Carcass size and ground substrate drive detection rates of avian carcasses by human surveyors and a dog team

Accurate avian mortality estimates are essential for understanding anthropogenic impacts to bird populations and informing conservation strategies. Carcass surveys are commonly conducted by human surveyors or by detection dogs, but the factors influencing surveyor detection abilities have not been fully explored. In this study, we conducted two years of detection trials in the semi-arid high desert of southern New Mexico, USA, testing 27 human surveyors and one conservation detection dog across 1096 trials with 238 carcasses representing 50 avian species. We directly compared detection abilities between surveyor types (human and dog) and identified key factors influencing detection probabilities. The conservation detection dog exhibited a significantly higher detection probability (mean = 0.87) than human surveyors (mean = 0.49, individuals ranged 0.25–0.71), consistent with previous studies. Detection probabilities for both surveyor types were influenced by carcass size and ground substrate; detection probability was higher for larger carcasses and areas with lower vegetative complexity. We discuss our results in the context of common tradeoffs faced by managers in designing carcass surveys and how guidance may vary under different scenarios. Broadly, our study provides valuable insight that can enhance wildlife mortality monitoring, ensuring more accurate mortality estimates to inform management and conservation efforts.