Reproductive dynamics and offspring allocation in fig wasps

Abstract

This thesis considers problems in sex ratio theory, especially those posed by fig wasps. In structured populations with male dimorphisms for dispersal, optimal sex ratios depend on the frequencies of the morphs. Species with relatively more dispersing males should have relatively less female biased sex ratios when foundress number is kept constant. The frequencies of male morphs are affected by two factors. Firstly, the proportion of females that can be mated by either. Secondly, the frequency of the non-dispersing male morph decreases as the likelihood of local mate competition against brothers increases. Females are envisioned to adjust their sex ratios to the average relatedness of daughters to mothers. If females can determine their relatedness to their mates, they can conditionally adjust their sex ratios to more accurate estimates of their daughter's relatedness to them. Submitted females should produce more female biased sex ratios than outbred females (except when very high inbreeding depression occurs) resulting in a split sex allocation pattern in the population. Split sex ratios are conducive to the evolution of eusociality. When female fig wasps simultaneously adjust their sex ratio and clutch size, a dichotomous oviposition pattern should evolve. Females either produce large very female biased clutches or one male egg clutches. This allocation strategy reduces the relatedness of males sharing figs, which in tum, allows the evolution of fatal fighting in certain wasp species. Local mate competition models normally assume that females can produce exact sex ratios. When females' sex ratios are imprecise and varies randomly it may affect the average predictions of models. It is shown that variation, unless extremely high, will not affect the predicted sex ratios. Pollinating females are able to oviposit sequentially. This can affect the predicted optimal sex ratios by skewing the relative reproductive contributions of initial and subsequent foundresses. Second females have more information than the first ones and can produce a conditionally optimal sex ratio. Contrastingly, the first females need to produce sex ratios that have to be optimal for a variety of foundress numbers. Such average strategies can explain the variation in single foundress data.