The evolution of sexually selected traits in male swordtail fishes (Xiphophorus: Poeciliidae)

Abstract

Sexual rather than natural selection can account for the evolution of exaggerated male ornaments that appear to be detrimental to survival. Such a sexually selected trait is the sword-like caudal fin extension in males of species that belong to the genus Xiphophorus. Swordtail fishes feature prominently in empirical work and theories of sexual selection. Recently, the pre-existing bias hypothesis was offered as an explanation for the evolution of swords in these fishes. Under this hypothesis, the female preference was postulated to predate the origin of the male's sword and to have driven its evolution directly. A reconstruction of the evolution of this trait, using a traditional, largely morphology-based phylogeny of this genus, and recent behavioural experiments suggest that female preference for mating with sworded males arose in ancestrally swordless species. The applicability of the pre-existing bias hypothesis is questioned, however, by a molecular phylogeny of Xiphophorus, which differs from the traditional one. It demonstrates that some features (e.g. possibly coloration and black ventral margin) of the character complex ‘sword’ are likely to have been present in the common ancestor of all species of Xiphophorus. The propensity to develop ventral caudal fin extensions is intra-specifically variable in several Xiphophorus species; this variation may be phylogenetically widespread and plesiomorphic for poeciliid fishes. It is shown that many male secondary sexual characters are evolutionarily labile. They are probably more easily lost and regained than are female preferences for them, which might be phylogenetically more resilient to change. Differences in the rates of evolution of female preference and male traits might be caused by stronger natural selection pressures on males than on females.