made at the footrope because the diameter of cookie and 

 roller gear were known and used as the reference standard. 

 The use of the roller and cookie gear as a reference standard 

 may also impart inaccuracies on the data. Consequently, 

 during future research, reference standards of known length 

 will be attached to the net at key locations. 



Apparent, or measured, distances in front of the 

 footrope were transformed to actual distance using the 

 formula: X = x' + [(B tan a) - A]; where: X is actual 

 distance in front of footrope, x' is measured distance in 

 front of footrope, B is camera height above fish, tan a is 

 tangent of angle at camera between vertical and camera line 

 of sight over cookie gear, and A is the distance camera is 

 located behind cookie gear. 



The quantity and quality of data obtained to date has 

 been impacted by several factors, most significant of which 

 was the viewing distance limitations imposed by water 

 clarity. Viewing distances between 10-15 feet were 

 encountered during the sampling program. This limitation 

 determined the number of replicates and accuracy of 

 measurements made on fish at the periphery of viewing 

 distance. Visual limitations and submersible positioning 

 also prevented an assessment of fish responses to net 

 groundlines and wing ends. 



Another factor which affected the amount and type of 

 data obtained from the video tapes was camera placement. Our 

 analysis was conducted using video tapes obtained by the NMFS 

 during their crab bycatch studies. These tapes had the 

 camera positioned to observe crab rather than finfish 

 behavior. This positioning provided only a limited view of 

 the footrope directly in front of the camera and no view of 

 the area behind the footrope or in the belly of the net. 

 This limitation severely restricted a complete analysis of 

 flatfish behavior. Data generated were mainly restricted to 

 the area in front of and over the footrope. A few 

 observations were obtained from the area behind the footrope 

 when the camera occasionally panned to either side. It was 

 not possible to follow individual fish throughout the capture 

 process into the belly of the net, the area which is probably 

 most important for species separation. 



The final factor which may have influenced the behavior 

 patterns observed was the effect of the submersible on 

 flatfish entering the trawl. Due to the limited data 

 obtained outside the submersible ' s influence, we have been 

 unable to quantitatively assess this problem. However, those 

 observations which were obtained did not indicate major 

 changes in flatfish behavior. The submersible appeared to 

 only influence the fish at a relatively close range. It was 



407 



