Persistence of natural populations during periods of climate change depends on both migration (range shifts) and adaptation. Organisms are locally adapted to their environment and risk becoming maladapted if rate of climate change exceeds their capacity to respond through migration or in situ evolutionary adaptation. Numerous paleoecological studies have documented species range shifts in response to Quaternary climate fluctuations based on fossil records. The strong evidence for range shifts has also contributed to the assumption that adaptation was not a major response to past climate change. The rationale is that if adaptation was an effective response to cope with past climate change, then species should have maintained their original geographic distribution without migrating or reducing populations size during major climate shifts. This assumption ignores the possibility that a changing environment capable of causing massive demographic shifts is also likely to induce major selection events. Here we assess ecological, biogeographic and evolutionary processes that allowed persistence of white spruce (Picea glauca) populations during the Last Glacial Maximum (LGM) in Eastern Beringia (Alaska and Yukon), a climate refugium north of the North American ice sheets. To this end, we deploy an integrative approach combining population genomics analysis of more than 4200 annotated genes and species paleodistribution modeling (paleoSDM).
Results/Conclusions
Genetic assignment tests along with paleoSDM indicate that modern-day white spruce range in Eastern Beringia includes five distinct glacial lineages. Four of these arose from local expansion of in situ refugial populations that persisted through the Last Glacial Maximum in Alaska, whereas another lineage, found in Yukon, likely originate from long-distance postglacial colonization from an extra-regional refugium located south of the ice-sheets. A genome scan identified genes putatively under selection that exhibit high correlation with LGM paleoenvironment (n = 68 genes), with modern-day environmental gradient (n = 162 genes), or with both environments (n = 41 genes). This result illustrates the concept of temporal conditional neutrality, whereby alleles can be favored in glacial environments while currently being neutral and vice-versa. Gene ontology enrichment analyses suggest that genes putatively adapted to glacial conditions are overrepresented for biological functions related to abiotic stress responses (e.g. drought), while those subject to contemporary natural selection are overrepresented for biological functions related to increased competitive ability (e.g. morphogenesis, development, reproduction, phenology, growth, response to fire). To our knowledge, this study represents the first empirical attempt to assess the role of local adaptation to Quaternary climate change.