Exploring Spatial Patterns of Genetic Differentiation

摘要

Genetic variation provides the raw materials for both local adaptation and the formation of new species, and understanding the processes generating and maintaining the diversity of living organisms is a fundamental aim across all areas and levels of biological research. The central question that motivated this dissertation research is: what are the patterns of variation within and between populations and species, and what processes, ecological and evolutionary, are generating those patterns? My dissertation is focused on developing statistical methods to infer and visualize spatial patterns of genetic differentiation, and provide clues about the processes that have generated them.;In my first chapter, I develop a statistical method to quantify the relative contributions of ecological and geographic distance to patterns of genetic differentiation. The method, BEDASSLE, models the allele frequencies in a set of populations at a set of unlinked loci as spatially correlated Gaussian processes, in which the covariance structure is a decreasing function of both geographic and ecological distance.;In my second chapter, I discuss the conceptual framework of Isolation by Environment (IBE, as opposed to Isolation by Distance, or IBD) in a review and synthesis. This work clearly defines IBE as a pattern, rather than a process, and details the different processes that can generate a pattern of IBE.;In my third chapter, I extend the statistical framework developed in the first chapter to quantify signals of IBD and IBE to a general method for inferring and visualizing patterns of population structure. This method, SpaceMix, infers, for a set of sequenced samples, a map in which the distances between population locations reflect genetic, rather than geographic, proximity.;Together, these chapters represent an advance in the conceptual and statistical framework for analyzing spatial patterns of population genetic structure across landscapes.