We propose a model for sex-ratio adjustment complementary to that of Trivers and Willard. In addition to the three basic assumptions of the Trivers-Willard model, our model assumes that the sex with more variable reproductive success (normally male) is also the sex less constrained for reproduction. This assumption seems realistic, because several studies have demonstrated that poor-condition males may adopt alternative mating strategies and sire some offspring, whereas females have physiological constraints for gestation or egg production that cannot be avoided. Thus, under these circumstances, sons of both poor and good condition would be more valuable for parents than daughters, whereas daughters would be relatively more valuable than sons at intermediate condition. This model predicts, therefore, a U-shaped relationship between parental condition and offspring sex ratio. We present a case study for the monogamous lesser kestrel (Falco naumanni) that fulfills the assumptions and predictions of the model. The minimum body condition for breeding, measured as pectoral thickness, was lower for sons than for daughters. Below this minimum, males had a higher chance of breeding than females. Above this minimum, however, the lifetime reproductive success was condition dependent in males but not in females. Thus, males in better body condition attain, on average, higher reproductive success than females. Offspring sex ratio varied with the size of the father's ornaments and mother condition according to the U-shaped pattern predicted by the model.

Corresponding Editor: B. Crespi