Interlocus sexual conflict (IRSC) occurs because of shared interactions that have opposite effects on male and female fitness. Typically, it is assumed that loci involved in IRSC have sex-limited expression and are thus not directly affected by selective pressures acting on the other sex. However, if loci involved in IRSC have pleiotropic effects in the other sex, intersexual selection can shape the evolutionary dynamics of conflict escalation and resolution, as well as the evolution of reproductive traits linked to IRSC loci, and vice versa. Here we used an artificial selection approach in Japanese quail (Coturnix japonica) to test if female-limited selection on reproductive investment affects the amount of harm caused by males during mating. We found that males originating from lines selected for high female reproductive investment caused more oxidative damage in the female reproductive tract than males originating from lines selected for low female reproductive investment. This male-induced damage was specific to the oviduct and not found in other female tissues, suggesting that it was ejaculate-mediated. Our results suggest that intersexual selection shapes the evolution of IRSC and that male-induced harm may contribute to the maintenance of variation in female reproductive investment.

Dataset collected from captive Japanese quail (Coturnix japonica) artificially selected for high (H-line) and low (L-line) maternal egg investment. After four generations of selective breeding, H- and L-line females differed in egg size by > 1 SD, but they did not differ in the number of eggs laid. Both L-line and H-line females were randomly paired with either a L-line or H-line male (2x2 design; female / male pairs: N = 11 L / L, N = 10 L / H, N = 10 H / H, N = 10 H / L) and kept in breeding cages for two weeks.

From each female, a blood sample was obtained before breeding, and tissue samples from the oviduct, the liver and the spleen were obtained after breeding. Oxidative stress markers were analysed in the samples.

There are two datasets. The main information about the line, age, reproductive performance and oxidative stress markers is included in the first dataset. The information included in the second dataset was used only to build the correlation matrix (supplementary table S7), in other words, this dataset does not include new information non-included in the first dataset.

The explanation of the variables is included as a comment on the variable name. The R code for running the models is also available.