Population genetics of wild and managed pollinators: implications for crop pollination and the genetic integrity of wild bees

Loss of habitat and chemical use associated with agriculture can cause population declines of wild pollinators. Less is known about the evolutionary consequences of interactions between species used in commercial agriculture and wild pollinators. Given population declines of many wild bee species, it is crucial to understand if commercial queens become established in natural areas, if wild bees visit agricultural fields and have the potential to interact with commercial bees, and if gene flow occurs between commercial and wild bees. We drew on a long-term data set that documents commercial bumble bee (Bombus impatiens) use in New England, and we conducted genetic analyses of foraging B. impatiens from areas with varying intensities of commercial bee use. In agricultural areas with a history of commercial bee use we also sampled bees directly from commercial hives. We found significant genetic differences among foraging B. impatiens and B. impatiens sampled directly from hives (average pairwise F′ST = 0.14), but not among samples of foraging bees from natural areas (average F′ST among foraging bees = 0.002). Furthermore, Bayesian analysis of population structure revealed that foraging bees caught in areas with a history of commercial bee use grouped with samples from natural areas. These results document an agricultural setting where there was no widespread introgression of alleles from commercial bumble bees to wild bumble bees, commercial bumble bees did not become established in natural areas, and wild bees were providing pollination services to crops.