How long to stay on a plant: the response of bumblebees to encountered nectar levels

This field study shows that the number of flowers visited per bee per plant (Anchusa officinalis) increases with the instantaneous nectar level at the plant. Observations during the season showed that a bee visits more flowers per plant of given nectar level, the lower the overall mean nectar level in the study area. These results agree with predictions from a model based on the 'marginal value theorem', but with assumptions and constraints adapted for nectar-foraging bees. It suggests that bumblebees assess the nectar level at a plant by sampling one or a few flowers, which is possible because within-plant nectar volumes are correlated. The bees compare encountered gains to an optimal plant switching threshold equal to the overall mean nectar level and leave an unrewarding plant as soon as possible, but continue to visit the flowers on a rewarding plant. However, the bees leave before having visited all flowers due to a searching constraint. The bees' response to plant nectar levels results in systematic flower visitation, because visitation to recently depleted flowers is reduced, which reduces the variation of the inter-visit time per flower. Systematic flower visitation implies that the overall mean encountered gain per flower is higher than the overall mean standing crop, as predicted by a model of systematic foraging. However, the sampling and searching constraints on the bees' response to plant nectar levels increase the variation of the inter-visit time per flower, and thereby limit the degree of systematic flower visitation and the effect on the mean encountered gain.