Abstract

Various sizes of wire mesh were tested as escape panels in fish traps in the New South Wales demersal trap fishery. The mesh selectivities of these escape panels were estimated using a 'parlour' trap, which is analogous to the 'covered codend' designs used to test selectivity in trawls. The results showed that we could estimate the selection size (l₅₀) for most species based on their maximum body height and the maximum aperture of the escape panel mesh. This enabled us to make escape panel mesh that selected the main species in the fishery, snapper (Pagrus auratus), at a predetermined size. Placing the escape panel meshes with the longest axis orientated either vertically or horizontally showed that fish behaviour is important in determining mesh size selectivity. However, the selection size of some species was unaffected by the orientation of the mesh, suggesting that they were able to turn on their sides to escape. Silver trevally (Pseudocaranx dentex) formed tight schools within traps and were selected at smaller sizes when the meshes were oriented with the longest axis horizontally, suggesting that they do not turn on their sides to escape. Selectivity ogives were used with estimates of the sizes of retained and discarded fish captured in the fishery to estimate the impacts of using the escape panels tested on current landings. The results show the difficulty in designing effective escape panels in a multispecies fishery.