RAPID DIVERGENT EVOLUTION OF SEXUAL MORPHOLOGY: COMPARATIVE TESTS OF ANTAGONISTIC COEVOLUTION AND TRADITIONAL FEMALE CHOICE

Male structures specialized to contact females during sexual interactions often diverge relatively rapidly over evolutionary time. Previous explanations for this pattern invoked sexual selection by female choice, but new ideas emphasize possible sexually antagonistic coevolution resulting from male-female conflict over control of fertilization. The two types of selection have often not been carefully distinguished. They do not theoretically exclude one another, but they have not necessarily had equally important roles in producing rapid evolutionary divergence. To date, most recent empirical studies of antagonistic coevolution have emphasized only a few taxa. This study uses the abundant but little-used data in the taxonomic literature on morphology to evaluate the roles of antagonistic coevolution and traditional female choice over a wide taxonomic spectrum (61 families of arthropods, mostly insects and spiders). Groups with species-specific male structures that contact females were checked for coevolution of species-specific female structures that are contacted by the male and that have mechanical properties that could potentially defend her against the male. Facultatively deployable, species-specific female defensive structures, a design that would seem likely to evolve frequently under the sexually antagonistic coevolution hypothesis, were completely absent (0% of 106 structures in 84 taxonomic groups). Although likely cases of sexually antagonistic coevolution exist, using conservative criteria, 79.2% of the 106 structures lacked even potentially defensive female coevolution. A common pattern (53.8% of 106) was a nearly complete absence of female change in areas contacted by species-specific male structures. Post-hoc arguments invoking possible coevolution of defensive female behavior instead of morphology, or of female sensitivities and responses to male sensory traps, could enable the sexually antagonistic coevolution hypothesis to explain these data. No case of such coevolution of female behavior or sensitivities has been demonstrated, and there are additional reasons to doubt that they are general explanations for the data presented here. Detailed studies of female resistance behavior could help illuminate several issues. The possibility of a greater role for antagonistic coevolution in reproductive physiology than in morphology and the possibility that female choice and sexually antagonistic coevolution have both been important in some lineages are discussed.