Biodiversity loss among pollinating insects has precipitously increased due to anthropogenic environmental changes. Among these taxa, the most comprehensive estimates of decline are for bees, for which human land use is the predominant driver of decline. Prior studies have demonstrated that human-modified environments can structure bee communities interspecifically, based on the matching of functional traits to local environments. However, little is known about whether bee functional traits can be structured intraspecifically across human-modified landscapes. Here, we ask the following questions: (1) Can bumble bees (Apidae: Bombus) exhibit intraspecific spatial structuring of body size, a developmentally plastic and ecologically consequential functional trait of bees? And, if so, (2) does this body size structure coincide with population genetic structure? We additionally estimate genetic diversity and inbreeding of our focal species - two factors that can affect extinction risk. To answer these questions, we studied five focal bumble bee species (B. auricomus, B. bimaculatus, B. griseocollis, B. impatiens, B. pensylvanicus) across an urban gradient in the greater Saint Louis, Missouri region in the North American Midwest. Across all species and sites, we collected >800 bees, genotyped individuals at 18 polymorphic microsatellite loci, and took thorax width measurements as a proxy for body size.
Results/Conclusions
We find that two of our studied species (B. impatiens and B. pensylvanicus) exhibit intraspecific spatial structuring of worker body size, despite a lack of population genetic structure. The direction of this body size structuring across the urban gradient differs between these two species: in a rural nature reserve, B. impatiens is smaller and B. pensylvanicus is larger when compared to intraspecific populations in urban areas. Genetic structure was not identified in any of our studied species, with all FST < 0.002; a power analysis confirmed that our data had sufficient statistical power to detect true genetic differentiation, if it were present. We also find evidence of inbreeding in B. griseocollis and reaffirm reports of low genetic diversity in B. pensylvanicus, results that have implications for the conservation of these species in environments altered by human activity. Collectively, these findings (1) are informative for conservation management practices of threatened species and (2) suggest that human-modified environments can induce landscape-level structuring of developmentally plastic functional traits.