Abstract
Empirical estimates of trawl capture efficiency are essential for converting survey data into absolute abundance. A mark-recapture experiment was conducted to quantify capture efficiency of Atlantic cod (Gadus morhua) in the School for Marine Science and Technology (SMAST) video trawl survey using a custom passive integrated transponder (PIT) tag detection system installed in the trawl codend. Directed tows on cod aggregations at Jeffreys Ledge yielded 1094 tagged releases and six PIT-tag detections. Laboratory calibration indicated 79% detection efficiency, and post-release survival was estimated to be 59% based on vitality scores. Assumptions of equal catchability, random mixing, population closure, tag retention, and survival were tested, and no evidence of violation was detected given the available data. Capture efficiency was estimated directly using a hierarchical probability model that propagated uncertainty in survival, detection probability, and swept-area coverage. The posterior distribution indicated a median capture efficiency of 12.5% (95% credible interval: 5.3%–25.5%). These results indicate that area-swept trawl surveys capture a small fraction of available cod and demonstrate the feasibility of integrating mark-recapture methods and hierarchical models into survey operations to empirically estimate gear efficiency.