Broadhurst et al.: Effects of square-mesh panels in codends and of haulback delay on bycatch reduction 4 1 7 



the weight of bycatch (mean reduction of 37%) and 

 the number and weight of red spot whiting (by 68% 

 and 58%, respectively) (Figs. 5, A-C; Table 4). This 

 codend did not significantly reduce the weight of 

 prawns (although the mean catch was 3% lower) nor 



1.0 



0.6- 



02- 



I) 



Oe 



Bo • o I 



No delay 



10-15 s delay 



Figure 3 



Differences in mean catch (±SE) be- 

 tween the control and 85-mm-long 

 codends in experiment 2 (effects of 

 haulback delay I for no delay and a 10- 

 15 s delay in haulback (<, >, and = in- 

 dicate direction of differences): (Al the 

 weight of discarded bycatch; (Bi the 

 weight of prawns (Penaeus plebejus); 

 (C) the number of discarded red spot 

 whiting <Sillago flindersi); (D) the 

 weight of discarded red spot whiting 

 iSi/lago flindersi); and (El the weight 

 of retained cuttlefish tSepia sp.). * = 

 P<0.05; ** = P<0.01. 



any of the other variables (Fig. 5E; Table 4). Two- 

 sample Kolmogorov-Smirnov tests failed to detect 

 any significant difference in the size compositions of 

 prawns between the 85-mm-long and control codends 

 (Fig. 6A). The sizefrequency distributions for red spot 

 whiting, however, were significantly different be- 

 tween these codends (Fig. 6B). 



Discussion 



The results of these experiments show that codends 

 with square-mesh panels have the potential to re- 

 duce bycatch of nontarget individuals (see also 

 Robertson and Stewart, 1988; Briggs, 1992; Fonteyne 

 and M'Rabet, 1992; Broadhurst and Kennelly, 1994). 

 By experimentally examining the effects of haulback 

 delay, we have also quantified, for the first time, the 

 effects that such operational procedures can have on 

 the escape of some individuals. 



In previous experiments, Broadhurst and Kennelly 

 ( 1994, 1995) showed that square-mesh panels in the 

 anterior sections of codends were effective in releas- 

 ing small fish (i.e. mulloway, Argyrosomus holo- 

 lepidotus) from prawn trawls in the Hawkesbury 

 River. These results were attributed to the differences 

 in behavior of fish and prawns in their response to 



 Control. n= 193 

 □ 85 mm long, n = 136 



I I I . I_fcL 



jOlJJ 



Control, n = 228 

 Q 85 mm long, n- 118 



fcLtl. 



Length (cm) 



Figure 4 



Length-frequency distributions of red spot whiting (Sillagv 

 flindersi) from the control and 85-mm-long codends for experi- 

 ment 2: (A) during no delay in haulback and (B) a 10-15 s delay 

 in haulback. 



