How males and females contribute to joint reproductive success has been a long-standing question in sexual selection. Under postcopulatory sexual selection (PSS), paternity success is predicted to derive from complex interactions among females engaging in cryptic female choice and males engaging in sperm competition. Such interactions have been identified as potential sources of genetic variation in sexually selected traits but are also expected to inhibit trait diversification. To date, studies of interactions between females and competing males have focused almost exclusively on genotypes and not phenotypic variation in sexually selected traits. Here, we characterize within- and between-sex interactions in Drosophila melanogaster using isogenic lines with heritable variation in both male and female traits known to influence competitive fertilization. We confirmed, and expanded on, previously reported genotypic interactions within and between the sexes, and showed that several reproductive events, including sperm transfer, female sperm ejection and sperm storage, were explained by two- and three-way interactions among sex-specific phenotypes. We also documented complex interactions between the lengths of competing males’ sperm and the female seminal receptacle, which are known to have experienced rapid female-male co-diversification. Our results highlight the non-independence of sperm competition and cryptic female choice and demonstrate that complex interactions between the sexes do not limit the ability of multivariate systems to respond to directional sexual selection.

Data were collected experimentally by staging competitive matings between males and females. Flies were observed, measured, dissected and processed as described in the Methods section of the article.

The dataset contains missing data because not all variables could be measured for each individual because, for example, not all females ejected a sperm mass (or it could not be found), or the precise start or end time of copulation or female sperm ejection were missed. If the ejected mass was missing, the number of first-male sperm residing in the female reproductive tract at remating, or the number of sperm transferred by the second male, were also missing by extention. However, the numbers of sperm stored by the female after sperm ejection would have been affected. Very few further specimens had missing data due to accidents during dissection (e.g. losing a spermatheca or bursting the female reproductive tract, preventing assignment of leaked sperm to specific storage organs). Rather than including only those females with complete data, we included all females that mated twice, but excluded those records with missing data for any given analysis.