Results

Over thirteen weeks, we observed 116 pollinators across at least three genera visiting S. lycopersicum and 330 pollinators across at least five genera visiting C. pepo flowers. Of the visitors toS. lycopersicum , bees we could not identify to genus made up 55.7% of visits, followed by Bombus (40.2%), Agapostemon(2.4%), and Apis mellifera (1.7%). Squash bees (Peponapis pruinosa) made up 54.8% of visits to C. pepo , followed by insects we could not identify to genus (18.1%), European honeybees (Apis mellifera ) (13.3%), and bees in the generaMelissodes (9.8%), Bombus (2.9%), and Agapostemon(1.1%).
Pollen limitation in S. lycopersicum increased with impervious surface across all measures of reproduction: fruits per flower, seeds per fruit, fruit mass, and fruit diameter (Table 3; Fig. 2-3). All four reproductive measures were significantly pollen-limited across all sites, except for fruit mass and diameter at our suburban site with the least surrounding impervious surface (19.7%, Chicago Botanic Garden) (Table 2, Fig. 2-3).
Across sites, the magnitude of pollen limitation in C. pepoincreased marginally with impervious surface for fruit set, but there was no effect for seed set (Table 3; Fig. 4a). Cucurbita pepofruit set was not significantly pollen-limited at any of the six sites. Seed set was significantly pollen-limited at one of the six sites, the site with the highest level of impervious surface (77.2%, Kenneth) (Table 2, Fig. 4b).
For S. lycopersicum , the rate of pollinator visitation decreased with greater surrounding impervious surface (R2m = 0.073, p = 0.015, estimate ± SE = -0.023 ± 0.01; Figure 5a). The rate of visitation by Bombus spp. did not relate to impervious surface (R2m = 0.052, p = 0.174, estimate ± SE = -0.019 ± 0.01; Figure 5c), while visitation by small bees we could not identify to genus decreased marginally with impervious surface (R2m = 0.089, p = 0.063, estimate ± SE = -0.026 ± 0.01; Figure 5e). In contrast, the rate of visitation for C. pepo did not change significantly with impervious surface (R2m = 0.007, p = 0.56, estimate ± SE = 0.007 ± 0.013; Figure 5b). The rate of visitation by P. pruinosa increased marginally with impervious surface (R2m = 0.148, p = 0.093, estimate ± SE = 0.049 ± 0.03; Figure 5d).