An assessment of the accuracy and reliability of hair identification of south-east Australian mammals

Abstract

The aim of this experiment was to quantify the accuracy and reliability of identifying mammals from hair samples, using two highly skilled practitioners. Hair samples were collected from 37 mammal species occurring in south-eastern Australia. This material was divided into 252 samples, which were then used in a blind test to determine the accuracy of the technique. The taxa were then grouped into reliability categories based on the accuracy and consistency of the practitioner’s identifications. In all, 23 taxa, including 19 species, were regarded as being reliably identified from hair analysis. Identification of the remaining 18 species involved at least some level of error. Several factors influenced the accuracy of identifications in this study, principally (1) the need to identify samples to species level wherever possible, rather than not making an identification, and (2) the diverse range of species used (from across Victoria) and the lack of collection locality. Target species have been grouped into reliability categories, as a guide to aid evaluation of field-derived data. We emphasise the need for practitioners to gain considerable personal experience of the technique, the diagnostic characteristics used to identify hair of different species and intra-specific variation, in order to maximise the reliability of identification results.