Sex identification based on biometrics in four long-distance migratory tern species mist-netted along the East Atlantic Flyway

David Monticelli,Hugues Defourny,Emile Degros,Bruno Portier,Lara R. Cerveira,Jaime A. Ramos,Vítor H. Paiva,Ulf M. Berthelsen,Thomas Bregnballe,Jan van der Winden,Sandra Bouwhuis,Patricia Szczys
DOI: https://doi.org/10.1007/s10336-024-02241-5
2024-11-28
Journal of Ornithology
Abstract:In most tern species, males and females look alike, complicating the sexing of individuals in the field. We mist-netted, measured and molecularly sexed 494 terns of four species (Common Sterna hirundo , Roseate S. dougallii , Black Chlidonias niger and Little Sternula albifrons Terns) at two locations in Senegal, during spring migration, to evaluate the direction and magnitude of sexual size dimorphism (SSD) for four morphological traits and used a species-specific stepwise discriminant function (DF) analysis to establish sexing rules. In addition, we assessed the effect of geographic origin on the ability of DFs to correctly predict sex in two species, and tested the validity of the sexing rules established from our sample of non-breeding individuals using datasets of molecularly sexed birds obtained from European breeding colonies in three species. Our results indicate SSD to be relatively weak, male-biased and greater for head-bill length and body mass compared to tarsus length and wing length. Body mass varied among trapping sites in Senegal in all species, and also differed between age classes and among birds with a different region of origin in Common, but not Roseate, Terns. Head-bill length was the most powerful predictor of sex for all species, being retained in all DFs, while wing length and body mass were additionally predictive in Black and Little Terns, respectively. Overall, sexing accuracy of DFs was lowest in Little Terns (79%), intermediate in Roseate and Black Terns (81%) and highest in Common Terns (83%). When excluding parts of the samples where morphometric measurements overlap between the sexes, this accuracy could be increased to > 90%, such that the developed DFs prove applicable.
ornithology
What problem does this paper attempt to address?