Data from: Multivariate sexual selection on ejaculate traits under sperm competition

Full description

Raw data used in multivariate selection analyses. These data include: experimental block number, male ID, number of successes (competitive fertilizations), number of failures (total fertilizations in block minus number successfully fertilized by the male), proportion of successful fertilizations, flagellum beat cross frequency (BCF), curvilinear velocity (VCL), path linearity (LIN), percentage of motile sperm, mean sperm head length and mean sperm flagellum length. The widespread prevalence of sperm competition means that ejaculates face intense sexual selection. However, prior investigations of sexual selection on gametes have been hampered by two difficulties: (1) deriving estimates of relative fitness from sperm competition trials that are comparable across rival male and female genotypes; and (2) obtaining measures of competitive fertilization success that are not confounded by post-zygotic effects. Here, we exploit the experimental tractability of a broadcast spawning marine invertebrate to overcome these challenges and characterise multivariate sexual selection on sperm traits when multiple ejaculates compete. In multi-male spawning events, we tracked real-time success of sperm using fluorescent tags that are visible inside fertilized eggs. We then used multivariate selection analyses to identify patterns of linear and non-linear sexual selection on multiple sperm morphology and motility traits. Specifically, we found non-linear selection against divergent combinations of sperm length, velocity and swimming path linearity. These patterns likely reflect the way different swimming strategies allow sperm to locate and track eggs. Our results demonstrate that there are overall patterns of selection on ejaculates across a biologically realistic range of ejaculate-ejaculate and ejaculate-female interactions; therefore, there is the potential for adaptive evolution of ejaculate traits under sperm competition.