterfly trapping. Biodiversity and Conservation. 4: 35 55. Dai ly, G.C. 1998. Food production, population growth, and the environment. Science. 281: 1291 1292. Daily, G.C., P.R. Ehrlich, and G.A. Sanchez Azofiefa. 2001. Countryside biogeography: use of human dominated habitats by the avifauna of southern Costa Rica. E cological Applications. 11: 1 13.
Djupstrom, J.A. 1997. Island biogeography and arthropod diversity in disturbed forest fragments. CIEE Fall 1997. 119 142. Duncan, R.S., and C.A. Chapman. 1999. Seed dispersal and potential forest succession in abandoned ag riculture in tropical Africa. Ecological Applications. 9: 998 1008. Dunn, R.R. 2000. Isolated trees as foci of diversity in active and fallow fields. Biological Conservation. 95: 317 321. Ehrlich, P.R., A.H. Ehrlich, and G.C. Daily. 1995. The stork and the plow: The equity answer to the human dilemma. Putnam, New York, New York. USA. Estrada, A., R. Coates Estrada, D. Meritt, Jr., S. Montiel, and D. Curiel. 1993. Patterns of frugivores species richness and abundance in forest islands and in agricultural hab itats at Los Tuxtlas, Mexico. Vegetatio. 107/108: 245 257. Estrada, A., R. Coatees Estrada, and D.A. Meritt, Jr. 1997. Anthropogenic landscape changes and avian diversity at Los Tuxtlas, Mexico. Biodiversity and Convservation. 6: 19 43. Fogden, M. 1993. An Annotated Checklist of the Birds of Monteverde and PeÃ±as Blancas. Michael Fogden, Monteverde, Costa Rica. Greenburg, R., P. Bichier, A.C. Angon, and R. Reitsma. 1997. Bird populations in shade and sun coffee plantations in central Guatemala. Conservation Biology. 11: 448 459. Guevara, S., and J. Laborde. 1993. Monitoring seed dispersal at isolated standing trees in tropical pastures: consequences for local species availability. Vegetatio. 107/108: 319 338. Halffter, G. 1994. Conservacion de la bioversidad y areas protegidas en los paises tropicales. Ciencias. 36:4 13. Holdridge, L. 1967. Life zone ecology: revised edition. Tropical Science Center, San Jose, Costa Rica. Hughes, J.B., G.C. Daily, and P.R. Ehrlich. 2002. Conservation of tropical forest birds in countryside habitats. Ecology Letters 5: 121 129. Luck, G.W. and G.C. Daily. 2003. Tropical countryside bird assemblages: richness, composition, and foraging differ by landscape context. Ecological Applications. 13(1): 235 247. Magurran, A.E. 1988. Ec ological Diversity and Its Measurement. Princeton Univ. Press, Princeton, New Jersey. USA. Musters, C.J.M., H.J. de Graf, and W.J. ter Keurs. 2000. Can protected areas be expanded in Africa? Science 287: 1759 1760. NA. . Global Forest Resources Asses sment 2005. FAO. Available [online]: http://www.fao.org/forestry/site/fra/en Palmer, M., E. Bernhardt, E. Chornesky, S. Collins, A. Dobson, C. Duke, B. Gold, R. Jacobson, S. Kingsland, R. Kranz, M. Mappin, M.Luisa Martinez, F. Micheli, J. Morse, M. Pace, M . Pascual, S. Palumbi, O.J. Reichman, A. Simons, A. Townsend, and M. Turner. 2004. Ecology for a crowded planet. Science. 304: 1251 1252. Perfecto, I., R.A. Rice, R. Greenburg, and M.E. Van der Voort. 1996. Shade coffee: a disappearing refuge for biodivers ity. BioScience. 46: 590 608. Peterson, B.S. 1998. The distribution of Danish farmland birds in relation to habitat characteristics. Ornis Fennica. 75: 105 118. Sekercioglu, C., P.R. Ehrlich, G.C. Daily, D. Aygen, D. Goehring, and R.F. Sandi. 2002. Disappe arance of insectivorous birds from tropical forest fragments. Proceedings of the National Academy of Sciences (USA). 99: 263 167. SoulÃ©, M.E., and M.A. Sanjayan. 1998. Conservation targets: do they help? Science. 279: 2060 2061. Southwood, T.R.E. 1966. Eco logical Methods. Chapman and Hall, New York, New York. USA. Stiles, F.G., and A.F. Skutch. 1989. A Guide to the Birds of Costa Rica. Comstock Publishing Associates, Ithica, New York. USA. Terborgh, J. 1999. Requiem for nature. Island Press, Washington D.C. , USA. Tucker, G.M. 1997. Priorities for bird conservation in Europe: the importance of the farmed landscape. Page 79 116 in D.J. Pain and M. W. Pienkowski, editors. Farming and birds in Europe: the common agricultural policy and its implications for bird conservation. Academic Press, San Diego, California, USA. ecosystems. Science 277: 494 499. Wunderle, Jr., J, and S. Latta. 1998. Avian resource use in Dominican shade coffee plantations. The Wilson Bulletin. 110: 271 281.