Gear selectivity for three by-catch species in the shallow-water shrimp trawl fishery at the Sofala Bank, Mozambique

Abstract

The reduction of shrimp by-catch when fishing with shrimp trawlers can constitute a strategy to minimise discards. A research cruise was conducted in February 1995 in shallow waters off Sofala Bank, Mozambique to investigate the sorting (selection) efficiency of cod-ends (55mm and 60mm meshes) compared to the top-grid (14mm space bar) systems using covered cod-end method. The selectivity parameters were obtained for Otolithes ruber, Johnius dussumieri and Thryssa vitrirostris through the maximum likelihood estimate (MLE) method, using logit function and pooled samples through the variance component model (VCM) . The top-grid was mounted at a theoretical angle of 33°, one metre aft of the retriever strap, in a downward direction. The overall shrimp to by-catch ratio was 1:1.7. On average, the cod-end selection system gave a smaller L50% and wider selection range than the top-grid for all three species. According to the study, the top-grid had better selection properties than the cod-ends relative to the species mentioned above. Differences in the selectivity of 55mm, 60mm meshes and 55mm mesh combined with grid are discussed as well as cod-end and grid selection systems.