Abstract
Environmental DNA (eDNA) is increasingly used to noninvasively monitor aquatic animals in freshwater and coastal areas. However, the use of eDNA in the open ocean (hereafter referred to OceanDNA) is still limited because of the sparse distribution of eDNA in the open ocean. Small pelagic fish have a large biomass and are widely distributed in the open ocean. We tested the performance of two OceanDNA analysis methods-species-specific qPCR (quantitative polymerase chain reaction) and MiFish metabarcoding using universal primers-to determine the distribution of small pelagic fish in the open ocean. We focused on six small pelagic fish species (Sardinops melanostictus, Engraulis japonicus, Scomber japonicus, Scomber australasicus, Trachurus japonicus, and Cololabis saira) and selected the Kuroshio Extension area as a testbed, because distribution of the selected species is known to be influenced by the strong frontal structure. The results from OceanDNA methods were compared to those of net sampling to test for consistency. Then, we compared the detection performance in each target fish between the using of qPCR and MiFish methods. A positive correlation was evident between the qPCR and MiFish detection results. In the ranking of the species detection rates and spatial distribution estimations, comparable similarity was observed between results derived from the qPCR and MiFish methods. In contrast, the detection rate using the qPCR method was always higher than that of the MiFish method. Amplification bias on non-target DNA and low sample DNA quantity seemed to partially result in a lower detection rate for the MiFish method; the reason is still unclear. Considering the ability of MiFish to detect large numbers of species and the quantitative nature of qPCR, the combined usage of the two methods to monitor quantitative distribution of small pelagic fish species with information of fish community structures was recommended.