Scandol et al Experiments to reduce bycatch in an estuaiine squid-trawl flsheiy 



535 



Length in 

 meshes 



Length in 

 meshes 



Top panel ,157 



; Bottom panel 



1 4-mm braided 



60-mm mesh 



throughout 



175 T 



30 



100T 



20 T 



Wing 

 panel 



I 

 Zipper 



Figure 2 



Plan of the 4-seam squid trawl used in the study. N 

 B = bars. 



20 T 



normals; T = transversals. 



and 100 meshes, respectively, and posterior sections 

 with a length of approximately 1 m, but different mesh 

 circumferences and sizes (all meshes were made from 

 3-mm braided twine; Fig. 3, A-D). The first codend 

 (41/150) represented commercially used designs with a 

 posterior section of 41-mm mesh and a circumference of 

 150 meshes (Fig. 3A). The second codend (41/1001 had 

 a posterior section made from the same mesh, with a 

 circumference of 100 meshes (Fig. 3B). The third and 

 fourth designs (45/150 and 45/100, respectively) had 

 posterior sections with the same circumferences as those 

 above, but were made from 45-mm mesh (Fig. 3, C and 

 D). The fifth codend design (termed the "square-mesh 

 codend") was made entirely of 30-mm mesh (1.4-mm 

 twisted twine) hung from the bar (i.e., the meshes were 

 oriented so that they were square shaped) and had a 

 total length and fishing circumference of 144 and 90 

 bars, respectively (Fig. 3E). 



Four variations of behavioral-type bycatch reduction 

 devices (BRDs) were tested in combination with certain 

 codends (Fig. 3, F-I). All BRDs were installed in the 

 top anterior section of the relevant codend at a distance 

 starting 1.1 m anterior to the last row of meshes in the 

 posterior section according to the specifications provided 

 by Broadhurst et al. (2002). The first BRD (the diamond 

 BRD; see Broadhurst et al. [2004]) was a diamond- 

 shaped opening (11x11 bars) cut into a conventional 

 41/150 codend (Fig. 3F). The second and third BRDs 



were composite square-mesh panels (CSMP; Broadhurst 

 and Kennelly[1996]) with 90- or 75-mm mesh, respec- 

 tively, hung from the bar in the main escape panels 

 and were inserted into the conventional 41/150 and 

 41/100 codends. These modified codends were called 

 the "90-CSMP" and "75-CSMP" codends (Fig. 3, G and 

 H). The fourth BRD was inserted into the square-mesh 

 codend (termed the "75-panel codend"; Fig. 31); it was 

 a simple panel of 75-mm mesh hung from the bar (and 

 having the same dimensions as that used in the CSMP 

 codend). 



Experimental procedure 



Three independent experiments were completed during 

 2002 to identify an appropriate combination of codend 

 and BRD that minimised bycatch whilst maintaining 

 catches of squid. Experiment 1 (25 February-1 March 

 2002, five days, two replicate hauls/day) tested pre- 

 dictions about the effect of 1) changing the posterior 

 codend circumference within the maximal and minimal 

 codend mesh sizes legally permitted in the fishery; and 

 2) using the diamond and CSMP BRDs located in the 

 commercially used 41/150 codend. Experiment 2 (29 

 July-9 August 2002, 10 days, two replicate hauls/day) 

 repeated experiment 1 above (but with additional repli- 

 cation to increase the power to differentiate between the 

 codends) and also tested the effectiveness of the square- 



