Urbanization is associated with shifts in bumblebee body size, with cascading effects on pollination

Abstract

Urbanization is a global phenomenon with major effects on species, the structure of community functional traits and ecological interactions. Body size is a key species trait linked to metabolism, life-history and dispersal as well as a major determinant of ecological networks. Here, using a well-replicated urban-rural sampling design in Central Europe, we investigate the direction of change of body size in response to urbanization in three common bumblebee species, Bombus lapidarius, Bombus pascuorum and Bombus terrestris, and potential knock-on effects on pollination service provision. We found foragers of B. terrestris to be larger in cities and the body size of all species to be positively correlated with road density (albeit at different, species-specific scales); these are expected consequences of habitat fragmentation resulting from urbanization. High ambient temperature at sampling was associated with both a small body size and an increase in variation of body size in all three species. At the community level, the community-weighted mean body size and its variation increased with urbanization. Urbanization had an indirect positive effect on pollination services through its effects not only on flower visitation rate but also on community-weighted mean body size and its variation. We discuss the eco-evolutionary implications of the effect of urbanization on body size, and the relevance of these findings for the key ecosystem service of pollination.

Keywords: Bombus spp.; fragmentation; intertegular distance; land use; road density; temperature.

Figure 1

(a) Mean body size (± 95% CI) per species across all rural (black) and all urban (red) sites (n.s., not significant; ***p < .001). Relationships between Bombus body size and (b) ambient temperature when sampling, (c) road density at the 250 m scale and (d) road density at the 1,000 m scale, both as length of roads within a site for that scale. Plotted lines show predicted relationships

Figure 2

(a) Mean body size variation (± 95% CI) per Bombus species across all rural (black) and all urban (red) sites (n.s., not significant). (b) Relationship between Bombus body size coefficient of variation (%) and ambient temperature when sampling. Plotted lines show predicted relationships

Figure 3

Representation of the structural equation model of urbanization and ambient temperature when sampling, their relationships with bumblebee community‐weighted mean (CWM) body size and CWM of the coefficient of variation (CV) of body size, and the effects of visitation rates and body size on pollination. T. Urban, urban treatment of this categorical variable. Black solid arrows show positive and red arrows negative effects, as derived from the piecewise SEM analysis. Standardized path coefficients are reported next to the bold arrows and R ² values (conditional R ² _c and marginal R ² _m) are reported for all response variables. ns not significant; *p ≤ .05; **p ≤ .01; ***p ≤ .001

Figure 4

Relationships between Trifolium pratense seed set per plant and (a) visitation rates of all visitors per 30 min per plant; (b) CWM Bombus spp. body size (in mm) and (c) CWM CV of Bombus spp. body size (%). Plotted lines show predicted relationships