Sex allocation and local mate competition in Old World non-pollinating figwasps

The populations of many species are structured such that mating is not rand om and occurs between members of local patches. When patches are founded by a single female and all matings occur between siblings, brothers may compe te with each other for matings with their sisters. This local mate competit ion (LMC) selects for a female-biased sex ratio, especially in species wher e females have control over offspring sex, as in the parasitic Hymenoptera. Two factors are predicted to decrease the degree of female bias: (1) an in crease in the number of foundress females in the patch and (2) an increase in the fraction of individuals mating after dispersal from the natal patch. Pollinating fig wasps are well known as classic examples of species where all matings occur in the local patch. We studied non-pollinating fig wasps, which are more diverse than the pollinating fig wasps and also provide nat ural experimental groups of species with different male morphologies that a re linked to different mating structures. In this group of wasps, species w ith wingless males mate in the local patch (i.e. the fig fruit) while winge d male species mate after dispersal. Species with both kinds of male have a mixture of local and non-local mating. Data from 44 species show that sex ratios (defined as the proportion of males) are in accordance with theoreti cal predictions: wingless male species < wing-dimorphic male species < wing ed male species. These results are also supported by a formal comparative a nalysis that controls for phylogeny. The foundress number is difficult to e stimate directly for non-pollinating fig wasps but a robust indirect method leads to the prediction that foundress number, and hence sex ratio, should increase with the proportion of patches occupied in a crop. This result is supported strongly across 19 species with wingless males, but not across 8 species with winged males. The mean sex ratios for species with winged mal es are not significantly different from 0.5, and the absence of the correla tion observed across species with wingless males may reflect weak selection to adjust the sex ratio in species whose population mating structure tends not to be subdivided. The same relationship is also predicted to occur wit hin species if individual females adjust their sex ratios facultatively. Th is final prediction was not supported by data from a wingless male species, a male wing-dimorphic species or a winged male species.