334 
Fishery Bulletin 96(2), 1 998 
caught by angling at the surface at night gave val- 
ues ranging from 0.03 to 0.09 (rough approximations 
from fish length and swimbladder dimensions). Be- 
cause these fish had ascended several meters to take 
artificial bait at the surface, we chose to use the more 
accurate determination of 0.03 for pollack and saithe, 
Pollachius virens, calculated from data in Foote 
( 1985). For converting fish length to weight we used 
W = 0.0057 L 3 026 , determined from lengths and 
weights of hake captured in the 1992 NMFS trawls, 
males and females combined. 
Calculated swimbladder size distributions were 
modified to account for two sources of variability ex- 
pected in the natural hake population. First, for vari- 
ability of bladder radii due to the above mentioned 
effects of food and gonadal products, we approxi- 
mated the distribution of swimbladder volume to fish 
volume as follows: 30% of the fish had the mean value 
of 0.03 whereas 10% of the fish had swimbladders 
50% smaller, 20% had swimbladders 25% smaller, 
30% had swimbladders 25% larger, and 10% had 
swimbladders 50% larger than the mean. These val- 
ues were based on rough approximations of informa- 
tion contained in Ona ( 1990). Second, for variability 
in the fish weight (W) to length (L) relationship, we 
approximated a normal distribution of W on L, where 
20% of the fish had the mean W given by the regres- 
sion equation W = 0.0057L 3 026 , and 17% were 10% 
heavier (lighter), 12% were 20% heavier (lighter), 7% 
were 30% heavier (lighter), and 4% were 40% heavier 
(lighter) than the mean. 
Abundance and biomass estimates 
Abundance data for hake were determined by sum- 
ming n, the column vector of fish number to obtain 
0. Biomass was determined by summing the prod- 
uct of n, the column vector of fish number, and the 
expected weight of each fish (r to L and L to W, rela- 
tionships above). Biomass values were averaged for 
the NRL stations, grouped according to NMFS fish- 
eries assessment regions: Canada South (Sta. 1); 
Southern Columbia (Sta. 4, 5, 6, and 7); and Eureka 
(Sta. 8, 9, 10, and 11) and compared with reported 
surface biomass of hake obtained from the NMFS 
high frequency acoustic survey (Fig. 10 in Dorn et 
al„ 1994). 
Results 
Volume scattering data 
Stations 1 (Fig. 3), 7 (Fig. 4), and 11 (Fig. 5) were 
selected to show representative volume scattering 
because they are on the slope in 900 to 1500 m of 
water and encompass the latitudinal range of sta- 
tions covered by the NRL survey (Fig. 1). Stations 7 
and 11 had day and night data; station 1 had only 
day data. In each image, S v is shown as a function of 
frequency and depth. For shallow stations, bottom 
echoes are black. In all images, what appears to be 
intense scattering near the surface is actually the 
direct blast from the shot and close-in surface scat- 
ter. Most stations show three layers of scatterers. 
All stations had a strong scattering layer between 
100 and 500 m both day and night. During the day 
the layer was strongest at 200 to 300 m with a reso- 
nance peak at about 1,000 Hz. This layer becomes 
progressively deeper to the south, being centered at 
200 m at station 1, 220 m at station 7, and 320 m at 
station 11. This layer was presumed to be composed 
of Pacific hake on the basis of a prior consideration 
of the most likely abundant scatterers. 
A second layer occurred at about 50 to 200 m at 
night and scattered sound at 3 to 5 kHz, frequencies 
well above the strongest (peak) scattering from the 
hake (Figs. 4B and 5B). The depth and frequency of 
occurrence of this layer matches the findings of 
Kalish et al. (1986) who provide a detailed study of 
the composition and scattering strengths of mesope- 
lagic fish off Oregon. Therefore, we conclude this 
nighttime layer to be mesopelagic fishes that, dur- 
ing the day, should be deeper (near 200 to 300 m), 
resonant above 5,000 Hz, and not detectable by our 
measurements. This is supported by weak daytime 
scattering above 4,000 Hz at 250 m (Fig. 5B), which 
may be the low-frequency end of this daytime meso- 
pelagic layer. 
A third layer, beginning at about 1,000 m and ex- 
tending to the end of the data record was evident in 
all data extending beyond 1,000 m. On the basis of 
this layer’s resonance near 2,000 Hz and its location 
in depth, Nero et al. (1997) concluded that this scat- 
tering layer was caused by pelagic grenadiers. In 
subsequent analysis we considered only the low-fre- 
quency scattering from the hake — those scattering 
layers that are above 600 m and have a resonance 
below 2 kHz. 
The acoustic resonance of the hake swimbladders 
is evident in the layer strength data for both day and 
night (points, Figs. 6 and 7). During the day, the 
majority of the scattering occurred in a fairly nar- 
row depth range (Figs. 3-5); therefore only one set 
of layer strength data was calculated at each station 
(Fig. 6). During the night, the fish were more dis- 
persed, and therefore the hake layer was separated 
into sublayers, and layer strengths were calculated 
for each (Fig. 7). Day layer strengths in the north at 
station 1 peaked at a slightly lower frequency (1,050 
