Stained Glass Windows On The Evolutionary Process

The outcome of speciation is put to the test in regions where incipient species' ranges overlap and hybrid offspring are produced, known as hybrid zones. Delineating the parental species that contributed each part of a hybrid genome, the ancestry, can provide insights into the history of hybrid individuals and the progression of speciation between these lineages. The genomic locations where regions of different ancestry meet are called junctions. Junctions are historical recombination events that occurred in hybrids, and thus are records of the history of the zone in which they are found. In this dissertation, I explore the effect of demography on junctions and then investigate junctions in a classic hybrid zone. I find that junctions are very responsive to demographic history, particularly the age of the hybrid zone, the population size, and the rate of migration within a stepping-stone model. I also find that population substructure uniquely impacts ancestry in a way that is not captured by single population models. I also introduce a novel metric to describe the sharing of junctions between individuals-the junction frequency spectrum. I then review my attempt to integrate the junction frequency spectrum and other metrics of ancestry into an inference method using approximate Bayesian computation. My approach is unable to accurately estimate parameters, but I outline suggestions that might allow this inference to be successful in the future. Finally, I identify hundreds of thousands of junctions in mice from two populations within the European house mouse hybrid zone between Mus musculus domesticus and M. m. musculus. I show that these junctions are variable between chromosomes and across the genome. I present the first estimates of the junction frequency spectrum in a hybrid population. The populations that I survey show unique patterns of ancestry and junction sharing, indicating that they have distinct histories. I use junctions along with the site frequency spectrum to infer the age of the hybrid zone as a few thousand years. Finally, I identify several regions of the genome that are outliers for ancestry metrics as candidate regions for incompatibilities between M. m. musculus and M. m. domesticus.