Determining an effective sampling method for eDNA metabarcoding: a case study for fish biodiversity monitoring in a small, natural river

In recent years, biodiversity loss has become one of the most serious environmental issues worldwide, especially in aquatic ecosystems. To avoid diversity loss, it is necessary to monitor biological communities, and environmental DNA (eDNA) metabarcoding has been developed as a rapid, noninvasive, and cost-effective method for aquatic biodiversity monitoring. Although this method has been applied to various environments and taxa, a detailed assessment of the efficient sampling methods for monitoring is still required. In this study, we explored eDNA metabarcoding sampling methods for fish at a single site to maximize the number of detected species using realistic effort in a natural, small river. We considered the following three parameters: sample type (water or sediment), sample position at a site (right and left shore and center of the river), and water volume (10–4000 mL). The results suggested that the number of detected species from sedimentary eDNA was equivalent to that from aqueous eDNA, although the species composition was different. The number of detected species could be saturated by collecting a 1000 mL water sample, regardless of sampling position within a survey site. However, sedimentary eDNA showed a spatially heterogeneous species composition between sampling positions within a survey site despite the short distance (5 m) between positions, without apparent differences in physical properties such as velocity and sediment particle distribution. By completing eDNA biodiversity monitoring of fish with 1000 mL water samples across the whole river, we detected more fish species than in previous traditional surveys conducted at the same sites. Thus, the aqueous eDNA metabarcoding method is as efficient as traditional surveys, while sedimentary eDNA metabarcoding could complement the results of aqueous eDNA metabarcoding.