The susceptibility of tropical forest bird communities to habitat fragmentation
Description
Habitat fragmentation impacts on bird communities are relatively well studied in the temperate zone, where brood parasites and non-forest predators are the main causes of declining avian biodiversity. However, life history traits of tropical species suggest different and more diverse effects of habitat fragmentation on tropical bird communities. This dissertation used historical data and quantitative comparisons of fragmented and control sites to infer causes and patterns of tropical avian community response to fragmentation in Central America. Particular ecological guilds, namely forest understory insectivores, mixed-species flock participants, and birds that nest on or near the ground declined or disappeared from La Selva Biological Station, Costa Rica since the 1950s as the site became increasingly isolated from other forest. Spot-maps of twenty-two insectivorous bird species at La Selva in the 1970s indicate that insectivore populations were probably never large enough to be viable, emphasizing the need for larger reserves given the low density of many tropical forest species. Comparisons of four forest sites (24.6-1200 ha) to an non-fragmented control site in southeastern Nicaragua show significant reduction in biodiversity, corresponding decline in total bird abundance, and changes in community composition, even in the larger sites. Lack of change in the Nicaraguan control site suggests landscape changes rather than climate change is the most parsimonious explanation for community changes in the fragments. Finally, comparison of community changes in two large isolated forest reserves, La Selva and Barro Colorado Island, Panama, demonstrate similar patterns of losses of insectivores and low nesting species. However, landscape level and climatic differences between the two sites also affect the ability of some species to persist in remaining forest. Overall, this dissertation suggests that deterministic factors, especially ecological characteristics, put particular species at risk to landscape changes. No single mechanism appears to explain all the patterns, but small population size, poor dispersal ability, and increase in nest depredation are implicated by this study. Further investigation is needed to assess the relative importance of these and other mechanisms. Moreover, trophic cascades involving other taxa are predicted as a result of the lack of ecological replacement of disappearing specialized forest insectivores