322 
Fishery Bulletin 99(2) 
0.0 — ' ~ T — 1 — — >- 
25 30 35 40 45 50 55 60 
Fork length (cm) 
Figure 6 
Commercial bottom trawl (472-inch mesh codend) fishery 
selectivity estimates for sablefish (sexes combined) in the 
northern and southern subareas during 1978 and 1996. 
Selectivity curves for other years were intermediate. Esti- 
mates for 1978 are biased owing to commercial trawls in 
the fishery with smaller mesh codends. Smooth lines show 
trends and were fitted to estimates by locally weighted 
regression smoothing (LOESS). 
1.0 
0.8 
^ 0.6 
> 
O 
Q) 
CD 
0.2 
0.0 
20 25 30 35 40 45 
Total length (cm) 
Figure 7 
Commercial bottom trawl (472-inch mesh codend) fishery 
selectivity estimates for Dover sole (sexes combined) in the 
northern and southern subareas during 1978 and 1996. 
Selectivity curves for other years were intermediate. Esti- 
mates for 1978 are biased due to commercial trawls in 
the fishery with smaller mesh codends. Smooth lines show 
trends and were fitted to estimates by locally weighted 
regression smoothing (LOESS). 
as Bayesian priors for selectivity parameters estimated by 
stock assessment models (Metliot, 1990). 
Our estimates of depth distributions reflect conditions 
during the autumn (October-December). Our study did 
not include data collected during other seasons that could 
be used to test hypotheses about seasonal migrations be- 
tween deep and shallow water (Alverson, 1960). Addition- 
al bottom trawl survey data collected at different times 
of the year with a variety of vessels and trawl gears are 
available (Lauth 3 ) and could be used to measure seasonal 
differences in depth distributions. 
Our analysis used data from bottom trawl surveys to 
estimate depth distributions. Our approach might be ap- 
plicable to other types of surveys as long as densities of 
organisms in each strata can be calculated on an relative 
or absolute basis. Although survey gear selectivities can- 
cel out in calculations, it is important that the survey 
gear be relatively efficient for length groups used in cal- 
culations. Otherwise, density estimates used to calculate 
depth distributions may be too variable. 
The assumption that survey bottom trawl selectivities 
are constant with depth is important in calculating depth 
distributions p(d\L) because the proof that survey selec- 
tivities cancel out depends on the assumption. It is possi- 
ble for example, that small fish might evade bottom trawls 
by hiding in rubble or depressions. If there were more 
rubble or depressions in deep water than in shallow wa- 
ter, then bottom trawl survey selectivities would change 
with depth, and depth distribution estimates (as well as 
commercial fishery bottom trawl selectivity calculations) 
would be affected. The magnitude of any possible problem 
would depend on a variety of factors (e.g. the relative 
abundance of small fish at depths with more or less rub- 
ble) and cannot be predicted in general. However, factors 
that affect selectivity of survey bottom trawls (including 
herding, escapement under the footrope, escapement over 
the top of the net, and escapement through meshes) are 
not important in calculating depth distributions, even if 
they depend on fish size, as long as they remain the same 
for all depth strata. 
Sensitivity of selectivity estimates to errors in fishing 
effort data 
Our estimator for fishery selectivities does not depend on 
knowing total effective commercial fishing effort (E y d ) for 
each depth stratum. It does depend on knowing the pro- 
portion of total effective fishing effort (which is propor- 
tional to fishing mortality) in each stratum. This means 
that unreported fishing effort would not affect our calcu- 
lations unless there were differences in depth of fishing 
among fishermen who did and did not turn in log data, 
differences among states in depth of fishing and propor- 
tion of fishermen who submit log data, or differences in 
logbook reporting rates among fishermen who fish in dif- 
ferent areas or at different depths. 
:l Lauth, R. 1998. Personal commun. Alaska Fisheries Sci- 
ence Center, National Marine Fisheries Service, 7600 Sand 
Point Way, BIN C 15700, Seattle, WA 98115-0070. 
