No abstract

Bumble bees (Bombus spp.) are important pollinators, yet rapidly declining globally. In North America some species are thriving while others are nearing extinction. Recognizing subtle differences in species’ biology and responses to environmental factors is required to illuminate key threats and to understand their different population trajectories. We intensively surveyed bumble bees in Ohio, USA, along the receding southern boundary of many species’ ranges, to evaluate current conservation status of the state’s species. In 318 90-min field surveys across two consecutive years we observed 23,324 bumble bees of 10 species visiting 170 plant species. Habitat, landscape, latitude, and their interactions significantly influenced bumble bee abundance, species richness, and community composition during peak season. Sites planted with flowers yielded more bumble bee individuals and species than did sites not planted with bee food plants. Bombus impatiens, B. griseocollis, and B. bimaculatus comprised 93% of all observations. Their abundances all peaked in habitats planted with wildflowers, but there were species-specific responses to local and landscape factors. Three less common species (B. fervidus, B. vagans, and B. perplexus) were more likely to be found in forested landscapes, particularly in the northeastern portion of the state. Bombus perplexus was also affiliated with planted urban wildflower patches. These results provide a strong starting point for future monitoring and conservation intervention that targets less common species. A quantitative synthesis of detailed state-level and regional datasets would allow additional insight into broad scale patterns of diversity in bumble bee communities and species conservation trajectories.

No abstract

Regionally specific flower preference data are needed to optimize conservation habitat plantings for at-risk pollinators such as bumble bees (Bombus spp.).Current tools for selecting flowers for plantings rely on raw bee flower visits, which can be biased toward abundant flowers. To assist in planning habitat enhancements for bumble bees, we quantified genus-and species-level floral preferences using a selection index that accounts for floral availability. Through 477 h of observation in Ohio, USA during the summers of 2017 and 2018, we recorded 22,999 observations of eight Bombus species visiting 96 flowering plant taxa. As a genus, Bombus selected flowers nonrandomly; the most strongly preferred plants included Asclepias spp., Cirsium spp., Convolvulaceae, Dipsacus spp., Echinacea purpurea, Monarda fistulosa, Penstemon digitalis, and Silphium spp. Only a few Fabaceae were highly selected (Baptisia spp., Trifolium pratense, and Vicia spp.), while some were preferred only during their peak bloom (Securigera varia), and others were not preferred by bumble bees (T. hybridum and Melilotus spp.). Diets differed among habitats, and in restored meadows, bumble bees selected for native planted species such as Monarda fistulosa, Asclepias syriaca, Echinacea purpurea, Penstemon digitalis, and Silphium spp. Diets and preferences shifted over the season, largely driven by changes in plant phenologies (e.g., in June, Penstemon was strongly selected, in July, Asclepias, and in August, Verbena). For the three most common Bombus (B. impatiens, B. griseocollis, and B. bimaculatus), rarefaction analysis indicates that we were able to detect almost all plants in their summer diets. However, for five less common species, even our extensive sampling was insufficient to fully characterize their diets. The common Bombus species differed in their feeding niches, perhaps reducing interspecific competition. In contrast, we found high diet overlap between three rarer species-B. vagans, B. fervidus, and B. pensylvanicus, suggesting that these at-risk species might benefit from different floral communities than would the common species. Five of eight species