245 
Sogard and Berkeley: Movement, growth, and survival of Anoplopoma fimbria off Oregon 
Table 4 
Results of nonlinear regression for estimating recapture size of sablensh (Anoplopoma fimbria) as a function of 
initial size, time at large, sex, depth at initial capture, depth at recapture, and recapture gear. Growth data were 
available for 398 females and 203 males. Estimated parameter coefficients are listed along with their standard 
errors (SEs) and 95% confidence intervals (CIs). 
Parameter 
Estimated coefficient 
SE 
Lower 95% Cl 
Upper 95% Cl 
Initial size (fork length) 
-0.002651 
0.000098 
-0.002843 
-0.002459 
Time at large (days) 
-0.000223 
0.000016 
-0.000254 
-0.000193 
Sex (0=female, l=male) 
-0.539742 
0.048204 
-0.634437 
-0.445048 
Initial depth 
-0.001105 
0.000108 
-0.001317 
-0.000893 
Recapture depth (m) 
-0.000565 
0.000090 
-0.000741 
-0.000389 
Recapture gear (0=fixed, l=trawl) 
-0.251627 
0.049858 
-0.349570 
-0.153684 
Table 5 
Coefficient of determination (r^) and Mallows’s Cp statistics for nonlinear regressions in 
estimating recapture size of sablefish (Anoplopoma fimbria) as a function of models includ¬ 
ing 1-6 independent variables. The model with the highest and lowest Cp values for 
each subset number of input variables is noted with an X, indicating variables included 
in the model. 
Number of input variables 
6 
5 
4 
3 
2 
1 
Cp 
6.0 
33.9 
81.3 
332.5 
687.4 
900.5 
r2 
0.789 
0.776 
0.756 
0.693 
0.577 
0.503 
Independent variable 
Initial size 
X 
X 
X 
X 
X 
X 
Sex 
X 
X 
X 
X 
X 
Initial depth 
X 
X 
X 
X 
Time at large 
X 
X 
X 
Recapture depth 
X 
X 
Recapture gear 
X 
For fish initially captured in depth zones 1 and 2, 
recapture depths indicated extensive movement within 
the upper slope region, but limited recaptures at depths 
within the OMZ. Only 10% of the total recaptures from 
zones 1 and 2 were at depths >600 m. Low oxygen en¬ 
vironments may contribute to reduced growth rates but 
enhance longevity through reducing oxidation damage. 
Cailliet et al. (2001) found that within the speciose 
rockfishes iSebastes spp.), deeper dwelling species had 
much longer life spans than shallower dwelling species. 
They suggested that the reduced metabolic rates of fish 
living in deep, low oxygen waters promote longevity by 
reducing exposure to oxidative stress. Thus, there may 
be a trade-off between growth and longevity for fish 
living above the OMZ in contrast with those fish living 
within it. This trade-off is suggested by the contrast 
between size and age distributions with depth observed 
for female sablefish by Head et al. (2014). 
In our study, the sex ratio became increasingly bi¬ 
ased toward females as depth increased, reaching 81% 
in depth zone 3 (1112-1225 m). Beamish et al.^ ob¬ 
served a further skewed sex ratio in extremely deep 
waters off British Columbia, with 93% females in wa¬ 
ters deeper than 1800 m. On the basis of our results 
and those of Beamish et al.^ and Head et al. (2014), 
sablefish residing along the deep slope are primarily 
females with minimal growth, minimal propensity to 
disperse, and extended life spans. A disproportionate 
contribution of older or larger fero.ales to reproductive 
success has been documented for many long-lived spe¬ 
cies (Hixoe et al., 2014). Rodgveller et al. (2016) found 
that older female sablefish in Alaska produced larger 
eggs and had an earlier start to the spawning season, 
although relative fecundity did not increase with age. 
In addition to these maternal effects on reproduction, 
there is clearly a potential for long life spans to buffer 
against longevity overfishing, or the removal of large 
numbers of older fish as described by Beamish et al. 
(2006). The relative inaccessibility and lower fishing ef¬ 
fort in deep slope habitats suggests that they provide 
