Social structure and split sex ratios in the antPheidole pallidula

We have investigated the social structure and reproductive allocation of the ant Pheidole pallidula. Allozyme analyses indicate that (1) queens are singlymated, (2) both monogynous (single-queen) and polygynous (multiple-queen) colonies co-occur in the same population and (3) polygynous colonies may have anything from unrelated to full-sister queens. Sex ratios in this species were shown to be extremely split and we have tested the extent to which this colonylevel sex ratio specialisation could be explained by (1) the resource level hypothesis (RLH) and (2) the relatedness asymmetry hypothesis (RAH). Our data show that the proportional investment in females was positively associated with total sexual productivity, which supports the RLH. In contrast to the expectation, relatedness among nestmate females was negatively correlated with relative investment in female reproductives. However, no evidence was obtained for a concomitant decrease in relatedness asymmetry; several trends in our data suggest that relatedness of females to male nestmates decreases at least at the same rate as relatedness among female nestmates, which implies that RA is either not correlated or positively correlated with the degree of female bias in the sex ratio. The RA hypothesis cannot therefore be refuted as an account of the split sex ratio in this species. Our study shows that the empirical testing of both the RLH and the RAH may become unexpectedly complex in social systems where both queen number and relatedness among queens vary, and that estimates of maleto- worker relatedness are essential for testing the RAH in such populations.