662 
Fishery Bulletin 95(4), 1 997 
[1 Control, n = 510 
Length (cm) 
Figure 6 
Size-frequency distributions of discarded stout whiting caught with control and com- 
posite-panel codends from (A) Yamba, (B) Ballina, and (C) size-frequency distributions 
of retained red mullet from Ballina. 
prawns at Port Stephens, Yamba, and Ballina (by 
14%, 5.5%, and 6%, respectively) and for the nonsig- 
nificant increase of 4% at Southwest Rocks (Fig. 3 A). 
By reducing the amount of total discarded bycatch 
and therefore the weight and drag in the codend, the 
trawl with the composite-panel may have achieved 
greater spreads between the otter boards (i.e. an in- 
creased swept area) than did the control, thereby 
covering more of the seabed and capturing more 
prawns. These prawns were probably the same sizes 
as those that we sampled, because Kolmogorov- 
Smirnov tests failed to detect any significant differ- 
ences in prawn sizes between the codends for any of 
the ports (Fig. 4). 
In support of the theory discussed above, there was 
also an increase (although not statistically signifi- 
cant) in the mean numbers of retained octopus at 
Southwest Rocks and Yamba (by 11% and 14%, re- 
spectively), retained red mullet (by 17%) at Ballina, 
and discarded eastern blue spot flathead at Yamba 
and Ballina (by 19.5% and 14.5%, respectively) with 
the composite-panel (Fig. 3, C-D, and I; Table 1). 
