Abstract The European adder ( Vipera berus ), classified as critically endangered in Germany, faces severe threats from habitat loss, particularly the decline of primary habitats such as peatlands. Peatland restoration is essential but may pose risks when hibernacula are not considered during rewetting, potentially leading to flooding, site loss, and population declines. Similar risks arise from construction activities in secondary habitats. Protecting these key sites is therefore vital for species conservation but requires reliable methods to locate them. Traditional monitoring approaches, such as visual surveys and telemetry, are often invasive, imprecise, or resource-intensive. This study tested whether a trained adder detection dog could identify adder scent—using sheds and fecal swabs—beneath substrate in controlled line-up experiments and under field conditions. In the line-up tests, we assessed how sample type, burial depth, and substrate influenced detection performance, whereas field searches evaluated detection success under more realistic conditions. In a pilot study, we further examined whether the dog could generalize from collected odor samples to live adders by testing its ability to locate the hibernacula of two previously tagged individuals. In the line-up experiments, the dog identified both sheds and fecal swabs with high sensitivity, performing slightly better with sheds. Detection remained effective at depths of up to 1.2 m, although detection success declined with increasing depth. Under field conditions, the dog reliably detected samples buried at a depth of 50 cm without any false indications. In the pilot study, the dog successfully located hibernating adders, demonstrating that training non-invasively with collected odor samples rather than live animals can be sufficient for this purpose. These findings suggest that detection dogs represent a promising, non-invasive, and efficient tool for locating adder hibernacula. Future studies should include multiple dogs, larger sample sizes, and additional tagged individuals to further validate detection performance under field conditions.
Schemel et al. (Sat,) studied this question.