It has been suggested that sex ratios dependent on host size are unlikely to evolve in parasitoid wasps developing in growing hosts because future host quality would not be predictable at the time of oviposition by the female. We used logistic regression to estimate the primary sex ratio of a solitary parasitoid of aphid nymphs, using data on offspring sex at the time of adult eclosion. The method corrects for any differential preadult mortality between male and female offspring. Male preadult mortality is estimated separately from data on male-only offspring produced by unmated females. This information is built into the statistical analysis of data on offspring sex at the time of eclosion of progeny of mated females to estimate female preadult mortality and the primary sex ratio. The method was used to demonstrate manipulation of offspring sex by females of the parasitoid Ephedrus californicus parasitizing large (third-instar) and small (first-instar) pea aphids, Acyrthosiphon pisum. Mated E. californicus females fertilized nearly 50% of their eggs laid in large host aphids but only about 20% of those laid in small hosts, the difference being highly significant. Parasitoid survival from oviposition to adult eclosion was about 10% higher in small hosts than in large hosts, but for a given host size, the incidence of survival did not differ significantly between the sexes. Therefore, differential preadult mortality had no effect on modifying the primary sex ratio under the experimental conditions. The results also indicated that female decisions about the sex of offspring were affected by previous experience with the range of host sizes available and were more flexible with large than with small hosts, which were allocated male progeny almost invariably. We discuss the evolution of sex-ratio manipulation in solitary hymenopterous parasitoids of the koinobiotic type, which develop in growing host stages, such as aphid nymphs, as opposed to eggs and pupae. Our data indicate that a growing host can represent a reliable resource that is predictable from its initial size, even though it has not reached its potential size at the time of parasitoid oviposition. At least in species such as E. californicus that attack a range of host instars differing widely in size and thus in potential for parasite growth, the ability to effect sex-ratio adjustments based on host size at the time of oviposition may help to maximize female reproductive success, despite any uncertainty about future host quality.
Limited Dispersal, Deleterious Mutations and the Evolution of Sex
