Bumblebees flexibly adjust learning and decision strategies to sensory information content in a foraging task
Johannes Spaethe,Selma Hutzenthaler,Alexander Dietz,Karl Gehring,James Foster,Anna Stoeckl
DOI: https://doi.org/10.1101/2024.12.13.628313
2024-12-17
Abstract:When animals and humans make decisions based on their sensory input, they often rely on multiple cues at once. These provide complementary sources of information, which can help overcome ambiguity and noise, and increase the accuracy of decisions. While most studies have focused on the benefits of multimodal and within-modal integration for learning and decision making, their costs have received less attention. Processing multiple inputs, and learning and memorizing them requires more neural resources than single cues, and might also require more time. When a decision can be made with high accuracy based on a single cue alone, adhering and memorizing multiple ones simultaneously might thus concur higher costs than benefits, and not be an optimal decision-making strategy. In this study, we tested if insects trade-off the costs and benefits of learning multiple cues in a foraging task, using the buff-tailed bumblebee Bombus terrestris. To maximize comparability between cues, we presented combinations of visual-only features, such as are found on natural flowers: colours of varying discriminability, paired with shapes or patterns. We found that the bees relied exclusively on colours when these were easy to discriminate, and did not learn pattern or shape features presented simultaneously. With harder to discriminate colours, the bees learned both colour and shape or pattern features. Our results demonstrate that bumblebees flexibly adjust their learning strategies when presented with visual features of varying discriminability, to balance the costs and benefits of multi-cue learning. Our analysis of the learning rates of the bees with multi- and single attribute stimuli suggests that blocking could serve as a mechanism to implement this strategy switch. These results shed light on trade-offs in learning and decision making with multiple cues, and can directly be compared to studies in other insects, animals and humans.
Biology