Summary: | Abstract Accurate data characterizing species distribution and abundance are critical for conservation and management of aquatic resources. Inventory methods, such as gillnet surveys, are widely used to estimate distribution and abundance of fish. However, gillnet surveys can be costly in terms of material and human resources, may cause unwanted mortality in the fish communities being studied, and is subject to size and species selection bias. Detecting allochthonous DNA released by species in their environment (i.e., environmental DNA, hereafter eDNA) could be used as a noninvasive and less costly alternative. In this study, we directly compare eDNA metabarcoding and gillnets for monitoring freshwater fish communities in terms of species richness and relative species abundance. Metabarcoding was performed with the 12S Mifish primers. We also used species‐specific quantitative PCR (qPCR) for the most abundant species, the walleye (Sander vitreus), to compare estimated relative abundance with metabarcoding and gillnet captures. Water sample collection, prior to gillnet assessment, was performed on 17 sites in the hydroelectric impoundment of the Rupert River (James Bay, Canada), comparing two water filtration methods. After controlling for amplification biases and repeatability, we show that fish communities’ complexity is better represented using eDNA metabarcoding than previously recorded gillnet data and that metabarcoding read count correlates with qPCR (r = 0.78, p < .001) in reflecting walleye abundance. Finally, based on partial redundancy analysis, we identified alpha chlorophyll, pH, and dissolved oxygen as environmental variable candidates that may influence differences in fish relative abundance between metabarcoding and gillnets. Altogether, our study demonstrates that the proposed eDNA metabarcoding method can be used as an efficient alternative or complementary technique adapted to the biomonitoring of the fish communities in boreal aquatic ecosystems.
|