RESULTS
Pollinators at urban sites carried more conspecific pollen than pollinators at natural sites (Est. = 0.33; P = 0.02; Figure 1). Relative to natural sites, individual pollinators at urban sites carried pollen from fewer plant species (Est. = -0.52; P = 0.027), and the species diversity of pollen on individual pollinators was lower in urban sites (Est. = -0.17; P = 0.03). At the site level, urban and natural sites did not differ in the total number of plant species from which pollen was found on pollinators (t = -0.71, df = 8.1,P = 0.5), the number of plant species on which pollinators were caught (t = 0.55, df = 6.2, P = 0.61), or plant species richness (t = 0.87, df = 6.8, P = 0.42), diversity (t = 1.3, df = 7.2,P = 0.24); or evenness (t = 1.1, df = 36, P = 0.29). Across sites, pollinators exhibited negative frequency dependent foraging, with plant relative abundance predicting pollinator preference for a given plant species (Est = -0.34, P = 0.001). Across sites there was an inverse relationship between conspecific pollen amount and plant species richness (Est. = -0.057, P = 0.0496), diversity (Est = -0.38, P = 0.002), and evenness (Est = -0.79; P = 0.013).
Urban sites had lower insect species richness than natural sites (mean urban = 2.8; mean natural = 7.5; t = -3.5, df = 8.3, P = 0.008). Urban sites also had lower insect species diversity than natural sites (t = -3.9, df = 9.0, P = 0.0036). However, the amount of conspecific pollen was predicted neither by insect species richness (Est. = -0.029; P = 0.26) nor species diversity (Est = -0.22,P = 0.11).
Floral abundance at the site level was positively associated with the amount of conspecific pollen (Est = 0.13, P = 0.0014), and the number of flowers per plant was higher in urban areas (Est. = 1.8,P = 0.046). Urban sites also tended to have a higher total number of flowers but this trend was not significant (Est. 1.5, P = 0.06). There was substantial variation among sites in the species richness of invasive plants (Table 1). Urban sites contained fewer invasive plant individuals (Est. = -1.6, P = 0.028), but contained more flowers from invasive plants (Est. 1.2, P< 0.001). Furthermore, the increase in the number of flowers per plant individual in urban sites was greater for invasive plants (Interaction est. = 1.2, P < 0.001; Figure S2). Pollinators carried more conspecific pollen when caught on invasive plants in both types of sites (Est = 0.07, P < 0.001; Figure 2), and the difference in conspecific pollen amount between urban and natural sites was lower for invasive plants (Interaction est. = -0.058; P< 0.001).
Common pollinators comprised a greater amount of the pollinator community at urban sites than at natural sites (Est = 4.7, P< 0.001). Differences between urban and natural sites in the amount of conspecific pollen carried by pollinators was different for common and rare pollinators. Common pollinators carried more conspecific pollen than rare pollinators (Est. = 0.38; P < 0.001), and the increase in the amount of conspecific pollen carried in urban environments was greater for common pollinators (Interaction est. = 0.12; P < 0.001; Figure S3).