Assessment of the cover effect in trawl codend selectivity experiments

The covered codend technique is widely used to estimate codend selectivity curves. The codend cover may, however, influence the codend selectivity. In this experiment, a codend fitted with a square mesh window and surrounded by a cover was tested against a similar uncovered codend in a twin trawl rig in the Baltic cod fishery. Kites were attached to the cover to maintain a good clearance from the codend meshes. The analysis of the data followed the general SELECT approach adapted to fit the particular conditions of the experiment. Selection curves for the two codends were fitted simultaneously to the catch data and compared. The selectivity of the codend with cover was determined by the covered codend method and that of the uncovered codend effectively determined by the twin trawl method. It was estimated that 21% fewer cod entered the codend with cover than the codend without cover. The estimated 50% retention length was 2.4% lower for the codend with cover but not significantly different. The parameter variance and the between-haul variance in the 50% retention length were significantly lower for the covered codend. The selection range was 15.8% lower for the codend with cover but not significantly different. Retention rates of cod in the 35-55 cm length classes were estimated to have been higher in the covered codend.