Abstract .-Some combinations of 
trawler, trawl, and crew catch fish bet- 
ter than others. Systematic error en- 
ters catch per unit of effort (CPUE) data 
from trawl surveys through such rela- 
tive differences in efficiency among 
sampling instruments. Correcting rela- 
tive fishing power differences can re- 
move bias due to systematic error but 
may also increase the variance of the 
mean CPUE estimate. As a result, the 
overall error of the estimate may actu- 
ally become worse, even when the fish- 
ing power difference is statistically sig- 
nificant. A decision rule specified by the 
mean square error (MSE) of the mean 
CPUE estimate avoids this mistake: 
namely a correction that reduces the 
error in the mean CPUE estimate 
would be applied, a correction that in- 
creases the error would not be applied. 
I describe and demonstrate an algo- 
rithm, based on minimizing the MSE, 
for deciding to correct a fishing power 
difference. The strategy requires that 
a probability density function exist that 
models the CPUE data reasonably well. 
Manuscript accepted 3 March 1998. 
Fishery Bulletin 96:538-546 (1998). 
A decision rule based on the mean 
square error for correcting relative 
fishing power differences 
in trawl survey data 
Peter T. IWHunro 
Resource Assessment and Conservation Engineering Division 
Alaska Fisheries Science Center 
National Oceanic and Atmospheric Administration, NOAA 
7600 Sand Point Way NE, BIN C- 1 5700 
Seattle, Washington 981 15-0070 
E-mail address: peter.munro@noaa.gov 
Trawl surveys are often used to col- 
lect data on catch observations stan- 
dardized by fishing effort (the catch 
per unit of effort [CPUE]), and the 
mean CPUE is often interpreted as 
an index of abundance. Ideally, fish- 
ing power, or fish catching effi- 
ciency, must be held constant in 
trawl surveys lest altered catch 
rates be confounded with changes 
in abundance of fish or inverte- 
brates. Unfortunately, the sampling 
instrument is a complex system 
that includes the vessel, vessel op- 
erators, and fishing gear, all of 
which may vary from survey to sur- 
vey, introducing changes in fishing 
power (Gulland, 1956). Correcting 
fishing power differences seems 
necessary for proper interpretation 
of mean CPUE. However, methods 
have not been established for deter- 
mining if an improved estimate of 
mean CPUE actually results from 
such correction. 
Relative differences in fishing 
power make standardization in 
trawl surveys difficult. Technologi- 
cal changes in fishing gear, as well 
as replacement of older research 
vessels, may affect fishing power 
(Azarowitz, 1981; Byrne et ah, 
1991). Fishing power differences are 
an inherent part of multiple vessel 
surveys. Examples of this type of 
survey are annual and triennial 
surveys in several regions of the 
north Pacific Ocean and Bering Sea 
conducted by the Alaska Fisheries 
Science Center (AFSC) of the Na- 
tional Marine Fisheries Service 
(NMFS) (Harrison, 1992; Weinberg 
et al., 1994; Munro and Hoff, 1995; 
Goddard and Zimmermann 1 ) and 
the International Young Fish Sur- 
vey in the North Sea (Anonymous 2 ). 
Koeller and Smith (1983) reported 
change in a single vessel’s ability to 
measure speed over a 3-year period 
and hypothesized that this may 
have altered its fishing power from 
year to year. Operator effects have 
also been shown to account for fish- 
ing power differences among vessels 
(Munro and Hoff, 1995) and thus 
may be inferred for change in op- 
1 Goddard, P., and M. Zimmermann. 1993. 
Distribution, abundance, and biological 
characteristics of groundfish in the east- 
ern Bering Sea based on results of the U.S. 
bottom trawl survey during June-Septem- 
ber 1991. U.S. Dep. Commer., NOAA, 
Natl. Mar. Fish. Serv., Alaska Fish. Sci. 
Cent., 7600 Sand Point Way NE, Seattle, 
WA 98115-0070, Proc. Rep. 93-15, 324 p. 
2 Anonymous. 1986. Manual for the In- 
ternational Young Fish Surveys in the 
North Sea, Skagerrak, and Kattegat. In- 
ternational Council for the Exploration of 
the Sea, Palaegade 2-4, DK-1261, Copen- 
hagen K, Denmark. ICES Council Meet- 
ing 1986/H:2. 
