The Genomics of Speciation
MetadataShow full metadata
Speciation, the process by which reproductive isolation evolves between diverging lineages, is pivotal to our understanding of evolution. Across multiple wild populations I explored the genetic architecture of reproductive isolation and adaptive traits, the interaction between gene flow and genetic architecture of traits and their impact on the process of speciation, and finally I assessed the repeatability of genetic differentiation and absolute diversity across the genome, across multiple species pair comparisons. My dissertation includes investigations of hybridization between pitcher plants (Sarracenia sp.), a repeated trophic polymorphism within the Cuatro Ciénagas cichlid fish (Herichthys minckleyi), and a species complex of blue butterflies (Lycaeides sp.) that have a complicated evolutionary history that includes repeated, independent evolution of hybrid species. I generated genome-wide population genetic data to quantify patterns of genomic differentiation in all of these case studies. I used a combination of analyses to dissect the relationships between trait architecture, adaptation, and reproductive isolation. Bayesian clustering was used to describe patterns of variation and identify areas of admixture. Bayesian Sparse Linear Mixed Models (BSLMM) were used to map the genetic architecture of a variety of traits and I compared estimates of introgression for genomic regions that contribute to trait variation to understand if these traits are associated with fitness in admixed individuals. Bayesian Genomic Clines models were used to identify patterns of introgression and excess ancestry in admixed individuals. Patterns of differentiation measured along chromosomes was used to assess the repeatability of differentiation and potential adaptation. I found remarkable variation in trait architecture, ranging from very simple to highly complex. Many genomic regions were associated both with trait variation and patterns of strong selection, though this was not universal. Repeatable patterns were detected in some regions of the genome which suggests that evolution can be predictable, yet there are also instances of unrepeated differentiation suggesting a role for historical contingency. Overall, my results contribute to our understanding of the process of speciation and highlight the power of genome-wide data to resolve important questions in evolution.