The evolution of alarm calling: a cost-benefit analysis

Abstract Three arguments for the likelihood of alarm calling in colonial animals are presented. The first, a game-theoretic portrait, predicts that alarm calling should be less probable per animal in larger groups and also that actual fitness benefits and costs are not quantitatively important in determining the evolutionarily stable probability of calling in large populations. The second is a geometric model of the dependence of risk upon position within a colony of sedentary animals. The two models are integrated to form a third, which predicts that the probability that an animal gives an alarm call is independent of both group size and the ratio of benefits to costs, above some minimum group size. Some of the available data are consistent with the models, and some are not.