432 
Fishery Bulletin 113(4) 
Starboard Middle Port 
Figure 1 
(A) Schematic illustration of the bottom trawl used to collect video images of yellowtail flounder 
(Limanda ferruginea) in June 2010 on the southern Grand Bank off eastern Newfoundland and (B) 
an example of the grid used for the analysis of video footage. The gray square in panel A indicates 
the area observed by the camera (3 m highx4 m wide field of view), which was ~5 m from the 
center of the footgear (and 3 m from seabed; the headline was 4 m in front of the footgear). Video 
frames, as shown in panel B, were divided into 100 squares, with the middle 4 columns represent- 
ing the center of the footgear and the 3 columns on either side representing port and starboard 
of the footgear. 
and durations of tows were in a range of 2-3 h. Tow 
direction was decided by commercial operations and 
was different for each tow. The height of the headline 
was approximately 3 m and was recorded with a sensor 
(Marport Stout Inc., Snohomish, WA) attached to the 
headline on all tows without the use of camera docu- 
mentation during the cruise. The door spread ranged 
from 113 to 123 m and was measured with spread sen- 
sors (Marport Stout Inc.) placed on the trawl doors. 
During 5 tows, more than 12 h of video footage of 
flatfishes were collected at depths of 70-82 m, at bot- 
tom temperatures ranging from 0.6°C to 1.2°C. A new 
high-definition, self-contained underwater camera sys- 
tem without artificial lights (72% accuracy for identi- 
fication of yellowtail; for details, see Underwood et al. 
[2012]) was used to observe approximately a quarter 
of the footgear (i.e., the central region, excluding the 
wings; Fig 1). Observational techniques, such as the 
use of artificial lights, may influence the behavior for 
some species (Walsh and Hickey, 1993; Weinberg and 
Munro, 1999). Therefore, time of year and day was cho- 
sen to optimize natural underwater light for the cam- 
era and so that artificial lights were not needed. The 
camera system was attached to the inside of the cen- 
ter of the headline (Fig. 1A). The straight-line distance 
from the camera to the center of the footgear was ~5 
m, providing a field a view 3 m high by 4 m wide of the 
lower first belly and the central part of the footgear. 
