REVIEW OF ASPECTS OF SOUTHERN BLUEFIN TUNA BIOLOGY, POPULATION, AND FISHERIES

A.E. Caton (editor)
Fisheries Resources Branch
Bureau of Resource Sciences
P.O. Box E11, Queen Victoria Terrace
PARKES A.C.T. 2600, Australia

1. INTRODUCTION

Scientists of the southern bluefin tuna (SBT) working group commenced a review for the Food and Agriculture Organization Expert Consultation on Interactions in Pacific Tuna Fisheries. When the need arose for a background review on SBT for the May 1990 World Meeting on Bluefin Tunas at the Inter-American Tuna Commission in La Jolla, California, drafts of completed chapters of the Expert Consultation SBT review were distributed to participants. A more complete and edited version was included in the world meeting report¹. The review presented contained a considerable amount of new data on SBT and its fisheries, updating material in the earlier comprehensive reviews of Shingu (1978)² and Olson (1980)³. It included a comprehensive reference list, but other references may also be located in the bibliography of Eckert et al. (1987)⁴.

The SBT review from the World Bluefin Meeting report was not distributed in its entirety for the purposes of the Expert Consultation; instead, the ‘Overview’ chapter from the review was provided as a background document, with a listing of the review's contents (Appendix 1). That overview follows below, updated slightly and including figures and tables.

¹ [Caton, A.E. (editor). 1991. Review of aspects of southern bluefin tuna biology, population and fisheries. In World meeting on stock assessments of bluefin tunas: strengths and weaknesses, edited by R.B. Deriso, and W.H. Bayliff. 1991. Spec.Rep.I-ATTC, (7):181–350.]

² [Shingu, C. 1978. Ecology and stock of southern bluefin tuna. Japan Association of Fisheries Resources Protection. Fisheries Study 31:81 p. (In Japanese. English translation in Rep.CSIRO Div.Fish.Oceanogr., 131:79 p.)]

³ [Olson, R.J. 1980. Synopsis of biological data on the southern bluefin tuna, Thunnus maccoyii (Castelnau 1872). Spec.Rep.I-ATTC, (2):151–212.]

⁴ [Eckert, G.J., J. Kalish, J. Majkowski, and R. Pethebridge. 1987. An indexed bibliography of the southern bluefin tuna (Thunnus maccoyii [Castelnau, 1872]). Rep.CSIRO Mar.Lab., 185:49 p.]

2. OVERVIEW

Southern bluefin tuna (SBT), Thunnus maccoyii (Figure 1), are large (up to 200 cm and 200 kg), long-lived (more than 20 years) and highly migratory pelagic fish. The species has a general circum-global distribution between 30°S and 50°S, with a spawning ground between 7°S and 20°S in the northeastern Indian Ocean south of Java (Figure 2).

Figure 1. Southern bluefin tuna, Thunnus maccoyii (Castelnau, 1872). (Line drawing courtesy FAO; from: FAO species catalogue, Vol.2.Scombrids ofthe world; an annotated and illustrated catalogue of tunas, mackerels, bonitos and related species known to date; by B.B. Collette and C.E. Nauen, FAO Fisheries Synopsis No. 125 (1983), 137 p.)

Figure 2. Known worldwide distribution of southern bluefin tuna (hatched) and the spawning ground (shaded). (After Collette and Nauen, 1983, and Nishikawa et al., 1985.)

Southern bluefin tuna can be distinguished in the field from the bluefin tuna Thunnus thynnus, mainly found in the northern hemisphere, by its yellow median caudal keel (dark in T. thynnus and other thunnids) and the relatively longer adult pectoral fin (20–23% of length to caudal fork in T. maccoyii, compared with 17–22% in T. thynnus).

A single spawning area, morphological uniformity, and tag-return data suggest a single population. Because spawning is restricted in time and space, there is little chance for different populations to develop through reproductive isolation. High vagility (as evidenced by the numerous long-distance tag returns) in relation to the geographical distribution (indicated by the range of longline catches) would also suggest limited potential for population differentiation.

Southern bluefin tuna with an elevated gonad index, are found only in the Indian Ocean. In spite of many years of worldwide tuna longline fishing, no data suggesting spawning have been reported from any waters other than the area of the Indian Ocean south of Java. On the spawning grounds, only large, mature SBT with well-developed gonads are caught. Maturity can occur at about 120 cm fork length, but more commonly at 130 cm (about 8 years old). Fecundity of one fish 158 cm long with a gonad index of 4.31 was estimated to be 14–15 million. Ripe eggs remaining in a 143 cm, spent fish were 0.66–1.05 mm in diameter. Examinations of sex ratio in catches in various fishing areas show that, as juveniles, females outnumber males; the situation is reversed in adults.

Little detailed research has been carried out on the reproductive biology of SBT. To remedy this, biological sampling on the Oka (spawning area) and Oki (adjacent area to the south) fishing grounds is required. It is not known whether spawning has a particular periodicity (such as a lunar cycle), or how spawning intensity is spread throughout the spawning season; ichthyoplankton sampling in the area has not been carried out at the appropriate spatial and temporal scales. However, recent ichthyoplankton surveys suggest that the main spawning period is from January to February, as the highest larval catches have occurred then. The fishing ground to the south (Oki) appears to be the staging area where large SBT with ripening or spent gonads accumulate prior to, during, and after the spawning period (September to March). It is believed that these SBT migrate southward from the Oki ground after the spawning season.

The limits of the spawning ground have been established by the distribution of adults with ripe gonads and the capture of larvae. The identification of larvae is, however, unreliable, and confusion with other Thunnus species can occur. Larval distribution within the spawning ground has not been defined because of insufficient sampling relative to the patchy occurrence of larvae. Juvenile tuna six to eight cm long, and thought to be SBT, have been collected by Bouke-ami net on the Australian northwest shelf. Larger juveniles (20–60 cm fork length) have been caught by gillnet and trolling further south (22°S – 34°S) on the shelf.

Off the southwest coast of Western Australia (WA; Figure 3), SBT between 1 and 3 years old are common in surface catches. Eastward, originally as far as New South Wales (NSW; Figure 3), SBT in surface catches commonly range from 2 to 6 years old, with individuals to at least 9 years old also taken at times. The younger-aged individuals (1 to 3 years old) seem more closely associated with waters of the coast and continental shelf than are older individuals. Some immature fish (from 3 years old, and progressively for older ages) leave the warmer waters of coastal areas for the wide region of the West Wind Drift (40° – 45°S) where feeding adults are mainly found.

Figure 3. Southern bluefin tuna spawning ground and migration pattern within and out of Australian waters. The 200-mile Australian fishing zone is indicated by the solid line and the horizontal hatching indicates the composite distribution of the Australian surface fishery. The general distribution of Japanese longline fishing is inset. (Modified from Majkowski et al., 1988.)

A promising technique for the study of migration patterns, involving the correlation of physical environmental features (particularly temperature) with otolith microchemistry, is under investigation. Also, the adoption of new, high technology tagging techniques - in particular the use of archival tags - could provide new data on continuous three-dimensional movements.

Information about the nutrition of SBT is largely anecdotal. The larval diet, which consists of mainly microcrustacea and macrozooplanktonic crustacea, is similar to that of related tuna species. It includes a degree of cannibalism. Studies show that survival and growth are dependent on the densities of larvae and other prey. Juveniles and adults are opportunistic feeders, chiefly on cephalopods, crustacea, fish, and salps. In general, smaller SBT feed mainly on crustacea. Adults feed mainly on fish. Sharks, other tunas, other teleosts, seabirds, whales, and seals are possible competitors, whilst various species of sharks and the killer whale are possible predators.

Tag-recapture data and length-frequency modal analysis have traditionally provided information on SBT age structure and growth rates, and length-at-age estimates for the fished population. There are different age-at-length relationships because of the low numbers of recoveries of fish at liberty for long periods, and the questionable reliability of data provided by some recoveries. The von Bertalanffy growth parameters derived have ranged from 171.5 to 207.6 cm for L∞, 0.187 per year for K and -0.554 to 0 years for t_o. They produce lengths at 2, 8 and 14 years old ranging between 46 and 56 cm (2 years old), 124.7 and 136.7 cm (8 years old), and 157.4 and 174.7 cm (14 years old). The von Bertalanffy growth parameters which have been adopted for stock assessments in recent years have been 207.6 cm for L∞, 0.128 per year for K and -0.394 years for t_o (Figure 4). Growth variations among individual fish, and seasonal growth (faster during summer and early autumn) have been reported, apparently synchronised with ambient water-surface temperature. Recoveries of SBT tagged in 1983–84 all indicated significantly faster growth than previously tagged fish, the increase persisting with increasing time at liberty.

Figure 4. Growth in length* and weight of southern bluefin tuna. Values of von Bertalanffy growth parameters for the length curve area K=0.128 per year; t-0.394 years; and L =207.6 cm. Growth in weight is derived from the length-weight relationships in Figure5. (* The bars on the length curve represent the range of length at ages 2, 8, and 14 years for the different von Bertalanffy parameters derived for southern bluefin tuna.)

Annual banding has been reported on scales from fish up to 130 cm fork length (which had seven bands) but clarity decreased with increasing fish size. Bands on scales from larger fish (> 130 cm fork length) were unreadable. Daily growth of larvae has been determined from otolith-increment microstructure, validated using marginal-increment analysis. A burning technique, (validated for 2–4 years old, inclusive) showed annual growth-bands in otoliths from SBT up to 167 cm fork length. For fish younger than about 8 years old, the relationship of age and mean-length was similar to those derived previously by the methods mentioned above. Recently, X-ray examination of SBT otolith microchemistry has shown conspicuous episodic variations in strontium-calcium ratios for both small and very large SBT, apparently consistent with the expected age of a fish, based on its size and previous estimates of growth rates. This technique may enable reliable age determination of large (> 170 cm fork length) SBT, thus filling the gap left by previous aging techniques. Finally, trial use of computer-based, simultaneous analyses of a time series of length-frequency distributions to separate modes into age groups (MULTIFAN) has given results consistent with age-at-length relationships from other methods.

The age composition of the catch has been determined from length composition, using an age-at-length relationship. Where length composition is unavailable, catch-weight composition and a length-weight relationship have been used to derive length composition and, thereby, age composition. For fish less than 130 cm in length to caudal fork, the relationship between length and weight used during the later 1980s for stock assessments has been W=3.13088 × 10^-5 × L^2.9058, where W is whole weight and L is length to caudal fork (Figure 5). For fish greater than 130 cm in length, the relationship used has been W=1.4036 × 10^-6 × L^3.5399. It is more difficult to classify larger, older fish by this method however, raising doubt about the accuracy of estimates of age-distribution in the catch for such fish.

Figure 5. Relationships currently used to convert southern bluefin tuna weight to length; the relationship for southern bluefin tuna > 130 cm has been adjusted from a processed-weight; length relationship to a whole-weight:length relationship using a 1:1.15 ratio of processed to whole weight.

The natural mortality rate (M) of SBT is an input parameter to Virtual Population Analysis (VPA) and is required for yield-per-recruit analyses. Again, there is considerable uncertainty about the appropriate value for SBT and whether M varies by area or age. The most commonly used value has been 0.2. However, assessments have generally incorporated tests to examine their sensitivity to the uncertainty in the parameter. All reliable analyses indicate that the rate of natural mortality is within the range 0.2 ± 0.1 per year. Further analyses may yield closer bounds of M.

Australia, Japan and New Zealand have been the main nations fishing for or controlling access to fishing zones for SBT. Incidental troll catches of SBT had been taken off south-eastern Australia for many years before pole-and-live-bait fishing for surface aggregations of juveniles commenced in the early 1950s. Catches steadily increased, and peaked at 21,500 mt (about 2.37 million fish) in 1982 (Figures 6 and 7; Table 1). Quotas have progressively forced down catches to just over 5,000 mt at the end of the 1980s (Table 2). The fishery commenced off NSW, where SBT 2 to 5 years old were common at the edge of the continental shelf. South Australian (SA; Figure 3) activity, on SBT 2 to 4 years old, developed in the mid-1950s, and a WA fishery, mainly for SBT 2 years old, commenced at the end of the 1960s. In the NSW and SA fisheries, aerial spotting developed, activity spread widely and further offshore, purse-seining commenced, and combined poling and purse-seine sets on schools evolved. The age of fish taken increased to about 9 years old. Off WA, in contrast, vessels remained comparatively small (7–15 m) and the fishery was predominantly confined to the continental shelf and to SBT 2 years old. In the early 1980s the NSW fishery failed and until 1991 no surface schools were reported there. However, Japanese longline catches continue, albeit at lower hook rates than at the end of the 1970s. The WA fishery has virtually stopped, catches declining from more than one million fish in the early 1980s to about 5,000 in 1992 because of the transfer of quota holdings to SA. Areas of SBT availability off SA also dwindled during the 1980s but increased to some extent after 1987. Scientists had expressed concern in the early 1980s about declining parental biomass, and as a result a quota was set for the 1983–84 season. In order to generate finance within the industry by the sale and purchase of quotas, in 1984–85 transferable quotas, allocated to vessels individually, were introduced. Another measure at the time required fishermen to re-direct activities towards older fish. This was only temporarily successful because the declining availability of surface aggregations of large fish over several years forced resumption of SA operations on predominantly 3-year-old and 4-year-old SBT, despite their lower value and the increasing economic pressure of reduced quotas. After 1990, there was a major shift in emphasis in the Australian fishery from surface operations to longlining in joint ventures with Japan. In addition, experimental cage rearing of SBT was successful to the extent that four commercial ventures commenced in 1992.

Figure 6. Japanese and Australian southern bluefin tuna catch (tonnes), 1952 to 1991 caledar years [Japanese catch for 1991 is not known, so quota for that year (see Table 2) is used].

Figure 7. Western Australian, South Australian, and New South Wales southern bluefin tuna catches, 1951–52 to 1991–92 quota years (1991–92 catch is an estimate using data from Australian Fisheries Management Authority).

The Japanese SBT longline fishery commenced in 1952, the catch peaking sharply by 1960–61 at more than 75,000 mt (about 1.2 million fish), then declining steadily to 11,000 mt by 1988 (Figure 6 and Table 1). Quotas reduced the catch to less than 7,000 mt in 1990 (Table 2). Japanese fishing operations developed on the spawning ground and later off eastern New Zealand, spreading southward in the Indian Ocean to the West Wind Drift region. They eventually ranged from the east of New Zealand, westward throughout the West Wind Drift region, to the southern Atlantic (Figure 8). Fishing effort increased steadily from commencement of the fishery, slowed between 1970 and 1980, then declined because of a 20% reduction in the number of vessels. It subsequently increased slightly as a result of an increasing trend in number of hooks per set, then declined again (Figure 9). The longline grounds east of New Zealand and off NSW are not as important for SBT as they used to be. The former area is a geographically marginal area for SBT, and the latter has been subject to closures and has seen an increase in fishing activity directed at yellowfin tuna (T. albacares), bigeye tuna (T. obesus), broadbill swordfish (Xiphias gladius), and marlin (Makaira spp.). Activity on the spawning ground and the Oki ground to its south is now directed mainly at bigeye tuna. The most important area for the SBT fishery since the 1960s has been the fishing ground south of Africa, where effort reached 50 million hooks in 1979. The hook numbers there remained around 30–40 million annually until 1990 when less than 25 million hooks were set. Globally, effort levels were in the order of 100–125 million hooks annually between 1980 and 1990, when the level declined to 63 million hooks. The hooking rate peaked soon after commencement of the fishery, then declined during and after the expansion of fishing grounds in the 1960s. In 1987 it was about 25% of the early 1970s rate. In contrast to the Australian fishing operations on surface aggregations of juveniles, the Japanese target medium-sized and large-sized SBT (70–180 cm fork length), with the proportion of medium-sized fish increasing when the fishing grounds expanded from the spawning ground. There has been a partial overlap in size composition of the surface and longline fishery catches (Figure 10). In 1971, because of increasing catches of small SBT, voluntary area-closures were adopted in the longline fishery. An area south of the spawning ground was also closed seasonally for the second half of the spawning season to protect migrating adults (Figure 11). Following recommendations by scientists for catch restraints, and several subsequent years of trilateral management discussions among Australia, Japan and New Zealand, Japan introduced a catch limit of 23,150 mt for its 1986–87 season. The Japanese catch limit was reduced, progressively, to 6,250 mt for the 1990–91 season (Table 2).

Table 1. Calendar year catch (whole weight, mt) of southern bluefin tuna by country, 1952 to 1991.

Sources: Australia, Japan, and New Zealand: data provided to trilateral scientific discussions on southern bluefin tuna; data for Japan are values which have been amended from the historical catch series, taking account of a 1990 revision of Japanese statistical areas for the southern bluefin tuna fishery; Koreak, China (Taiwan): Indian Ocean and Southeast Asian tuna fisheries data summary for 1990; IPTP Data Summary No. 12, May 1992, Colombo, Sri Lanka (note that it modifies the Data Summary 11 series); Indonesia: Fishing ground and distribution of southern bluefin tuna (Thunnus maccoyii) in southJava and Nusatenggara waters. (Doc. No. 16) Third Southeast Asian Tuna Conference, Bali, Indonesia, 22–24 August 1989; n.a. =not available.
^A Total of 5,435 mt from domestic surface fishery; 684 mt from Australia/Japan joint-venture longline vessels.
^B Indonesia catch is uncertain; tonnages in parentheses are Japanese imports of bluefin tuna from Indonesia.
^C This tonnage had been calculated using a different length:weight relationship to that for previous years; use of the previous relationship would generage a tonnage close to 10,500 mt for 1989, but this is inconsistent with recent length:weight observations. Currently there is no basis for revising the time series to take account of changes in the relationship over time.
^D Total of 134 mt from domestic handline/troll fishery; 290 mt from NZ/Japan joint-venture longline vessels.
^E Values in parentheses were provided to the Australian Fisheries Service by the Taiwan Fisheries Bureau as southern bluefin tuna catch, but appear to include all bluefins. For 1989, statistics on catch by ocean and gear suggest that 879 mt of the 1,486 mt were southern bluefin tuna.
^F Provisional total; ingnores values in parentheses for China (Taiwan) and Indonesia.
^G Total of 4,319 mt from domestic fishery; 400 mt attributed to Australia/Japan joint-venture longline vessels.
^H Total 247 from domestic fishery; 233 mt from New Zealand/Japan joint-venture longline vessels.
^I Total 2,871 mt from domestic fishery; 1,290 mt from Australian Japanese-style charter, and Australia/Japan joint-venture longline vessels.
^J Values in parentheses are taken from Fisheries Yearbook Taiwan Area, Taiwan Fisheries Bureau; they are a continuation of the series in parentheses for 1981 to 1989.
^K Total of 35 mt from domestic fishery; 125 mt from New Zealand/Japan joint-venture longline vessels.

Table 2. Southern bluefin tuna catch limits, periods of catch-limit operation, and catches realised for Australia, Japan, and New Zealand since 1983.

Definitions:

Negotiation Year - the year when the catch limit indicated was determined.
Operating Period - the period for which the catch limit applied. The Australian limit commences on 1 October of the negotiation year, the New Zealand limit on 1 January of the subsequent year, and the Japanese limit on 1 March of that subsequent year.
Catch Limit and Catch Realised are in tonnes unless indicated otherwise.

Abbreviations: s.n.c. = season not completed; n.a. = not available.

Keys to Table 2:

Figure 8. Distribution of southern bluefin tuna catch in numbers by Japanese longline tuna fishery, 1970. (From: Fisheries Agency of Japan, 1972. Annual report of effort and catch statistics by area on Japanese tuna longline fishery 1970.)

Figure 9. Trends of annual fishing effort (hooks) of Japanese longline fishery in the southern bluefin tuna fishing grounds. (Source: Nishida, T. and Y. Ishizuka, 1992. Japanese southern bluefin tuna (Thunnus maccoyii) fishery in recent years (1985–90). Working document for the eleventh meeting of Australian, Japanese, and New Zealand scientists on southern bluefin tuna, National Research Institute of Far Seas Fisheries, Shimizu, Japan.)

The small, New Zealand domestic troll and handline fisheries for adult SBT, operating alongside trawlers taking demersal species off the west of the South Island, commenced in 1980 (Figure 12). Catches reached 305 mt in 1982, declined until 1987, increased following the diversification of some vessels to small-vessel longlining, then fluctuated (Table 1). From 1989, domestic activities included the operation of four to five chartered Japanese longliners. Since the 1950s, Japanese longliners had fished in areas adjacent to New Zealand (Figure 12), and their operations there came under New Zealand control when the New Zealand 200-mile exclusive economic zone (NZEEZ) was established in 1978. Japanese catches, number of hooks set, and hooking rate in the NZEEZ declined steadily from 1980 to 1988. Catches declined from 7,600 mt to 762 mt. Although the number of hooks set halved, the hooking rate declined to one fifth over the same period. Vessels have increased the average number of hooks per set from 2,400 to 2,900 in an attempt to counteract this decline. Between 1980 and 1989 the average size of SBT taken increased from 63 kg to 95 kg, with a concurrent decline in abundance of 40–50 kg SBT. However, there has been an increase in relative abundance of 20–40 kg SBT since 1989, with a reduction in operations in the south east of New Zealand.

Figure 10. Length-frequency distribution of southern bluefin tuna caught by Australian surface fishery (dotted line) and Japanese longline fishery (solid line) in 1960, 1970, 1980, and 1987.

The access areas for Japanese longliners in the 200-mile Australian fishing zone (AFZ) have decreased progressively since the zone came into effect in 1979. The region around Tasmania (Figure 3) has become the main area of longliner operation. With the continued decline of hooking rates in the area to the south west of Tasmania, which supported a summer fishery for adults, effort has been re-directed to a winter fishery off eastern Tasmania. In this area there has been an increase in the proportion of pre-adult SBT in the longline fishery (Figure 13), probably reflecting increased escapement from the Australian surface fishery because of the progressive reduction in Australian quotas. However, the increase has been later than anticipated. In keeping with Australia's objective of diverting SBT fishing effort to larger fish, an Australia-Japan longline joint venture involving Japanese longliners and Australian industry and quotas commenced in 1989.

Figure 11. Regions and periods of the southern bluefin tuna fishing grounds closed by voluntary restrictions. (Source: Singu, 1978.)

Southern bluefin tuna catches by countries other than Australia, Japan and New Zealand have, in the past, been considered relatively trivial (Table 1). However, the SBT by-catch of driftnetters from Taiwan, and of longliners from Taiwan, Indonesia, and previously, Korea may be more significant than previously thought, especially in comparison with the considerably reduced catches of Australia, Japan, and New Zealand. Imports of bluefin to Japan (Figure 14) and anecdotal information suggest that the 1991 SBT catch by Indonesia may have been in the order of 250 mt, and that the Taiwan catch was possibly in the order of 1,300 mt.

Landed catch by boat, date of landing, port of landing, and destination (purchaser) have been recorded for the Australian surface fishery from its inception, and is currently recorded in conjunction with the monitoring of cumulative catch, during a season, in relation to quota. A logbook (Figure 15) to record details of fishing location and effort was not firmly established until the early 1980s. Southern bluefin tuna length composition has been sampled routinely in the south-eastern fishery from 1963–64, the sampling taking account of the sorting of fish at sea. Length frequency of the catch, by one cm length class and for each half-month, is available for NSW, SA, Albany and Esperance (Figure 3) (or WA, when these could not be discriminated). Spotter aircraft have been used to some extent in the NSW and SA sectors of the Australian SBT fishery since the very early days. Attempts to introduce an aircraft logbook met with little success until 1989 but some spotters' private records are available for recent years. By means of radio reports and log books (Figure 16), catch and effort data for SBT have been collected from Japanese longline vessels operating in the AFZ since 1979. From 1989 longliners have been required to measure length to caudal fork of each SBT taken in the AFZ. Observers occasionally join vessels to check that catch-reporting procedures are understood and properly maintained.

Figure 12. Locations of catch of southern bluefin tuna by New Zealand domestic (triangles) and foreign (crosses) vessels; positions respresent locations of daily position reports by domestic vessels and start-of-set locations for longliners, for the period 1980 to 1988.

Japanese tuna longliners maintain logbooks of daily activity, detailing fishing location, catches (in number of fish by species) and hooks set (Figure 17). Data tabulated by five degree grid squares, were published from 1962 to 1980. Since 1981, data have been provided directly from the Far Seas Fisheries Research Laboratory to CSIRO Australia and Fisheries Research Centre, New Zealand. At first, catch length composition had to be sampled at Yaizu and Tokyo markets but subsequently lengths were measured at sea by cooperative fishermen or derived from individual-weight data from fishing masters' private logbooks. The data have been used to generate catch length frequency for each statistical area and for each yearly quarter in two cm length classes. For up-to-date stock assessment and fishery management, timely collection and processing of catch, effort and length data are needed. A lag of 1.5 to 2.5 years had occurred routinely before Japanese longline fishery catch and length distribution data became available. However, for 1992 assessment work about 70% of the 1991 data was available, representing a major improvement in timeliness. Additionally, a real time catch, effort and length/weight monitoring programme was introduced, involving 12 longliners in 1991 and 17 in 1992. It is still desirable to improve the recovery rate of length data from longliners.

New Zealand vessels are required to complete a catch, effort and landings logbook (Figure 18). Catch data are verified independently against records maintained by licensed fish receivers. The logbook records fishing method, area or position of fishing, number of hours fishing, number of lines used and number of hooks or lures on each line, target species, estimated total unprocessed weight of all fish caught and the estimated unprocessed weight of the five most important species caught by weight in decreasing order. At the time of landing, all fish on board must be accounted for. Data on individual fish length, weight, and sex are collected by a research logbook distributed to the freezer vessel and filled in voluntarily by the crew. Japanese vessels fishing in the NZEEZ record catch and effort statistics on a set-by-set basis in a logbook (Figure 19) considerably more detailed than the logbook required by Japan. This includes information on environmental factors, target species, size composition and by-catch. A limited observer programme begun in 1987 was formalized in 1989. It extends from April to September for an estimated total of approximately 240 observer-days per year. Observers record details of the fishing method and catch composition, they measure and weigh all tuna and billfish species caught and they collect various biological samples for ongoing research programmes.

Figure 13. Southern bluefin tuna length-frequency distribution in the southeastern Australian winter (Southern Hemisphere) Japanese longline catch, 1988 to 1992.

Figure 14. Annual bluefin exports to Japan from Taiwan, Republic of Korea, Indonesia, South Africa, and Singapore. [Source: Japan Tariff Association. (monthly). Japan Exports and Imports Commodity by Country.]

Scientists at trilateral discussions since 1982 have continued to express concern about the biological state of the SBT stock, in particular its capacity to continue to maintain satisfactory levels of recruit production. Assessments of stock condition have been carried out by virtual population analysis (VPA) but attention has also been drawn to a range of direct indicators from the commercial fishery. Both sources point to a severe decline in the population. The indicators identified at the 1988 trilateral scientific discussions were: the extent and persistence of the decline in catches and hooking rate in conjunction with high fishing effort; contraction in the area of the surface fishery and the extent of productive areas of the longline fishery; and major declines in the abundance of pre-adult SBT. Catch and hooking rates have declined further, but there has been a slight increase in the extent of productive areas in the longline fishery, apparently linked to an increase in the pre-adult SBT population, mainly off the coast of Tasmania.

Figure 15. Prescribed logbook form for use by southern bluefin tuna pole-and-line fishermen operating off southeastern Australia. (Source: Australian Department of Primary Industries and Energy.)

Figure 16. Prescribed logbook form for use on Japanese-longline fishing vessels operating in the 200-mile Australian fishing zone. (Source: Australian Department of Primary Industry and Energy.)

Figure 17. Logbook form used globally on Japanese longline fishing vessels. (Source: Japan Fisheries Agency.)

Figure 18. Catch, effort, and landing return form used by New Zealand fishermen. (Source: New Zealand Ministry of Agriculture and Fisheries.)

Figure 19. Prescribed logbook form for use on Japanese longline fishing vessels operating in the 200-nautical-mile New Zealand exclusive economic zone.
(Source: New Zealand Ministry of Agriculture and Fisheries.)

Assessments based on VPAs have generally concentrated on trends in the parental biomass and in recruitment at age 1 estimated from VPAs. In particular, attention has focussed on the size of the current parental biomass in relation to its initial level, and on projected stock trajectories based on a fitted stock-recruitment relationship and a range of possible future catches. Annual variations in the Australian fishery for juvenile fish probably prevent reliable use of backwards VPA from ages less than 7 years old to estimate recruitment. Consequently, while parental biomass can be estimated from the VPAs from 1960 to the latest year for which catch data are available, estimates of recruitment at age 1 are not available for the last six years.

Results of VPAs are subject to the uncertainty inherent in input parameters such as natural mortality rate and terminal fishing mortality. Use of a range of feasible values of natural mortality rate generate quite different interpretations on state of the stock. The catch-age structures and terminal ages adopted have been questionable because of uncertainty about growth parameters used for determining age-at-length. Again, the consequences for assessments of state of the stock are significant. In the 1990 assessments, values of terminal fishing mortalities adopted were ‘tuned’ using fishing mortality rates estimated independently from tag-recapture experiments, by taking account of change over time in the concentration of effort in areas of operation of the longline fishery and by making assumptions about levels of recruitment in the early period of the fishery. In 1992, after more detailed analyses of CPUE and catch length composition data, a broader range of indices were used for tuning, in conjunction with the use of an ‘ADAPT’ VPA procedure. Even so, it has still not been possible to develop a satisfactory series of effective effort data for the fishery; therefore the tuning remains questionable and the stock assessments uncertain. Despite these shortcomings, VPAs have consistently indicated continuing decline of parental biomass to historically low levels, with the decline persisting to at least 1991 (Figure 20). The nature of trends in recruitment has been less clear, but it appears that recruitment declines had already occurred during the 1970s (Figure 21). The uncertainty in the parent stock-recruit production relationship, and the variability introduced by uncertainties in VPA input parameters, have prevented reliable projection of stock response under various catch regimes. In turn they have led to considerable uncertainty as to which advice to give to managers about appropriate quota levels. In summary, some projections have indicated that the stock will recover with the current quota levels, whereas others have indicated a continuing decline (Figure 22). Even the projections showing recovery indicate that it would be in the order of 20 years before the parent stock returned to levels at which recruitment decline commenced.

Estimated population parameters from the Australian tagging programme do not refer to the global population, but rather to the sub-population that is vulnerable to the Australian fisheries. Tagging of SBT in Australian waters began in 1959. In the 1960s and 1970s, more than 50,000 fish were tagged in the three main fishing grounds (off NSW, SA and WA; Table 3). The purpose was to delineate stock boundaries, show migration paths, confirm growth rates and assist stock assessment. New releases in 1983–84 were directed more at quantitative aspects, such as mortality rates, interactions among fisheries, estimates of the local population and survival to maturity. Special attention was also given to factors such as effects of shrinkage on growth estimates, short-term and long-term effects of tagging on growth, and consequences of tag-shedding.

Figure 20. Indicative virtual population (VPA) estimates of southern bluefin tuna spawning stock tonnage, 1960–1991. Absolute tonnages are not shown because they vary with VPA input parameters and tuning procedures. (Source: Background documentation; Australian, Japan, and New Zealand Scientific Discussions on Southern Bluefin Tuna, Shimizu, Japan, 1992.)

Figure 21. Indicative virtual population (VPA) estimates of southern bluefin tuna recruitment at age one, 1060–1986. Absolute recruit numbers are not shown because they vary with VPA input parameters and tunin procedures. (Source: Background documentation; Australian, Japan, and New Zealand Scientific Discussions on Southern Bluefin Tuna, Shimizu, Japan, 1992.)

Comparative recovery rates suggest that 2 year old SBT off WA migrate at a higher rate into ocean waters than those off SA, supporting the hypothesis that a significant proportion of fish do not travel from the WA fishing grounds to the more eastern, surface-fishery grounds. Less than 15% of SBT tagged between 1959 and 1980 have been recaptured and reported, the component from the Japanese longline fishery being 0.7% In comparison, over 40% of the fish tagged in 1983–84 were reported as recaptured by Australian fishermen, indicating very high fishing activity during the early 1980s. A substantial fraction of the fish then passing through the Australian fishing grounds as juveniles would not have survived to reproductive maturity. The year-classes of fish tagged in 1983–84 are now becoming part of the parental stock, which has already been reduced to a level where there is concern about adequate young fish production, so the situation is likely to deteriorate further. Quota reductions since 1984 should have reduced this problem dramatically. Nevertheless, the decline in the parental biomass may not be reversed if recruitment has declined at a rate faster than catches have been reduced, or if SBT escaping Australian fishing grounds are subsequently caught at a higher rate than usual in order to maintain the catches of the longline fishery.

Figure 22. Indicative trajectories of future southern bluefin tuna spawning stock tonnage to year 2010 under the current quota regime (121,500 mt globally) and assuming continuation of the historic range of variability in recruit production relative to spawning stock tonnage. Absolute spawning stock tonnages are not shown because they vary with VPA estimates generated and with the stock-recruitment relationship assumed.

The number of young fish recruited to Australian fisheries was estimated from the 1983–84 tagging data by the Peterson method. This estimated number could be an underestimate, however, because some recruits may not pass through the WA fishing grounds. The need for a real-time index of recruit abundance has been highlighted. The Japan Marine Fishery Resource Research Center (JAMARC) will complete a 5-year troll and pole-and-line survey in 1992 along the west and south coasts of Australia, which was based on the assumption that all juvenile SBT travel along the WA coast, mainly in November-December. The objectives were to assess if troll and pole-and-line surveys are effective as a means of developing a synoptic recruit abundance index; to determine patterns of occurrence-distribution of recruits; to determine what association exists between physical environmental factors and recruits, and to tag recruits to determine the level of recruitment. It is still unclear as to whether a sufficiently precise index of recruit abundance will be obtained. For the last 3 years of the JAMARC survey, a tagging programme was incorporated, in collaboration with Australia, and extending to the SA fishery region. There is optimism that recruit abundance indices may be estimated with a shorter time lag than estimates derived from VPAs. An aerial survey in progress in SA may provide an additional index if factors influencing fish occurrence at the surface can be clarified. Even so, comprehensive research surveys to determine if there are recruits which migrate directly to the central part of the Indian Ocean from the spawning ground off Java or from the west coast of Australia, are needed.

Table 3. Numbers of juvenile southern bluefin tuna that were tagged in Australian waters and those subsequently recaptured, according to fishery, year and location of tagging and region of recapture. The Australian fishing grounds are (i) Western Australia (WA), (ii) South Australia (SA) and (iii) Eastern Australia (EA). The longitudes of operation of the Japanese longline fisheries are (iv) West ≤ 60°E, (v) 60°E < Mid ≤ 120°E and (vi) 120°E ≤ East.