The shift from outcrossing to selfing is one of the most common and evolutionarily significant transitions in flowering plants. Breeding system affects several evolutionary processes, which result in important differences between closely related selfing and outcrossing taxa. However, while the genetic and phenotypic differences between selfers and outcrossers have been documented often, the adaptive processes responsible for these changes are still not well understood.
In this dissertation, I compare several populations of the outcrossing Lupinus nanus and the selfing L. bicolor, to investigate the genetic, selective, phenotypic, and adaptive consequences of mating system. I predicted that the selfing mating system of L. bicolor would result in low levels of genetic diversity and thus an inability to respond to selective pressures. The outcrossing nature L. nanus, on the other hand, should maintain sufficient intrapopulation genetic diversity to allow for natural selection to take place.
First, I used microsatellites to examine genetic diversity and population differentiation in both species to test the hypothesis that L. bicolor had less intrapopulation diversity and more interpopulation isolation than L. nanus. Second, I calculated the selective landscapes of both species and their ability to respond to selection. Third, I measured local adaptation in L. nanus and L. bicolor to answer the question: Do L. nanus or L. bicolor populations exhibit higher levels of local adaptation? Finally, I compared variance-covariance matrices to examine past evolutionary trajectories.
I found that L. bicolor has consistently lower intrapopulation genetic diversity and higher levels of isolation by distance than L. nanus. Additionally, L. nanus showed a consistent floral shift in the direction predicted by selection whereas L. bicolor did not, suggesting that L. bicolor is unable to respond to selective pressures due to its lack of genetic diversity. My measurements of local adaptation and home-site advantage were consistent with this finding, as all L. nanus populations showed evidence of home-site advantage, and three of the four populations were locally adapted, while L. bicolor showed neither home-site advantage nor local adaptation for any of my study populations. The variance-covariance matrix comparisons were less conclusive, showing moderate similarities among floral suites of traits for both L. nanus and L. bicolor.
We conclude that, because L. bicolor is strongly selfing, it lacks the genetic diversity needed for selection to work on. Evolution in this species is likely dominated by neutral evolution, such as genetic drift. L. nanus, on the other hand, is not genetically depauperate and thus can respond to selective pressures. We also discuss the conservation and restoration implications of these findings.