Premise: Evidence suggests that bees may benefit from moderate levels of human development. However, the effects of human development on pollination and reproduction of bee-pollinated plants are less well understood. Studies have measured natural variation in pollination and plant reproduction as a function of urbanization, but few have experimentally measured the magnitude of pollen limitation in urban versus non-urban sites. Doing so is important to unambiguously link changes in pollination to plant reproduction. Previous work in the Southeastern US found that urban sites supported twice the abundance of bees compared to non-urban sites. We tested the hypothesis that greater bee abundance in some of the same urban sites translates into reduced pollen limitation compared to non-urban sites.

Methods: We manipulated pollination to three native, wild-growing, bee-pollinated plants: Gelsemium sempervirens, Oenothera fruticosa, and Campsis radicans. Using supplemental pollinations, we tested for pollen limitation of three components of female reproduction in paired urban and non-urban sites. We also measured pollen receipt as a proxy for pollinator visitation.

Results: We found that all three plant species were pollen-limited for some measures of female reproduction. However, opposite to our original hypothesis, two of the three species were more pollen-limited in urban relative to non-urban sites. We found that open-pollinated flowers in urban sites received less conspecific and more heterospecific pollen on average than those in non-urban sites.

Conclusions: These results suggest that even when urban sites have more abundant pollinators, this may not alleviate pollen limitation of native plant reproduction in urban landscapes.

Data are fruit and seed data from pollen supplementation experiments (hand-pollinated and open-pollinated control flowers) and pollen deposition data (both conspecific and heterospecific pollen receipt).

No specific software is required.