Analysis of Hawaii Longline Catch Data Shows Changes at Top of the Subtropical Gyre Ecosystem

During the past decade, species composition of the catch in the deep-set sector of the Hawaii longline fishery has changed. While the catch proportions for most species have remained fairly steady, those for mahimahi and monchong have increased substantially and the proportion of albacore has declined noticeably.

The changes were documented in a study by Melanie Abecassis, Evan Howell, and Jeffrey Polovina in the PIFSC Ecosystem and Oceanography Division. Having previously examined changes in the bottom of the subtropical gyre's trophic web (e.g., phytoplankton), they recently looked for changes at the top of the pelagic ecosystem, namely the fish species harvested by Hawaii longline vessels. They focused on the sector of the longline fleet that uses deep-set gear. These vessels target primarily bigeye tuna, but also catch many other species.

The number of hooks deployed in the deep-set sector of the Hawaii longline fishery increased significantly during 1996-2006. Monthly data show seasonal patterns in tuna fishing effort.

Abecassis, Howell and Polovina examined trends in the deep-set fishery using daily logbook data submitted to NMFS by longline vessel captains. Information on the number of hooks per float, used to identify deep sets, is available in logbook records since 1996, so data from 1996 through 2006 were used for the analysis. While the fishery catches many species, the 10 most common species account for about 90% of the catch in numbers. Average monthly fishing effort in the deep-set fishery, as measured by the number of hooks set, has increased as much as threefold from 1996 to 2006. Catch-per-unit-effort (CPUE), measured as the number of fish caught per 1000 hooks, exhibited statistically significant linear trends for many of the top 10 species over the 1996-2006 period. Abecassis, Howell and Polovina used a generalized additive model to fit both the linear trend and seasonal components of monthly CPUE data for each species. Over the last decade, five species, namely blue shark, bigeye tuna, shortbill spearfish, striped marlin, and albacore had declining linear trends in CPUE of -3.1%/yr, -3.3%/yr, -3.3%/yr, -4.8%/yr, and -9.9%/yr, respectively. Three species — ono, skipjack tuna, and yellowfin tuna — showed no statistically significant linear trend in CPUE. For two species, pomfret (monchong) and mahimahi, CPUE increased linearly during the period by 10.0%/yr.

Statistics for the 10 most common species in the deep-set longline catch: percent representation in the total catch in 1996 and 2006, annual percent change in CPUE during 1996-2006 (linear trend), ratio of production to stock biomass, and trophic level (higher values indicate a position higher in the food web).
Species % of total catch % percent change in CPUE (annual) P/B TL
1996
2006
Albacore tuna
24
3
-9.9
0.6
4.0
Striped marlin
7
5
-4.8
0.5
4.3
Shortbill spearfish
3
3
-3.3
0.5
4.3
Bigeye tuna
24
24
-3.3
0.8
4.0
Blue shark
14
14
-3.1
0.3
4.0
Skipjack tuna
3
4
0.0
1.9
3.9
Yellowfin tuna
5
7
0.0
1.2
4.0
Ono
2
5
0.0
2.0
3.9
Mahimahi
4
10
10.0
3.0
3.9
Monchong
5
12
10.0
0.3
3.2

The increase in mahimahi CPUE is not surprising as the CPUEs of some of its predators, marlins and sharks, have declined. It is interesting that ono and skipjack don't show increases even though they have the same trophic level as mahimahi and fairly high productivity. The strong increase in CPUE of monchong, a lower trophic level, deepwater species with low productivity, is surprising from an ecological perspective. However, monchong ex-vessel prices have increased at a rate of about 18% per year since 2002, so the increasing CPUE for monchong may be due at least in part to an increase in targeting.

Averaged over the 10 most commonly caught species in the deep-set Hawaii longline fishery, annual mean trophic level for the catch complex declined during 1996-2006, whereas the mean ratio of production to biomass increased.

In addition to documenting changes in catch composition and CPUE, we calculated some multispecies fishery indicators of ecosystem change. Averaging results for the 10 top species, we computed overall annual means of trophic level (TL) and ratio of production to stock biomass (P/B) for the catch complex. When we analyzed these indicators over the 1996-2006 period, we found there was a modest decline in the overall mean TL of the catch (all 10 species combined) from 4.0 to 3.9, a decline of about 3%. The mean P/B value increased by about 37.5% from about 0.8 to 1.10.

In summary, if changes in the fishery CPUE statistics are indicative of changes in the ecosystem, then the pelagic ecosystem has more productive species (higher values of P/B) now than it did a decade ago. High P/B species track environmental variation more closely than low P/B species, so the current ecosystem is likely to exhibit more environmental driven variation.

For more information contact: Jeff Polovina