A meta‐analysis of haddock size‐selection data

Many jurisdictions have introduced management regimes prohibiting or limiting discarding, and a key objective of the European Union Common Fisheries Policy is the gradual elimination of discards. One way of reducing the catch of unwanted fish is to use more size‐selective gears. Gear‐based management options can best be explored using models that predict size selection across a wide range of variables related to gear design. Such a model is developed for haddock (Melanogrammus aeglefinus) through a meta‐analysis of 21 trials of codend selection and 19 trials of the combined selection of the codend and a square mesh panel. Individual‐haul estimates of the 50% retention length (l₅₀) and the selection range (SR) are related to a set of explanatory variables through a structural model that describes the dual process of panel and codend selection. Codend l₅₀ and SR are positively related to codend mesh size and negatively related to the codend twine diameter. Codend l₅₀ is also negatively related to the number of open meshes around the codend circumference. Panel l₅₀ increases with panel mesh size. The panel contributes more to gear selection as it is moved closer to the codline. The panel is most effective between November and January and least between May and July, periods which broadly coincide with peak and poor haddock condition. The results are illustrated for a typical trawler targeting haddock in the Scottish whitefish sector, and the utility of the model for gear design and legislation is discussed.