Use of AI-mediated image analysis to investigate the influence of environmental factors and recreational scuba diving on fish activity patterns in a marine protected area

Traditional intrusive survey methods are limited in their capacity to conduct long-term monitoring of fish behavior and to accurately assess the impacts of human activity. Using an unmanned underwater camera system and YOLOv5 artificial intelligence recognition technology, we assessed the temporal changes of fish activity patterns in relation to the diurnal cycle and water temperature, and examined whether scuba divers alter fish behavior in the Keelung City Wanghaixiang Chaojing Bay Aquatic Plants and Animals Conservation Area. Although designated as a Marine Protected Area (MPA) and a no-take zone, the area remains a popular site for recreational diving. We confirmed a clear diurnal pattern in fish activity, and identified temperature as a key factor influencing species-specific activity patterns. The presence of scuba divers was associated with a significant avoidance response in fishes, but this acute response subsided quickly as soon as the divers had left. Hour-by-hour data nonetheless indicated that the number of fish present at the site may be influenced by the number of visitors at any given time. An increase in visitors was found to suppress the activity of damselfishes (Pomacentridae) for a short time, but the number of visitors did not significantly affect the number of fish observed per day and had no effect on their overall diel activity patterns. In sum, recreational divers appear to have short-term effects, but probably no lasting impact, on fish activity patterns in this MPA.