Biometric sex discrimination is unreliable when sexual dimorphism varies within and between years: an example in Eurasian Oystercatchers Haematopus ostralegus

Abstract

Molecular sexing of birds has been possible for over a decade, but for practical reasons many studies still use biometric data for sex discrimination. In some species, the sexes are easy to distinguish but sexual dimorphism is often more subtle, requiring the use of statistical analyses of biometric measurements to discriminate sexes. These models are usually parameterized and validated using data from a limited number of sites and years. However, the resulting discriminant functions are often applied to other populations and periods. A crucial, but usually untested, assumption is that sexual dimorphism does not vary in time and space. Here we illustrate the consequences of violation of this assumption in Eurasian Oystercatchers Haematopus ostralegus, a species for which most studies have relied on biometric sexing. Using biometric data from captures of known-sex birds, we show that sexual dimorphism varied substantially in time and even reversed in some months and years. Furthermore, some biometric traits used in sexing changed gradually over time, causing a reduction in sexual dimorphism. We show that the consequences of this variation on sex discrimination in Oystercatchers are subtle and easily overlooked, but can result in inaccurate and strongly male- or female-biased sex-ratio estimates. We recommend that biometric sexing should be avoided in Oystercatchers unless specific calibration for each month, year and area is carried out. This recommendation also applies to other species where biometric traits may depend on environmental conditions. We argue that this condition might apply to many bird species and therefore advise caution when interpreting results based on biometric sex discrimination.