Kimura et al. : Stock structure and movement of Anoplopoma fimbria 
473 
Figure 5 
Average movement of tagged and recovered sablefish aggregated by area of tagging. 
Arrows point to the average location where sablefish tagged in a specified area were 
recovered (i.e. where fish were going). Tail of arrow originates in one of the 27 areas 
in which tagging occurred. Tagging did not occur in all areas. 
or vice versa, were excluded from Table 7. For Alaska 
sablefish, the depth of tag recovery generally had the 
same distribution regardless of the depth of tagging. 
For west coast sablefish, recoveries tended to be in 
the same depth zone as the depth of tagging. How- 
ever, few recoveries were made from the relatively 
large number of tags released in the 400-1000 m 
depth zone. 
Results related to ENSO events and 
migration from Alaska to the west coast 
Global-scale El Nino-Southern Oscillation (ENSO) 
events are the strongest climatic feature occurring 
in the Pacific Ocean. Along the west coast it weak- 
ens upwelling and is known to lower food availabil- 
ity and to stunt growth of pelagic species (Bakun, 
1996). To the north, off Alaska, ENSO may have the 
opposite effect, enhancing growth and biological pro- 
ductivity (Beamish and Bouillon, 1993; Bakun, 1996). 
A significant negative effect on sablefish growth due to 
ENSO events was detected (P[ 1 1 | >3.973] < 7.2xl0 -5 , 
df=3666) off the west coast, and a significant posi- 
tive effect on growth was detected (P[ \ t | >2.604] < 
0.01, df=6030) for Alaska sablefish (Table 8). In this 
analysis growth increment data were used only if 
tagging and recovery occurred in the same region. 
Residuals histograms suggested a normal pattern of 
errors for the model fits (Fig. 7). 
We found that sablefish that migrate from Alaska 
waters to the west coast appear to concentrate in 
areas off Vancouver Island, Washington, and Oregon. 
Area 20, between 47°30' and 50°N latitude (Fig. 1), 
is of particular interest because it is an area of mix- 
ing for Alaska and west coast sablefish (Table 6, col- 
umn 20). The question arises whether we can detect 
factors that might influence the migration from 
Alaska to areas 20 and 21 off the west coast. One 
factor that seems to have a positive correlation with 
this migration south is the strength of upwelling 
in these southern areas (Fig. 8, two-tailed, sign. a= 
0 . 01 ). 
Length frequencies from sablefish trap index sur- 
veys (1978-91), aggregated over all years, strongly 
distinguish between fish from Alaska (areas 17 and 
18) and the west coast (areas 20-27) (Fig. 9A). How- 
ever, the length frequencies for fish greater than 60 
cm (Fig. 9B) suggest three groups: Alaska (areas 17 
and 18), west coast north (areas 20-24), and west 
coast south (areas 25-27). The west coast north (ar- 
