Bumble bees are among the best‐studied bee groups worldwide, yet surprisingly we know almost nothing about their overwintering habitats nor the microsite characteristics that govern selection of these sites. This gap represents a critical barrier for their conservation, especially if preferred overwintering habitats differ from foraging and nesting habitats. Current conservation plans focus on foraging habitat, potentially creating a problem of partial habitats where improved forage might fail to prevent population declines due to limited overwintering sites. We provide the first data on the overwintering habitat for any western North American bumble bee. Our data suggest that overwintering and foraging habitats are likely distinct, and queens’ selection of overwintering sites may be shaped by environmental stressors of the year. In our study area, queens overwintered in litter beneath cypress trees, where no floral resources exist. Whether this separation of overwintering and foraging habitat holds for other bumble bee species remains to be discovered. Our data highlight the need to consider the whole life cycle for understanding population dynamics and conservation planning. This need is underscored by growing evidence for the decline of multiple North American bumble bee species.
Many animals provide essential ecosystem services in the form of plant pollination. A rich literature documents considerable variation in the single visit pollination effectiveness of different plant visitors, but this literature has yet to be comprehensively synthesized. We conducted a hierarchical meta-analysis of 193 studies and extracted 1716 single visit effectiveness (SVE) comparisons for 252 plant species. We paired SVE data with visitation frequency data for 75 of these studies. Given the global dominance of honeybees in pollinator communities, we used these data to ask: 1) Do honeybees ( Apis mellifera ) and other floral visitors vary in their SVE?; 2) To what extent do plant and pollinator attributes predict the difference in SVE between honeybees and other visitors?; and 3) Is there a correlation between floral visitation frequency and SVE? We found that honeybees were significantly less effective than the most effective non-honeybee pollinator. Although not significantly different, honeybees also tended to be less effective than the mean community effectiveness. Honeybees were less effective as pollinators of crop plants and when compared to birds and other bees. Visitation frequency and pollination effectiveness were positively correlated, but this trend was largely driven by data from communities where honeybees were absent, suggesting that honeybees generally combine high visitation frequency and lower SVE. Our study demonstrates that non-honeybee floral visitors are highly effective pollinators of many crop and non-crop plants. While the high visitation frequency typically displayed by honeybees undoubtably makes them important pollinators, we show that honeybees are slightly less effective than the average pollinator and rarely the most effective pollinator of the plants they visit. As such, honeybees may be imperfect substitutes for the loss of wild pollinators and safeguarding global crop production will benefit from conservation of non-honeybee taxa. Open Research Statement: Although we are fully committed to data transparency, we are also aware of different research teams working on related meta-analyses. As such, we prefer to wait until our paper is accepted to make data publicly available but are happy to share data upon request. Data will be permanently archived on Figshare following acceptance.