- Human alteration of natural landscapes leads to biodiversity loss, often from a combination of area effects and fragmentation effects. Smaller habitat patches support fewer species than large ones and incur additional consequences from isolation. Efforts to preempt biodiversity loss from insular habitat fragments are complicated by individualistic species responses and time-delayed extinctions. Understanding how human activity affects bird communities in species rich, disturbance sensitive tropical forests is a conservation priority. Nevertheless, tropical ecosystems remain under-studied; long-term species inventory data from tropical forests are rare.
This dissertation combines a unique set of spatially extensive avian inventories from lowland forests in central Panama with a long history of bird surveys from Barro Colorado Island (BCI), a well-studied land-bridge island isolated within the Panama Canal, to better inform our understanding of how human-altered environments drive long-term changes in tropical forest bird communities. First, I evaluated trait predictors of species extinctions from BCI. I assessed to what degree changes in BCI’s bird community can be explained by loss of species sensitive to fragmentation-associated environmental drying. In my second data chapter, I examined the pattern of delayed extinctions on BCI among different species groups. I used species-area models and extinction trends to predict how many species BCI might still lose and how long these remaining species losses could take. Lastly, I focused on lowland forest patches within the Canal zone to evaluate how increasing urbanization influences the use of forest patches by tropical birds, with a focus on species composition, traits, and phylogenetic diversity.
Birds are likely to have disappeared from BCI if they had small populations in the 1920s, specialize on terrestrial arthropods, and are sensitive to forest moisture conditions. As a consequence of
extreme, persistent declines among understory insectivores associated with wet forests, the bird community on BCI has significantly shifted to resemble forest bird communities on the drier portion of the rainfall gradient. Extinctions accelerated 40-60 years following isolation and the island no longer supports the number of species expected for its size and amount of annual precipitation. From six to 92 additional species may be lost from BCI over the next one to nine centuries. Enduring species losses on BCI appear to be a product of habitat loss, edge effects, and negative consequences of isolation for dispersal-limited birds. In lowland forests of the Canal zone, urbanization is associated with community simplification and decreased compositional similarity without subsequent loss of functional diversity. Urban forests broadly favor good dispersers with short development periods, and recent evolutionary histories.
My results reveal the important roles of connectivity, dispersal ability, and sensitivity to local habitat conditions structuring avian community composition in tropical forests of central Panama. Limited human activity and access to large, protected forest tracts appear to be key drivers of avian community composition for both BCI and urban forest fragments. Dispersal-limited tropical birds with small populations, especially habitat specialists sensitive to moisture conditions and human disturbance, may be at greatest risk of extinction in insular rainforest fragments.