The effects of a bacterial challenge on reproductive success of fruit flies evolved under low or high sexual selection

Abstract

The capacity of individuals to cope with stress, for example, from pathogen exposure, might decrease with increasing levels of sexual selection, although it remains unclear which sex should be more sensitive. Here, we measured the ability of each sex to maintain high reproductive success following challenges with either heat‐killed bacteria or procedural control, across replicate populations of Drosophila melanogaster evolved under either high or low levels of sexual selection. Our experiment was run across four separate sampling blocks. We found an interaction between bacterial treatment, sexual selection treatment, and sampling block on female reproductive success. Specifically, and only in the fourth block, we observed that bacterial‐challenged females that had evolved under high sexual selection, exhibited lower reproductive success than bacterial‐challenged females that had evolved under low sexual selection. Furthermore, we could trace this block‐specific effect to a reduction in viscosity of the ovipositioning substrate in the fourth block, in which females laid around 50% more eggs than in previous blocks. In contrast, patterns of male reproductive success were consistent across blocks. Males that evolved under high sexual selection exhibited higher reproductive success than their low‐selection counterparts, regardless of whether they were subjected to a bacterial challenge or not. Our results are consistent with the prediction that heightened sexual selection will invoke male‐specific evolutionary increases in reproductive fitness. Furthermore, our findings suggest that females might pay fitness costs when exposed to high levels of sexual selection, but that these costs may lie cryptic, and only be revealed under certain environmental contexts.

Genomic pedigree reconstruction identifies predictors of mating and reproductive success in an invasive vertebrate

“…Finally, we used PLINK 1.9 to derive the genome‐wide estimate of inbreeding ( F hat3 , Yang et al, ) for each individual from the final filtered set of SNPs ( N = 654). After running the global models, we used the drop1{stats} function in R (R Core Team, ) to test the significance of fixed effects using likelihood ratio tests of the global model against a null model lacking the predictor of interest ( χ 2 , α = .05; see Nystrand, Cassidy, and Dowling () and Sales et al () for similar approaches). To avoid possible bias of effect sizes, we only report estimates and standard errors for parameters for the global models (Harrison et al, ; Tables and ).…”