Abstract
The catch composition of bottom trawls is commonly refined and improved through changes in codend design. Measures like reducing the number of meshes in codend circumference or turning diamond netting by 90 degrees are well known to improve the size selectivity of fish species with rounded cross-sectional shape. Based on this we speculated whether the same measures, if applied in other parts of a bottom trawl, would provide similar benefits as in the codend. Therefore, experiments were carried out by deploying these changes to the trawl extension piece in a Mediterranean bottom trawl fishery. However, for European hake and monkfish, results showed no indication of improved selectivity or catch pattern compared to the standard extension piece in the trawl. Contrary, for red mullet, one of the most important species in this fishery, reducing the number of meshes in the circumference of the extension piece jeopardized the size selection obtained in the trawl with a standard extension piece. The lesson learnt from this study was that the design changes that work for the codend do not necessarily work for other parts of the trawl. In fact, they can even have negative effects.