Sexing animals using biometry: Intra-pair comparison is often superior to discriminant functions

In sexually monomorphic animal species it is difficult to sex unknown individuals. The most widely used method of sexing in those cases is discriminant analysis. In species that are socially monogamous and have biparental care, however, there exists another method which utilises the intra-pair difference in the trait used in sexing. Equations for calculating the fraction of wrongly sexed individuals using both methods are given and used to show that the latter method is more exact than discriminant analysis. The advantages and limits of both methods are compared. Field data from the kittiwake (Rissa tridactyla) are used to exemplify the intra-pair comparison method, and to test the assumptions underlying this method for this species. Among other things, it is shown that kittiwakes display random mating with respect to head+bill length. Sexing kittiwakes at Hornøya, northern Norway, by intra-pair differences in head+bill length gives correct results in 96% of the cases.