Abstract
Abstract
Memory allows animals to utilize ecologically relevant information, yet whether it endures through prolonged metabolic shutdown is poorly understood. We tested spatial memory retention in a long-term, free-living hibernator, the Columbian ground squirrel (Urocitellus columbianus), using a field-based spatial maze over two consecutive active seasons. Ten individuals were tested before and after a 9-month hibernation period and compared with eight naive individuals. Maze-solving performances—time to exit, path sinuosity and navigational errors—were analysed over 10 consecutive trials using Bayesian nonlinear learning curves to separate retained memory from relearning. Strikingly, post-hibernation individuals with prior exposure to the task significantly outperformed naive individuals by 41–98% across all metrics, with initial performance after hibernation approaching pre-hibernation asymptotes. We show that spatial memory survives prolonged hibernation in the wild, supporting the hypothesis that neural plasticity during hibernation reflects adaptive retention rather than memory loss.