Five hypotheses of cache recovery behaviour in Clark's nutcracker (Nucifraga columbiana) were examined experimentally. Most caches were made in soil within 5 cm of conspicuous large objects. Both seed-caching and non-seed-caching nutcrackers were able to locate caches. Seed-caching nutcrackers relocated caches using large objects as remembered visual cues. Soil microtopography and small (<2 cm diameter) objects may be used as cues to facilitate cache recovery but are not essential. Non-seed-caching nutcrackers located caches by using soil disturbances at cache sites as visual cues and by searching preferentially near objects where caches were concentrated. Success rates of seed-caching nutcrackers ranged from 52 to 78% and those of [...]
