236 
Fishery Bulletin 116(3-4) 
region. This region encompasses the oxygen minimum 
zone of the eastern tropical Pacific Ocean. Across much 
of the SE region, the oxygen-concentration threshold of 
1.0 mL/L occurred at depths shallower than the depths 
of both the preferred thermal habitat of bigeye tuna 
and the depths of deep-set gear (Fig. IB), rendering it 
poor habitat for bigeye tuna and poor longline fishing 
grounds. 
Effort-weighted trends in the depth of preferred hab¬ 
itat of bigeye tuna indicate that the fishery has moved 
into the more favorable oceanographic conditions of the 
NE region. At the beginning of the time series, the fish¬ 
ery was operating largely in waters where the median 
depth of preferred thermal habitat for bigeye tuna was 
roughly 320-340 m below the surface. However, by the 
end of the time series, the fishery was operating in wa¬ 
ters where the median depth of preferred thermal habi¬ 
tat was 270-315 m below the surface and more closely 
aligned with the median depth of deep-set gear (250 
m; Boggs, 1992; Bigelow et al., 2006). These trends in 
depth weighted by total quarterly effort indicate that 
fishermen were targeting regions where either pre¬ 
ferred thermal habitat was more closely aligned with 
their gear or thermal habitat shoaling was greatest or 
possibly employing a combination of these 2 tactics. 
Across the entire fishing ground, the preferred thermal 
habitat of bigeye tuna shoaled by only about 12-15 m. 
Yet, when weighted by quarterly effort, the shoaling 
increased to roughly 37-65 m. Without information on 
depth of capture, it is difficult to determine the degree 
to which this shoaling actually influenced the fishery 
yield. However, given the vertical distributions of both 
deep-set gear and preferred daytime thermal habitat 
of bigeye tuna, shoaling could increase the degree to 
which these distributions overlap and could compress 
the total vertical habitat that bigeye tuna occupy. Both 
scenarios should increase the catchability of bigeye 
tuna, and in turn, fishery yield. 
Effects of fishery expansion on catch composition 
During the period studied, the spatial expansion and 
seasonal shift of the fishery influenced the seasonal 
timing of both the catch and catch composition. Al¬ 
though the primary target species, bigeye tuna, con¬ 
sistently was about 20% of the total annual catch, the 
bulk of the annual catch shifted from the first and 
fourth quarters to the third and fourth quarters. A 
combination of factors could have contributed to this 
shift. The foremost factor was the increase during the 
third quarter in effort deployed in the NE region (Fig. 
2), where catch rates of bigeye tuna were consistently 
high over time (Fig. 3B). Additionally, by the end of 
the period examined, less effort was deployed in the 
SW region in the first quarter than in the CW region 
during the fourth quarter (Fig. 2). First quarter catch 
rates of bigeye tuna in the SW region declined over 
the past 2 decades (Fig. 3B), whereas fourth quarter 
catch rates of bigeye tuna in the CW region remained 
consistently high. In summary, by 2015, the fishery 
deployed most of its effort in the regions and during 
the quarters when catch rates of bigeye tuna were 
highest. It is interesting to note that these regions 
are also those where preferred thermal habitat for 
bigeye tuna completely overlaps with deep-set gear 
(NE region) and where preferred thermal habitat for 
bigeye tuna is most compressed (CW region) (Suppl. 
Fig. 3) (online only). 
It is possible that the shift in time and place of the 
bulk of bigeye tuna catch each year can be attributed 
to changes in fishing gear, although we found no evi¬ 
dence that this shift was the cause. Using the number 
of hooks per float as a proxy for hook depth, we found 
no significant differences between gear set in the SW 
region during the first quarter, in the NE region dur¬ 
ing the third quarter, and in the CW region during the 
fourth quarter (5% significance level, Wilcoxon-Mann- 
Whitney rank-sum tests). 
The shift in the annual timing of catch of bigeye tuna 
could also be attributed to fish movement or changes 
in population dynamics. Stock assessments from both 
fishing convention areas (west of 150°W, WCPFC, Har¬ 
ley et al., 2014; east of 150°W, IATTC, Aires-da-Silva 
and Maunder, 2015), along with tagging data (Schae¬ 
fer et al., 2015), indicate that there is extensive zonal 
movement by bigeye tuna. At low latitudes (e.g., 15°S- 
15°N), there is more eastward movement than west¬ 
ward movement (Aires-da-Silva and Maunder, 2015; 
Schaefer et al., 2015). However, a lack of tagging data 
for areas farther north makes it difficult to determine 
whether bigeye tuna make the same directional move¬ 
ment in our study area. If they do, the high catch rates 
in the NE region noted in our study may have been 
fueled in part by fish moving into the region. The role 
of population dynamics is also unclear. Although it is 
likely that large-scale population dynamics affect inter¬ 
annual changes in CPUE of bigeye tuna (Harley et al., 
2014; Aires-da-Silva and Maunder, 2015), size struc¬ 
ture (and presumably age structure) of bigeye tuna was 
fairly consistent across the fishing ground (Suppl. Fig. 
5) (online only), echoing earlier work (Kume, 1969). 
For other commercially valuable species, such as 
yellowfin tuna and striped marlin, the spatial shift 
in effort exacerbated declining catch rates. Although 
CPUEs for both species declined across the fishing 
grounds, catch rates for these species were greatest 
in the SW and CW regions despite the movement of 
the fishery away from these regions (Fig. 3B). Catch 
rates for skipjack tuna, although not declining, were 
generally highest in the SW and CW regions (Fig. 3B). 
Therefore, the fishery’s changing footprint likely con¬ 
tributed to an overall decline in the contribution of 
skipjack tuna to total annual catch (Fig. 4). 
Discard rates also were influenced by the spa- 
tiotemporal shift in effort. In the core region of the 
fishery (12-27°N), rising discard rates were linked to 
increased fishing effort (Polovina and Woodworth-Jef- 
coats, 2013). At the same time, catch rates of longnose 
lancetfish in particular rose as a result of the fishery’s 
northward expansion and increased focus on the third 
