248 
Fishery Bulletin 115(2) 
bia waters but have varied widely over time and lo¬ 
cation within Alaska waters (Hanselman et ah, 2015); 
they have not been estimated for the U.S. West Coast 
(California, Oregon, and Washington). Our results may 
have been biased toward higher or lower dispersal than 
indicated by tag returns, depending on spatial differ¬ 
ences in reporting rates. Third, geographic or temporal 
variation in fishing effort can bias the interpretation 
of migration patterns. As noted above, the recently 
increased effort in deep slope waters off Oregon ap¬ 
peared to impact the temporal likelihood of recaptures 
from the deeper sites of tagging set 2. However, recap¬ 
tures in a nearly continuous contour along the slope 
from northern California to the Aleutian Islands refiect 
the geographic continuity of fishing effort throughout 
sablefish habitat. Finally, the tagging process in itself 
can impact fish biology. McFarlane and Beamish (1990) 
found that the presence of a Floy tag reduced growth 
and delayed maturity in sablefish. In our study, most 
of our comparisons were between different groups of 
tagged fish—comparisons that could be assumed to be 
valid if all were impacted similarly by the presence of 
a tag. Our calculated growth rates may have been un¬ 
derestimates, however. 
Additional caveats concern the sampling design of 
our study. Fish were not collected and tagged from a 
broad region of depths between zones 2 and 3 (depths of 
650-1110 m). It is possible that some of our conclusions 
regarding depth-related movements could be modified 
if these depths had been included in sampling. In ad¬ 
dition, with the growth model that we developed, there 
was the assumption of a smooth transition of growth 
rates from shallower to deeper habitats. It is possible 
that the decline in growth with depth instead follows a 
step function that matches transitions in habitat, such 
as those transitions associated with the OMZ or other 
parameters of habitat quality. 
Implications for management 
Our results concur with those of prior comparisons of 
northern and southern stocks of sablefish, particularly 
in the strong ‘resident’ behavior of Oregon fish. How¬ 
ever, fish that did disperse often travelled thousands 
of kilometers away from Oregon, primarily into Gulf 
of Alaska waters. Within the northern stock there is 
extensive movement throughout the Gulf of Alaska 
and Bering Sea (summarized by Echave et ah, 2013, 
and Hanselman et ah, 2015), but there is more limited 
migration to waters within the range of the southern 
stock. Thus, overall, the 2 stocks are geographically 
distinct, which supports the current stock separation 
for management, but there is sufficient interchange to 
prevent genetic distinction. Because tagging studies to 
date have resulted in removal of fish from the popula¬ 
tion, movement patterns of individuals over their life¬ 
times and the extent to which dispersers return to the 
location of tagging are unknown. Likewise, movement 
patterns are unknown for the many fish that were 
caught near their tagging location after many years 
at large. It might be hypothesized that the northward 
dispersers in this study were northern fish that had 
made a temporary excursion to the south, but the dis¬ 
persers tended to be smaller in size at tagging, sug¬ 
gesting they were spawned in Oregon. Future studies 
with archival and satellite tags (Echave et ah, 2013) 
can provide insight into the extent to which fish make 
repeated movements from one region of the population 
to another. If the movement is a one-time event, then 
those fish will be lost to their respective stocks. 
The recaptures from tagging set 2 indicated that 
warm surface temperatures may increase discard mor¬ 
tality rates for small fish. This outcome may be impor¬ 
tant in a management context because smaller fish are 
more likely to be discarded at sea in favor of retention 
of more valuable larger fish. Our results are likely to 
reflect a best case scenario, in which fish were captured 
by pots, carefully handled and released, and exposed 
to relatively moderate temperatures even in the high 
treatment (15.3-17.8°C). Surface temperatures during 
extreme El Nino-Southern Oscillation conditions as 
well as those predicted with climate change will likely 
exceed those of our study. Effects of capture depth on 
discard mortality rates were confounded in our study 
with fishing effort. The reduced recapture rates from 
depth zone 3 in relation to depth zone 2 may have 
largely reflected the much lower fishing effort within 
depth zone 3, particularly during the initial post-tag¬ 
ging period. A more evenly balanced effort to recover 
tagged fish would be necessary to discern a depth effect 
on discard mortality. 
The potential trade-off between growth and longevi¬ 
ty along the slope depth gradient may be altered by fu¬ 
ture climate change effects, such as predicted increases 
in hypoxic conditions in coastal habitats and shoaling 
of the oxygen minimum layer (Pierce et ah, 2012; Gilly 
et ah, 2013). The capacity of sablefish to tolerate low 
oxygen conditions may benefit their future survival but 
at an unknown cost to individual growth rates. Low 
exploitation rates in very deep slope habitats may help 
to provide a spatial refuge, allowing accrual of the eco¬ 
logical benefits of a long life span. 
Acknowledgments 
T. Rippetoe assisted with tagging. R. Miller created the 
map of recapture locations. W. Satterthwaite provided 
statistics advice. N. Maloney, W. Mitton, K. Echave, 
S. Malvitch, S. Flores, and N. Wilsman assisted with 
recovery of recapture data. G. Hettman assisted with 
project design. M. Freeman provided Oregon Depart¬ 
ment of Fish and Wildlife data on fishery effort by 
depth. Funding was provided in part by NOAA Fisher¬ 
ies and the Pacific States Marine Fisheries Commis¬ 
sion. We are indebted to the fishermen who provided 
the tagging cruises and to the crew of over 300 com¬ 
mercial fishing vessels who reported tag recaptures. We 
especially thank B. Eder for additional support in all 
aspects of this study. 
