Leaf-damaging behavior by queens is widespread among bumblebee species

Abstract

Phenological mismatches and resource limitations resulting from ongoing environmental change can have severe impacts on pollinator fitness. Recent findings show that bumblebee workers respond to pollen scarcity by damaging plant leaves in ways that can accelerate flowering, suggesting a mechanism by which direct information transfer from bees to plants might influence the timing of flower production. However, the ecological and adaptive significance of this interaction remains uncertain. Here we report that mated and unmated queens of Bombus terrestris also damage leaves, with similar effects on flowering. Furthermore, we document leaf damage by wild-caught queens from 12 species, spanning seven subgenera, indicating damaging behavior is widespread among Bombus species. Leaf damage by bumblebee queens may have particular relevance in the context of colony founding and early development, where the timely availability of local floral resources can be critical for colony success and fitness.

Fig. 1. Typical leaf damage by unmated queens of Bombus terrestris.

a Typical example of leaf damage by unmated B. terrestris queens on B. nigra plants (scale bar = 10 mm). b Mechanical reproduction of queen damage using needles on B. nigra plants (scale bar = 10 mm).

Fig. 2. Effects of queen damage on time to flower of Brassica nigra and Solanum lycopersicum.

Reproductive development was monitored for a B. nigra and b S. lycopersicum plants assigned to two treatments: (i) unmated queen damage and (ii) mechanical damage (control). Following damage treatment, plants were monitored daily for macroscopic signs of flowering (indicated on the vertical axes, days post treatment). Damage by unmated B. terrestris queens accelerated the average vegetative-to-reproductive switch of B. nigra plants by 17 days relative to mechanically damaged controls (n = 8 biologically independent samples) and the flowering time of S. lycopersicum plants by 5 days (n = 7 biologically independent samples). Data is shown as boxplots, where the line represents the median and the box the interquartile range (IQR). A pairwise comparison (Exact Wilcoxon-Mann-Whitney Test) was used to determine statistical significance. The p-value is indicated above brackets.

Fig. 3. Distribution of queen damaging behavior across Bombus subgenera and species.

Simplified phylogenetic tree of Bombus subgenera, adapted from Williams et al. . Subgenera are indicated in bold text, with collected species listed below (gray text indicates subgenera not found in Switzerland). For each species, yellow bars represent numbers of damaging individuals and gray bars represent individuals that were not observed to damage. Information about host-plant specialization/ dietary breadth (as indicated by tongue length) and nesting ecology for each subgenus are presented in brackets (social parasitism is highlighted in red). Bombus sp. = identification only to subgenus level. B. terrestris queens came from a commercial colony and were artificially hibernated.

Fig. 4. Variation of leaf-damage shape in different Bombus species.

Representative scans of leaves damaged by spring queens (mated and hibernated) of each species (scale-bar = 3 mm). Plants used for this experiment were Brassica nigra and Solanum lycopersicum.