The first step in conserving a rare species is to determine a detailed association between environmental factors and the species’ occurrence. The use of environmental DNA (eDNA) analysis for macro-organisms in aquatic ecosystems has recently undergone rapid progress. This method enables a quick survey in many research sites. Here we show the results of ecological niche modeling based on a large-scale eDNA survey for the Japanese giant salamander (Andrias japonicus). This survey revealed an association between environmental factors and the species’ occurrence. The Japanese giant salamander is listed as a near-threatened species in the IUCN Red List, and its distribution is shrinking. Because the species lives underwater throughout its lifetime and is nocturnal, field observation can require significant effort and is often time-consuming. Therefore, to detect the species more rapidly by eDNA analysis, we collected water samples across Western Japan and extracted eDNA from each sample. The DNA of the Japanese giant salamander was detected by real-time PCR targeting the ITS1 region. In tandem with eDNA evidence, past observation records of this species from the 1960s were gathered. The potential distribution maps based on eDNA and past records were constructed by maximum entropy modeling, and the two maps were compared.
Results/Conclusions
We collected 454 water samples from 366 river sites in 13 river systems over 21 total days of sampling trips. The DNA of the Japanese giant salamander was detected at 127 of the 366 sites. The eDNA-positive sites were distributed throughout Western Japan, and were especially concentrated in mountain areas. The area under the curve (AUC) value of constructed niche models based on eDNA data (0.946) was better than that based on past records (0.839). The predicted habitat potentials of the Japanese giant salamander from the two models correlated significantly (r = 0.52, p < 0.001). These results suggest that the distribution of the species can be estimated on the basis of eDNA data. Although we used all reliable past records of the species, eDNA data obtained in the present study provide a better distribution model. By applying eDNA analysis, we can collect a large and reliable dataset in a short period. The use of ecological niche modeling based on eDNA data is a promising tool for continual monitoring of rare species.