Measured in species richness and abundance, ants are globally successful in temperate to tropical latitudes. Explaining the origin, success, and maintenance of diversity at both global and local scales has proven challenging because patterns of diversity and processes that drive patterns of diversity often differ between global and local scales. In this thesis, I study the patterns and processes driving ant diversity over a gradation of geographic levels. Chapter 1 compares and describes the density and dispersion of nut-nesting ants in southeastern United States temperate deciduous forests under nut producing trees. Species diversity and nut occupancy rates do not differ among sites or states and that ant-occupied nuts are spatially aggregated across ant species, a pattern inconsistent with spatial segregation of species that might arise in a competition-assembled community. In Chapter 2, through the development of new rapid sampling methods, I determine Strumigenys ant communities are feasible to study. I found 0.20 ha habitats and 1.0 m2 microsites are appropriate spatial scales for investigating abiotic factors and general habitat characteristics important for Strumigenys communities. I concluded that an even finer scale (< 1.0 m2) would be necessary to investigate community patterns on spatial scales on which Strumigenys are likely to interact or compete.

Chapter 3 focuses on broad scale diversity patterns and describes the biogeographic origins of Nearctic Strumigenys. I produce a molecular based phylogeny to describe phylogenetic relationships of species and infer likely biogeographic histories of Nearctic species. I tested two alternative hypotheses for the assembly of biogeographic patterns in Nearctic Strumigenys, the adaptive radiation hypothesis verses the evolutionary conservatism hypothesis. Ranges of migrant or introduced species within the U.S. are consistent with average annual temperature and rainfall of ranges they occupy outside of the U.S. Results of climate and phylogenetic comparisons support predictions of the evolutionary conservatism hypothesis.

Chapter 4 examines patterns and processes in the ant genus Strumigenys. I examined the phylogenetic and trait relationships of co-occurring Strumigenys species at very small to very large geographic spatial scales to test predictions of a competition hypothesis against three alternative community assembly hypotheses. The largest geographic scale included 60,000km2 bioregions across North America North of Mexico, to southeastern U.S. local communities of 2000m2 and 25m2, and 0.1m2 microsites. Patterns of biodiversity of Strumigenys differed as a function of scale and suggested processes are differentially important depending on the geographic scale of pattern investigation. However, there was no evidence that competition had influenced assemblage patterns of Strumigenys communities at any geographic scale.