Predictive models and comparison of the selectivity of standard (TO) and turned mesh (T90) codends for three species in the Eastern Mediterranean

The main objective of this study was to compare and predict the size selective properties of 900 turned mesh codends (T90) with those of similar standard diamond mesh codends (TO) without other design changes. In particular, we wanted to investigate whether a T90 codend can be a simple way to improve size selectivity in the multi-species Mediterranean demersal trawl fisheries. Data were collected and analyzed for three species, while considering the codend design parameters mesh size (m) and netting direction (TO or T90) as potential fixed effects for size selection parameters 50% retention length (L50) and selection range (SR). A special model was constructed and model choice was performed separately for red mullet (Mullus barbatus), common pandora (Pagellus erythrinus) and annular sea bream (Diplodus annularis) based on the AIC value (Akaike, 1974) while considering every possible simpler sub-model following the procedure described in Wienbeck et al. (2011). L50 and SR predictions were made based on this model for TO and T90 codends with mesh sizes from 38 mm to 52 mm for each species separately. The findings revealed that T90 codends significantly improve size selectivity of red mullet and common pandora compared to conventional TO codends of the same mesh size. However, L50 and SR values for TO and T90 codends were found to be similar for annular sea bream. This difference in codend size selectivity between species with regard to turning codend netting orientation by 900 may be related to the difference in body shape. Red mullet and common pandora have rounded shapes while annular sea bream has a laterally compressed body shape. Increase in mesh size was found to increase L50 for all three species for both codend types. (C) 2013 Elsevier B.V. All rights reserved.