DNA from archaeological, paleontological, and museum samples (ancient, or aDNA) provides a unique opportunity to trace eco-evolutionary history of populations affected by environmental shifts on geological time scale. Yet it is still unclear how climate-driven environmental change and biogeographical barriers affect diversification, population size and population structure of large-bodied herbivores inhabiting northern regions of the Northern Hemisphere. The goal of this dissertation is to fill in this gap by utilizing ancient DNA techniques and population genetic analysis to reveal the demographic and population history of extinct and present-day equids, genus Equus, focusing on their key ancient dispersal corridor - the Bering Land Bridge.

In the following chapters, I explore the links between paleoenvironments and population history of various equid groups using high coverage paleogenomes recovered from fossil horse specimens sampled across Beringia. In my first chapter, I use in-solution DNA capture enrichment and mitochondrial genome assembly to reconstruct a whole mitochondrial genome of a specimen found in Western Beringia and initially identified as E. hemionus, or an Asiatic wild ass. With molecular phylogenetic analysis I demonstrate that the specimen belongs to a group of caballoid horses, E. ferus, rather than stenonid wild asses. The results obtained in Chapter 1 highlight the utility of ancient DNA studies in identification of incomplete, juvenile, or otherwise problematic museum specimens. In my second chapter I discover that Beringia was a key contact zone for populations of Late Pleistocene caballoid horses, E. ferus. I use new high coverage nuclear and mitochondrial paleogenomes, isolated from fossils of caballoid horses sampled across the Northern Hemisphere to infer that North American and Eurasian caballoid horse populations diverged around 0.8-1 million years ago. With coalescent simulations and genome-wide adamixtude inference I show that evolution of caballoid horses after this divergence continued in the presence of cross-continental gene flow. My demographic inference suggests that disappearance of the Bering Land Bridge likely exacerbated an already ongoing extinction of Beringian caballoid horse populations. In the third chapter, I recover the ~700,000 year old paleogenome of a previously unknown stenonid horse species inhabiting Klondike, Canada’s Yukon Territory - the oldest non-caballoid equid genome known to date. Using genotype likelihood approach on a dataset of present-day and ancient equid nuclear genomes, I show that the population of the newly discovered stenonid equid species was evolutionary close to the present-day zebras and Asiatic wild asses. I suggest that the new to genetics species likely represents an extinct branch of archaic stenonid ungulates that coexisted with “true”, or caballoid equids in the Early-Middle Pleistocene Yukon. In the fourth chapter I expand my study system to another iconic Bringian megafauna species - steppe bison, Bison priscus. Using molecular phylogeny reconstructed from new high coverage mitochondrial genomes, I explore the phylogenetic diversity of steppe bison in Western Beringia. I confirm the existence of the deeply divergent steppe bison clade and shed new light on the evolutionary history of bison during the Pleistocene to Holocene transition in ancient Siberia.