Inferring species phylogeny and gene flow from genomic data

Abstract

Species occasionally exchange genetic material. This process of gene flow between species, also known as introgression, can benefit the recipient species by facilitating adaptation to new environments. But how common is introgression in nature? What are its evolutionary consequences besides local adaptation? Over the past decade, genomic studies have documented numerous examples of introgression across a wide range of organisms. However, methods used to study introgression are often heuristic or approximate, limiting our understanding to coarse features of introgression. They rely on data summaries, from which information about past demography and species divergence has been lost or severely reduced. Here, we study methods that make better use of genomic sequence data to detect and quantify features of introgression in finer detail, including its intensity, timing, and direction. These methods, based on the multispecies coalescent framework, also enable the estimation of key population parameters such as effective population sizes and species divergence times. We first establish a theoretical correspondence between two idealized models of gene flow: pulse introgression and continuous gene flow, and the information needed to infer gene flow under each model. We then study how using an incorrect model of gene flow can impact inferences, including the ability to detect gene flow (Chapter 1). Next, we apply the pulse introgression and continuous gene flow models to analyze genomic data in two insect systems: Heliconius butterflies (Chapter 2) and Anopheles mosquitoes (Chapter 3). In Heliconius, we obtain a resolved species phylogeny and introgression history of major species groups within the genus for the first time, despite extensive historical and ongoing gene flow. In Anopheles, we study North American species in the subgenus Anopheles for which no genomic resources exist and there is no prior evidence for gene flow. Using newly generated genome assemblies and resequencing data, we find evidence for extensive gene flow among sibling species, leading to a hypothesis that introgression is a common feature of species divergence rather than an exception. Advances made in this dissertation contribute toward understanding the prevalence and evolutionary impact of introgression across the diversity of life and throughout the Earth’s history.