Developing predictive models of behaviour: Do rockwallabies receive an antipredator benefit from aggregation?

Many species receive an antipredator benefit from aggregation such that animals in larger groups are able to allocate more time to foraging and less to antipredator vigilance. These beneficial ?group size effects? must be traded-off against the costs of increased competition for limited resources that may result from aggregation. Certain species, or species living in certain habitats, may be predisposed to receive greater benefits from aggregation than others. Based on the results of a study of captive yellow-footed rock-wallabies (Petrogale xanthopus; a ?Vulnerable? macropodid marsupial), we predicted that because rock-wallabies must defend vital resources (the locations where they shelter by day), there are costs which reduce the overall antipredator benefits obtained from aggregation while foraging. We tested this prediction by observing three different species of free-living rock-wallabies as they foraged in aggregations of different sizes. Allied (P. assimilis) and unadorned (P. inornata) rock-wallabies received no obvious antipredator benefits from aggregation since there was no effect of group size on time spent vigilant by individuals. Mareeba rock-wallabies (P. mareeba) may receive antipredator benefits, since animals tended to forage more and looked less as group size increased. However, this result was influenced by two observations that had substantial leverage. Additionally, even if present, the specific shape of this group-size function suggests that intraspecific competition in P. mareeba increases with group size. As a clade, rockwallabies appear to have costs which reduce or eliminate antipredator benefits associated with aggregation. Conservation efforts to recover populations should consider the likely importance of intraspecific competition for these species, and generalizations about introducing or translocating social animals socially should rest upon their being demonstrable benefits from aggregation.