Files
Abstract
Species are finite in their abundances and distributions, and the processes that form distributional patterns are complex. In this dissertation, I investigated the factors that contribute to geographic range limits in a set of narrow endemic species. Specifically, I first constructed species distribution models to investigate the location of abiotic niches and their predicted stability under climate change for nine threatened/endangered species. This work demonstrated that suitable habitat is not predicted to remain stable under the predictions of climate change, and these species are at great risk of extinction under current climate projections. Second, I performed reciprocal transplant experiments to investigate local adaptation and niche constraints in two sister taxa with varying distributions, Polygonella americana (widespread) and P. fimbriata (narrow). Populations of the narrow species displayed no evidence of niche constraints, and very little evidence of local adaptation. In contrast, populations of P. americana appeared locally adapted to their home environments. Finally, I used microsatellite markers designed specifically for these same two species (P. americana and P. fimbriata) to investigate the role of genetic constraints in shaping their ranges. Ifound that populations of the narrow species are genetically depauperate, while the widespread species had populations with relatively higher levels of genetic diversity. Additionally, the structure of diversity in the widespread species demonstrated a strong geographic signal. These results indicate that an edge group of populations is diverging in this species as a result of low gene flow between it and the rest of the populations we sampled. Overall, my research indicates that the factors contributing to range limits in P. fimbriata adhere mainly to an evolutionary genetic constraints model of range limits, and that the added effects of dispersal limitation are causing this species to remain geographically restricted. My work has implications for the field of range limits, and I put forward specific improvements in methodology for investigating questions related to geographic distributions.