Environmental variation can maintain genetic variation in sexually selected traits if it affects the strength of directional selection. Specifically, environmental variation in sex-specific mortality will change the operational sex ratio (OSR), which predicts the intensity of mating competition. How the OSR affects selection for specific male traits is poorly understood; and it is unknown how often sexual selection is affected by interactions between the OSR and environmental factors that alter social variables such as mate encounter rates. Here, we experimentally manipulated the OSR and habitat complexity and quantified sexual selection on male mosquitofish (Gambusia holbrooki). In G. holbrooki there is high within-population variation in male size, which may exist because of a tradeoff between the ability to sneak copulate (favouring small males) and monopolize females (favouring large males). The success of each tactic is predicted to depend on the OSR, encounter rates and the ability to stealthily approach conspecifics. We show that, despite greater sharing of paternity under a male-biased OSR, neither the opportunity for selection, nor selection on male traits was affected by the OSR or habitat complexity. Instead, sexual selection consistently favored smaller males with high genetic heterozygosity (using >3000 SNP markers), and a relatively long gonopodium (intromittent organ).

16Although there are many correlational studies, unbiased estimates of inbreeding depression 17 only come from experimental studies that create inbred and outbred individuals. Few such 18 studies determine the extent to which inbreeding depression in males is due to natural or 19 sexual selection. Importantly, traits that are closely related to fitness are predicted to be 20 most strongly affected by inbreeding depression, so measuring fitness or key fitness 21 components, rather than phenotypic traits, is necessary to estimate inbreeding depression 22 accurately. Here, we experimentally created inbred and outbred male mosquitofish 23 (Gambusia holbrooki) by mating full-sibs (f=0.25). We show this led to a 23% reduction in 24 genome-wide heterozygosity. Males were then raised on different diets early in life. We 25 then allowed adult males to compete freely for females to test if inbreeding, early diet, and 26 their interaction affect a male's share of paternity. Early diet had no effect on paternity, but 27 outbred males sired almost twice as many offspring as inbred males. We also found that 28 males with a relatively long gonopodium (intromittent organ) had greater reproductive 29 success. We demonstrate that inbreeding has important consequences because it negatively 30 affects a key component of male fitness. Given there was no difference in adult mortality 31 this finding can only be due to inbreeding negatively affecting sexually selected traits. 32 33