Contradictory results in sex ratio studies: populations do not necessarily differ

In avian sex ratio studies, results often differ between species and between populations within species. Some researchers argued that positive results were simply statistical artefacts and that sex ratio adjustment did not exist. However, many of the proposed mechanisms of sex ratio adjustment result in costly laying gaps. In these cases, females laying large clutches may restrict the sex manipulation to the first egg of the clutch. Consequently, detectability of sex ratio adjustment on the level of the clutch can be low. Though obvious, this fact is often neglected in the literature. Using simulations, I show that the proportion of undetected sex manipulation can be surprisingly high when the manipulation is restricted to the first egg. If the sample size is 50 broods, there is 47% and 71% chance in 6- and 12-egg clutches, respectively, that sex manipulation is undetected. Even with large samples (n = 100), the figures are 15% and 46%. These data suggest that nonsignificant results for clutch sex ratios do not necessarily mean that sex is not manipulated in a portion of the brood, e.g. in first-laid eggs. Hence, whenever possible, data on laying order-specific sex manipulation should also be collected. Without such data, contradictory results on brood sex ratios should be interpreted cautiously.