Sexual selection may support phenotypic plasticity in male coloration of an African cichlid fish.

The expression of sexually selected traits, such as ornaments or body coloration, is often influenced by environmental conditions. While such phenotypic plasticity is often thought to precede evolutionary change, plasticity itself can also be a target of selection. However, the selective forces supporting the evolution and persistence of plasticity in sexual traits are often unclear. Using the cichlid fish Astatotilapia burtoni, we show that variation in the level of mate competition may promote plasticity in body coloration. In this species, males can change between yellow and blue colour. We found that experimentally increased competition over mating territories led to a higher proportion of males expressing the yellow phenotype. The expression of yellow coloration was found to be beneficial because yellow males won more staged dyadic contests and exhibited a lower level of oxidative stress than blue males. However, females were more likely to spawn with blue males in mate choice experiments, suggesting that expression of blue coloration is sexually more attractive. The ability to adjust colour phenotype according to the local competitive environment could therefore promote the persistence of plasticity in coloration.

Social instability influences rank-specific patterns of oxidative stress in a cichlid fish.

In many animal societies, dominant individuals have priority access to resources. However, defending high rank can be costly, especially in unstable social hierarchies where there is more intense competition. Oxidative stress has been proposed as a potential cost of social dominance, but few studies have examined this cost in relation to social stability. We studied the cost of social dominance in the cichlid fish Astatotilapia burtoni by manipulating social stability among males in replicate naturalistic communities for 22 weeks. We found that our social stability treatment influenced status-specific patterns in 3 out of 6 measurements of oxidative stress. Specifically, dominant males experienced increased plasma oxidative damage (measured as reactive oxygen metabolites, ROMs) compared with subordinate males in stable hierarchies only. Subordinate males in unstable hierarchies had higher ROMs than their stable community counterparts, but we found no effect of social stability treatment for dominant males. However, dominant males tended to have reduced total antioxidant capacity (TAC) in the liver when compared with subordinate males in unstable hierarchies, suggesting that the cost of social dominance is higher in unstable hierarchies. There were no effects of status and treatment on gonad TAC, muscle TAC or oxidative DNA damage. We conclude that the stability of the social environment influences the relative cost of social dominance in a tissue- and marker-specific manner.