Tests of community assembly across spatial scales in Neotropical birds

Species diversity varies dramatically across the surface of the Earth. A key step in the accumulation of species diversity is the ability of species to coexist in biological communities. Thus, identifying the mechanisms underlying community assembly is a major challenge for ecologists seeking to explain patterns in species diversity and composition. Recently some consensus has been reached on the set of processes that influence community assembly: speciation, demographic stochasticity, niche-based fitness trade-offs among species and dispersal. However, it is unclear how the importance of a particular process changes with spatial scale, which interactions exist among processes at large spatial scales and the extent to which niche-based resource partitioning among species explains differences in diversity among communities. Neotropical birds offer an ideal opportunity to address these uncertainties because of their high diversity and the existence of detailed information on their evolutionary history and ecology. In this thesis, I first use trait and phylogenetic metrics of community structure to show that both habitat filtering and interspecific competition shape community composition at the scale of individual bird territories (~1-2 ha). Second, I use simulations of community assembly to show that trait-based metrics of community structure outperform phylogenetic metrics for detecting niche-based community assembly, and that both sets of metrics often have low power when multiple processes influence community composition. Third, taking a trait-based, species-level approach, I show that both habitat filtering and interspecific competition influence species occurrence at regional scales (~75000 km2), and interact with dispersal ability so that their effect on species occurrence is increased for species with greater dispersal ability. Finally, using a combination of trait- and isotope-based methods to quantify resource partitioning, I show that species' niche widths do not change and niche overlap is reduced at high compared to low species richness. Taken together, these results suggest that both habitat filtering and interspecific competition (i.e. niche-based processes) influence community assembly from local to regional scales. However, at least at regional scales, the degree to which these processes are important for determining the occurrence of any given species depends on that species's dispersal ability. They also suggest, based on niche-based interspecific competition influencing community composition, that differences in species richness among communities are in part explained by differences among sites in the breadth of available niche space, not by increased ecological specialisation or niche overlap