A twin codend trawl was fished in the northern Baltic to study the size selectivity of square mesh and diamond mesh codends of 36-mm nominal mesh size. For each codend, 15 hauls were completed with a small mesh (20 mm) codend deployed on the other side of the trawl. The relative size of the catches in the two sides of the trawl varied considerably from haul to haul (the separator section was not operating properly) and selection curves were estimated from each individual haul using a method that incorporated the differences in catching efficiency of the two sides. The length of 50% retention decreased with increased catch for both the diamond and square mesh codends, although in neither case was this relationship statistically significant. Selection curves fitted to the combined haul data were asymmetric. The square mesh codend retained significantly less small herring than the diamond mesh codend, and for larger herring the two codends had similar selectivity. In both codends, most escapes occurred at the front of the catch bulge, from the upper side of the codend. At high catch rates, mesh blockage was observed for several metres ahead of the catch bulge during the later part of the tow.
Fish Vision: Size Selectivity in the Zebrafish Retinotectal Pathway
