Significant advancements in environmental DNA (eDNA) analysis technology have led to its widespread adoption for species monitoring. However, the low DNA concentration in eDNA often raises concerns about PCR bias, which is a primary issue in enhancing the reliability of eDNA survey results. This study aims to identify the suitable eDNA analysis method and detection strategy for obtaining the presence data of Eurasian otters in an urban stream and to investigate the seasonal distribution patterns of Eurasian otters and their potential prey species. The eDNA survey revealed the presence of six out of the 13 mammal species recorded in fieldwork and literature reviews, including Canis lupus, Felis silvestris and Mustela sibirica. Notably, Eurasian otters were not detected in the eDNA metabarcoding conducted in April and October, despite traditional surveys confirming their presence. In contrast, qPCR assays successfully amplified Eurasian otters from the same samples that were analysed using metabarcoding. When evaluating detection criteria, based on the number of positive samples in repetitions, Eurasian otters were detected at 42.9% to 85.7% of the sampling sites in April and at all sampling sites in October. This suggests a higher detection probability of Eurasian otters in October, indicating a potential expansion of their home range during that season compared to April. Metabarcoding results revealed similar findings regarding fish species as traditional surveys, with Cyprinidae accounting for the largest proportion of fish species at the family level (April, 54.57%; October, 43.58%), followed by Gobiidae (April, 16.90%; October, 22.88%). At the species level, P. parva was the dominant fish species in Saetgang, constituting 5.68% and 6.35% of relative abundance in April and October, respectively. This implies that Eurasian otters as opportunistic predators, may increasingly take advantage of the availability of species within the family Cyprinidae, notably Pseudorasbora parva, as a food source within the study area during the months of April and October. This study highlights that qPCR is the more effective approach in urban areas to offer insights into otters’ distribution patterns, while metabarcoding is useful to provide the properties of the biological environment. Furthermore, this study indicated that it is necessary to determine the suitable eDNA analysis methods depending on the research purpose to obtain detection results effectively.
