476 
Fishery Bulletin 96(3), 1998 
o 
o 
00 
o 
— o 
C CD 
D 
O 
o O 
o 
o 
o 
C\J 
...ill 
ii„ 
-20 0 20 40 
Alaska residuals 
o 
o 
00 
o 
o 
CD 
O 
o H 
o 
o 
C\J 
..ll 
[III,.. 
-40 -20 0 20 40 
West coast residuals 
Figure 7 
Histograms of residuals from fits to the growth model used to detect ENSO growth ef- 
fects suggest that residuals from the model fits are approximately normally distributed. 
Figure 8 
Plot showing correlation of the number of sablefish tagged off Alaska 
and recovered off the west coast (dash) with the strength of upwelling 
off the west coast (solid). One year was subtracted from the year of 
tag recovery, allowing time for the tag recovery to take place. The 
Bakun upwelling indices were mean coastal summer upwellings 
(April-August) from 42-48°N latitude (Mason and Bakun, 1986). 
to be a difficult and ineffi- 
cient for long migrations. 
Also, it would seem that 
such migrations would take 
time, but several authors 
have noted that for tagged 
sablefish, there appears to 
be little relation between 
time at liberty and distance 
traveled (Bracken, 1983; 
Dark, 1983; Beamish and 
McFarlane, 1988). A second 
possibility is that sablefish 
are using the dominant 
ocean circulation patterns 
to redistribute themselves. 
The dominant circulation 
patterns in the northeast 
Pacific — the counter-clock- 
wise Alaska Gyre (and Alas- 
ka Current) and the clock- 
wise Central Pacific Gyre 
(and California Current) — suggest both a 
physical basis for separation and a mecha- 
nism for partial exchanges between Alaska 
and west coast stocks (Fig. 1). North of 50°N 
latitude, all life stages of sablefish occur in 
the counterclockwise rotating gyre. In con- 
trast, south of 50°N latitude, adults live in 
a northward flowing undercurrent, and pe- 
lagic juveniles and benthic subadults live 
in the southward flowing surface current 
(i.e. the California Current). 
Tag recoveries from midocean seamounts 
come from all areas of tagging and appear 
to demonstrate that sablefish routinely mi- 
grate on (and mix on) open ocean currents 
from release areas along the continental 
slope (Fig. 3). The Alaska Current (and 
Gyre) could provide a passive means for 
sablefish to make long migrations through- 
out the Gulf of Alaska. Similarly, the Califor- 
nia Current (flowing offshore and southerly 
on the surface) and the California Undercur- 
rent (flowing nearshore and northerly at 
depth) could provide the physical conveyance 
for sablefish to migrate passively up and down 
the North American west coast. 
Our findings concerning exchanges among areas 
in Alaska stocks corroborate earlier findings (Sasaki, 
1985; Fujioka et al., 1988; Heifetz and Fujioka, 1991). 
Perhaps the relative strength and general timing of 
prevailing currents available to Alaska sablefish are 
what allow them to migrate farther and with greater 
frequency than west coast sablefish (Fig. 4). The lati- 
tudinal range of the Alaska population is only 50- 
60°N, half that of the west coast population which is 
30-50°N. Perhaps this narrower latitudinal range 
also facilitates migrations. 
Earlier, we described how migrations from Alaska to 
the west coast are reciprocated by migrations from the 
west coast to Alaska. Although this is true, sablefish 
