Fine-scale spatial distribution of a fish community in artificial reefs investigated using an underwater drone and environmental DNA analysis

ABSTRACT: Although artificial reef (AR) effect evaluation is useful for planning the installation of high-rise ARs and their management, few studies have investigated them quantitatively. The fine-scale 2-dimensional fish distribution in ARs was estimated regarding current fields and vertical structures of 2 high-rise ARs (20 and 30 m high at 62 and 72 m depths, respectively) in Tateyama Bay, central Japan, using underwater drone recordings with vertical line transects and environmental DNA (eDNA) metabarcoding. The species detected by video surveys (21 taxa were identified to species, and 1 to genus) were fewer than by eDNA analysis (103 species and 6 genera), especially for pelagic, small-sized, and cryptic fish. Video surveys revealed that demersal fish increased with decreasing horizontal distance from the AR surface within 20 m, and the richness and total fish density were significantly higher upstream of the ARs. Conversely, the fish eDNA concentration showed different patterns, with significantly higher concentrations downstream of the ARs. The richness peaked at horizontal AR surfaces (e.g. reef top), but density of the dominant species peaked near the bottom by video survey. In comparison, eDNA analysis indicated lower richness and higher eDNA concentration of the dominant species at the reef top. Such discrepancies may be explained by the influence of eDNA transport or its specific behavior or buoyancy. Video surveys revealed the growth stage and sex information of 4 species from their morphology, which is not possible using eDNA analysis. This study shows that the advantages of each evaluation method can complement each other.