An integrated approach to estimating the effective sampling area of baited underwater camera traps

Baited remote underwater video (BRUV) surveys are increasingly used to monitor fish, but estimating absolute abundance remains challenging because effective sampling areas (ESA) are unknown. We present an approach to estimate ESAs using an unmarked spatial capture-recapture (USCR) framework where observed counts arise from fish movement and space use. The model integrates telemetry data on movement and counts from independent remotely operated vehicle surveys. We applied this method to red snapper (Lutjanus campechanus) at two sites in the southeastern United States. At Chicken Rock, North Carolina, estimated ESAs were 8,806-10,532 m^2, consistent with estimates from previous studies and telemetry data. At Turtle Mound, Florida, estimated ESAs were unrealistically large (230,129-444,977 m^2), with low estimated abundance (241-528). Poor performance at Turtle Mound likely resulted from wide camera spacing and limited information on the spatial distribution of individuals from telemetry data. Our results demonstrate the potential of USCR models to estimate fish density, absolute abundance, and BRUV ESAs, while emphasizing the sensitivity of model performance to camera placement and robust auxiliary data in heterogeneous systems.