Body surface temperatures as biomarkers of physiological environmental adaptation in wild birds and mammals.

The ability of individuals to cope with their environment, and therefore the likelihood that they survive and pass on their genes (i.e. fitness), is largely determined by physiological state. Tracking physiological state in wild animals, however, is challenging. Predominant techniques rely on capture and invasive procedures, restricting research to trappable species and individuals. Additionally, natural behaviours are interrupted, results may be affected by surgery or carrying apparatus, and welfare constraints restrict repeated sampling. Also, the leading non-invasive alternative - faecal sampling - cannot detect rapid physiological changes. Thermal imaging offers an increasingly popular option for studying physiological state in homeothermic endotherms (birds and mammals). The method resolves many of the above concerns and can infer both fast and slow underlying physiological changes from body surface temperature dynamics. Nonetheless, the generalisability of results across settings and populations remains unclear because systematic synthesis is lacking. Correspondingly, important knowledge gaps may be currently overlooked for the same reason. To address these deficits, we performed a systematic review of research linking endotherm body surface temperatures and the four main physiological functions expected to influence surface temperatures - thermoregulation, metabolism, stress and immune responses. We combined outcomes into consensus profiles to ascertain whether responses are generalisable. We also evaluated article publication metrics, study subjects, and methods to characterise research trends and identify approaches most likely to drive progress. Consensus profiling suggested thermoregulatory, metabolic and acute stress (up to 3 min from stressor onset) body surface temperature responses are likely to be broadly generalisable. By contrast, body surface temperature dynamics during immune activation likely depend on discrete ranges of environmental conditions. However, the reviewed literature demonstrates that we still lack sufficient understanding of the mechanistic processes connecting body surface temperatures with underlying physiology. Therefore, further development of methods for inferring physiology from body surface temperatures in natural environments will require combinations of detailed laboratory validations and confirmatory field studies. Such research would also benefit from greater rigour than is evident in the currently available literature, in terms of routinely validating physiological challenges, avoiding use of stress-inducing methods, analysing life-history stage and sex differences, investigating effects of both challenge increase and decrease, and assessing responses across all possible thermoregulatory states. Assuming these knowledge gaps can be filled and technical challenges overcome, inferring physiology in the wild using thermal imaging will present a host of valuable eco-evolutionary research opportunities surpassing those available with invasive or integrating techniques.