Learning More than One Resource: An Analysis of Behavioral Flexibility in Bumble Bees

Abstract

Organisms face various challenges in changing environments, and learning to deal with these challenges can increase their fitness. However, learning one task or resource can impose cognitive constraints on learning another. Different memories can interact to inhibit or facilitate recalling of one another. Also, learning can affect further learning opportunities by changing the environment that organisms experience. However, the factors determining the positive or negative direction of interactions between multiple learned memories and consequently affecting the performance of organisms are still poorly understood. This dissertation addresses how generalist bumble bees (Bombus impatiens) learn diverse tasks and floral traits during foraging in multi-floral contexts, focusing on cognitive constraints and decision-making. Specifically, I asked three questions: 1) How does the similarity between learned “flower color-handling tactic” associations affect bees’ foraging performance? Is it enhancing (transfer) or hindering (interference) performance? 2) How do transfer and interference from learned associations influence bees' decisions regarding which resources to select? 3) How does previously learned behavior restrict the floral stimuli that bees experience and influence further learning? Laboratory experiments using artificial flowers with varying traits revealed that when bees experienced two types of flowers sequentially, flower color similarity and handling tactic similarity interacted to determine transfer and interference between learned associations. When handling tactics were similar, bees transferred learned tactics more easily if flower colors were similar, enhancing their foraging performance. However, when handling tactics differed, similar flower colors led to incorrect tactic transfer, causing bees to make more errors. In an experiment with two flower types presented simultaneously to bees, similar interaction patterns were observed. Despite the high error rates, bees switched frequently between two flower types, which contradicted the hypothesis that bees would specialize on one flower type to reduce costs associated with switching. In a separate experiment, learning how to handle a flower also affected bees’ subsequent learning of flower traits associated with rewards, potentially by restricting access to floral stimuli. Taken together, my studies have demonstrated several challenges bees face in learning to handle multiple flower types and have provided insights into the cognitive mechanisms of organisms faced with complex and changing environments.