Chickadees are selfish group members when it comes to food caching

Abstract

Many food-caching animals live in groups and cache pilferage may be one of the negative consequences of social living. Several hypotheses have been proposed to suggest that individuals may benefit from caching even when cache pilferage is high if all individuals can cache and pilfer equally. Stable groups may hypothetically support the evolution of such "reciprocal pilfering" because all group members may potentially have numerous opportunities to pilfer each other's caches. If that were the case, then we would expect animals cache openly in front of their group members, but to avoid caching in direct view of unknown conspecifics. We tested this hypothesis by allowing mountain chickadees (Poecile gambeli) to cache food in three experimental conditions: (1) with a familiar observer from the same group and with an unfamiliar conspecific observer present; (2) with a familiar observer from the same group only, and (3) without any observers. When presented with both a familiar and an unfamiliar observer, the caching chickadees treated both observers equally by choosing caching sites that were both farther away and out of sight of both observers. When only the familiar observer was present, chickadees shifted their choice of caching sites to the surfaces both away from and out of sight of the observer. When no observers were present, all available caching sites were used equally. Our results thus do not support the reciprocal cache sharing hypothesis and suggest that chickadees try to minimize cache pilferage from both familiar group members and unfamiliar conspecifics.

Figure 1

Experimental room design. Numbers indicate all available caching surfaces. All numbering always started from the cage with the familiar observer, whose location was alternated between trials.

Figure 2

Mean ± SE sum of caches made by mountain chickadees during 6 trials for each of the three experimental treatments on all 8 available caching surfaces (top panel, A, C, E) and on caching surfaces collapsed into 4 categories (1 + 2, 3 + 4, 5 + 6, 7 + 8) based on their orientation either towards or away from the observer (bottom panel, B, D, F). A and B – two observers present, familiar observer and an unfamiliar observer; C and D – only one known observer present; E and F – no observers present. Surface count always starts from the surface closest to the familiar observer.

Figure 3

Mean number of caches made by mountain chickadees during 6 trials for each of the three experimental treatments (two observers, one observer, no observers).