Plantings in residential neighborhoods can support wild pollinators. However, it is unknown how effectively wild pollinators maintain pollination services in small, urban gardens with diverse floral resources. We used a 'mobile garden' experimental design, whereby potted plants of cucumber, eggplant, and purple coneflower were brought to 30 residential yards in Chicago, IL, USA, to enable direct assessment of pollination services provided by wild pollinator communities. We measured fruit and seed set and investigated the effect of within-yard characteristics and adjacent floral resources on plant pollination. Increased pollinator visitation and taxonomic richness generally led to increases in fruit and seed set for all focal plants. Furthermore, fruit and seed set were correlated across the three species, suggesting that pollination services vary across the landscape in ways that are consistent among different plant species. Plant species varied in terms of which pollinator groups provided the most visits and benefit for pollination. Cucumber pollination was linked to visitation by small sweat bees (Lasioglossum spp.), whereas eggplant pollination was linked to visits by bumble bees. Purple coneflower was visited by the most diverse group of pollinators and, perhaps due to this phenomenon, was more effectively pollinated in florally-rich gardens. Our results demonstrate how a diversity of wild bees supports pollination of multiple plant species, highlighting the importance of pollinator conservation within cities. Non-crop resources should continue to be planted in urban gardens, as these resources have a neutral and potentially positive effect on crop pollination.
Keywords: Ecosystem service; Floral resources; Pollinators; Residential garden; Urban agriculture.