The genetic structure of species' geographic ranges: An evaluation using the coastal dune endemic Camissoniopsis cheiranthifolia (Onagraceae)

The development of molecular techniques has spurred thousands of population genetic studies on a wide variety of plant and animal species. Particularly important, but still relatively rare, are studies that properly test for geographic variation in genetic structure across species' ranges. This thesis investigates the effects of population density, mating system variation, distance between populations and hybridization on the genetic diversity, differentiation and structure across the range of Camissoniopsis cheiranthifolia (Onagraceae). By combining a transplant experiment with microsatellites, I also provide an empirical test of one of the most poorly resolved questions in evolutionary biology: Why do species exhibit limits to their distributions? I developed 24 species-specific nuclear microsatellites loci (nSSR) and used 13 of these and six variable chloroplast microsatellites (cpSSR) to investigate the genetic consequences of the transition from outcrossing to selfing in C. cheiranthifolia. As predicted, small-flowered, selfing populations had lower nSSR diversity (but not cpSSR) than large-flowered, outcrossing populations but they were not more differentiated. The reduction in diversity was greater than the expected from selfing alone, but could not be accounted for by indirect effects of selfing on population density. Five parapatric nSSR clusters and three groups of cpSSR haplotypes usually (but not always) differed in mating system, suggesting that selfing may often initiate ecogeographic isolation. However, lineage-wide diversity failed to support that selection for reproductive assurance initiated selfing in this species. I also investigated if hybridization with C. bistorta has influenced the genetic structure of C. cheiranthifolia in areas of closest parapatry. I confirmed that closet parapatry was associated with extensive genetic continuity between self-incompatible C. cheiranthifolia and C. bistorta and these were indistinguishable genetically within the broader context of geographic variation in C. cheiranthifolia, suggesting that C. bistorta could be an inland ecotype of C. cheiranthifolia Finally, fitness increased toward and beyond the northern range limit of C. cheiranthifolia but edge individuals were not best suited to conditions beyond the range. Despite strong genetic differentiation and very low, nondirectional gene flow, there was no evidence of local adaptation, except at the coarsest spatial scale, suggesting range limitation via constraints to dispersal.