Density-dependent sex ratio and sex-specific preference for host traits in parasitic bat flies
BACKGROUND Sex ratios are important empirical data in predicting sex allocation strategy and selection in populations. Therefore, they should be sampled at crucial developmental steps before and after parental investment. In parasites with free-living (off-host) developmental stages the timing and method of sampling is not trivial, because ecological niches are frequently poorly known. Consequently, information is scarce for sex ratios of these parasites between conception and sexual maturity. Often, only data from adult parasites are available, which usually were collected from the parasite's hosts. Generally, these ratios are assumed to represent operational sex ratios. METHODOLOGY/PRINCIPAL FINDINGS We here report three years of empirical data on population sex differentials from a bat ectoparasite (Trichobius frequens) with off-host developmental stages. At emergence these parasites exhibit a significant and seasonally stable female biased sex ratio. This bias is lost in the adult population on the roosting host, which shows sex ratios at equality. This is best explained by a behaviorally driven, sex-dependent mortality differential. Because consistently only subsets of females are available to mate, the operational sex ratio in the population is likely male biased. Host capture experiments throughout the day show a statistically significant, but temporary male excess in bat flies on foraging bats. This phenomenon is partly driven by the diurnal rhythms of female larviposition, and partly due to parasites remaining in the bat roost during foraging. Because most previous research in bat flies is based only on foraging bats, female contributions to physical sex ratios have been underestimated. CONCLUSION/SIGNIFICANCE Our results highlight the importance of detailed natural history observations, and emphasize that ignoring the spatial and temporal heterogeneity of reproduction in any organism will lead to significant empirical sampling errors of sex ratios, and may obscure operational sex ratios.