What controls diversity? The high diversity of many tropical taxa and subsequent decline in diversity away from the Equator has inspired many theories and much debate. Theories attempt to explain these patterns through a combination of one or several processes that may leave signatures in the functional, spatial and phylogenetic structure of species assemblages. Unfortunately, the testing of these theories has been hampered by a limited ability to separate the effects of multiple processes on observed patterns. In this thesis, I use recent advances in functional, spatial and phylogenetic methods in ecology to parse the contributions of multiple processes generating patterns of diversity in a hyper-diverse tree community in the Ecuadorian Amazon and the New World flora.

In Chapter 1, I identify drought stress and herbivore pressure as key drivers of habitat associations and community structure in Amazonian trees using trait-based null models of community assembly. To do this I collected and analyzed leaf drought tolerance for 80 species and leaf lamina toughness for 454 species, both functional traits that are more tightly linked to a single assembly process than integrative traits such as specific leaf area.

In Chapter 2, I ask if spatial aggregation in Amazonian trees is primarily determined by habitat associations, dispersal limitation or both processes using spatial point process modeling. I found that both processes were important drivers of aggregation for the majority of species, but that leaf traits such as drought tolerance were most predictive of habitat associations while seed mass predicted the size and density of species clustering independent of the abiotic environment.

In Chapter 3, I test whether one proximate cause of latitudinal gradients in alpha diversity is a gradient in taxonomic or phylogenetic turnover, or beta diversity, with a large dataset of ~81,000 New World vascular plants. I found for both taxonomic and tip-weighted phylogenetic beta metrics that higher tropical diversity was associated with higher turnover. However, I found the opposite pattern for basal-weighted phylogenetic beta diversity, suggesting these metrics capture distinct aspects of biotic change across space. The results of this thesis suggest multiple processes combine and interact to create diversity patterns, but their influence can be separated with appropriate metrics.