Estimation of biodiversity metrics by environmental DNA metabarcoding compared with visual and capture surveys of river fish communities

Information on alpha- (local), beta- (between habitats), and gamma- (regional) diversity is fundamental to understanding biodiversity as well as the function and stability of community dynamics. Methods like environmental DNA (eDNA) metabarcoding are currently considered useful to investigate biodiversity. We compared the performance of eDNA metabarcoding with visual and capture surveys for estimating alpha- and gamma-diversity of river fish communities, and nestedness and turnover in particular. In five rivers across west Japan, by comparison to visual/capture surveys, eDNA metabarcoding detected more species in the study sites (i.e. alpha-diversity). Consequently, the overall number of species in the region (i.e. gamma-diversity) was higher. In particular, the species found by visual/capture surveys were encompassed by those detected by eDNA metabarcoding. Estimates of community diversity within rivers differed between survey methods. Although we found that the methods show similar levels of community nestedness and turnover within the rivers, visual/capture surveys showed more distinct community differences from upstream to downstream. Our results suggest that eDNA metabarcoding may be a suitable method for community assemblage analysis, especially for understanding regional community patterns, for fish monitoring in rivers.