New Indices of Abundance Computed for Assessment of South Pacific Albacore

Temporal changes in annual stock status reference points for the South Pacific albacore stock from the 2008 assessment. Status indicators for fishing mortality and spawning stock biomass (relative to Maximum Sustainable Yield conditions), based on several alternative model configurations, are plotted from 1960 (palest blue dot) to 2006 (darkest blue dot), with labels at 5-yr intervals. Estimates of current status (average of 2004-2006) indicate that overfishing is not occurring and the stock is not overfished. From Hoyle, Langley, and Hampton (2008).

An important component of models used to assess the status of tuna stocks is an index showing changes in the stock's relative abundance over time. South Pacific albacore assessments have relied on abundance indices derived from aggregated nominal catch-per-unit-effort (CPUE) statistics reported in logbooks by distant-water longline fleets of Japan, Korea and Taiwan, and logbook data from domestic longline fisheries. The distant-water CPUE indices used in a 2005 stock assessment were based on data aggregated by month and 5° latitude by 5° longitude areas of the ocean.

Recently, these indices were improved by PIFSC scientist Keith Bigelow and his colleague Simon Hoyle of the Oceanic Fisheries Program, Secretariat of the Pacific Community (OFP/SPC). They developed a new standardized CPUE index for distant-water fleets targeting South Pacific albacore by analyzing detailed, operational-level data in the logbooks of vessels landing their catch at canneries in Pago Pago, American Samoa and Levuka, Fiji. The series of data for the Japan fleet was relatively short (~ 10 years) while the Korea and Taiwan series cover 35 years or more. Only Taiwanese vessels actively targeted albacore through most of the time period studied. Altogether, the data sets represented 1163 vessels conducting 8909 trips and making 475,019 longline sets. The data were spatially stratified into four regions centered at 25°S and 180°, and generalized linear models (GLMs) were developed within each region for each of the three distant-water fleets — 12 models in all.

The analysis was confined to data from vessels that fished in four quarters or more. In developing the GLMs, a suite of 10 variables were evaluated as potential predictors of albacore CPUE: year-quarter, location (latitude and longitude), vessel (unique vessel identifier), sea surface temperature, depth of the 15° C isotherm, and four interaction terms. There were substantial spatial differences in catch, effort and CPUE between the aggregated data and the operational-level data from albacore-targeting vessels. There was good coherence in nominal CPUE among fleets targeting albacore. However, the older, nominal CPUE index for albacore, derived from aggregated data, showed little coherence among fleets and regions, perhaps due to a change in targeting from albacore to yellowfin and bigeye tuna over time.

Four of the 10 variables were included in the final GLM models as useful predictors of CPUE: year-quarter, vessel, interaction between month and latitude, and interaction between latitude and longitude. Standardized CPUE derived from the GLMs differed from the corresponding nominal CPUE time series, especially in subtropical regions. In addition, the new GLM-standardized abundance indices for Japan and Korea generally had less variability than the indices used in the 2005 assessment based on aggregated data. For Taiwan, the new indices were similar to the ones used in 2005 because the Taiwan fleet has consistently targeted South Pacific albacore. Taiwan indices for all regions were highest in the 1960s, declined moderately until 1975, then continued to decline at a lower rate.

A new assessment of the South Pacific albacore stock was completed in 2008 by Hoyle, Adam Langley, and John Hampton (all of SPC/OFP). The results differed substantially from those in the previous assessment (2005) due to the standardized CPUE index of abundance for the distant-water longline fleets and inclusion of new parameters for gear selectivity, natural mortality and reproductive potential. These changes represent both refinements to the assessment model and substantive changes to the model's structure. The current South Pacific albacore assessment indicates lower levels of stock size and maximum sustainable yield than estimated in previous assessments; the new results appear to be more realistic.

The research by Hoyle and Bigelow on the standardized CPUE abundance index was partially funded through a project of the Pelagic Fisheries Research Program of the NOAA-University of Hawaii Joint Institute for Marine and Atmospheric Research - "Fishery Dynamics in the Samoan Archipelago".

A full report of the study by Hoyle and Bigelow is available at [albacore CPUE].pdf.

A report on the 2008 stock assessment by Hoyle, Langley, and Hampton is available at [ALB assessment].pdf.

For more information contact: Keith Bigelow