Investigating potential fish distributions in artificial and natural reefs using an optical method: Insights from subtropical offshore wind farms

With the goal of achieving clean energy and green transition, offshore wind farms (OWFs) have been rapidly established worldwide. The underwater pile foundations in OWFs have shown potential to serve as artificial reefs. To explore the ecological benefits in subtropical ecosystems, fish distributions were investigated at selected sites: two OWFs, a nearby natural reef, and a shipwreck reef. A low-cost optical method was developed to collect fish data, followed by deep learning techniques to identify four target species from underwater videos. To reconstruct distribution patterns from the sea surface to the bottom, fish densities were calculated along the recording path based on the camera view. In the OWFs, fish aggregation was confirmed by the high densities of round scad (0.025–0.26 per m${}^3$) and spinefoot (0.052–0.15 per m${}^3$). The target fish groups were observed near the individual wind turbine foundations throughout the entire vertical water column, with higher aggregation in the mid-water layers. Black scrapers exhibited the highest density at the shipwreck reef (0.024 per m${}^3$). Damselfish were dominant at the natural reef area (0.055–0.14 per m${}^3$); however, they were rarely found at the nearby OWF or shipwreck reef, indicating that the ecological function of artificial reefs may not fully cover that of natural reefs. Multiple regression analysis showed that the fish may exhibit specific preferences for structural features and environmental variables. This study provides insights into potential fish distributions to inform local managers and fishing enthusiasts, while supporting sustainable OWF development and contributing to marine conservation efforts.