Salthaug: Adjustment of commercial effort for Godus morhua 
341 
CD 
O' 
3 
U) 
in 
2,4e6 
2e6 
1 ,6e6 
1 ,2e6 
8e5 
4e5 
Year 
Figure 2 
Yearly CPUE in kilograms per hour and trawlable biomass from VPA estimates for 
original (A) and adjusted (B) CPUE. In B the effort is adjusted to 1972-level bv using 
the learning curve estimated in Figure IB. 
An important assumption in this 
work is that a constant fraction of 
the stock is caught when the fleet 
lias a constant efficiency. Many fish 
stocks, including the Atlantic cod 
stock in the Barents Sea, have a 
seasonal cycle in availability to 
the fishing fleet because of regular 
spawning and feeding migrations 
that cause large peaks in the den- 
sity of fish. The assumption above 
is therefore violated if the spatial 
and temporal distribution of effort 
in the fleet during the year chang- 
es between years. Correction for 
changes in the seasonal distribu- 
tion of effort between years are, 
however, described by e.g. Gulland 
(1983) and Gavaris (1980). Around 
1979 quotas and fishery regula- 
tions were introduced in the Nor- 
wegian fisheries. As seen from the 
outliers in Figure 1, it is obvious 
that something affected the epsf af- 
ter 1980. Dramatic responses are 
expected when a fishery is regu- 
lated (Hilborn and Walters, 1992). 
Respones to regulations may in- 
clude changes in the effort distri- 
bution during the year, changes 
in the patterns of competition 
and cooperation with other vessels, 
increased misreporting, and in- 
creased amounts of discards. 
In addition to uncertainties in the 
reported catch and effort from the 
vessels, there are also uncertainties 
in the VPA estimates of stock size 
(see Ulltang, 1977). The trawlable 
fraction of the youngest year class- 
es also changes within and among 
years due to individual growth 
and yearly differences in growth. 
The constant retention probabili- 
ties used in this work to select the 
youngest age classes from the VPA 
estimates, and the use of only one 
single stock estimate per year is 
thus, perhaps, an over simplifica- 
tion. To get a better estimate of the 
learning curve, epsf can be estimat- 
ed based on averaging effort and 
catch over shorter time periods than 
in this work. This requires, however, 
advanced modelling of the selection 
process, and possibilities for obtain- 
ing estimates of total abundance 
more than once a year. 
The VPA estimates and the catch and effort data used 
in this work are not totally independent from each other. 
A time series of commercial CPUE from the Norwegian 
trawler fleet was used, together with other CPUE time se- 
ries, to calibrate the most recent VPA estimates of the old- 
