26 
Fishery Bulletin 11 6(1) 
Figure 2 
Standard deviation of vessel heave (SDH), in centi¬ 
meters, at the starboard trawl block, plotted against 
wave height, in meters, estimated by the captain of FV 
Vesteraalen. Data were obtained from an experiment 
conducted in the eastern Bering Sea during September 
2003. The solid line shows the regression of SDH on 
wave height, excluding consideration of vessel heading 
relative to wave direction. 
wave height (R function lm) and tested the significance 
of the slope, using standard methods. However, it is 
possible that factors, such as a temporal change in 
population structure, could influence this relationship. 
Therefore, to better isolate interannual changes due to 
waves from more gradual changes, we calculated the 
deviations between the survey abundance estimates 
and the abundance estimates produced by the yellow- 
fin sole stock assessment model (Wilderbuer et al. 1 ), 
then modeled the deviations as a linear function of. 
mean wave height. The estimates of abundance from 
the models we used are slightly different from those 
reported in Wilderbuer et al. 1 because the version of 
the model used for stock assessment included an en¬ 
vironmental correlate (mean bottom temperature). 
Because we were examining the influence of another 
environmental correlate, we fitted the model without 
the temperature effect to produce the model estimates 
of survey abundances used in our analysis. 
Results 
Vessel motion and wave height 
Vessel motion, expressed as the SDH, increased signifi¬ 
cantly with the estimated wave height (Fig. 2, Table 1). 
However, SDH was not significantly related to either 
the numeric or the binary measure of the alignment of 
vessel heading and wave direction. 
Table 1 
The relationship between the standard deviation of 
vessel heave at the trawl block with the 3 predictor 
variables from results of an experiment conducted in 
the eastern Bering Sea during September 2003. The 
predictor variables are 1) wave height (summed val¬ 
ues of wave and swell heights estimated by the vessel 
captain), 2) alignment (alignment of vessel course and 
wave direction, i.e., absolute value of the cosine of the 
angle between the 2 directions) and 3) relative direc¬ 
tion (vessel course either away from, ±180°, or into the 
direction of wave travel). 
Coefficient 
P-value 
Intercept 
14.10 
0.071 
Wave height 
11.98 
<0.001 
Alignment 
-2.86 
0.259 
Relative direction 
0.07 
0.500 
Bridle and footrope distances off bottom 
Off-bottom distance measurements along the lower bri¬ 
dles and footrope oscillated with a period nearly identi¬ 
cal to that of vessel motion. For example, on one haul, 
which had a wave height of 5.3 m, the heave oscilla¬ 
tion period (5.13 s) was nearly identical to that for the 
periods measured at all trawl positions, both along the 
bridles (50 m, 5.16 s; 40 m, 5.12 s; 25 m, 5.12 s) and 
along the footrope (wing, 5.16 s; corner, 5.16 s; center, 
5.16 s). At low levels of SDH, however, spectral peaks 
were harder to identify, and the variability among po¬ 
sitions increased. This finding indicates that, at least 
for relatively high waves, vessel motion is detectable 
throughout the trawl and substantiates the notion that 
surface waves influence trawl performance. 
The magnitude of the trawl oscillatory movement, 
expressed as the standard deviation of off-bottom dis¬ 
tance, increased significantly with SDH at all trawl po¬ 
sitions, except at the wings. The magnitude of the in¬ 
crease (slope of the regression line) varied with position 
(Fig. 3, Table 2). The smallest increase in SDH occurs 
at the wings, consistent with the design of the survey 
trawl that produces the maximum bottom contact, and 
presumably strongest downward force at the wing tips. 
Away from the point of maximum bottom contact, both 
forward along the bridles and aft along the footrope, 
the standard deviation of off-bottom distance increases 
as the constraints on vertical motion, such as tension 
and bottom contact, lessen. At the 50-m bridle position, 
where the bridle rarely contacts the bottom, the oscil¬ 
lations were the largest (Table 2). 
The mean distance off bottom also varied among 
trawl positions; however, its increase with SDH was 
only significant at the 40-m bridle position (Fig. 4, Ta¬ 
ble 3). At the wing and the 25-m bridle positions, the 
downward forces keep the mean off-bottom distance 
