The maintenance of genetic polymorphism in sexually antagonistic traits

Abstract

AbstractSelection often favours different trait values in males and females, leading to genetic conflicts between the sexes, or sexual antagonism. Theory suggests that such conflict can maintain genetic variation by generating balancing selection. However, most of this theory is based on insights from models of single loci with fixed fitness effects. It is thus unclear how readily sexual antagonism drives balancing selection at loci where the fitness effects are not arbitrarily set but emerge from the genotype-phenotype map and the fitness landscape. Here we use mathematical models and computer simulations to investigate the evolution of traits that have different optima in males and females and so experience sexually antagonistic selection. We examine the consequences of such selection for the maintenance and nature of genetic variation under different assumptions of genetic architecture and for different fitness landscapes. We show that the conditions necessary to maintain variation are significantly more restrictive than previously appreciated. This is because the majority of typical sex-specific fitness landscapes generate stabilising selection on traits, and so favour the fixation of a single homozygote genotype that encodes the phenotype with the highest sex-averaged fitness. Instead, the maintenance of genetic polymorphism requires extremely, if not implausibly, strong conflict between males and females in order to disfavour extreme and intermediate phenotypes by generating diversifying selection. Furthermore, in all cases, we find that distinctive patterns of genetic variation produced by sexual antagonism arise rarely and are often transient, and so are difficult to observe in genomic data. Taken together, our results indicate that sex-specific selection is not a straightforward source of balanced genetic polymorphism and leaves few traces in the genome. Rather, we suggest that sexual antagonism is most likely to maintain genetic variation when traits are determined by a single large effect locus where genetic constraints lead to a limited range of possible alleles.