Pollination service provision
Urbanisation had an overall positive effect, although not statistically significant, on pollination service provision (d = 0.20; 95% CI = [-0.10, 0.51]; p = 0.19). The heterogeneity of the effect sizes was large and statistically significant (Qt = 1543.29, df = 115, P < 0.001). The pollinator groups sampled in the studies that estimated pollination service provision differed significantly in their contribution to plant reproductive success (Fig. 5, Table 2). Honey bees significantly boosted plant reproductive success (Fig. 5). Flowering plants with radial flowers had an increased reproductive success with increasing urbanisation (Table 2, Fig. 5). The origin of plant species (native vs. non-native) did not explain the effects of urbanisation on pollination services (Table 2). Urbanisation had a positive effect, although not statistically significant, on the pollination of both native and non-native plants (Fig. 5). The five reproductive success indexes significantly explain the effects of urbanisation on pollination (Table 2), with positive responses in studies that estimated seed set and visitation duration; positive but not statistically significant responses for studies that measured fruit set and visitation rate and negative responses, although not statistically significant, for studies that evaluated the number of flower visits (Fig. 5). In addition, the richness and abundance of pollinators were not significantly related to pollination service provision (Fig. S6, Table S4)