Using morphometrics to identify sex in the eastern blue-tongued skink (Tiliqua scincoides).

Abstract

Tiliqua scincoides coexists with human activity and is frequently presented for rehabilitation due to injury. The correct identification of sex is important as animals identified as female should be subject to a different decision-making matrix for rehabilitation. However, identification of sex is notoriously difficult in Tiliqua scincoides. We describe a reliable, safe and cost-effective morphometry-based method. Adult and sub-adult, wild Tiliqua scincoides dead on presentation or euthanased due to their presenting injuries were collected in South-East Queensland (SE Qld). Head-width to snout-vent length ratio (H:SV) and head-width to trunk length ratio (H:T) were measured and sex was defined at necropsy. Similar data were obtained from a previous study in Sydney, New South Wales (NSW). H:SV and H:T were assessed for accuracy of sex prediction by the area under the receiver operating characteristic curve (AUC-ROC). Optimal cut-points were identified. The AUC-ROC for the H:T test was for NSW adults, 0.99 (n = 29), NSW sub-adults, 0.95 (n = 10), Qld adults, 0.90 (n = 35) and Qld sub-adults, 0.79 (n = 25). In all cases, H:T was as good or superior to H:SV. H:T cut-points optimized for female sexing or both sexes ranged from 0.20 to 0.23 depending on State and adult status. Sensitivities and specificities of the test at suggested optimal cut-points ranged from 0.54 to 1.0. We describe how H:T can be used as an accurate method to determine sex in Tiliqua scincoides. However, it is more accurate in adults than sub-adults and more accurate in NSW skinks than in SE Qld skinks.