In Central America, shade coffee is an important agroforestry ecosystem that provides habitat for many tropical bird species, yet the effects of fragmentation on modified habitat are poorly studied. Patches of shade coffee and forest are frequently located in a matrix of open agriculture and pasture. Birds living in these patches may be subject to the same edge effects documented in temperate forest-agricultural systems, such as increased predation and nest parasitism. The goal of this study was to determine if edge effects negatively impact birds breeding in the modified habitat of shade coffee. I compared rates of brood parasitism by the striped cuckoo (Tapera naevia) and rates predation on nests of the rufous-and-white wren (Throythorus rufalbus) in forest and shade coffee patches at three different spatial scales: nest site, territory site, and local landscape. Nests were monitored for outcome over three breeding seasons, 2005-2007. Vegetation was measured at the nest site following a modified BBIRD protocol. Habitat was classified in ESRI ARCGIS 9.2 with quantitative vegetation transects, satellite and aerial photographs of the area. Territory habitat composition, nest density, and landscape configuration were analyzed using ESRI ARCView and ARCGIS 9.2.

Results/Conclusions

Nest Site: Brood parasitism, but not predation, was significantly higher in nests located in coffee fields. Proximity to edge did not affect rates of parasitism, but depredation was higher in nests closer to open area edges. Territory: Territories composed partially or entirely of coffee were significantly more likely to experience predation or brood parasitism. Territories closer to edges were more likely to be parasitized. Landscape:  Brood parasitism was correlated to the number of streams in an area, regardless of habitat type the stream was in. Brood parasitism was significantly higher in areas of high nest density, but predation showed no relationship to nest density. Nest density was highest in coffee fields despite equal proportions of forest patches in the landscape. Avian predation was positively correlated to the percent open agriculture in the landscape. A high degree of coffee-pasture edge may be the cause of  high parasitism rates in coffee. Conclusion: These data indicate that agroforestry habitats and stream corridors facilitate access to nests by predators and parasites, and the presence of edge effects depends on the scale of analysis. Bird species increasingly use modified areas as native habitats disappear, necessitating more research on fragmentation effects in modified landscapes.