Individual variation in spatial memory performance in wild mountain chickadees from different elevations

For food-caching animals inhabiting environments with strong seasonal variation, harsh winter conditions may limit access to naturally available food and favour the evolution of enhanced spatial memory. Spatial memory enables animals to remember the locations of food caches for overwinter survival, therefore animals in harsher conditions may benefit more from more accurate spatial memory than conspecifics living under relatively mild conditions. Despite numerous laboratory studies lending indirect support to the hypothesis that a harsh environment favours the evolution of spatial memory, there is no direct evidence showing fitness consequences of variation in spatial memory. As a step towards evaluating this hypothesis in natural populations, we established spatial arrays of programmable bird feeders equipped with radio frequency identification technology (RFID) to test for individual variation in spatial memory in food-caching mountain chickadees, Poecile gambeli, at two elevations differing in winter climate severity. Individuals could only access food at a single rewarding feeder within an array of eight, and each individual had to learn the location of their unique rewarding feeder. Birds at higher elevations visited the arrays significantly more frequently than birds at lower elevations, suggesting more intense food caching. They also performed better at locating their rewarding feeder than birds from lower elevations. Individuals showing better performance participated in more overall trials, suggesting a link between food-caching propensity and memory performance, but higher overall levels of activity within each array yielded less accurate memory performance. Finally, rotating the arrays showed that birds relied specifically on spatial memory in order to locate their rewarding feeder. To our knowledge, this is the first explicit test of spatial memory performance in food-caching birds under natural conditions. Our results corroborate earlier laboratory-based work showing large individual variation in spatial memory performance and lay the groundwork for future investigation into the fitness consequences of individual variation and the evolution of spatial memory.