Genomic analyses capture the human-induced demographic collapse and recovery in a wide-ranging cervid

Abstract

The glacial cycles of the Quaternary heavily impacted species through refugial contractions and post-glaciation expansions. Similarly, population demographic trajectories were intensely shaped by human pressures such as harvest and land use changes, particularly during the colonial-era. White-tailed and mule deer survived in different refugia during the last glacial maximum and were heavily impacted by the European colonisation as harvest and forestry practices drove white-tailed deer to extirpation in some regions, and severely reduced mule deer populations. Here, we analysed 73 re-sequenced deer genomes form across their North American range to understand the consequences of climatic and anthropogenic pressures on deer demographic history and adaptation. We found a strong signal of glacial-induced vicariance and demographic decline; notably, there is a severe decline in white-tailed deer effective population size (Ne) at the end of the Last Glacial Maximum. We found robust evidence for a colonial impact in the form of a recent and dramatic drop in Ne in all analysed populations, followed by a recovery in mainland white-tailed deer only. Historical census size and restocking data show a remarkable parallel to historical Ne estimates, while temporal Ne/Nc ratio shows patterns of conservation concern for mule deer. Signatures of selection highlight genes related to olfaction genes essential to deer biology and a cold receptor previously detected in woolly mammoth. We also detected numerous immune-genes that we surmise reflect the changing land-use patterns in North America. Our study provides a detailed picture of anthropogenic and climatic-induced decline in deer diversity, and clues to understanding the conservation concerns of mule deer and the incredible demographic recovery of white-tailed deer.

Genealogical asymmetry under the IM model and a two-taxon test for gene flow

“…, inferring low levels of gene flow via secondary contact between the two species (Kessler and Shafer 2024). We calculated A m across a range of block sizes for a pair of individuals with high coverage genome sequences data (Ov ON6 and Oh WA1).…”

“…We next perform simulations under two alternative demographic histories inferred by Kessler and Shafer (2024). The first of these demographies includes a low rate of bidirectional secondary contact gene flow, where the probability that a single lineage migrates at least once is low (P = 0.015).…”

“…As a result, blocks that span congruent genealogies due to hidden recombination events will have non-zero values of A m , even in the absence of migration. In the case of the Odocoileus deer it was helpful to perform coalescent simulations of previously inferred demographic histories (Kessler and Shafer 2024), as this suggested that A m is likely non-zero due to past gene flow, rather than due to differences in N e alone. Nevertheless, the influence of recombination on estimates of A m is certainly a weakness of our approach.…”