Asian pika populations track local glaciation events through the Pleistocene

Abstract Background The Pleistocene glaciation cycles (2.6 mya - 11 kya) were major climatic events that shaped diversity and community assemblages on Earth. Cold-adapted, high-elevation specialists are expected to have responded negatively to warm interglacial periods, with populations contracting and being pushed up the elevational gradient, leading to isolated populations on mountain tops. During the cool glacial maximums, they are expected to have been distributed at both low and high elevations, with populations expanding and facilitating gene flow across mountains. Pikas ( Ochotonidae ) are poorly studied high-elevation specialist Lagomorphs being extirpated at alarming rates due to climate change. Insights into their historical demography and current effective population size are crucial in determining their current population status to inform conservation planning. Results We use de novo assembly to construct partial genomes of three Asian pika species ( Ochotona ladacensis, O. nubrica, and O. macrotis ) that differ in their life history, lifestyle, and social behaviors. We use a combination of species distribution modeling (SDM) projected back in time, in addition to inferring historical demographic patterns using a pairwise sequentially Markovian coalescent (PSMC) approach. Our SDMs predicted the largest bio-climatic niches for all three species during the last glacial maximum (LGM ∼20 kya), and the smallest niches at the last interglacial (LIG∼120 kya), agreeing well with the elevational shift hypothesis. Our PSMC models for the populations in the Changthang Biotic province (CBP), revealed largely synchronous oscillations of populations sampled in geological timescales but displayed contrasting temporal patterns of population spikes when compared with SDMs. The largest effective population sizes of all species examined on the CBP were inferred to pre-date both the LGM and LIG and were placed around the MIS-6 (∼ 191 kya). After the MIS-6 local glaciation event, populations were inferred to have been through a deep bottleneck, correlating well with local glaciation patterns on this plateau. Conclusion Our study illustrates the power of using complementary approaches to infer historical demography in mountainous landscapes with complex glaciation histories. While SDMs appropriately describe species responses to historical climate across the geographic range of species, PSMCs are more informative about local population dynamics when subject to large degrees of population differentiation. All three species of pikas responded similarly to historical glaciation events on the CBP that differ from the rest of the world and other parts of the Himalayas and likely have very low effective population sizes at current timescales. This is one of the first few studies to examine the population demography of small mammals on the Indo-Tibetan Plateau and has the potential to inform species and population-specific conservation planning in the Himalayas. It is very likely that small mammals such as pikas, marmots and voles on the CBP have low effective population sizes due to historical inbreeding during the LIG and the lack of a glaciation rescue during the LGM. This highlights the need for conservation measures for small mammals on the landscape.