Sex ratios in geographically structured populations

In his book on sexual selection (1), Darwin documented evidence that the primary sex ratio (the proportion of males at conception) is about 1/2 in a wide variety of species. Otherwise, he explained, a newly conceived member of the rare sex will, on average, have more offspring than one of the common sex, since each offspring has one mother and one father; thus there is frequency-dependent selection in favour of parents producing the rare sex. Darwin formulated this explanation in the first edition (1871) for monogamous species, but he failed to extend it to polygamous species, and in the second edition (1874) he retracted it completely. It was left to Fisher (2) to develop the theory in the more general form that there should be equal parental expenditure on the two sexes, allowing for the possibility that one sex may cost more to produce than the other. Despite the wide applicability of Fisher's principle, recent work on sex ratio evolution has focused on situations where it breaks down (3). Hamilton (4) first pointed out that Fisher's argument assumes population-wide competition for mates, whereas most natural populations have a geographical population structure in which limited dispersal imposes constraints on mating patterns. What are the consequences for the sex ratio?