Count Me In: Bayesian N-Mixture Modeling of Aggregating Trout in Thermal Refuges Using Underwater Video Surveys.

Sullivan, Christopher J.*, Jason C. Vokoun, Wildlife and Fisheries Conservation Center, Department of Natural Resources and the Environment, University of Connecticut, Storrs, CT

Climate change is increasing periods of thermal stress for cold-adapted fishes. Riverine conservation and management are focusing on thermal refuge habitats in response, but traditional sampling approaches that remove fish from the water are stress-additive and problematic in refuges. Recent advances in underwater video sampling provide a promising alternative, particularly when fishes congregate as they do in thermal refuges. We video-recorded Brown, Rainbow, and Brook Trout in six thermal refuges along a 15 km stretch of the Housatonic River, CT, from July 27th to August 16th 2022 and subsampled the videos to derive estimates of abundance. We then used zero-inflated Bayesian N-mixture modeling to estimate trout abundance and influential environmental conditions across the thermal refuges. We detected at least one trout on each sampling occasion and estimated no fewer than 600 individuals in the aggregate occupying thermal refuges on a given sampling occasion. We approached estimating fish abundance during a period of high thermal stress with a non-capture, less-stressful method paired with a quantitatively rigorous analytical approach. Our results provide one of the first estimates of thermal refuge capacity in a managed trout fishery and provide insight into relationships between thermal refuge habitats and trout abundance that inform proactive management of riverine thermal refuges.