The influence of spatial memory on caching behaviour and reproductive success in the North Island robin (Petroica longipes)

Many bird species cache food items for later retrieval, a behaviour requiring spatial memory to accomplish. Several studies have now linked individual performance on a spatial memory task with reproductive success in wild food-storing birds, providing evidence that selection acts on spatial memory. However, it remains uncertain whether caching is the specific target of this selection, and whether individual differences in spatial memory ability are also reflected in caching behaviour. Furthermore, it is also unclear whether a single task can accurately quantify the spatial cognitive abilities used by birds during caching. In this thesis, I investigated the interface of spatial memory, caching behaviour, and reproductive success in a wild food-storing bird, the North Island robin (Petroica longipes), henceforth referred to by their Māori name, toutouwai. Using an established study population, I first quantified the birds’ caching behaviour, finding that individuals consistently differed in their storing along a ‘caching syndrome’. ‘Clump-caching’ birds rapidly created clusters of nearby cache sites, while ‘scatter-caching’ birds spent longer selecting a greater number of distant sites when storing food. I then ran a three-task spatial memory test battery on the same birds, investigating individual variation in 1) preference for spatial cues or local visual cues, 2) the ability to learn to discriminate a single rewarded position in an array, and 3) spatial working memory for recently depleted positions in an array. No convergent validity was found between any of the tasks, suggesting that spatial memory in toutouwai may consist of several modules rather than a single general ability. Finally, I compared individuals’ spatial memory task performance and individual caching behaviour with each other, and with three proxies for fitness: breeding season start date, number of independent offspring, and survival over the breeding season. Caching behaviour covaried with spatial discrimination learning, with ‘clump-caching’ males showing significantly worse spatial memory that ‘scatter-cachers’. However, there was no effect of caching or spatial memory on any fitness proxy. Overall, my findings represent the first field study to provide detailed spatiotemporal caching measures, multiple spatial memory measures, and fitness proxies have been integrated in a wild food-storing bird. The results suggest that individual differences in the spatiotemporal aspects of caching may not directly influence fitness in toutouwai, but do not rule out the influence of other caching measures, or other behaviours involving spatial memory.