Environmental DNA as a detection and quantitative tool for stream‐dwelling salamanders: A comparison with the traditional active search method

Abstract As amphibians are showing significant signs of decline, adequate information and understanding of target species are essential for taking appropriate conservation measures. In recent years, environmental DNA has seen notable growth as a monitoring tool and testing this emergent method with various species has become an important step toward a better understanding of its benefits and limits for studying specific taxa. This study focused on using species‐specific qPCR assays developed in our research group to test the eDNA method for three stream‐dwelling salamander species of the Plethodontidae family: the Spring salamander (Gyrinophilus porphyriticus), the Northern dusky salamander (Desmognathus fuscus), and the Northern two‐lined salamander (Eurycea bislineata). The traditional active search method and the eDNA method were compared for both their ability to detect species as well as to provide a quantitative assessment of populations in 24 headwater streams in Québec, eastern Canada. For all three species, eDNA was detected in every stream where the target species was observed during the active search method. Moreover, eDNA was detected in nine streams where the target species was not identified with the active search. A marginally significant association was found between eDNA concentration and salamander density for D. fuscus only. All species showed high variability for eDNA concentration between qPCR technical replicates and between samples of a given stream. The results of this study lead us to conclude that eDNA can be an excellent method for detection of stream‐dwelling salamanders. Given the inconsistent quantitative aspect of eDNA with the studied species, the future of these stream‐dwelling salamander monitoring most likely lies in the combined use of both eDNA and active search methods. Hence, active search could continue to offer useful small‐scale detection and reliable quantitative data while eDNA could be implemented as an efficient and promising tool for large‐scale detection