Adaptive sex allocation in anticipation of changes in offspring mating opportunities.

Sex allocation theory explains why most species produce equal numbers of sons and daughters, and highlights situations that select for deviation from this norm. Past research has, however, heavily focused on situations with discrete generations. When temporally varying generational overlap affects future mate availability, models predict cyclical shifts in sex allocation, but these predictions have not yet been appropriately tested. Here we provide evidence that mosquitofish (Gambusia holbrooki) populations possess a suitable life history: some autumn-born females bred alongside their own offspring, while such overlap was rare or absent for spring-born females and for all males. Our analytic model of sex allocation for these populations produced a perfect rank-order correlation between observed birth sex ratio biases and theoretical predictions, with stronger biases observed as the extent of female generational overlap increased. This is the first robust evidence that sex allocation theory accounts for cases when mating opportunities vary predictably over time.