Conservation detection dogs: A critical review of efficacy and methodology.

Abstract

Conservation detection dogs (CDD) use their exceptional olfactory abilities to assist a wide range of conservation projects through the detection of target specimens or species. CDD are generally quicker, can cover wider areas and find more samples than humans and other analytical tools. However, their efficacy varies between studies; methodological and procedural standardisation in the field is lacking. Considering the cost of deploying a CDD team and the limited financial resources within conservation, it is vital that their performance is quantified and reliable. This review aims to summarise what is currently known about the use of scent detection dogs in conservation and elucidate which factors affect efficacy. We describe the efficacy of CDD across species and situational contexts like training and fieldwork. Reported sensitivities (i.e. the proportion of target samples found out of total available) ranged from 23.8% to 100% and precision rates (i.e. proportion of alerts that are true positives) from 27% to 100%. CDD are consistently shown to be better than other techniques, but performance varies substantially across the literature. There is no consistent difference in efficacy between training, testing and fieldwork, hence we need to understand the factors affecting this. We highlight the key variables that can alter CDD performance. External effects include target odour, training methods, sample management, search methodology, environment and the CDD handler. Internal effects include dog breed, personality, diet, age and health. Unfortunately, much of the research fails to provide adequate information on the dogs, handlers, training, experience and samples. This results in an inability to determine precisely why an individual study has high or low efficacy. It is clear that CDDs can be effective and applied to possibly limitless conservation scenarios, but moving forward researchers must provide more consistent and detailed methodologies so that comparisons can be conducted, results are more easily replicated and progress can be made in standardising CDD work.