Understanding size selectivity of trawls using structural models: Methodology and a case study on fish sorting grids

Abstract

Fish behaviour affects the performance of selection devices in fishing gears. Traditionally, fish behaviour in relation to selection devices is assessed by direct observation. However, this approach has limitations, and the observations are not explicitly incorporated in the selectivity models. Further, underwater observations and quantification of fish behaviour can be challenging. In this study we outline and use an indirect method to explicitly incorporate and quantify fish behaviour in trawl selectivity analysis. We use a set of structural models, which are based on modelling the actual processes believed to determine the size selection of the device, to discern which behaviours are most likely to explain the selectivity process. By bootstrapping we assess how confident we can be in the choice of a specific structural model and on discerning the associated behavioural aspects. We collected size selectivity data in the Barents Sea demersal trawl fishery targeting gadoids, where the use of a sorting grid is compulsory. Using our modelling approach, we obtained deeper understanding of which behavioural processes most likely affect size selectivity in the sorting grids tested. Our approach can be applied to other fishing gears to understand and quantify fish behaviour in relation to size selectivity.