Peck orders and group size in laying hens: `futures contracts' for non-aggression.

We analyse a simple model of the establishment and maintenance of dominance hierarchies in hens. To be beneficial, dominance relationships require that the probability of meeting the same individual repeatedly is high, otherwise costs of establishing the dominance relation are never recouped. Winners and losers benefit from dominance relationships, not necessarily from changing the rate at which they acquire resources but by avoiding costly contests over them in future encounters. We show that so-called loser effects', in which animals base their strategies for contesting resources solely upon their past experiences of winning or losing dominance fights and not upon who their opponent is, cannot work - these strategies (pragmatists') must additionally involve either individual or status category recognition. As alternatives to dominance relationships, we show that signals of status or fighting ability that determine access to contested resources are expected to evolve in species with typically large groups because in such conditions the costs of establishing dominance relations are not recouped. Such signals do not depend upon recognizing others individually, but rather upon general category recognition. Status signals are not expected in small groups because dominance relationships are likely to be cheaper and just as effective. The results of the model have implications for the welfare of hens kept in large groups.