The Developmental Instability—Sexual Selection Hypothesis: A General Evaluation and Case Study

Abstract

Developmental instability results from small, random perturbations to developmental processes of individual traits. Phenotypic outcomes of developmental instability include fluctuating asymmetry (FA, subtle deviations from perfect bilateral symmetry) and phenodeviance (minor morphological abnormalities). A great deal of research over the past 18 years has focused on the role of developmental instability in sexual selection. A driving force behind this research has been the developmental instability-sexual selection hypothesis, which posits that symmetry and lack of phenodeviance in secondary sexual traits are assessed by mates and rivals because they provide a reliable cue of individual genetic quality. The present article tests this hypothesis by evaluating its five main predictions using published results: expressions of developmental instability in secondary sexual traits should be (1) negatively correlated with mating success; (2) directly assessed by mates and sexual rivals; (3) heritable; (4) condition-dependent; and (5) negatively correlated with ornament size. The first two predictions receive considerable, though not ubiquitous, support from a range of animal species. However, FA in secondary sexual traits is generally not significantly heritable, indicating that FA is unlikely to reveal genetic quality that can be transmitted to offspring. Similarly, there is little evidence to support the predictions that FA is condition dependent, and that it is negatively phenotypically or genetically correlated with sexual trait size. Based on an evaluation of the evidence overall, it is concluded that this hypothesis is unlikely to be viable; it appears unlikely that mate choice for symmetry evolves by “good genes” sexual selection. Hypotheses that do not require asymmetry and phenodeviance to reveal heritable genetic quality may explain observed links between FA/phenodeviance and mating success. Results of a case study of Drosophila bipectinata are summarized, which reinforce this general conclusion. It is suggested that nonadditive genetic variation arising from an interaction between trait-specific developmental genes and genetic background may drive sexual selection for reducing developmental instability in some cases. Levels of developmental instability variation in a population may need to surpass a critical threshold for sexual selection to operate, possibly explaining some of the pronounced heterogeneity in the effect of developmental instability on sexual selection reported in the literature.