Somerton et al,; Whole-gear efficiency of a benthic survey trawl for flatfisfi 



283 



Arrowlooth flounder 



20 40 60 



Rex sole 



Flatfiead sole 



20 30 



Dover sole 



Length (cm) 



Figure 2 



Proportion caught as a function of body length for four species of flatfish: arrowtooth flounder 

 (Atheresthes stomias), flathead sole {Hippoglossoides elassodon), rex sole (Glyptocephalus zachirus), 

 and Dover sole (Microstomas pacificus). For each species, the fit of the best fitting parametric 

 model is shown with a solid line. However, for arrowtooth flounder (Atheresthes stomias), a 

 nonparametric model, the cubic spline, fitted the data even better and is shown with a dashed 

 line. Relative abundance by size group is indicated by the radius of the circles. 



decreased slightly (0.4 m; Table 2) as bridle length 

 increased, but the decrease was not significant; bridle 

 angle (i.e., the angle between the bridles and the direc- 

 tion of travel (a) decreased significantly with increas- 

 ing bridle length. 



Estimates of /r^ 



Tests for length dependency in the herding process, 

 based on the linear regression of kf^ on fish length, 

 indicated that the slopes were positive in all four cases 

 (Table 3; Fig. 3), but significant only for arrowtooth 

 flounder. Because the significance of the relationship for 

 arrowtooth flounder is primarily due to the conspicu- 

 ously lower value ofk^ at the smallest length class (Fig. 

 3), we consider the evidence for an increase in kf^ with 

 size as credible, but still equivocal. 



Length dependency in the herding process should lead 

 to differences in the size distribution as bridle length 

 is changed; however, size distributions for each spe- 

 cies appeared quite similar for each of the three bridle 

 lengths (Fig. 4) and none of the species had a significant 



(P<0.05) difference in mean size among bridle lengths. 

 Consequently, k^^ was considered as length invariant in 

 the calculation of efficiency for all four species. 



Length-invariant estimates of ^^ were similar for 

 the three sole species, ranging from 0.22 for rex sole to 

 0.24 for Dover sole (Table 4). These values are slightly 

 larger than the estimated value of 0.17 for arrowtooth 

 flounder. The values of the herding coefficient (h), or the 

 herding efficiency in relation to the area swept by the 

 lower bridle, were considerably higher than the values 

 of kj^. For the three sole species, h ranged from 0.53 

 for rex sole to 0.58 for Dover sole. Again, these values 

 were higher than the h estimate of 0.39 for arrowtooth 

 flounder. Thus, roughly 40-50% of the flatfish encoun- 

 tering the lower bridle were ultimately herded into the 

 path of the net. 



Whole-gear trawl efficiency 



Trawl efficiency estimates for arrowtooth flounder, flat- 

 head sole, and rex sole increased with increasing fish 

 length and reached maxima of 0.45, 0.42, and 0.43, 



