Woodworth-Jefcoats et al.: Oceanographic variability, fishery expansion, and longline catches in the North Pacific 
235 
the proportion of catch that was discarded (Suppl. Fig. 
4) (online only). In general, bigeye tuna composed near¬ 
ly 20% of the total catch, although their contribution 
ranged from as low as 8% (CW region, second and third 
quarter averages) to over 21% (CW and SW regions, 
fourth quarter averages). Discard rates had more vari¬ 
ability; the lowest rates occurred in the SW region in 
the first and second quarters (<30% on average), and 
the highest rates in the third and fourth quarters 
across all regions (40-55% on average). 
In looking at catch composition, we found that each 
of the 11 species in Figure 4 accounted for at least 5% 
of the total annual catch at some point in the time se¬ 
ries. Their contribution to total catch is broken down in 
that figure by quarter for the beginning (Fig. 4A) and 
end (Fig. 4B) of the time series. These distributions in¬ 
dicate that the seasonal timing of catch of bigeye tuna 
shifted from the first and fourth quarters to the third 
and fourth quarters, but that the overall contribution 
of this species to the annual catch changed little. Con¬ 
versely, the contribution of longnose lancetfish to total 
annual catch increased by about 14%, primarily in the 
third quarter. The proportion of blue shark (Prionace 
glauca), yellowfin tuna, and striped marlin in total an¬ 
nual catch declined by 4-6%, whereas the proportion of 
sickle pomfret, snake mackerel, and escolar (Lepidocy- 
bium flavobrunneum) rose by 4-6% over the time series. 
Discussion 
Over the 21-year period examined in this study, the 
fishing effort of the Hawaii-based longline fishery in¬ 
creased more than 5-fold. A growing proportion of this 
effort occurred in the NE region of the fishing grounds, 
particularly during the third quarter of the year. The 
GODAS reanalysis and WOA13 data indicate that 
oceanographic conditions are favorable for bigeye tuna 
across much of the fishery’s footprint. Although increas¬ 
ing effort should correlate with the desire of fishermen 
to catch more fish, the shift in the seasonal and spatial 
deployment of effort raises several biologically perti¬ 
nent questions. Why did the fishery expand its spatial 
footprint, as opposed to it simply setting more hooks 
across the CW and SW regions? Why did it expand, 
primarily, into the NE region and only during the third 
quarter of the year? 
The expansion of the fishery into the NE region dur¬ 
ing the third quarter is likely the result of several fac¬ 
tors. One possibility is that the CW and SW regions 
were already supporting maximum effort. Effort was 
rather stable in these regions after about 2004 (Fig. 2), 
and previous work has documented that the catch rates 
of large, high-trophic-level, commercially valuable fish 
were declining in these waters as a result of increased 
fishing effort (Polovina et al., 2009; Polovina and Wood¬ 
worth-Jefcoats, 2013). Furthermore, competition from 
international fisheries may have precluded additional 
Hawaii-based effort in the SW region. In the NW and 
NE regions, on the other hand, there was comparably 
little Hawaii-based effort and little to no competition 
from international fisheries. 
Less than 10% of the total annual catch was caught 
during the third quarter of the year at the beginning 
of the time series (Fig. 4). Furthermore, in the CW and 
SW regions, target catch rates were lowest (9% and 
14% on average, respectively) and discard rates high¬ 
est (56% and 44% on average, respectively; Suppl. Fig. 
4) (online only) during the third quarter, possibly explain¬ 
ing why fishermen have been willing to change fishing 
locations. These low target catch rates may also ex¬ 
plain why effort was lowest in the third quarter at the 
beginning of the time series, and why, unlike in other 
quarters, effort was not concentrated in a specific re¬ 
gion (before the focus of the fishery on the NE region). 
Considering the distribution of fishing effort to¬ 
gether with catch rates, we found that trends in catch 
rates are strongly correlated with the shift in the loca¬ 
tion of effort. Comparison of quarterly CPUE of big¬ 
eye tuna with the proportion of annual effort in each 
region and quarter indicates that the third-quarter 
CPUE of bigeye tuna was strongly correlated with the 
proportion of effort in the NE region (coefficient of cor¬ 
relation [r]=0.66) and negatively correlated with third 
quarter effort in the CW and SW regions (r=-0.56 and 
-0.46, respectively). No other significant correlations 
were found (P<0.5). Given the above correlations and 
the trends in catch composition, we conclude that the 
fishery reaction was a response to low catch rates for 
the target species in the CW and SW regions during 
the third quarter. The NE region proved to be a par¬ 
ticularly effective fishing ground with high catch rates 
of target species, relatively low discard rates, and with 
little competition from international fishing fleets. As 
a result, a large portion of annual catch of bigeye tuna 
occurred in the third quarter by the end of the time 
series (Fig. 4). 
Although the movement of the fishery toward the 
NE region was greatest in the third quarter, the fishery 
does occur in this region throughout the year, although 
to a lesser degree (Fig. 2). As discussed above, catch 
rates of target species in the SW and CW regions were 
generally higher during the rest of the year, possibly 
explaining why there was less fleet movement outside 
the third quarter. 
The role of oceanographic variability in fishery expansion 
The enhanced fishery yield in the NE region can be 
explained by the oceanographic conditions of the re¬ 
gion. It has the largest area in which preferred thermal 
habitat of bigeye tuna closely overlaps vertically with 
both deep-set hooks (100-400 m) and with waters that 
have suitable oxygen concentrations (> 1.0 mL/L; Fig. 
IB). The time series of the depths of preferred thermal 
habitat shows that in the NE region, the preferred day¬ 
time habitat of bigeye tuna was consistently and com¬ 
pletely within the depth range of deep-set hooks (Sup¬ 
pl. Fig. 3) (online only). Oceanographic variability also 
explains why the fishery did not expand into the SE 
