location and dimensions of the net using Loran C fixes. We 

 documented important net characteristics and the catch in the nets 

 using still photography and color video with onsite narration by 

 the scientific observer. The primary observations made were on net 

 vertical profile, net integrity, fouling, species caught in the 

 net, catch frequency by species, and fate of those species caught. 



The second and third year work quantatively surveyed Jeffreys 

 Ledge and Stellwagen Bank, respectively. Jeffreys Ledge is a 

 fishing area that is well-defined on a chart by depth contours. We 

 identified eight survey areas on Jeffreys Ledge as sites that 

 commercial gillnet fishermen traditionally fished over the past 10- 

 15 years (Fig. 1). These sites were identified through interviews 

 with gillnet, dragger, and recreational fishermen who fish 

 Jeffreys. Stellwagen Bank is not easily defined by depth 

 contours; therefore we identified this area only by the location of 

 the traditional fishing grounds. Five survey areas were identified 

 as traditional commercial gillnet fishing areas on or near 

 Stellwagen Bank (Fig. 1). 



The number of submersible transects in each area was 

 determined by the number of submersible dives available in a given 

 year and the relative size of each specific traditional gillnet 

 area. Each submersible transect (ideally) was planned to be two 

 nautical miles in length, but segmented into four half-mile legs, 

 each requiring a change of direction. Each transect began at the 

 periphery of the study area and went from a deeper depth to a 

 shallower depth for efficient observational use of the submersible. 

 The periphery of each area was divided into 1/4 nautical mile 

 segments that were numbered consecutively (Fig. 2). We used a 

 table of random numbers to determine at which segment the transects 

 would start. 



The submersible followed the ideal transect as close as 

 possible. Position and courses were provided by the surface 

 support vessel through frequent fixes of the submersible location. 

 We decided that if we found many gillnets, we would modify the 

 submersible 's course alternatively left or right to survey, using 

 full video documentation, that portion of the gillnet to its 

 termination. We would then assume that we had surveyed 50% of each 

 gillnet found. The submersible would then resume a course parallel 

 to its original course and complete the transect. If one or two 

 gillnets were found, the submersible would survey the total net and 

 then resume a parallel transect. 



During the 1985 and 1986 surveys, we returned to two nets 

 found the first year (1984) on Jeffreys Ledge so that more data 

 could be gathered on their profile, catch, and likely impact. 

 These were ghost nets whose locations were known to sport 

 fishermen. They were not included in the quantitative survey. 

 Another survey was carried out in July 1986 using the unmanned 



432 



