Urban community gardens are increasingly popular, and play an important role in small-scale food production in American cities. However, many urban garden crops require pollination, and urbanization may reduce the diversity and abundance of key pollinators through loss of ground-nesting habitat. We asked: (1) Is crop production in community gardens limited by pollination services? (2) If so, do pollination services vary spatially across cities in response to landscape variables that could influence pollinator nesting success?
We used a combination of citizen science and a manipulative experiment to address this question in Seattle community gardens. We distributed cherry tomato starts (pollinated only by bumble bees in the Pacific Northwest) to gardeners for transplantation at ten garden sites and supplemented these with ten additional replicate potted cherry tomatoes at each site. We helped participants impose three pollination treatments: (1) pollinator exclusion; (2) open to pollinators; (3) semiweekly supplemental pollination with a hand-held tuning fork, which mimics sonication by bumble bees. We observed pollinator visitation rate weekly and recorded tomato volume. Simultaneously, we collected USGS landscape cover data at radii of 250, 500, 750, and 1000 m around each garden, including proportions of impermeable surface, herbaceous cover, wetland, and forest cover.
Results/Conclusions
Unexpectedly, more than 90% of visits to cherry tomato flowers were carried out by a single bumble bee species, Bombus vosnesenskii. We detected visits by three additional bumble bees during the first two weeks of flowering; for the remainder of the eight-week season, the only pollinating visitors were B. vosnesenskii and Apis mellifera. In the absence of bee visits, tomato fruit size was strongly reduced. There was mixed evidence for pollen limitation, but we interpret this with caution because our supplemental sonication did not provide cross pollen. A smaller follow-up study found that hand sonication with pollen transfer strongly increased fruit and seed set over ambient pollination. Visitation rate and fruit size were both reduced in gardens surrounded by more impermeable surface at multiple radii. Herbaceous vegetation, forest cover, and wetland had no detectable influence on either visitation or fruit size.
Our results suggest that the type of development surrounding urban gardens influences the productivity of the crops grown there. These results are especially important for gardens located in industrialized or highly-developed areas of cities where parks and potential nesting habitat are limited. In addition, our results underscore the potential fragility of pollination services in urban gardens where pollinator diversity is low.