Broadhurst et al.: Use and success of square-mesh codends in reducing bycatch and improving size-selectivity of prawns 



441 



100 

 80 

 60 

 40 

 20 

 



80n 

 60 

 40 

 20- 



- -nJlilll 



100 

 H Composite square 2 n := 413 80 



□ Control n = 630 60 



40 



20 



•   



rM»"i- 



B 



H Composite square 2 n = 53 

 □ Control n - 421 



n-n-n-.-. 



80 

 60 

 40 

 20 

 



H Composite square 2 n = 298 80 



Q Control n = 400 gQ 



^ 



 Composite square 3 n = 355 

 n Control n = 476 



tlll»i^ -^ ~ ^ 



 Composite square 3 n = 9 

 □ Control n = 422 



B Composite square 3 n = 209 

 n Control n = 241 



Figure 5 



Size-frequency distributions of the composite square-mesh codends and control codend for lAi sand trevally [Pseudocaranx 

 wrighti). (B) red mullet iUpeneichthys porosus), (C) Degen's leatherjacket iTIiamnaconus degeni). iDi school whiting iSillago 

 bassensis). and (E) southern sand flathead iPlatycephaliis hassensisl 



flow anterior to the catch. The location of the panel 

 (approx. 1.1 m from the end of the codend. Fig. 3) 

 was based on a similar design (termed the compos- 

 ite square-mesh panel) that is currently used com- 

 mercially in oceanic prawn trawls in NSW where it 

 has been shown to be very effective in reducing the 

 bycatch of large numbers of small fish (Broadhurst 

 and Kennelly, 1996, 1997). In previous experiments, 

 Broadhurst and Kennelly (1996) and Broadhurst et 

 al. (1999) determined that at this position, there is 

 some displacement of water forwards owing to the 

 twine area and build-up of catch in the posterior sec- 

 tion of the codend. Because small fish are probably 

 using anaerobic muscle power to maintain position 

 in the moving trawl and are quite fatigued when they 

 enter the codend, this displaced water may be suffi- 

 cient 1) to assist them to swim forwards and out 

 through the larger meshes in the panel; and (or) 2) 

 to enable them to reduce their tail-beat frequencies 



and maintain their position in the vicinity of the 

 larger mesh for a longer period, increasing their 

 chances of randomly escaping. The extent to which 

 such a flow facilitated the escape offish in the present 

 study probably depended on their relative size or 

 physiology (or both), because these factors largely 

 influence swimming speed and endurance (Beamish, 

 1978). For example, the bycatch of fast-swimming 

 species, such as sand trevally, and relatively large 

 individuals of school whiting and southern flathead 

 was greatly reduced in the two modified designs. 

 However, although statistically significant, there was 

 only a SO'^ reduction in the numbers of the relatively 

 small Degen's leather jackets in the composite-square- 

 2 codend and a lower nonsignificant 24% reduction 

 in the composite-square-3 codend (Fig. 41; Table 1). 

 A possible explanation for the significant reduc- 

 tion of Degen's leather jacket from the composite- 

 square-2 codend may be the configuration of mesh 



