O'Malley et al.: Effects of exploitation evident in age-based demography of Pristipomoides flavipinnis and P. auridlla 
333 
small (<3°) and unlikely to affect growth, the Mariana 
Archipelago spans 7° of latitude. An experiment properly 
designed (with use of replicate and adjacent areas that 
have different fishing exposure) to evaluate the effects of 
fishing is required to investigate the potentially confound¬ 
ing environmental factors across the Mariana Archipelago. 
Despite the benefits of the direct measurement of natu¬ 
ral mortality, this estimate should still be met with some 
caution because of some assumptions of our modeling 
approach. Along with assumptions of equilibrium and of 
natural mortality being constant over age and time, we 
also assumed that the natural mortality estimated for the 
unfished areas can be transferred as the value of natural 
mortality in the fished areas. There are ecological reasons 
why the transferability assumption may be untenable. In 
the unfished areas, predation and competition may differ 
from that in the fished areas, resulting in a larger estimate 
of natural mortality (Walters et al., 1999; Macpherson 
et al., 2000; Martell et al., 2005). This larger value for nat¬ 
ural mortality, in turn, may result in underestimation of 
fishing mortality for the fished areas if the back-calculation 
is made from total mortality. There will always be a risk of 
providing precise but inaccurate values of natural mortal¬ 
ity from a direct method; however, treating natural mor¬ 
tality as a constant would provide overly confident total 
mortality intervals for the fished areas and bear the same 
risks of inaccuracy and transferability. 
Fishing effects, assessment, and management 
The most prominent effect from fishing on these species is 
the truncated age composition. This is not unexpected in 
commercial fisheries and is not necessarily an indication 
of overfishing by itself. However, there are consequences 
of age truncation, such as a reduction in a population’s 
reproductive potential due to a reduced spawning stock 
and age-dependent maternal effects (Trippel et al., 1997; 
Berkeley et al., 2004; Hsieh et al., 2010). Older females 
spawn higher quality eggs, resulting in higher larval sur¬ 
vival than that of their younger counterparts (Vallin and 
Nissling, 2000; Bobko and Berkeley, 2004; Sogard et al., 
2008). They also spawn over a longer period, increasing the 
probability of their larvae matching optimal environmen¬ 
tal conditions that would promote the resiliency of a pop¬ 
ulation (Hsieh et al., 2008; Rodgveller et al., 2012; Rouyer 
et al., 2012). Loss of older individuals that have experi¬ 
enced several spawning events is detrimental to the repro¬ 
ductive potential of a population, and failure to account for 
these maternal effects on recruitment may introduce bias 
into stock assessments (Murawski et al., 2001). 
Despite the lack of reliable catch information from the 
1980s, ancillary information substantiates the notion that 
catches of deepwater snappers from the offshore banks of 
American Samoa were higher prior to 1989 (Itano, 1996; 
WPRFMC 2 ; Moffitt 3 ). During recent research cruises, the 
approximate age of the oldest goldeneye jobfish caught was 
the same as the number of years since fishing ceased at the 
offshore banks and contrasts starkly with that of the oldest 
fish captured in the fished areas. This result indicates that, 
despite intensive fishing over a short period (10 years) in the 
1980s, populations of goldeneye jobfish were able to attain 
a broader age composition once intense fishing stopped. 
However, the maximum age may have yet to be realized at 
the offshore banks if they remain unexploited. 
Why growth varied between the fished and unfished 
areas differently for the goldeneye jobfish and goldflag 
jobfish is difficult to interpret. The small sample size of 
goldeneye jobfish from the fished areas is a suspected 
issue; however, the size and age distributions do not 
indicate any obvious problems that would bias growth 
estimates. The absence of smaller and younger fish from 
the unfished areas could have confounded growth rate 
comparisons; for this reason, among others, we used the 
constrained growth models. Density dependence is a 
strong driver of growth variability (Beverton and Holt, 
1957), but evolutionary effects of fishing on growth rates 
are also plausible (Conover and Munch, 2002; Jprgensen 
et al., 2007; Enberg et al., 2012). It is difficult to defin¬ 
itively determine or quantify the significance of either, 
especially the evolutionary effects of fishing on growth 
rates (Rose et al., 2001; Hilborn and Minte-Vera, 2008). 
In this case, the apparent differences in growth likely do 
not indicate an actual fishery-induced change in growth 
through density dependence or evolutionary changes but, 
perhaps, are a function of the fishery primarily removing 
the individuals that have already reached their but 
not removing significant numbers of fish below their L x . 
This is another example of fishing having a greater effect 
on age than on size. 
Goldeneye jobfish and goldflag jobfish are assessed and 
managed as part of a 17-species complex in the U.S. terri¬ 
tories of American Samoa, Guam, and the Commonwealth 
of the Northern Mariana Islands. Assessing species on the 
level of a species complex has been criticized because the 
specific species that are potentially being overharvested 
or underutilized cannot be distinguished. The results of 
the assessment conducted in 2016 indicate that the spe¬ 
cies complex was not being overfished or experiencing 
overfishing in any of these U.S. territories (Yau et al., 
2016). The mortality estimates and truncated age compo¬ 
sition from this study indicate that the current status of 
goldeneye jobfish in fished areas of the Samoan Archipel¬ 
ago and of goldflag jobfish in fished areas of the Mariana 
Archipelago may not be as optimistic as the status of the 
species complex. It is plausible that the offshore banks 
act as a source for the fished areas. This is an avenue for 
future research. Finally, it must be noted that this study 
examined goldeneye jobfish throughout the entire Samoa 
Archipelago, including Wallis and Futuna, whereas the 
assessment conducted in 2016 used data from only fished 
areas of American Samoa. 
In conclusion, goldeneye jobfish and goldflag jobfish 
exhibited growth and longevity similar to those of other 
Pristipomoides species with statistically significant but 
minimal sexual dimorphism in growth. Estimates of age 
structure and mortality for goldeneye jobfish and goldflag 
jobfish indicate that these populations are affected by fish¬ 
ing pressure around populated islands, but this impact is 
