Abstract
Models developed to geolocate individual fish from data recorded by electronic tags often require major modification to be applied to new regions, species, or tag types due to variability in oceanographic conditions, fish behavior, and data resolution. We developed a model for geolocating Atlantic cod (Gadus morhua) off New England that builds upon an existing hidden Markov model (HMM) framework and addresses region-and species-specific challenges. The HMM framework contains a likelihood model that compares tag-recorded environmental data (depth, temperature, tidal characteristics) with those derived from an oceanographic model and a behavior model that constrains the horizontal movement of the fish. Validation experiments were performed on stationary tags, double-electronic-tagged fish (archival and acoustic tags), and simulated tracks. Known data, including fish locations and activity metrics, showed good agreement with those estimated by the modified approach and improvements in performance of the modified method over the original. The modified geolocation approach will be applicable to additional species and regions to obtain valuable movement information that is not typically available for demersal fishes.