Honeybees solve a multi-comparison ranking task by probability matching

Abstract

Honeybees forage on diverse flowers which vary in the amount and type of rewards they offer, and bees are challenged with maximizing the resources they gather for their colony. That bees are effective foragers is clear, but how bees solve this type of complex multi-choice task is unknown. Here, we set bees a five-comparison choice task in which five colours differed in their probability of offering reward and punishment. The colours were ranked such that high ranked colours were more likely to offer reward, and the ranking was unambiguous. Bees' choices in unrewarded tests matched their individual experiences of reward and punishment of each colour, indicating bees solved this test not by comparing or ranking colours but by basing their colour choices on their history of reinforcement for each colour. Computational modelling suggests a structure like the honeybee mushroom body with reinforcement-related plasticity at both input and output can be sufficient for this cognitive strategy. We discuss how probability matching enables effective choices to be made without a need to compare any stimuli directly, and the use and limitations of this simple cognitive strategy for foraging animals.

A Markovian dynamics forC. elegansbehavior across scales

“…Interestingly, such a strategy, termed probability matching [71], or Thompson sampling [103], is a well-studied heuristic solution for the multi-armed bandit problem, a game in which different actions have variable rewards that are a priori unknown to the player, whose goal is to maximize total pay-out. Evidence for probability matching in decision making tasks has been previously demonstrated in experiments in animals and humans [104, 105], and is an active area of research in cognitive science of decision making [106].…”

How honey bees make fast and accurate decisions

“…To improve the accuracy of the model in acceptance responses we added learning cells ( L 1 and L 2) to the model (Figure 6C) that receive input from the sensory cells on the identity of the colours and send different inhibitory outputs to the accumulator cells (Figure 6C). Following a model approach by (MaBouDi et al, 2020b) L 1 is activated when the low rewarded colours were presented to the model. L 2 is activated by the high rewarded colour.…”

“…The fight arena is fully described in MaBouDi et al (2020). The method of attracting bees into the arena is described in (MaBouDi et al, 2020b).…”

Switching costs in stochastic environments drive the emergence of matching behaviour in animal decision-making through the promotion of reward learning strategies

“…Additionally, foraging bees have evolved to use only a subset of decision-making strategies that are most adaptive to environmental stochasticity, as this allows bees to track variation in both the quality and availability of food sources 40 . Thus, PM may represent an ecologically optimal foraging solution for animals such as budgerigars if reward learning probabilities are highly variable 43 , whilst also evolving in less competitive environments for the birds as a direct result of near-optimal reward learning 39 .…”