Population genetics of divergence and admixture in bumble bees across complex landscapes
Understanding how climatic variation across complex landscapes affects dispersal has important implications for conservation, as well as for revealing how populations diverge and adapt to local conditions. Population genetics studies in bumble bees (Bombus) have commonly revealed weak population structure over large geographic scales, presumably due to a high capacity for dispersal and lack of gene flow barriers. However, in species that occupy montane habitats, more complex patterns of genetic structuring have emerged. Bombus bifarius is widespread throughout western mountains of North America, where its range is determined by availability of suitable environmental conditions. B. bifarius also comprises a complex of abdominal pile color variants, with easternmost populations exhibiting “red-banded” coloration, westernmost populations exhibiting “black-banded” coloration, and geographically intermediate populations typically exhibiting some degree of intermediate coloration but have historically been considered closely related to the black-banded bees. Together, the heterogeneous geographic range and color pattern variation is likely to reflect a complex evolutionary history. I present data on B. bifarius that combines traditional population genetic sampling across this species’ range with climate-based niche models to test how abiotic heterogeneity and landscape connectivity may impact patterns of population connectivity at large spatial scales. I also present new data that focuses on intensively sampling the genome with RNAseq and RADtags to test the resulting hypotheses in three populations selected to represent the two geographic and color pattern extremes, as well as geographically and phenotypically intermediate bees that could reflect a potential history of admixture.
Microsatellite data suggest that B. bifarius can be grouped into population clusters that roughly reflect geography and color pattern, however differentiation is fairly weak, and most individuals contain non-negligible contributions from multiple ancestral populations. Isolation-by-resistance analyses derived from niche model predictions support the hypothesis for ongoing gene flow restricted by regions of narrow climatic connectivity over a scenario of pure isolation. However, analysis from 10,000s of loci from population genomic data find little support for ongoing gene flow across the B. bifarius range, instead suggesting relatively deep divergence of red-banded bees from the black-banded and intermediate color forms, while the latter show substantial genetic similarity. Results suggest that B. bifarius may represent populations in the process of speciation, as opposed to a patchwork of populations connected by gene flow. I discuss possible causes of discrepancies between data sets, and the advantages of incorporating genomic data into landscape genetic studies of bumble bees.