Assessing the relative role of evolutionary processes on genetic diversity is critical for understanding species response to climatic change. However, many processes, independent of climate, can lead to the same genetic pattern. Because effective population size and gene flow are affected directly by abundance and dispersal, population ecology has the potential to profoundly influence patterns of genetic variation over microevolutionary timescales. Here, we use aDNA data and simulations to explore the influence of population ecology and Holocene climate change on genetic diversity of the Uinta ground squirrel (Spermophilus armatus). We examined phylochronology from three modern and two ancient populations spanning the climate transitions of the last 3000 years. Population genetic analyses based on summary statistics suggest that changes in genetic diversity and structure coincided with the Medieval Warm Period (MWP), c. 1000 years ago. Serial coalescent simulations allowed us to move beyond correlation with climate to statistically compare the likelihoods of alternative population histories given the observed data. The data best fit source–sink models that include large, mid-elevation populations that exchange many migrants and small populations at the elevational extremes. While the MWP is likely to have reduced genetic diversity, our model-testing approach revealed that MWP-driven changes in genetic structure were not better supported for the range of models explored. Our results point to the importance of species ecology in understanding responses to climate, and showcase the use of ancient genetic data and simulation-based inference for unraveling the relative roles of microevolutionary processes.