How do insects pollinate flowers? Behaviour during visitation determines high pollination efficiency by a generalist insect

Abstract Ecosystem services provide essential benefits to human survival and well‐being. Animal behaviour is central to the delivery of many ecosystem services, yet we know little about how behaviour drives services' provision. Pollination is often mediated by insect foraging; however, despite extensive research, our understanding of how pollinators' behaviour during a single flower visit determines plant reproductive success remains limited. Using single visit, pollination efficiency experiments with marmalade hoverflies ( Episyrphus balteatus ) and strawberries ( Fragaria × ananassa ), we tested the hypothesis that longer time spent performing behaviours relevant to pollen transfer leads to higher pollination efficiency. To this end we compared fruit weight (g) and the number of fertilised seeds of virgin flowers following a single uninterrupted visit by one fly with that of hand‐pollinated flowers, a measure of maximal pollination efficiency and two negative controls, that is flowers that were not cross‐pollinated. A linear mixed effect model in a Bayesian framework with treatment group as fixed effect and plant identity as random effect revealed that flowers visited by hoverflies produced fruits that were 70–85% heavier and with 64–82% more fertilised seeds than pollinator‐excluded controls, and with as many fertilised seeds as hand‐pollinated flowers. Among all behaviours that flies performed on flowers, pollination efficiency was predicted by the durations of nectar and pollen feeding and stigma‐oriented grooming. The durations of these behaviours were stronger predictors of fruit weight and seed count than total visit duration. We conclude that marmalade hoverflies deliver highly efficient pollination in just a single visit through behaviours that promote stigma contact and pollen transfer, challenging the assumption that these generalist insects contribute significantly to pollination only when abundant. Ultimately, our study reveals how pollinators' behaviour during flower visitation drives service provision and explains why visits are not functionally equivalent. Ultimately, our study shows that specific behaviours during flower visitation determine pollination efficiency and explain why individual visits are not functionally equivalent. Because these behaviours can be altered by anthropogenic stressors, a behavioural perspective will improve predictions of pollination service provision across species and ecological contexts. Read the free Plain Language Summary for this article on the Journal blog.