Associative learning without lasting memory: a recency bias in Dinoponera quadriceps

This study investigated associative learning, memory retention and individual learning profiles in the solitary foraging ant, Dinoponera quadriceps , using a visual guided bifurcated maze. We hypothesized that workers could retain a learned route for a period considered long-term memory. Eighty-two workers from ten colonies underwent four days of individual training to locate a food reward in one arm of a bifurcated maze, followed by an unrewarded retention test conducted at 24, 48, or 72 h after training. Thirty-eight individuals were classified as “learners” based on a consistent reduction in latency across training sessions, with performance significantly improving by the final session. Notably, learners were significantly faster than non-learners in their average latency to find the reward, indicating that the learner subset was not drawn from individuals with poorer overall foraging ability. Furthermore, a significant negative correlation was found between an individual’s daily foraging activity and its learning performance, suggesting an activity-learning trade-off in which more active foragers were slower to master the maze. Critically, performance across all retention tests (24 h to 72 h) was static and statistically equivalent to the initial, naïve performance in the first training trial. Our results indicate that while D. quadriceps workers can acquire a visually guided route, the behavioral expression of the learned information is not maintained in delayed unrewarded tests. We suggest that this pattern may reflect rapid forgetting, extinction-like responding, retrieval failure, or interference from post-training experience. Instead, the observed retrieval deficit at intermediate intervals, consistent with non-monotonic retention, may constitute an ecological adaptation that prioritizes up-to-the-minute information in their highly dynamic, ground-dwelling forest habitat. Together, our results highlight individual variation in learning performance and suggest that memory expression in solitary foraging ants may be strongly shaped by ecological context and individual foraging profiles.