PS 47-122 - The rich get richer: North American red squirrels with larger food caches acquire more resources, but timing matters

Background/Question/Methods

Despite increased energetic resources generally having a positive effect on survival and reproductive success, there remains substantial variation in resource accumulation by individuals. Food caching is an energetically costly activity, such that variation in success is thought to result from individual and environmental constraints that influence the energy and time budgets that can be allocated to caching. We hypothesized that animals adjust caching effort relative to the resources they defend presently. North American red squirrels (Tamiasciurus hudsonicus) in the southwest Yukon, Canada harvest and cache conifer cones within individually-defended territories from late summer through early autumn. We investigated whether existing cache size corresponds to caching effort using 12 years of annual caching records indexed at the level of the individual. We disentangled whether this relationship is attributable to individual differences through an overwinter food supplementation experiment that altered cache size prior to the caching season. Finally, we investigated whether knowledge of cache size is regularly assessed in time, or acquired through long-term processes by manipulating cache size (cone addition n = 12, removal n = 12) during caching season. Using high-resolution biologging data, we investigated whether changes in cache size elicited changes in behavioural time budgets within individuals.

Results/Conclusions

Existing cache size of old conifer cones (distinguished as being brown in colour) positively correlated with newly harvested cones (bright green and purple in colour) in unmanipulated caches. Food-supplemented squirrels maintained larger caches of cones cached in previous seasons than control individuals, presumably because they diverted their consumption to supplemental peanut butter from cone caches over winter. In turn, these larger caches were associated with more newly harvested cones later in the year after caching had ceased. Cache sizes of squirrels that received additional cones in the cache manipulation experiment remained larger than cache sizes of squirrels that had cones removed, indicating that squirrels that had cones removed did not increase caching effort to compensate for the loss. Biologging data revealed that prior to midden manipulations in which cones were added or removed, activity and cache size tended to be positively correlated, although this was not significant. Mean activity decreased post-manipulation in both treatment groups while mean time spent in the nest increased, indicating neither treatment elicited a facilitated or compensatory behavioural response to shifts in existing cache size. This study serves to disentangle energetic influences on resource acquisition and provides insight into annual timing of decisions around energy balances.