Sex-Ratio Selection in Species with Helpers at the Nest: The Repayment Model Revisited.

We reexamine the theoretical and empirical basis for the "repayment model" of Emlen, Emlen, and Levin, which predicts that the overall sex ratio of offspring in cooperative-breeding species with helpers at the nest should be biased toward the more helpful sex. We conclude that in theory the model may yield a biased sex ratio but only under restricted conditions that are unlikely to be met in most populations. In those cases where the model is potentially applicable, estimates of both direct and indirect fitness of sons and daughters are necessary in order to avoid double accounting. As a result, extensive demographic data are required to determine the expected sex-ratio bias. Empirically, none of the three examples originally suggested as supporting the repayment model withstand critical scrutiny. Using long-term demographic data from two populations of one of these species, the red-cockaded woodpecker (Picoides borealis), we determine that the model predicts a biased sex ratio not observed in the data. This suggests that confounding factors may counter the sex-ratio bias predicted by the model, that the extensive empirical data available for this species may be insufficient to test the model, or that there are as yet unknown theoretical difficulties with the model. Finally, we discuss the recently reported case of the Seychelles warbler (Acrocephalus sechellensis). The extraordinarily biased sex ratios reported in this species are in accord with those predicted by the model when applied at the scale of individual groups but not at the population level. Additional empirical work is necessary to determine what, if any, the generality and usefulness of the repayment model may be.