Shifting expectations? Bumblebees show limited adjustment based on recent reward history
摘要
Models of decision-making often assume that choices reflect objective payoffs, yet humans and other animals frequently evaluate options relative to prior experiences or available alternatives. This can lead to incentive contrast effects, where the perceived value of a reward is shaped by recent exposure to higher- or lower-quality rewards. Consequently, the same option may be judged unfavorably after better rewards but favorably following poorer ones. Relative evaluation may be especially important for generalist foragers such as bumblebees, which visit thousands of flowers each day that differ in nectar quality and quantity. Although bumblebees exhibit incentive contrast effects when assessing nectar concentration, it remains unclear how they evaluate other nectar attributes. Here, we investigated whether bees use reference-based evaluation to assess nectar reliability. We trained one group of bees on highly rewarding flowers and then reduced nectar availability to measure behavioral responses. When compared to a group that received lower reward rates throughout, these downshifted bees showed minimal disruption in preferences for previously trained flowers in response to reductions in reward reliability. We argue that bees may receive fundamentally different information from changes in nectar concentration and reliability. Given that nectar rewards vary within and across plant species, our findings suggest that bees may be slower to abandon flowers that decline in nectar reliability relative to other aspects of nectar quality. More broadly, when animals must repeatedly sample a resource to estimate its quality or quantity, they may perceive changes in ways that reduce the likelihood of reference-based evaluation.
Significance statementUnderstanding how animals evaluate changing resources is essential for predicting their foraging decisions in natural environments. Bumblebees routinely encounter flowers that vary in how reliably they provide rewards. Here, we find that, unlike their strong contrast responses to shifts in nectar concentration, bees respond weakly when the reliability of nectar delivery declines across foraging bouts. Bees previously exposed to highly reliable flowers did not sharply reduce their preference for those flowers after rewards became less dependable, behaving similarly to bees that did not experience a change in reward reliability. These findings suggest that bees weigh changes in nectar reliability differently from other nectar traits, potentially slowing or preventing them from abandoning these resources. More broadly, our findings highlight that animals may show limited reference-based evaluation when resource attributes must be assessed through repeated sampling.
Graphical Abstract