THE EVOLUTION OF SELECTIVE AGGRESSION CONDITIONED ON ALLORECOGNITION SPECIFICITY

Abstract

Many sessile cnidarians deploy specialized structures while competing aggressively for living space. The initiation of aggression is often contingent on the relatedness of the interacting contestants; clonemates and close kin generally behave passively toward one another, whereas more distant relatives generally behave aggressively. Behavioral specificity of this sort requires that there be 1) an allorecognition system that can discriminate among subtle differences in cell-surface determinants and 2) a highly polymorphic genetic system that provides specific labels of relatedness (haplotypes or allotypes). The evoutionary models analyzed in this paper show that a population of individuals that behave aggressively only against haplotypically distinct individuals (discriminating phenotypes) will not be evolutionarily stable in the face of either unconditionally aggressive or unconditionally nonaggressive phenotypes. Furthermore, even if the discriminating trait were somehow fixed, the rare recognition alleles necessary to confer allotypic specificity could not become established through natural selection. Thus, allotypic specificity is unlikely to be maintained by individual selection acting directly through aggressive behavior. Other selective mechanisms might account for the evolution of allorecognition specificity. Allotypic polymorphism could be maintained by pleiotropic mechanisms in which rare alleles are favored by natural selection acting either on gametic incompatibility, pathogen resistance, or somatic fusion, rather than aggressive behavior per se. However, these mechanisms do not explain the maintenance of selective aggression based on allotypic differences. Alternatively, if aggressive members of a clone indirectly enhance the reproductive output or survival of the entire clone (or close relatives), then kin selection acting directly through aggressive behavior could favor allorecognition specificity. Choosing among these alternatives will require the development of more sophisticated theory and empirical analyses of the costs and benefits of aggression.