UC San Diego

A fundamental controversy in the study of speciation is whether allopatry is the dominant mode of speciation. Understanding the genetic architecture of reproductive isolation between incipient species may be important in resolving this question. One approach to examining reproductive isolation is to estimate divergence times at multiple loci to determine if gene flow was restricted instantaneously among species (allopatry) or whether the divergence is better explained by stages of reproductive isolation (parapatry). The first two chapters of my dissertation develop a model to test these modes of speciation in sister species and to predict which genes were important in the early stages of speciation. The last chapter of my thesis focuses on genomic scans for these speciation factors to determine the degree of positive selection acting on genes with enriched expression in the female reproductive tract. This class of genes is thought to be the target of early genomic incompatibilities driving reproductive isolation in many incipient species. In Chapters 1 and 2, I use two swallowtail butterfly species, Papilio glaucus and Papilio canadensis, as a system to study the genetic basis of reproductive isolation. The two species make a compelling system because they differ in many ecological traits, like mimicry, that are linked to their sex chromosomes. Yet the species are still capable of forming fertile hybrids. Using markers linked to the Z chromosome, I ask whether a single divergence time estimate (allopatry) fits the polymorphism data better than a complex speciation process (parapatry). I develop an approximate Bayesian coalescent- based method to estimate the ancestral population size of the species and the per locus and joint divergence times. Using this framework, one can also identify candidates for reproductive isolation factors. These loci are expected to have deeper divergence time estimates than randomly selected loci. I establish that allopatric speciation is unlikely in P. glaucus and P. canadensis, and identify two genes that may be linked to speciation factors. In Chapter 3, I examine the effects of demography on scans for positive selection. Reproductive isolation between incipient species may arise as a by-product of divergent selection on a small number of phenotypic traits. If these traits have a simple genetic basis, then a relatively small number of genes may be responsible for reproductive isolation. In scans for positive selection, researchers often focus on genes preferentially expressed in the reproductive tract because interactions between male and female proteins may be the earliest drivers of reproductive isolation. Using a variety of single and multilocus tests for selection in ancestral African Drosophila melanogaster, I compare polymorphism in 9 female reproductive genes that are candidates for positive selection to a control set of 137 randomly chosen genes. Though I find evidence for recent positive selection at 2 of the 9 candidate loci in D. melanogaster from Zimbabwe, I find no evidence supporting the notion that the candidate loci are more frequent targets of adaptive evolution. These results demonstrate that a previous study identifying elevated rates of positive selection on female reproductive genes as a group may be incorrect because the study examined a population that recently underwent a severe bottleneck. This study highlights the importance of incorporating demographic parameters into scans for positive selection. Taken together, my dissertation describes a method for systematically evaluating the prevalence of allopatric speciation and shows the demographic considerations necessary for identifying candidate genes underlying reproductive isolation