The evolution of population differentiation of autogamous selfing ability in Collinsia verna

Pollinator service may be one of the most important agents of selection on selfing. When pollinator service is relatively good, selection may decrease selfing ability to minimize the number of low quality selfed offspring that are produced. When pollinator service is poor, selection may increase selfing ability because many selfed offspring provide more fitness than producing no offspring, or very few outcrossed offspring. Therefore, differences in pollinator service across populations may lead to differences in selection on selfing ability, and produce adaptive differentiation of the trait. This hypothesis may explain why population differentiation in selfing ability is often observed in species that can reproduce through a mix of selling and outcrossing. However, there are few empirical studies of this hypothesis.;In my dissertation research, I assayed differentiation of selfing ability across populations of the mixed-mating Collinsia verna from throughout its geographic range. I used three approaches to test the hypothesis that differences in selfing ability evolved as an adaptive response to differences in pollinator service across populations. I found a significant negative correlation between selling ability and pollinator service, which was consistent with the hypothesis that adaptive differentiation of selfing ability arose from pollinator service. However, I discovered that genetic drift alone could have produced the amount of genetic differentiation in selfing ability across populations that I observed, and therefore, differential selection may not have played a role in differentiation of setting ability. Additionally, by measuring selection on selfing ability in a population under natural and experimentally reduced levels of pollinator service, I found no evidence that differences in pollinator service produce differences in selection on selfing ability though female seed production. There was strong positive selection on selfing ability in both pollination environments, which suggests that in many pollination conditions, populations may experience selection to improve selfing ability.;This research demonstrates that differences in selfing ability across populations of mixed-mating species may not always be adaptive. Further, this research highlights the importance of measuring selection on selfing ability through a measure of fitness that incorporates inbreeding depression and male fitness. Selection on selling ability though female seed production may commonly be positive, but it remains untested whether differences in pollinator service would produce differences in selection on setting ability if a more comprehensive measure of fitness were used.