338 
Fishery Bulletin 96(2), 1 998 
cies of mesopelagic fish ( Symbolophorus calif orn- 
iensis, Tarletonbeania crenularis, and Diaphus theta ) 
and northern anchovy ( Engraulis mordax ) off the 
coast of Oregon at depths corresponding to our acous- 
tic layer depths and had ESR between 0.1 and 0.7 
cm. Thus, we expect that the fish with ESR between 
0.25 and 0.7 cm in the present study are most likely 
these species. 
For a more complete comparison of the trawl and 
low-frequency acoustic data, the lengths of hake 
caught in trawls were converted to ESR by using the 
previously mentioned weight-length regression and 
the swimbladder-volume-to-fish-volume ratio. The 
conversion included the variability in both relation- 
ships so that these ESR would be equivalent to the 
measured ESR. Figure 9 shows the resulting trawl 
estimates of ESR. The estimated distributions agree 
well with the measured ESR in the north at La 
Perouse, Heceta North, and Heceta South but not at 
the southern stations of Sebastian and Mendocino. 
The trawl data show a clear north-south trend in 
fish size (Figs. 2 and 9), whereas the acoustic data 
(Fig. 8) indicate that ESR, and presumably hake body 
size, do not change with latitude. The NRL low-fre- 
quency acoustic measurements did not show the 
same north-south trend in fish size that was appar- 
ent in the trawl data. 
A summary of the number of swimbladders for 
small, 0.25-0.7 cm, and large fish, 0.8-2. 5 cm, de- 
termined from the low-frequency acoustic data is 
given in Table 2, which includes information on the 
peak depth at which scattering occurred and the 
depth range of each layer. The density of 
small fish was much greater at night than 
during the day because these fish probably 
migrated to below 200 m during the day 
where they were resonant well above 5 kHz. 
For the large fish (presumably hake), at all 
stations where day-night measurements 
were made, densities were determined to be 
higher during day than at night. These 
higher daytime densities likely occurred 
because some hake migrated above 50 m 
depth at night and were missed because of 
the signal saturation caused by the explo- 
sion. (Measurements with an upward-look- 
ing system indicated that hake migrated to 
within several meters of the surface at some 
stations at night 2 ). 
The quasi-synoptic NRL and NMFS mea- 
surements provided an opportunity to com- 
pare two independent measures of the sur- 
face density of hake for several large regional 
areas of the U.S. West Coast. Only daytime 
density estimates were compared because 
data suggested that hake near the surface at 
night were missed by NRL acoustic mea- 
surements. Daytime surface density esti- 
mates for both surveys are given in Table 3. 
The estimates differ most for Canada South 
( NRL station 1 ) where NRL saw an abundance 
of fish over bottom depths near 1,000 m 
whereas the NMFS acoustic-midwater trawl 
survey saw approximately 1/4 that number. 
NRL station 1 was within the region cov- 
ered by the NMFS survey tracklines (Dorn 
et al., 1994) and was sampled by NRL 5 days 
2 1992. Naval Oceanographic Office (NAVOCEANO), 
Acoustics Div., Stennis Space Center, MS, 39522- 
5001. Unpubl. data. 
Figure 6 
Layer strength data (open circles) for the daytime. Curves show in- 
verse analysis fit to the data. 
