A comparison of concept learning ability in honeybees and bumblebees using matching-to-sample tasks

Reports of honeybees demonstrating abstract concepts like sameness and difference marked a pivotal development in comparative psychology. Subsequent studies expanded the scope of concept learning in honeybee cognition, yet most evidence relies on a single method: the delayed-matching-to-sample task using a Y-maze. Whether this setup is uniquely effective or if alternative approaches could yield similar results remains unresolved. Additionally, the failure of bumblebees (Bombus spp.) to complete this task, despite honeybees demonstrating success, remains unexplained. This study compared the performance of honeybees (Apis mellifera) and bumblebees (Bombus terrestris) across matching-to-sample tasks with varying degrees of physical continuity between sample and target stimuli. The objectives were twofold: to evaluate an alternative method for assessing concept learning in both species and to investigate potential species differences in such tasks. Contrary to prior findings, neither species succeeded at the reported proficiency levels in simultaneous matching-to-sample tasks. Moreover, bumblebees outperformed honeybees in one task. These results are consistent with an explanation based on species-specific differences in visual attention mechanisms, and underscore the need for further research on concept learning in social bees.