4. The purse-seine fishery for tunas in the Western and Central Pacific Ocean

In this section, we focus attention on the purse-seine fishery for tunas in the WCPO. We find that fishing capacity exceeds observed catches for all major fleets, i.e. the purse-seine fleets of Japan, the Republic of Korea, the Philippines, Papua New Guinea, the Taiwan Province of China and the United States, and for the other fleets combined. We estimate that, on average, during 1998-2002 purse-seine skipjack-fishing capacity, purged for TE, was around 138 000 tonnes per annum greater than the actual catch levels. For yellowfin and bigeye combined we estimate that, on average, during 1998-2002 purse-seine fishing capacity, purged for TE, was around 29 000 tonnes per annum greater than actual catch levels.

4.1 Data

Vessel level purse-seine catch (by species and set type) and effort (by days fished and searched^[22] and number of sets made by set type) data by vessel flag for the WCPO tuna fishery were obtained from the Secretariat of the Pacific Community (SPC) for 1980-2002.^[23] These data, which covered the operations of the purse-seine fleets of China, the Federated States of Micronesia, Kiribati, the Republic of Korea, the Marshall Islands, New Zealand, Papua New Guinea, the Philippines, the Solomon Islands, the Taiwan Province of China, the United States and Vanuatu throughout the WCPO, were obtained from vessel logbooks.

Data were also provided by the SPC for the purse-seine fleet of Japan; however, these data include only fishing activity in the Exclusive Economic Zones of countries other than Japan, which is only a portion of Japanese purse-seine operations. Data covering the fishing activities of the Japanese purse-seine fleet throughout the WCPO for 2000-2002 was obtained from the National Research Institute of Far Seas Fisheries (NRIFSF) of Japan. These data included catches by set type, the combined number days spent fishing and searching, and carrying capacities of the vessels, provided in ranges of 800 to 900 tonnes, 900 to 1000 tonnes, 1000 to 1100 tonnes, 1100 to 1200 tonnes, 1200 to 1300 tonnes, and 1300 to 1400 tonnes.^[24]

Data on the activities of the Spanish and Australian purse-seine fleets were not available at the time that the analyses were undertaken.

While, for confidentiality purposes, the data provide by the SPC could not be attributed to individual vessels the SPC matched vessel characteristics taken from the Regional Register of the Forum Fisheries Agency (FFA) to the logbook data to allow for a data set for each vessel that included both catch and effort data and data relating to the characteristics of the vessel. The vessel characteristics provided were: GRT, storage capacity, length overall (LOA) and power of main engine. These data pertain to the characteristics of the vessel at the time that they were obtained from the Regional Register, and thus do not capture changes in these characteristics of the period for which the analysis was conducted. Finally, in some cases the vessel characteristics for a vessel were not available^[25] or were incomplete. The data set used in the analysis is based on the sample of vessels for which complete data sets were available. The number of vessels that formed the data set for each fleet grouping used in the analysis is provided in Table 4.2.

TABLE 4.1
Exploitable biomass and sea-surface temperatures

Exploitable biomass estimates for the purse-seine fishery for yellowfin, bigeye and skipjack tunas, which were provided on a quarterly basis by the SPC^[26], are based on stock assessments undertaken for the 16th meeting of the Standing Committee on Tuna and Billfish of the SPC. The quarterly estimates were converted to annual estimates by averaging over a given year. Sea-surface temperatures taken at the time of each set of each vessel, in degrees Fahrenheit, are taken from the logbooks of United States purse-seine vessels. These data are collected jointly by the U.S. National Marine Fisheries Service and the Forum Fishery Agency. These temperatures are averaged (a simple or unweighted arithmetic average) over all sets, vessels and areas to provide a mean annual sea-surface temperature for the area fished in the WCPO. These temperatures are used for all fleets in the analysis, rather than just the United States vessels. The exploitable biomass and sea-surface temperature data used in the analysis are shown in Table 4.1.

For the analysis the data were grouped by fishing nations for the fleets of Japan, the Republic of Korea, Papua New Guinea, the Philippines, the United States, and the Taiwan Province of China. The remaining fleets were combined in a single group, as there were insufficient observations to allow the analysis to be undertaken at the individual fleet level.^[27]

Average vessel data across groups for each of the variables used in the analysis are reported in Table 4.2.

4.2 Methodology

Capacity output, capacity output adjusted for TE and CU rates (observed output divided by capacity output or observed output divided by capacity output adjusted for TE) are estimated by DEA. We estimate fishing capacity using the output-oriented non-radial method of Russell (1985), assuming variable returns to scale.

For the reasons outlined in the EPO analysis, capacity output estimates, with the exception of those for the Japanese fleet, were made for the five-year period of 1998-2002.^[28] For the Japanese fleet estimates were made for the three-year period of 2000-2002, as these were the only years for which data were obtained from the NRIFSF.

TABLE 4.2
Averages for vessel data used to estimate capacity in the purse-seine fishery of the WCPO

Note: For Japan catches reported by the NRIFSF as "free" sets are listed under "unassociated" sets in the table and those reported as "associated" sets are listed as "floating objects" sets.

Estimates of capacity outputs, allowing for variable returns to scale^[29], were made at the set and vessel level by mode of fishing. Reports of yellowfin often include bigeye, as the two species are difficult to distinguish during the juvenile stages of their life cycle, so the data on the catches of the two species were aggregated. For all fleets, with the exception of the Japanese purse-seine fleet, output or catches in the analysis was specified by species and method of harvest as follows: (1) yellowfin and bigeye tuna caught in sets on unassociated schools; (2) yellowfin and bigeye tuna caught in sets on floating objects (FADS or flotsam); (3) yellowfin and bigeye tuna caught in other set types such as anchored FADs; (4) skipjack caught in sets on unassociated schools; (5) skipjack caught in sets made on floating objects and (6) skipjack caught in other set types. For the Japanese fleet the catches were specified as (1) yellowfin and bigeye tuna caught in free sets; (2) yellowfin and bigeye tuna caught in associated sets; (3) skipjack caught in free sets; (4) skipjack caught in associated sets. Free sets are the same as unassociated sets and associated sets are the same as floating-object sets as used previously. The reported catches of other fish were negligible, and hence not considered in the analysis. The analysis estimated capacity output for all six (four for the Japanese fleet) outputs and three (two) types of fishing specifying a common harvesting frontier (i.e. the DEA models were run with all six (four) outputs at once, rather than separately for each of the three (two) types of fishing). To be able to accurately estimate capacity output by individual vessel for each of the different types of fishing, each of the six (four) outputs in the DEA model was specified as catches per vessel per year.

The capital stock or capacity base of an individual vessel was specified by its GRT, storage capacity, LOA and engine power, except for vessels of the Japanese fleet, for which it was specified by the mid-point of the carrying capacity band in which it fell.

Biomass estimates for yellowfin and skipjack were used to specify stock conditions with sea-surface temperature used to account for environmental conditions. Both of these variables were specified as non-discretionary or fixed (constrained) inputs.

The technological-economic measure of capacity output specifies full utilization of variable inputs. However, estimates of TE by DEA were made by using the annual numbers of days fished and searched as the variable inputs.

Annual capacity output on a per-vessel basis was estimated and subsequently converted to total annual fleet activity for each fleet group by summing over the individual vessels.

Finally, it should be noted that the catch estimates obtained from the logbook data and the data for the Japanese fleet provided by the NRIFSF differ from the SPC estimates of catches published in its Tuna Fishery Yearbook, as the logbook data are only one of many sources that are used to derive the published estimates. Given this, in Section 4.3.2 we combine the estimates derived from the DEA analysis and the SPC Tuna Fishery Yearbook catch estimates to obtain estimates of excess fishing capacity in the WCPO tuna purse-seine fishery.

4.3 Results

4.3.1 DEA estimates

The results of the analysis indicate that substantial excess fishing capacity, defined as fishing capacity output minus observed output (landings), when measured as: (1) potential catch minus actual catch or (2) potential catch, purged for TE, minus actual catch exists for:

• Skipjack for all of the major fishing nations and for other fishing nations as a group for all set types; and

• Yellowfin and bigeye combined for all of the major fishing nations and for other fishing nations as a group for all set types.

TABLE 4.3
Reported catches, estimated capacities and capacity purged for technical efficiency for the purse-seine fishing of the WCPO

Note: For Japan catches reported by the NRIFSF as "free" sets are listed under "unassociated" sets in the table and those reported as "associated" sets are listed as "floating-object" sets.

During 1998-2002 excess capacity for skipjack, purged for TE, ranged from eight percent of the observed catch for the Philippines' fleet to 35 percent of the observed catch for the Papua New Guinea fleet (Table 4.4). Excess capacity for skipjack, purged for TE, for the Korean, Taiwanese and United States fleets during 1998-2002 averaged 19, 10 and 20 percent of the observed catches, respectively. The estimates for the United States fleet are likely to be biased upward because some vessels of this fleet also operate in the EPO, and this is not accounted for in the analysis. For the other vessels as a group, it was estimated that excess capacity for skipjack, purged for TE, averaged 17 percent of their observed catch. For the Japanese fleet it was estimated that during 2000-2002 the excess capacity for skipjack, purged for TE, averaged six percent of the observed catch.

During 1998-2002 the excess capacity for yellowfin and bigeye combined, purged for TE, ranged from nine to 17 percent, a significantly narrower range than that for skipjack (Table 4.4). The group of other vessels was estimated to have the lowest relative excess capacity for yellowfin and bigeye combined, purged for TE, at nine percent of the observed catch, while the United States fleet was estimated to have the greatest, at 17 percent of the observed catch. Once again, the estimates for the United States fleet are likely to be biased upward for the reason given above. Excess capacity for yellowfin and bigeye combined, purged for TE, for the Taiwanese, Philippine, Korean and Papua New Guinea fleets during 1998-2002 averaged 11, 13, 14 and 15 percent of their observed catches, respectively. For the Japanese fleet it was estimated that during 2000-2002 the excess capacity for yellowfin and bigeye combined, purged for TE, averaged six percent of the observed catch.

The CU rates for all species combined also indicate significant excess capacity, defined as capacity output minus observed output, regardless of whether TE is purged (Table 4.5). (CU is defined as observed output divided by capacity output.) CU ranges between 0 and 1, where 0 indicates no observed output and 1 indicates that observed output equals capacity output. CU, purging TE from capacity output, during 1998-2002 averaged across all vessels in the respective fleets ranged from 0.81 for the Papua New Guinea fleet to 0.95 for the Philippines fleet. CU for all species combined, purged for TE, for the Korean, Taiwanese and United States fleets during 1998-2002 averaged 0.86, 0.91 and 0.86 percent respectively. For the reason previously mentioned, the estimates for the US fleet are likely to be biased downward. For the other vessels as a group it was estimated that CU for all species combined, purged for TE, averaged 0.88. For the Japanese fleet it was estimated that during 2000-2002 that CU, purged for TE, for all species combined averaged 0.95.

4.3.2 Using the DEA results and SPC catch estimates to estimate total excess fishing capacity in the WCPO

In this section the results of the DEA analysis are combined with the catch estimates in the SPC Tuna Fishery Yearbook to provide estimates of purse-seine fishing capacity and purse-seine fishing capacity, purged for TE, for all species combined and individually for skipjack and for yellowfin and bigeye combined. This is done because the DEA estimates derived in the previous section are based on logbook data, rather than raised catch data.

These estimates were derived as follows. For each fleet and species grouping covered in the analysis the estimated catches from the SPC Tuna Fishery Yearbook were multiplied by the estimated excess capacity, expressed as percentages, obtained from the DEA analysis for the given fleet and species grouping. This was done for both fishing capacity and fishing capacity purged for TE. The resulting excess capacity estimates were then divided by the SPC Tuna Fishery Yearbook total catch for the fleets covered in the analysis and this percentage multiplied by the total WCPO purse-seine catch.

TABLE 4.4
Reported catch, estimated excess capacity and excess capacity purged for technical efficiency for the purse-seine fishery of the WCPO

Notes. As no estimates for the Japanese fleet are available for 1998 and 1999 all vessel figures for these years are exclusive of this fleet.

TABLE 4.5
Average vessel capacity utilisation and technical efficiency

The derived estimates indicate that on average during 1998-2002 excess capacity for skipjack, purged for TE, was 137 452 tonnes, and was at its highest in 2000 at 188 991 tonnes and was at its lowest in 2001 at 89 088 tonnes. As indicated in Figure 4.1, excess capacity for skipjack, purged for TE, trended upward during 1998-2000, before declining significantly in 2001 and then increasing again in 2002. A possible driving force behind this pattern is the skipjack price fluctuations experienced over the period, with Bangkok skipjack prices plummeting to record lows in the second half of 1999 and remaining at these depressed levels throughout 2000 (Catarci, this collection). This price decline resulted in revenues below operating costs for some fleets, which led to some vessels tying up for prolonged periods and fishing fewer days than they would normally, particularly in 2000, as reflected in the number of days vessels in most fleets spent searching and fishing in total (Table 4.2) In other words, there was a reduction in the level of utilisation of variable inputs. In early 2001 the prices recovered to some degree, and prices throughout 2001 and 2002 remained significantly above the levels of the second half of 1999 and throughout 2000, and the average time vessels spent fishing and searching in total increased for all fleets except that of Japan, that was higher in 2001 than in 2000.

The derived estimates also indicate that on average during 1998-2002 excess capacity for yellowfin and bigeye combined, purged for TE, was 31 278 tonnes, and at its highest in 1999 at 43 873 tonnes and lowest in 2002 at 16 977 tonnes. As indicated in Figure 4.3, excess capacity for yellowfin and bigeye combined, purged for TE, rose in 1999, before declining continuously to its 2002 level.

4.3.3 Estimated fishing capacity in the WCPO and sustainable fishing mortality on yellowfin and bigeye stocks

The Scientific Co-ordinating Group of the Preparatory Conference for the Commission for the Conservation and Management of Highly Migratory Fish Stocks in the Western and Central Pacific recommended that there be no further increases in fishing mortality rates on yellowfin (particularly juveniles) and bigeye. Based on this recommendation, we compare estimated fishing capacity and fishing capacity, purged for TE. Against a target catch level set at the average yellowfin catch by purse seiners during 2000-2002. Fishing capacity and fishing capacity purged for TE, for yellowfin and bigeye combined with average purse-seine catches during 2000-2002 are compared in Figure 4.5. From this it can be seen that there was significant excess fishing capacity for yellowfin and bigeye when measured as fishing capacity minus average catches from 2000-2002 in the fishery from 1998 to 2002, although this excess capacity existed primarily in 1998 and 1999 and was very low in 2002. During 1998-2002 the average fishing capacity for yellowfin and bigeye combined was in excess of average catches between 2000 and 2002 by 88 762 tonnes or 38 percent.

It can be seen in Figure 4.5 that excess capacity for yellowfin and bigeye, when measured as fishing capacity, purged for TE, minus average catches from 2000-2002 existed in the fishery from 1998 to 2001, but in 2002 there was no excess capacity in the fishery. During 1998-2002 the average fishing capacity, purged for TE, for yellowfin and bigeye combined was in excess of average catches between 2000 and 2002 by 47 666 tonnes or 24 percent.

4.4 Summary and conclusions

The analysis conducted for the WCPO suggests that excess fishing capacity exists for all major fleets, i.e. the purse-seine fleets of Japan, the Republic of Korea, the Philippines, Papua New Guinea, the Taiwan Province of China and the United States, and for the other fleets as a group.

It was estimated that on average during 1998-2002 purse-seine skipjack fishing capacity was around 240 000 tonnes (35 percent) per annum greater than actual catch levels. However, it noted that when purging for TE excess skipjack fishing capacity was only 99 000 tonnes (14 percent) per annum greater than the actual catch levels. In other words, only around 40 percent of the potential increase in catches could be realised through increases in variable input usage, given the biomass, environmental conditions and the state of technology that prevailed over this period. Estimated excess fishing capacity, purged for TE, was at its highest level in 2000. It was hypothesised that this may have been caused by low skipjack prices in the second half of 1999 and throughout 2000, resulting in vessels reducing the number of days spent searching and fishing (Catarci, this collection).

For yellowfin and bigeye combined it was estimated that during 1998-2002 excess purse-seine fishing capacity was around 50 000 tonnes (28 percent) per annum greater than the actual catches. However, it noted that when purging for TE excess yellowfin and bigeye fishing capacity was only 19 000 tonnes (11 percent) per annum greater than the actual catches. In other words, only around 40 percent of the potential increase in catches could be realised through increases in variable input usage, given the biomass, environmental conditions and the state of technology that prevailed over this period. It was also estimated that on average during 1998-2002 fishing capacity, purged for TE, for yellowfin and bigeye combined was in excess of the average catches between 2000-2002 by 47 666 tonnes or 24 percent, but that no excess capacity existed in the fishery in 2002, when measured against average 2000-2002 catch levels.