OOS 83-7
Context dependency in pollinator-mediated plant-plant interactions along an urban-to-rural gradient
While urban areas have been shown to support significant pollinator communities, less is known about the effects of urbanization on the provision of pollination. Several studies have suggested that competition for pollination may be higher in urban areas, but none have explicitly tested this hypothesis. Given the increased interest in urban food production, understanding how urbanization affects crop pollination is of vital interest. This study looked at pollinator visitation to experimental arrays of cucumber and sunflower plants in paired high- and low-floral density patches along an urban-to-rural gradient in SE Michigan, USA, to determine how visitation is affected by local floral density, and whether this varies with degree of urbanization.
Results/Conclusions
Pollinator visitation to the two focal plant species showed different responses to changes in both local floral density and degree of landscape urbanization. Cucumber experienced greater visitation with increasing floral density, though this effect quickly plateaued at moderate floral densities; there was no effect of urbanization on visitation to cucumber. Sunflower, on the other hand, experienced reduced visitation as both urbanization and floral density increased. Total bee abundance, measured by pan traps and netting, was not affected by urbanization. We hypothesize that these results reflect differential responses to urbanization from the most important pollinator species to each plant. Specifically, >70% of visits to sunflower were from ground-nesting bees, which have been shown to decline in abundance with urbanization; cucumber was visited by more cavity-nesting species that may fare better in urban settings. These results suggest that, while robust pollination services can be maintained in urban environments, differential responses among pollinators to urbanization strongly mediate the provision of this service to particular plant species. Moreover, measures of pollinator abundance cannot necessarily predict visitation, even to broadly attractive flowers.