FEMALE-BIASED SEX RATIOS: INDIVIDUAL OR GROUP SELECTION?

Abstract

The evolution of biased sex ratios in a randomly structured population stems from individual selection acting through local parental control (LPC) of the sex ratio and hence of the mating success of the sons and/or daughters. As a general rule, the sex ratio is biased away from the sex whose fitness is most affected by changes in the local sex ratio. This is the sex whose fitness is subject to the most effective parental control. The bias acts to increase the fitness of the rarer, controlled sex and to increase parental productivity. In the specific case of the evolution of the female-biased Hamiltonian ratios, LPC can affect the mating success of sons but has no effect on the success of daughters. It is argued here and elsewhere (Nunney, unpubl.) that group selection can only promote the spread of a genotype through the maintenance of a positive association of individuals of that genotype. The importance of positive association is well established in the special case of kin selection. Given such a definition, group selection plays no part in the evolution of the Hamiltonian sex ratios, although it is possible to conceive of circumstances under which group selection could favor an even more extreme sex ratio bias. In general, such circumstances involve kin selection. It is argued that the examination of differences in group productivity is not a useful way of looking at the process of natural selection, since (i) by dividing up almost any evolving population into random groups, some groups (those with the highest frequency of the fittest individuals) will be more productive than others; and (ii) in the specific case of the evolution of the Hamiltonian ratios, it is possible to develop models either with or without a group structure and get the same result. Hamilton (1967) originally suggested that a female-biased sex ratio arose in his model because of the advantage of reducing local mate competition (specifically, reducing competition between brothers for mates). This possibility was eliminated by developing a model in which competition between the brothers was prevented regardless of the sex ratio. It was found that the optimum sex-ratio strategy was unaffected. On the other hand, the idea of local parental control has, in each case examined, been able to account for the predicted optimum strategy.