Hair hormone measurement is becoming a widely-used tool to measure welfare in wild mammals because it is well suited to differentiate acute adaptive stress responses from chronic maladaptive ones (distress). However, critical knowledge gaps on steroid incorporation into hair need to be addressed and adequate validation is required for the method to be robust. Firstly, binding of enzyme immunoassay antibodies to unknown compounds in hair extracts has been shown to overestimate glucocorticoid levels and to lead to erroneous results. Secondly, it remains to be demonstrated exactly which steroid hormones experience increased concentration in hair samples in response to chronic stress. Unfortunately, with notable exceptions, the general trend has been to ignore these issues. Meanwhile, wildlife management decisions that impact the welfare of populations of wild animals are increasingly based on assessments using such tools, making it urgent to address these shortcomings to safeguard the credibility of welfare assessments and the welfare of the animals involved.

We aim to develop and validate novel physiological indicators of animal welfare in hair samples using an untargeted metabolomics approach to compare hair steroid profiles of mammals (Rattus norvegicus) subject to validated stress models with those of controls In the pilot study, steroid profile data will be compared with behavioral indicators of welfare (ultrasonic vocalizations) for cross-validation. (objective 1).  Once the metabolomics workflow has been established in the pilot study, it will be applied to hair samples of wildlife species and used to identify fingerprints or biomarkers (objective 2) for the development of new enzyme immunoassays (EIAs) and to validate existing EIAs (objective 3). By allowing us to pinpoint steroids of interest, this tool will not only enable the precise and accurate measurement of stress responses in hair, but will also provide a workflow to identify, quantify and benchmark steroids of interest and select and validate more widely accessible EIAs.