Future climate change may outpace the ability of many species to either adapt or migrate in response to new stresses, resulting in higher rates of extinction. Paleorecords are key to understanding range shifts in response to climatic fluctuations, but are prone to false-negatives when detecting small populations in climate refugia. Genetic data can improve inferences by providing evidence of migration pathways and population structure. The mesic forests of the Pacific Northwest have two disjunct ranges at present: an interior range and a coastal range, separated by a vast area of fragmented landscapes in the rain shadow of the Cascade Range. The post-glacial development of this distribution has been the focus of longstanding debate, with numerous studies providing conflicting evidence. Its formation provides an important biogeographical model for understanding species’ ability to tolerate climate change. We conducted Genotyping-by-Sequencing using RADseq on two wide-ranging Pacific Northwest mesic conifers, mountain hemlock (Tsuga mertensiana, n=149), and western redcedar (Thuja plicata, n=144). Sequencing produced tens of thousands of genome-wide markers between the two study species, which were used to infer their post-glacial establishment in the interior range.
Results/Conclusions
Climate modeling and pollen records suggest that mesic forests were only recently able to disperse into the interior range. However, several mesic species demonstrate deep genetic divergence between coastal and interior populations, suggesting interior refugial persistence during the last glaciation. Our mountain hemlock data reveal that its coastal and interior ranges are divided into two distinct groups along a north/south axis, with the northern and southern interior each clustering with their respective coasts - a pattern inconsistent with an interior refugium. These results indicate mountain hemlock was able to disperse over the rain shadow via both a northern and a southern route, each independently populating a respective portion of the interior range. This pattern stands in contrast to pollen data that imply establishment after a single, rare dispersal from the coast. Dual migrations across a putative migration barrier suggest these species are more capable of tolerating habitat fragmentation and climate-induced migration than expected. As pollen presence alone cannot establish the number of independent dispersal events during expansion, our data offer new insights into the origin of the interior range and species resilience during past periods of climate change.