Phylogeography within the Peromyscus maniculatus species group: Understanding past distribution of genetic diversity and areas of refugia in western North America.

The effects of anthropogenic climate change on biodiversity have been recognized on every continent, ocean, and across different taxonomic groups. Here, we study the range dynamics and demography of a cosmopolitan species: the deer mouse, Peromyscus maniculatus. We generated a multilocus SNP dataset using the ddRADseq protocol for 218 individuals across the geographic range within three western North American lineages of this species group. We evaluated population structure using several methods and explored the correlation between geographic and genetic distances. We modeled the demographic history using a site frequency spectrum approach and used a machine learning algorithm to infer current and past (Last Glacial Maximum; LGM) environmental suitability. Lastly, we explored the origin of population expansion for the identified lineages. The genome-wide SNP dataset was able to identify-three regionally distinct groups- 1) P. m. gambelii (southern California); 2) P. keeni (Pacific Northwest); 3) P. m. sonoriensis (a broad population spanning the Pacific Northwest through central California and across the Rocky Mountains into the Great Plains). Demographic analysis indicated the splits between the three populations occurred within the last 500 thousand years, with one very recent (late Holocene) split. Ecological niche models for each of these lineages predicted suitable environment present throughout their known ranges for current conditions, and a severe reduction of northern habitat in the past. The deer mouse has responded to past climate changes by expanding its range during interglacial periods and contracting its range during glacial periods leading to strong population differentiation. But lower magnitude climate change or other processes within the Holocene interglacial period led to population differentiation as well, which is likely still ongoing today given the substantial anthropogenic climate change and other landscape transformations caused by humans during the Anthropocene. By understanding the historical processes that led to the contemporary geographic distribution of biodiversity, we can determine the relative importance of different factors that shape biodiversity, now and into the future.

Genome-wide genetic variation coupled with demographic and ecological niche modeling of the dusky-footed woodrat (Neotoma fuscipes) reveal patterns of deep divergence and widespread Holocene expansion across northern California.

Understanding how species have responded to past climate change may help refine projections of how species and biotic communities will respond to future change. Here, we integrate estimates of genome-wide genetic variation with demographic and niche modeling to investigate the historical biogeography of an important ecological engineer: the dusky-footed woodrat, Neotoma fuscipes. We use RADseq to generate a genome-wide dataset for 71 individuals from across the geographic distribution of the species in California. We estimate population structure using several model-based methods and infer the demographic history of regional populations using a site frequency spectrum-based approach. Additionally, we use ecological niche modeling to infer current and past (Last Glacial Maximum) environmental suitability across the species' distribution. Finally, we estimate the directionality and possible spatial origins of regional population expansions. Our analyses indicate this species is subdivided into three regionally distinct populations, with the deepest divergence occurring ~1.7 million years ago across the modern-day San Francisco-Bay Delta region; a common biogeographic barrier for the flora and fauna of California. Our models of environmental suitability through time coincide with our estimates of population expansion, with relative long-term stability in the southern portion of the range, and more recent expansion into the northern end of the range. Our study illustrates how the integration of genome-wide data with spatial and demographic modeling can reveal the timing and spatial extent of historic events that determine patterns of biotic diversity and may help predict biotic response to future change.