Interactions between surface and longline fisheries for southern bluefin tuna based on recent tagging results: the implications of reporting rates

Tom Polacheck, William Hearn and Wade Whitelaw
CSIRO Division of Fisheries
GPO Box 1538
Hobart, Tasmania 7001 Australia

ABSTRACT

Little direct evidence exist for large interactions between surface and longline fisheries for tunas. Previous tagging experiments suggest low rates of mixing and interaction among the fish caught in these two types of fisheries. Results are presented from recent and on-going tagging experiments on southern bluefin tuna (SBT) and compared with tagging results from the 1980s. In contrast to results obtained in the 1980s, the most recent tagging experiments suggests rapid and significant interchange of fish between the two fisheries as evidenced both by the relative recovery rates from the two types of gear and by the spatial and temporal distributions of the recoveries. Results are also presented for a limited number of SBT tagged and released from longline vessels indicating that substantial mixing occurs not only from surface fisheries into the longline fisheries but from longline fisheries into surface fisheries. The results do not support any previous speculations that the surface and sub-surface fish may constitute biologically or behavioural different components of the stock.

The difference in the results for this most recent tagging program and previous ones appears to be due to increased reporting rates from longline vessels as the result of very substantial tag recovery efforts which have included extensive direct contact between vessels and scientists, observers or tag liaison personnel. Thus, over 85% of the tag recoveries from Japanese longliners are the result of such direct contacts prior to vessels returning to Japan. However, comparison of tag recovery rates from longline vessels with and without observers in the same area and season suggests that recovery rates are still below 100%. Overall, the results emphasize the critical need for well developed and extensive tag recovery programs if results from large scale tagging programs are to be used to investigate and measure interactions between fisheries. Low return rates from longline vessels should not be used as an indication of lack of interaction or lack of potential for interactions between surface and longline fisheries without a careful assessment on reporting rates.

1. INTRODUCTION

The major commercial fisheries for tuna can be divided into two basic types based on the depth at which fish are captured--those that catch fish near the surface (i.e., by purse seine, pole and line, and troll gear) and those that concentrate on sub-surface fish, generally below 75 m (i.e., longline fisheries). Surface fisheries generally focus on schooling aggregations of juvenile and sub-adult fish while longline fisheries generally harvest larger and adult fish. However, fish over a wide size range can and are captured by both surface and longline gear. The catch from surface fisheries has been generally targeted for high volume markets (e.g., canning) and a single successful operation usually captures large quantities of fish (tens to hundreds of tonnes). In contrast, longline fisheries tend to target the lower volume but high valued sashimi market with a primary emphasis on fish quality. Successful operations generally catch small numbers of fish (a few to a few hundred fish per day).

Given the difference in the size range and market strategies between surface and longline fisheries, intense competition and interactions would be expected in situations in which there were moderate to high exploitation rates by the surface fisheries based on simple yield per recruit considerations (e.g., Suzuki, 1986; Fonteneau, 1987; Hampton, 1989; Miyabe, 1994). Moderate to high exploitation rates by longline fisheries could also be expected to lead to negative interactions via recruitment over-fishing, particularly in cases where there were also significant surface fisheries (e.g., Polacheck, 1994).

Somewhat surprisingly, there is little direct evidence for large interactions between surface and longline fisheries. Large scale tagging programs have been conducted for many of the main tuna stocks of various species in different regions (see Bayliff, 1993). For practical reasons (e.g., the number of fish that can be tagged per day, access to longline vessels, etc.), these tagging programs have almost exclusively tagged and released fish captured in surface operations. While large numbers of tagged fish have been recaptured by surface fisheries, the subsequent reported number of recaptures by longline vessels from these various programs has been small, both in absolute numbers and relative to the tag recapture rates from surface fisheries. Where quantitative analyses have been performed based on the tag return results, they suggest low rates of mixing and interaction among the fish caught in surface and longline fisheries (Lenarz and Zweifel, 1979; Fonteneau, 1988; Kleiber and Baker, 1987; Hampton, 1989; SPC, 1994). This has lead to speculation that the surface and sub-surface fish may constitute biologically different components of the stock at least over those size ranges caught by both surface and longline gears (Fonteneau, 1987; Polacheck, 1988; Hampton, 1994; SPC, 1994).

A major collaborative tagging program for southern bluefin tuna (SBT) involving the Commonwealth Scientific and Industrial Research Organization of Australia (CSIRO) and the National Research Institute of Far Seas Fisheries of Japan (NRIFSF) was begun in 1990. This program is on-going with from 6,900 to 16,500 tags released in each fishing season since 1990/91. Plans are in place to continue tagging for several more seasons. The results to date suggest a substantially different picture of the interaction between SBT caught in surface and longline fisheries than the general picture described above that has emerged from other tagging studies, both on SBT and other tuna species. The purpose of the present paper is to present some preliminary results documenting the interactions that are being observed and to compare these results with those from an earlier tagging experiment conducted in the 1980s. The likely reason for the difference in the results from this most recent experiment are also discussed. Overall, the results indicate the need for caution when inferring interactions or potential interactions (or lack thereof) between surface and longline tuna fisheries based on the results from past tagging experiments.

2. MATERIALS AND METHODS

2.1 Description of Tagging Programs

The basic tagging procedure used in the 1983/84 and 1990s tagging experiments were similar (see Williams, 1982, 1992; Hearn et al., 1991). Most fish tagged were caught by pole and line with a barbless hook while small numbers were caught by troll lines. In addition, a pilot experiment involving the release of longline caught fish is being conducted in the 1990s (Williams, 1993) and results from this pilot experiment are considered separately below.

After a fish had been hooked, it was hauled aboard the vessel and placed in a vinyl cradle¹. The fish were than doubled tagged by experienced operators with 12 cm dart tags inserted about 4 cm to the rear of the second dorsal fin on either side of the fish. During the 1980s a small percentage of the tagged fish were injected with oxytetracycline marking the calcareous tissue to assist in the validation of ageing from hard parts. In the 1990s approximately 50% of the tagged released fish were injected with a small dose of strontium chloride for the same purpose. Normally tags are yellow, except that orange tags are attached to injected fish. Recovery rates of tags from injected and non-injected fish have been similar (Gunn et al., 1994), but there is a need to increase the recovery rate of hard parts from orange tagged fish. Fork lengths were measured and tagged fish were returned to the water within 30 seconds of being captured.

¹ In case of the pilot longline experiment and the tagging of a small number of troll-caught fish off eastern Tasmania in the 1990s, foam mattress were used instead of vinyl tagging cradles.

² In the results below, it will be noted there are almost no recoveries from Western Australia in the 1990s. This lack of recoveries reflects the almost total cessation of commercial surface fishery operations for SBT in this area as the result of the introduction of restrictive quotas and a change in markets toward high valued sashimi products. The small size of fish that were previously caught in the Western Australian surface fishery are not suitable for the sashimi market. However, in recent years there has been some SBT caught in the general Western Australia areas by Australian/Japanese joint venture longline vessels.

Figure 1. Release and recapture locations of fish tagged in Western Australia by surface gears with times of liberty less then 1 year; (a) 1983/84 releases.

Figure 1. Release and recapture locations of fish tagged in Western Australia by surface gears with times of liberty less then 1 year; (b) 1990s releases.

In addition to these releases off the southern coast of Australia, a small number of SBT (86) were tagged and released in May and June of 1991 from nearshore waters off Eagle Hawk Neck in eastern Tasmania from troll-caught fish. In 1993, an additional 34 fish were tagged from trolling operations around Eagle Hawk Neck. In 1992, a trial release of 88 small fish was conducted from a Japanese longline vessel fishing in nearshore waters off eastern Tasmania and, in a continuation of these trials, an additional 364 were released in 1993 (Williams, 1993).

2.2 Description of Recovery Activity

Historically, the major fisheries for SBT have been the Australian surface fishery and the Japanese longline fishery with smaller scale New Zealand surface and longline fisheries (see Caton, 1994, for more detail). Recently, there has been increased longline catches of SBT by Taiwan and Indonesia plus small catches reported by Korea. The Australian surface fishery catches predominantly juvenile fish using primarily purse seine and pole-and-line gear. While historically, there were fisheries off New South Wales, South Australia and Western Australia, the juvenile fishery off NSW collapsed in the late 1970s/early 1980s due to lack of fish and the Western Australian fishery decline rapidly after 1985 with the introduction of catch quotas. In the 1990s, over 90% of the surface Australian catch has come from South Australia. While the Japanese longline fishery began on the spawning grounds south of Indonesia in the 1950s, by the late 1960s the fishery had shifted away from the spawning grounds and expanded to areas encompassing the southern feeding grounds (i.e., areas 3-9, Figure 2). The longline fishery catches both juvenile and adult fish. Effort in the fishery increased through the early 1980s but has decreased in recent years as the result of the catch quotas.

Figure 2. Large scale statistical areas used for reporting Japanese longline SBT catch and effort data (Note that the definition of areas 6 and 7 was revised in 1994).

2.2.1 Australian Fisheries

Tag recovery from Australian fishermen in both the 1983/84 and 1990s experiment depended to a large extent on field officers who had close and frequent contact with fishermen and processing workers. For each tag returned, the finder was given a reward (monetary in the 1983/84 and a choice of monetary, t-shirt, cap or coffee mug in the 1990s) and a letter or certificate with the corresponding fish release and recapture information. For both sets of experiments, it was possible to achieve a high level of awareness of the programs because there were relatively few vessels fishing for SBT and few companies processing them (the number of vessels in the 1990s were fewer than in the 1980 period) and there was an active liaison officer who had extensive contact and interactions with the industry.

For the 1983/84 experiments, meetings were held with fishermen and articles published in trade magazines explaining the need and purpose of the programs (Majkowski and Murphy, 1983; Williams, 1983). In addition, rewards posters, pre-paid postage envelopes and recapture forms were sent to all quota holders and processing establishments. In addition, a lottery in which all tag returners during the 1986/87 season were entered was well advertised and winners drawn in 1987. The total monetary rewards could be substantial for individual tag returners in the 1980s because of the high recovery rates (see below).

Table 1. Number of tags released by area, month and year.

Year	Month	Area	Number of releases
1983	January-March	W. Australia	4591
	August-October	W. Australia	2365
1984	January-February	S. Australia	3232
1990/91	November-January	W. Australia	3307
1991	January-February	S. Australia	4352
	May	Tasmania (trolling)	86
1991/92	December-January	W. Australia	3052
1992	January-February	S. Australia	4998
	June	Tasmania (longline)	88
1992/93	November-March	W. Australia	4941
1993	January	S. Australia	5873
	May-June	Tasmania (longline)	364
	April-June	Tasmania (trolling)	34
1994	January-March	W. Australia	8679
	January-February	S. Australia	7629
	June	Tasmania (longline)	196

Table 2: Number of tags released by estimated cohort and release year.

Cohort	Release Year1
	1983	1984	1991	1992	1993	1994
1980	3	8
1981	322	40
1982	5667	692
1983	945	2478
.....
1988			384
1989			3405	618	5	2
1990			3956	5458	2367	94
1991				2061	3824	2852
1992				1	5016	4261
1993						9281
no length	19	14				14
TOTAL	6,956	3,232	7,745	8,138	11,212	16,504

1) From November of the previous year through October of the current year

2.2.2 Japanese longline fisheries

In the 1983/84 experiment, CSIRO scientists depended primarily on their contacts with Japanese SBT research scientists through the SBT assessment process for promoting and recovery of tags, although the cooperation was also sought from the Federation of Japan Tuna Fisheries Cooperative Associations at various fora. Rewards and tag certificates were sent to tag finders in Japan, but this often involved a long delay because of the distance and logistic problems. Furthermore, the reward often went to the vessel owner or fishing master and not the actual tag finder. In 1987, a poster promoting the program was produced in Japanese for distribution to vessels in Japan. There is little information available on the extent of direct contact that was made with vessels and crews to publicize the program or on the distribution of promotional materials. After 1987, shipping agents and inspectors were asked to inquire for tags when Japanese vessels visited the Australian ports of Hobart and Fremantle. When tags were returned in Hobart, rewards were generally immediately given to tag finders by CSIRO scientists. In addition, Australian observers aboard Japanese vessels fishing with in the AFZ were aware of the program, but the actual observer coverage was very low during the 1984 to 1987 period (no more then a total of two weeks of coverage on a single vessel in any given year).

In the 1990s, a much more extensive effort has been developed to promote the tagging program and encourage the reporting of tags from recaptured fish from Japanese longliners. This has been made possible, in part, by the collaborative framework under which the research is being done and by the closer personal links that have developed between Australian and Japanese scientists and industries. It has also been made possible by the appointment within CSIRO of a full time industry/science liaison officer for tuna fisheries. A major components of this job includes responsibilities for tag promotion and recovery activities. The more extensive effort to promote the current tagging program has included:

1) production of posters and informational material in Japanese;

2) personal port visits to nearly 100% of all Japanese longline vessels during port calls in Australia. Since most vessels are required to undergo both a pre- and post- vessel inspection when fishing within Australian waters, this means that most vessels are visited twice during any season. Visits in Hobart and Fremantle are conducted by CSIRO personal while visits in Sydney and Brisbane are conducted by contracted Japanese speaking liaison officers (Table 3). We have avoided using marine police or inspectors for this role, although they have been informed about the program and a small number of tags have been returned directly to them;

3) promotion and direct recovery of tags by Australian Fishing Zone (AFZ) observers while on board vessels. In the 1990s, observer coverage within Australian waters has been about 10-20% of the total effort within the zone during the 1990s compared to very low levels of coverage in the 1980s. There has also been limited high seas observer coverage in the 1990s under the collaborative international Real Time Monitoring Program (RTMP);

4) personal port visits by contracted Japanese speaking liaison officers in Cape Town similar to those conducted in Australia (Table 3). Because of administrative constraints, these officers are only allowed to visit approximately 20-30% of the Japanese longline vessels stopping in Cape Town;

5) rapid (often immediate) and direct feedback to tag finders in terms of rewards and tag certificates with increased release/recapture information. Thus, rewards are given directly at time of tag collection during vessel visits and tag certificates are being faxed directly to vessels while at sea when the recovery of tags is reported by observers via radio;

6) direct presentations and discussions during pre-season meetings in Hobart with fishing masters and captains in 1992 and 1993 (about 30 attended each year);

7) direct contact with Japanese industry port agents from both the Federation of Japan Tuna Fisheries Cooperative Associations and National Ocean Tuna Fishery Association (ZENGOYREN). This has resulted in their cooperation in the distribution of promotional material, arranging contacts with vessels and collection of tags;

8) placement of advertisements in Japanese in the fax newspaper which is distributed regularly to Japanese longline vessels. These were prepared and sponsored directly by the Federation of Japan Tuna Fisheries Cooperative Associations;

9) visits in Japan by CSIRO scientists with Japanese researchers and scientific port agents that have direct contact with Japanese longline vessels when they return to Japan. This has resulted in increased cooperation in tag promotion and collection;

10) direct contact with New Zealand scientist and fishing organizations which has also resulted in their cooperation in the distribution of promotional material and collection of tags.

Year	CSIRO Scientist (Hobart and Fremantle)	Contracted Japanese Speakers
		Cape Town	Sydney	Brisbane
1991	8	0	0	0
1992	105	0	41	19
1993	115	72	37	10
1994*	93	111	30	2
Total	321	183	108	31

* through September 1994

2.2.3 Other fisheries

No promotional activities were carried out in the 1980s for other fisheries that are known to harvest or are suspected of harvesting SBT (see Caton, 1994). In the 1990s, the port visitation program in Cape Town has also included 143 visit to Taiwanese longline vessels using that port. Also, Indonesian authorities have been informed about the program and CSIRO scientists have direct contact with Indonesian tuna industry and research organizations through a joint research program being conducted on SBT in Indonesia. Tagging posters have been translated into Indonesian and distributed. To-date, no tags have been recovered from either Taiwanese or Indonesian vessels. However, it should be noted that the Indonesian vessels are fishing on the spawning grounds and that the size and age of fish being caught are greater than the size and expected age of the largest tagged fish currently at liberty from the 1990s experiments.

2.3 Estimation of the Age Compositions of the Commercial Catch and Tagged Fish

In the results presented below, estimates of the reported recapture rate by cohort per 1000 fish harvested are made for various components of the SBT fisheries. The values used for the estimated number of fish caught per cohort in a given year were based on the procedures developed for estimating the age-composition of the catch for input into the virtual population assessments of the SBT stock (Polacheck et al., 1994; Hearn, 1994; Klaer and Polacheck, 1993). The estimation of age was based on a knife-edge partitioning of estimated length frequency distributions of the catch using a two stanza von-Berlalanffy growth curve. The growth parameters used here are for the "base" case and take into account increased growth rates of SBT in the 1980s compared to the 1960s (Anon, 1994, Hearn, 1994). The same age estimation procedure was used to assign tagged fish to cohorts based on their measured length at release. Results by area of the estimated reported recapture rates per 1000 fish harvested are also presented for the Japanese longline fishery. These results use the main statistical areas that have been defined for reporting SBT catches (Figure 2).

The statistical properties of these reported recapture rates are not straight forward. If one assumes that the recapture process is essentially random within a fishery (i.e., complete mixing of tagged and untagged fish), then the recovery process can be considered to be approximately a Poisson process. In this case, the coefficient of variation in the recovery rates conditional on the estimated total number of fish harvested is a simple function of the actual number of tags recovered. The problem in determining an unconditional variance is that variance and sampling distribution associated with the estimates of the total number of fish harvested from a cohort (i.e., the denominator in the estimated recovery rate) are not known. Table 4 provides the calculated coefficient of variation assuming that the recovery process is Poisson and conditional on the estimated total number of fish harvested as a guide to the minimum level of uncertainty associated with these recovery rate estimates.

Table 4. Calculated coefficient of variation for estimates of tag recovery rates based on the number of tags recovered conditional on the estimated number of fish harvested and assuming that the recovery process is Poisson.

Number of tags returned	Coefficient of variation
1	100%
2	71
3	58
4	50
5	45
10	32
20	22
30	18
40	16
50	14
75	12
100	10

3. RESULTS - EVIDENCE FOR INTERACTIONS

3.1 1983/84 Tagging Results

Of the 10,188 fish tagged in the 1983/84 tagging experiment 4,135 tag recaptures were reported to CSIRO. Ninety-nine percent of these were caught in the surface fishery and only 57 (1.4%) were reported by Japanese longline vessels (Table 5). Overall, there were very few reported longline recaptures in the waters around Australia and the first reported longline recaptures in combined statistical areas 4 and 7 occurred in 1987 (Table 6).

Some 94% of all the reported recaptures were estimated to belong to the 1982 and 1983 cohorts (Table 7). Comparison of the estimates of the tag recovery rates from these two cohorts for the period from 1984 through 1987 between the Japanese longline fishery and Australian surface fishery indicate a substantially greater recovery rate in the surface fishery than in the Japanese longline fishery (Tables 8 and 9). Thus, tag recovery rates per 1000 fish caught were generally between 3 and 10 times greater in the surface fishery. Furthermore, as the tagged fish aged and became less available to surface fishery, there is no indication of a decline in the discrepancy in the recovery rates between 1984 and 1987 in the two fisheries during this period. If anything, the discrepancy became greater. Overall, only about 17% of the expected number of tags were recovered from the Japanese longline fisheries for the 1982 and 1983 cohorts between 1984 and 1987 based on the estimated recovery rates from the Australian fishery and estimated number of longline caught fish from these cohorts³.

³ The total expected number of recaptures was calculated by multiplying the observed recovery rate for a cohort in given year in the domestic surface fishery by the corresponding estimated catch by the Japanese longline fleet and summing these across the 1982 and 1983 cohort for the years 1994-87.

Year	Number of Recaptures
	Longline	Surface
1983	0	1409
1984	7	1393
1985	11	808
1986	6	410
1987	9	55
>1987	24	3

Table 6. Number of reported recaptures of tagged fish from the 1983/84 tagging experimented by Japanese longliners by statistical area and year.

Year	Number of Recaptures
	Area 4	Area 5/6	Area 7	Area 8	Area 9
1984	0	0	0	6	1
1985	0	0	0	5	6
1986	0	1	0	2	3
1987	2	0	0	0	7
1988-90	1	0	5	1	3
1991-94	0	2	6	4	2

Table 7. The percentage of recaptured tagged fish from the 1983/84 tagging program which are estimated to have belonged to the 1982 and 1983 cohorts based on their measured length at release.

Year of Recapture	Cohort		Total number of recaptures with release lengths
	1982	1983
1983	94.3%	3.2%	1,404
1984	48.9	49.3	1,494
1985	43.9	53.3	816
1986	33.0	65.8	415
1987	46.9	50.0	64
>1987	40.7	44.4	27

Table 8. Tag recovery rates (i.e., number of reported tag fish recovered per 1000 fish harvested from that cohort) that belonged to the 1982 cohort based by year of recapture and gear.

Year	Surface	Longline
1984	1.41	0.94
1985	2.69	0.21
1986	3.75	0.41
1987	4.20	0.87

Table 9. Tag recovery rates (i.e., number of reported tag fish recovered per 1000 fish harvested from that cohort) that belonged to the 1983 cohort based by year of recapture and area.

Year	Surface	Longline
1984	0.86	0.00
1985	1.43	0.50
1986	1.41	0.15
1987	0.88	0.14

There were no longline tag recoveries between 1983 and 1987 from any year class in areas 3 and 7 (see Figures 2 and 3, and Table 6). These statistical areas adjoin the surface fisheries and tagging areas in Western and South Australia (note that there was no longline effort or catch reported in area 3 in these years). The longline fishery in these areas was small during this period and only a total of an estimated 1,935 and 1,289 fish were harvested during this period from these two areas by Japanese longliners from the 1982 and 1983 year classes. Nevertheless, based on the recapture rates in the surface fishery, approximately 7 tags would have been expected to have been recovered in these area if in fact the tagged fish were uniformly mixing between the two fisheries. It is worth noting that during the period since 1988 a total of 11 tags have been recovered from area 7 (more than from any of the other statistical areas). While the estimated number of fish harvested in these areas from the 1982 and 1983 cohorts was relatively small, the recovery rate from 1988 through 1990 remained low (0.46 per thousand fish caught based on five recoveries from an estimated catch of nearly 11,000 fish). Coinciding with the commencement of the current tagging program in late 1990, there has been a marked increase in the recovery rate of 1983/84 tagged fish (2.84 based on five recoveries from an estimated catch of nearly 1,800 fish).

Figure 3. Release and recapture locations of tagged fish from the 1983/84 experiment by Japanese longline vessels; (a) recaptures through 1987.

Figure 3. Release and recapture locations of tagged fish from the 1983/84 experiment by Japanese longline vessels; (b) recaptures after 1987.

3.2 1990s Tagging Results

The results with respect to implications for interactions between surface and longline fisheries are substantially different for the 1990s experiment compared to the 1980s. Of the 42,832 fish that were tagged off Western and South Australia, 1498 of these have been recaptured as of 21 November 1994 (i.e., 3.5% to date compared to the 40.5% total for the 1983/84 releases). The large difference in the magnitude of the return rate reflexes both the imposition of restrictive catch quotas and a shift in the fishery towards larger fish. However, the relative distribution of recaptures between surface and longline gear does not show the large discrepancy that was observed in the 1980s. Thus, approximately 44% of the reported returns have been recaptured using longline gear (Table 10) and these have come from all of the Japanese longline statistical areas in which significant numbers of juvenile SBT are currently harvested (Table 11). There have been major changes in the SBT fishery that have contributed to this difference (see Caton, 1994; Caton et al., 1995; Nishida, 1994, for more detail). These include:

1) a large decrease in the absolute size of the surface fishery (from around 20,000 mt to about 2,000 mt) including the near cessation of any commercial catches in western Australia;

2) a decrease in the size of the surface catches relative to the global longline catches, particularly for juvenile fish. (Thus, in 1993, the surface fishery caught only 46% of the total number of age 2 to 4 year old fish compared to 90% in 1984);

3) the development of a small (approximately 250 mt) Australian domestic longline fishery, primarily off New South Wales;

4) the introduction of restrictive catch quotas for the Australian, Japanese and New Zealand fleets (there has been a subsequent increase in catches by other parties -particularly by Taiwan and Indonesia);

5) seasonal and area changes in the distribution of Japanese longline effort with a major concentration of effort in the winter fishery around Tasmania.

6) the development of joint venture arrangements among Australian quota holders and Japanese longliners resulting in increased summer and winter longlining by Japanese vessels within Australia's AFZ, especially adjacent to Tasmania and the southwest coast of Australia.

Table 10. Number of tag recoveries to date from releases off Western Australia and South Australia from the 1990s tagging experiments by year, gear and vessel nationality.

Year	Australian		Japanese Longline	Total
	Surface	Longline
1991	165	3	52	220
1992	206	18	131	355
1993	229	33	223	485
1994	251	47	152	450
Total	8521	101	558	1,512

1) one tag was recovered in the Australian surface fishery for which it was impossible to ascertain even an approximate date of recapture.

Year	Statistical Area
	2	4	5	6	3 & 7	8	9
1991	0	1	0	1	36	10	3
1992	0	11	0	5	89	22	3
1993	0	38	3	0	135	18	26
1994*	1	15	0	2	80	27	28
Total*	1	65	3	8	340	77	60

* through November 1994

Data on the release and recapture position of tagged fish provide one of the only sources of direct information on fish movements and thus the potential interactions between spatially disjoint fisheries. Results on the movement of SBT tagged in 1990-94 are presented in Figures 1 and 4 to 8. Each arrow indicates the release and recapture position of one recaptured fish. The level of fishing effort and catch in different areas needs to be considered when interpreting these maps. Fishing effort is highly seasonal for all areas where tagged fish have been recovered. In the 1990s, the only substantial surface fishery for SBT occurs off South Australia in the Great Australian Bight in the austral summer and autumn with small amounts of trolling activity off eastern Tasmania.

As was found in previous tagging studies, these maps indicate that there is a relatively rapid movement of fish from the surface fishery in Western Australia to the surface fishery in South Australia and that these fish remain vulnerable for up to four years after being tagged (Figure 4). However, a feature not found in previous studies (in part due to the lack of a substantial fishery off Tasmania) is the very rapid southeast movement of fish from both South Australia and Western Australia into the longline fishery off eastern and southern Tasmania (Figure 5). Fish tagged off South Australian in February are caught by offshore Japanese longline vessels fishing off Tasmania in May and June of the same year that they were released. (Note that the longline season off Tasmania has begun somewhere between early May and early June during 1991-94.) There also appears to be a relatively rapid movement of fish from Western Australia to Tasmania with a number of fish being caught off Tasmania within six months from their time of release in Western Australia. The number of short term recaptures from Western Australia is less than the number from South Australia. This probably reflects size related differences in movement patterns and possibly gear selectivity.

Figure 4. Release and recapture locations of surface tagged fish from Western and South Australia recaptured by surface gear from the 1990s tagging experiments;

(a) recaptures with time at liberty less than 1 year

(b) recaptures with time at liberty between 1 year and two years

(c) recaptures with time at liberty greater than 2 years.

Figure 5. Release and recapture locations of surface tagged fish from Western and South Australia recaptured by Japanese longline vessels (including joint venture operations) from the 1990s tagging experiments;

(a) recaptures with time at liberty less than 7 months

(b) recaptures with time at liberty between 7 and 12 months

(c) recaptures with time at liberty between 1 year and two years

(d) recaptures with time at liberty greater than 2 years.

Figure 6. Release and recapture locations of surface tagged fish from Western and South Australia recaptured by Domestic Australian longline vessels from the 1990s tagging experiments;

(a) recaptures with time at liberty less than 7 months

(b) recaptures with time at liberty between 7 and 12 months

(c) recaptures with time at liberty between 1 year and two years

(d) recaptures with time at liberty greater than 2 years.

Figure 7. Release and recapture locations of longline tagged fish recaptured from the 1990s tagging experiments;

(a) recaptures with times at liberty less than 1 year by longline vessels

(b) recaptures with times at liberty less than 1 year by surface gear.

(c) recaptures with times at liberty greater than 1 year by longline vessels

(d) recaptures with times at liberty greater than 1 year by surface gear

Figure 8. Release and recapture locations of troll-caught fish tagged of eastern Tasmania from the 1990s tagging experiments;

(a) recaptures with times at liberty less than 4 months

(b) recaptures with times at liberty greater than 4 months.

The rapid movement of surface tagged fish into the Japanese longline fisheries is not confined to waters around Tasmania. A large number of the fish tagged off South Australia have also been caught by longliners fishing to the southwest of Western Australia (Figure 5). However, few fish have been caught in this area by Japanese longliners from the releases of predominantly younger aged fish in Western Australia. There have also been a number of short-term recaptures from releases in South Australia caught by the Japanese fleet operating off South Africa. For example, two fish (released in 1991) were caught off South Africa less than 6 months after their release off South Australia. A total of seven recoveries have been reported off South Africa within one year of release from Western or South Australia.

The spatial distribution of tag returns from Japanese longliners for fish that have been at liberty greater than one year compared to tagged fish which were recovered within one year of release are similar except for an increase proportion being recovered off South Africa for longer times at liberty.

The development of the domestic Australian longline fishery has resulted in a substantial number of recaptures. This is somewhat surprising given the small size of this fishery (less than 350 mt per year or between 1 and 22% of the annual surface catch in the 1991-93 period). There is a rapid movement of surface tagged fish within the Great Australian bight into the longline fishery just off the shelf of the bight (Figure 6). Thus, there have been 12 recaptures of surface tagged fish by domestic Australian longliners within 3 months of their release with the shortest time at liberty being 16 days. Also, 12 out of the 56 recaptures of surface tagged fish by Australian longliners fishing off the east coast of Australia have occurred during the domestic longline fishing season during the same year (May to November) in which the fish were tagged (Figure 6).

The tag recaptures from the pilot longline release off-eastern Tasmania also indicate a movement of tagged fish into the surface fishery off South Australia (Figure 7, Table 12). Thus, tagged fish were recovered during the next season in South Australia after tagging (i.e., longline fish were released in late May and June and the South Australia season begins in late December/early January). Previous to these pilot longline releases, the movement of SBT has generally been considered to be uni-directional such that once fished entered the longline fishery there was no substantial movement back into the Australian surface fisheries (e.g., Majkowski et al., 1988; Caton, 1991; Hampton, 1991). The results from these longline releases are the first evidence of substantial movement of SBT from longline fishing grounds into surface fisheries. However, results reported by Shinju (1978) based on the recovery of two tags in the surface fishery off South Australia from 52 release in 1977 from a longliner operating southwest of Australia provided some previous indication that such movements were possible. The prior lack of evidence was due to the lack of opportunities to tag fish aboard longline vessels and the results illustrates the importance of being able to tag fish from all areas in which fisheries occur.

There has also been a single recapture in both the troll fishery off eastern Tasmania and the New South Wales recreational fishery from these longline releases. In addition, there has also been the recaptures of longline released fish by both Japanese and Australian longline vessels. However, to date, all of these have been from the waters around southeast Australia with no longline recaptures from either the southeast Indian Ocean or South Africa.

Table 12. Number of tags recovered from pilot tagging experiments off eastern Tasmania based on recoveries through November 1994.

		1991 Releases from Troll Caught Fish	1992 Releases from Longline Caught Fish	1993 Releases from Longline Caught Fish	1993 Releases from Troll Caught Fish
Number released		86	88	364	33
Total Number recaptured		15	11	45	12
Fleet	Time at Liberty
Japanese Longline	<30 days	12	0	5	5
	>30 days	1	2	5	5
AustralianLongline	<30 days	0	0	2	0
	>30 days	0	3	8	1
AustralianSurface	<30 days	0	0	0	0
	>30 days	2	6	25	1
Mean release length		89.1 (s.e. = 0.21)	90.4 (s.e. = 0.56)	93.6 (s.e. = 0.30)	94.3 (s.e. = 0.96)
First release date		28/5/91	10/6/92	29/5/93	28/4/93
Last release date		19/6/91	14/6/92	11/6/93	2/6/93

The small scale tagging of troll-caught fish in the inshore waters off eastern Tasmania resulted in a relatively large proportion of these surface caught fish being caught in the adjoining Japanese longline fishery off eastern Tasmania (Figure 8, Table 12). Thus, 12 out of 86 fish tagged in 1991 were recaptured by Japanese longline vessels within 30 days of their being released and 5 out of the 34 releases in 1993. This suggests a high degree of interaction and movement between the surface and longline fishery in this area and time period. Similar to the recapture of longline tagged fish from eastern Tasmania, all recaptures to date from these surface releases off eastern Tasmania have been from waters around Australia. Overall, the recoveries of tagged fish released off Tasmania indicate rapid and significant interaction between the inshore and adjoining offshore fishery around Tasmania.

3.4 Comparative Recovery Rates

The estimated cohorts for which tag recoveries have come from the 1990s experiments are spread across the 1988 to 1992 cohorts (Table 13) with the largest fraction (43%) coming from the 1990 cohort. Estimates of the recovery rate of tagged fish from these various cohorts by area and year can be calculated for 1991 to 1993 (at the time of drafting this paper, the catch data were not yet available for 1994 while the catch statistics for 1993 were preliminary estimates).

Comparison of the tag recovery rate per thousand fish caught for the Japanese longline fleet and the Australian domestic surface fishery for the 1990s tagging experiments suggest a very different picture than the rates observed in the 1980s (Tables 14 and 15). Thus, the rates for the two fisheries in the 1990s are of the same order of magnitude and the surface recovery rates are generally less than twice as large as rates in the longline fishery. The one exception is for the recovery rate of two year old fish in 1993 (16.86 versus 0.0). However, this rate is based on only 13 recoveries of small fish. Overall, the reported number of longline returns is between 55% and 67% of the expected numbers based on the recovery rates for the Australian surface fishery and the estimated number of longline fish caught from each cohort and each year for the 1989 to 1991 cohorts. In addition, the relative recovery rates between the Japanese longline fishery and the Australian domestic surface fishery are similar for tagged fish recovered in the year that they were release and for those with longer times at liberty.

Table 13. The percentage of recaptured tagged fish from the 1990s tagging program which are estimated to have belonged to various cohorts based on their measured length at release.

Cohort	Recapture Year
	1991	1992	1993	1994*
1988	17.4	1.1
1989	68.9	49.3	1.6	0.6
1990	13.7	48.7	14.5	6.5
1991		0.8	56.1	37.1
1992			25.4	38.8
1993			2.5	14.9
no length				2.0

* recoveries through November 1994

Cohort	Year recovered
	1991	1992	1993
1988	0.30 (25)	0.00 (0)	0.00 (0)
1989	1.57 (99)	1.84 (101)	2.24 (12)
1990	0.47 (29)	4.16 (104)	3.26 (141)
1991	0.00 (0)	0.19 (2)	2.02 (63)
1992	-	0.00 (0)	18.76 (13)

Table 15. Tag recovery rates for the 1990s tagging program (i.e., number of reported tag fish recovered per 1000 fish harvested) from the Japanese longline fisheries by year and cohort. The numbers in parenthesis are the number of tags reported.

Cohort	Year Recovered
	1991	1992	1993
1988	0.30 (13)	0.08 (4)	0.18 (7)
1989	1.50 (49)	1.16 (66)	1.25 (56)
1990	0.67 (1)	1.67 (61)	2.13 (122)
1991	0.00 (0)	1.28 (1)	1.48 (46)
1992	-	0.00 (0)	0.00 (0)

An examination of the recovery rates by area from the Japanese longline fishery suggests that the tagged fish are dispersing and are vulnerable to being harvested in all of the major longline fishing areas. Thus, the recovery rates in all of the major SBT statistical areas in which juveniles are harvested are of the same order of magnitude as those in the surface fishery (Table 16). The recovery rates tend to be the lowest in area 9 and highest in area 8, although for the later the actual number of tags recovered is small reflecting the low catch rate of fish from cohorts of fish which have been tagged. (Note that the small number of tags recovered from New Zealand waters (areas 5 and 6 in Table 11) reflects the small number of juveniles that are caught in this area rather than any evidence for lack of interactions between surface and longline fisheries.

The recovery rates from the longline component of the Australian domestic fishery are generally higher than those from other areas of the Australian domestic fishery and from the Japanese longline fishery (Table 17 compared to Tables 14 and 15). The recovery rates from the domestic longline fishery exceed those from the Japanese longline fishery in the same areas (areas 3,4 and 7) by a factor of two to six. The domestic longline fishing season continues after the end of the Japanese season. As such, returns rates might be expected to be greater during the first season after release if mixing/diffusion of tag fish into the New South Wales area was still occurring. However, similar differences in the recovery rates also exist for tags which have been recovered with times at liberty exceeding one year (Table 18).

Table 16. Tag recovery rates for the 1990s tagging program (i.e., number of reported tag fish recovered per 1000 fish harvested) from the Japanese longline fisheries by year, cohort and statistical area. The numbers in parenthesis are the number of tags reported.

Area	3	and	7	Area	4	Area	8	Area	9
1991	1992	1993	1991	1992	1993	1991	1992	1993	1991	1992	1993
1988	0.74 (11)	0.15 (2)	0.42 (3)	0.00 (0)	0.00 (0)	0.11 (1)	0.00 (0)	0.33 (1)	0.00 (0)	0.13 (2)	0.00 (0)	0.05 (1)
1989	2.11 (35)	1.70 (47)	2.06 (27)	0.50 (2)	0.56 (4)	1.46 (14)	1.15 (9)	2.07 (9)	2.40 (5)	0.26 (1)	0.18 (3)	0.41 (8)
1990	0.00 (0)	1.57 (41)	2.60 (72)	0.00 (0)	1.15 (7)	1.74 (22)	1.08 (1)	3.62 (12)	2.85 (10)	0.00 (0)	0.00 (0)	1.27 (16)
1991	-	0.00 (0)	1.86 (39)	0.00 (0)	0.00 (0)	0.73 (2)	-	0.00 (0)	0.99 (4)	-	0.00 (0)	0.99 (1)
1992	-	0.00 (0)	0.00 (0)	-	-	0.00 (0)	-	-	0.00 (0)	-	-	0.00 (0)

4. DISCUSSION

The results from the most recent SBT tagging experiments suggest very different conclusions about the potential for significant interaction between longline and surface fisheries than the results from previous SBT tagging studies⁴ or from tagging studies on other tuna. The results suggests rapid and significant interchange of fish between the two fisheries as evidenced both by the relative recovery rates from the two types of gear (e.g., Tables 14 and 15) and by the spatial and temporal distributions of the recoveries (Figures 1 and 4 to 8). The mixing between surface and longline fisheries in essentially the same area is apparently very rapid based on the results off eastern Tasmania and on the returns from domestic Australian longliners off South Australia. However, even for spatially disjoint fisheries, there is significant interchange occurring within 3 to 6 months of surface tagged fish being released. Substantial number of surface tagged fish from Australia are being recovered in longline fisheries as distant as South Africa. In addition, the limited results from longline tagging indicate that mixing occurs not only from surface fisheries into the longline fisheries but from longline fisheries into surface fisheries. The results do not support any previous speculations that the surface and sub-surface fish may constitute biologically or behaviourally different components of the stock.

⁴ It should be noted that a large tagging program for SBT was also conducted throughout the 1960s. A thorough analysis and presentation of the results from this programs was beyond the scope of the present paper and any analyses of the 1960s tagging results in terms of the recovery rates by cohort for the Japanese longline fishery would have greater uncertainty than those from the 1983/84 and 1990s programs presented here. There is greater uncertainty in interpretation because (1) a large fraction of the longline effort (over 50%) took place either in the spawning or spawning staging grounds (areas 1 and 2 in Figure 3) prior to 1966 and(2) concerns exist about the reliability of the length frequency data for the commercial catch prior to 1966 while 56% of the 1960s tag releases occurred prior to 1966. However, comparisons of the overall relative recovery rates by cohort from the Australian surface and Japanese longline fishery yielded similar results and a similar overall view of the potential interactions between these two fisheries as those from the 1980s.

Cohort	Year Recovered
	1991	1992	1993
1988	0.00 (0)	0.00 (0)	0.86 (1)
1989	-	10.40 (10)	2.57 (5)
1990	-	18.74 (8)	4.62 (17)
1991	-	-	4.62 (11)
1992	-	-	0.00 (0)

Table 18. Comparison of tag recovery rates (i.e., number of reported tag fish recovered per 1000 fish harvested) between the Australian domestic longline fishery and the Japanese longline fisheries in the same areas for recoveries which occurred subsequent to the year in which the fish were tagged by year and cohort.

Fishery	Cohort	Year Recovered
		1992	1993
Australian	1988	0.00	0.86
Domestic	1989	8.32	2.57
Longline	1990	9.37	4.08
	1991	-	2.10
	1992	-	0.00
Japanese	1988	0.15	0.42
Longline	1989	1.31	1.99
Areas 3 & 7	1990	0.46	2.13
	1991	0.00	0.24
	1992	0.00	0.00
Japanese	1988	0.00	0.11
Longline	1989	0.28	1.46
Area 4	1990	0.49	1.34
	1991	0.00	0.00
	1992	-	0.00

Furthermore, recent results on the swimming and diving of SBT also suggest that the variation in behaviour exhibited by individual fish would indicate that individuals would be vulnerable to capture by both surface and sub-surface gear. Thus, both sonar tracking and archival tag results have shown that within a single day fish make frequent and large changes in the depth at which they are swimming (Davis, 1993, 1994a, 1994b; Gunn et al., 1994). Periods of prolong times at the surface are occur almost exclusively in the late mornings and afternoons. Figure 9 is an example of the depth distribution within a single day for a juvenile SBT off the shelf of the Great Australian Bight based on data collected with an archival tag. The fish utilizes and "explores" a wide range of depths from the surface to over 100 m. This fish was caught twice (i.e., when first tagged and subsequently when recaptured) by surface gear. Data from the archival tag strongly suggests that during the interim period that it was feeding at depth.

The critical question in interpreting the difference in the 1983/84 tagging results with those from the 1990s is what is the reason for the much higher recovery rate from longliners in the 1990s. The fundamental difference does not appear to be biologically or fishery related but appears to be the result of a much higher reporting rate of recaptured tagged fish as a result of improvements in the promotion of tag recoveries. Thus, there has not been any fundamental change in how the Japanese longliners set their gear (e.g., depth of setting, hook size, baits, etc.) and it is hard to conceive of any factor that could result in a major behavioural change in feeding/vulnerability of the tagged SBT to longlining. As noted above, there have been seasonal and area shifts in the distribution of longline efforts. However, for this to explain the difference in the recovery rates would require a substantially disproportional higher occurrence of tagged fish in these areas/seasons relative to untagged fish of the same size/age. In contrast, there have been very substantial changes in the efforts and methods used to increase the reporting rate from Japanese longliners.

Figure 9. Depth distribution for 13 March 1994 for a juvenile SBT recorded by an archival tagged fish that had been at liberty for three months. The position of the fish was near but off the shelf of the Great Australian Bight.

Prior to the 1990s tagging experiment, there was little contact between CSIRO scientists conducting the tagging program and Japanese longline vessels. CSIRO scientists depended primarily on their contacts with Japanese SBT research scientists through the SBT assessment process for the promotion and recovery of tags. These Japanese scientists had no direct vested interest or involvement in the program and their work did not involve substantial direct contact with vessels and fishermen. Some promotional material was produced in Japanese but this did not occur until 1987 and there was not a good system for distribution of this material to the actual vessels. A further fact hindering recoveries was that the main location and time for reporting recoveries during the early years of the program was in Japan at the end of extended fishing campaigns. This meant that the time between the actual recoveries and the time when there was a chance for the information to be reported could be over one year. Under such circumstances, even if fishermen are well informed and have the intentions of reporting recoveries, it is all too easy just to forget or for the tags to get lost.

Promotion in the 1990s as discussed above has involved extensive direct contact with longline vessels as well as extensive preparation of tag recovery promotional material in Japanese followed by distribution of these materials directly to SBT longline fishing vessels. We believe that it is the combination of the following four factors that have made the main difference in the recovery rates: (1) the increased awareness of the program (e.g., a large fraction of the vessels now have tagging posters hung in their galleys), (2) visible and immediate provision of rewards, (3) the direct personal contacts and (4) the presence of observers on a significant fraction of vessels within the AFZ plus a small number of RTMP observers on the high sea. Thus, over 85% of the tag recoveries from longliners in the 1990s are the result of direct contact between the vessel and scientist, observers or tag liaison personal prior to the vessel returning to Japan. Only 12% of the tags have been recovered from vessels after they have returned to Japan.

It is impossible to estimate what fraction of the tags that were collected outside Japan would eventually have been recovered if they had not been returned prior to the vessel's return to Japan. However, some indication of the effectiveness of in-port vessel liaison visits can be ascertained from the tag recoveries off South Africa. In South Africa, the contract Japanese speaking liaison officers can only visit about 20-30% of the Japanese longliners fishing for SBT because of administrative reasons. Since these liaison personnel started visiting vessels, they have recovered 30 of the 60 tags recovered off South Africa which suggests that the recovery rate for vessels which are not visited in South Africa is only 25-40% of the rate for those which are visited. A further six of the 60 tags recovered from fishing operations off South Africa were reported during port visits in Australia or by Australian RTMP observers (see Table 19).

However, there are indications that reporting rates from Japanese longliners are less than 100% even in areas with high personal contact and extensive publicity. The SBT fishery off Tasmania receives the greatest publicity because of the location of the CSIRO laboratory on the docks in Hobart and the extensive use of Hobart as a port of call by Japanese longline vessels. All Japanese longline vessels fishing within the AFZ are required to have their holds inspected prior to and after the completion of fishing operations. Nearly 100% of the vessels are visited by CSIRO staff during these visits to Hobart and a large number of tags are reported in the process. The Tasmanian fishery has also had some of the most extensive observer coverage with over 30% of the vessels having an observer on board for part of their trips in 1991-93. Since observers are rotated among vessels while at sea, the overall observer coverage in terms of both fishing effort (i.e., number of hooks set) and number of SBT caught has been between 12 and 21% in 1991-93 (Table 20). Overall, around 27% of the reported tags that have been recovered from this fishery have occurred while observers are on-board. Observer coverage is spread throughout the season and fleet and is not targeted at any particular component and can be used to get an indication of reporting rates. The difference between the percentage of tags reported and the percentage of the catch/effort when observers are aboard vessels can be used to generate estimates of reporting rates. Such estimates for the Japanese and joint venture vessels fishing around Tasmania range from 37 to 80% and tend to be lower based on the coverage rate for effort than on the coverage rate for the number of SBT caught (Table 20). The reporting rate also appears to have increased in 1993 and may be some indication of the increased effectiveness of the promotional and recovery activities.

Table 19. The number of tags recovered from Japanese longline vessels by the various types of tag collectors. Note that this data has not been stored on the data base prior to 1993. Further note that the total number of Japanese recaptures in this table exceed the number in Table 10 since recovery from all releases are included and only surface releases from South and Western Australia in the later.

Year	Port Visit			Observers	In Japan		Not Recorded
	CSIRO Scientist	Japanese Speaking Contractor	Inspector		Industry	Scientist
1993	92	37	5	58	1	24	6
1994*	60	28	6	51	2	19	0

*through November 1994.

Year	Total number of recaptures	% of recaptures when observers aboard	% of SBT caught when observers aboard1	% of hooks when observers aboard	Estimated tag reporting rate
					based on catch	based on hooks
1991	35	34	21	16	50%	37%
1992	76	24	12	12	43	42
1993	149	25	21	17	80	62

1) since observers only observe about 70% of hooks retrieved during a set, the number of SBT caught when observers were aboard was calculated as the product of the number of SBT observed caught times the ratio of the total number of hooks set while observers were on board times the total number of hooks observed.

The overall results from the current SBT tagging program emphasize the critical need for well developed and extensive tag recovery programs if results from large scale tagging programs are to be used to investigate and measure interactions between fisheries. Previous tuna tagging studies have resulted in very low longline recovery rates of tagged tuna relative to surface recoveries. The differential has been so great that researchers have been lead to suggest that the difference could not be attributed solely to differentials in reporting rates and that there was a biological, stock or behavioural component to the difference. The comparison of the results from the 1983/84 and the 1990s tagging program clearly indicate that such a large differential in recovery rates can be due solely to differences in reporting rates. In cases where recovery rates are low, there is clearly a need to get verifiable estimates of the relative magnitude of different reporting rates if any qualitative, much less, quantitative inferences are to be made about potential interactions.

The need for estimating reporting rates has been recognised and recent work by the South Pacific Commission tagging program has demonstrated the need to get reasonable estimates from the large scale industrial surface tuna fisheries (SPC, 1994). In this case, most tags are discovered at the time of unloading and or processing, and tag seeding experiments by on-board observers provide a useful approach. For longline fisheries, designing experiments for measuring reporting rates is not a trivial matter. Since all fish are individually handled at the time of capture, tag seeding experiments are not a viable option. On board observers appear to be the only viable option at this point, but the logistic and financial implications are large for large scale observer programs in many longline fisheries.

In conclusion, the results from the current SBT tagging program indicate that low return rates from longline vessels should not be used as an indication of lack of interaction or lack of potential for interactions between surface and longline fisheries without a careful assessment on reporting rates. In the case of SBT, the current results indicate rapid and significant interchange of fish between the surface and sub-surface fishery. The results raise questions about previous suggestions/conclusions that the fish vulnerable to surface and longline gear constitute two distinct biological components and the lack of interaction between these two fisheries. However, at least in the cases of some yellowfin stocks, lack of interaction between these two fisheries is also supported by analyses of catch per unit effort and catch data (e.g., Fonteneau, 1988; SPC, 1994; Suzuki, 1994). In these cases, declines in catch rates by longliners and production model estimates of sustainable yields prior to the development of substantial surface fisheries are not consistent with subsequent catches taken in surface fisheries. As such the question of longline/surface interactions remains an open question for these species.

5. ACKNOWLEDGMENTS

Large scale tagging programs such as the ones described here require the enthusiasm, support and cooperation of a very large number of people to be successful. The list of individuals is too long to acknowledge all of them separately but we wish to thank them for their support and effort. Special thanks go to K. Williams, A. Caton, C. Stanley, J. Gunn, T. Carter, N. Clear, Y. Ochi, A. Iwasawa, I. Warashima, T. Itoh, D. Waddington, K. McLoughlin, A. de Fries and M. Ryba for their roles in getting fish tagged, tag recovery and data processing. The 1990s tagging program is a collaborative program involving CSIRO and NRIFSF with funding support at different times from JAMARC, AFMA, AFMA-SBTMAC and an Environmental Research Grant from the Commonwealth of Australia. K. Katsuyama from Japan Fishery Agency and Dr. M. Takahashi from JAMARC deserve a special thanks for their support and help in establishing and maintaining collaborative arrangements. We also wish to thank the Japanese and Australian tuna industries for their continued support, with special thanks to Federation of Japan Tuna Fisheries Cooperative Association, National Ocean Tuna Fishery Association (ZENGYOREN), Tuna Boat Owners Association of Australia, Y. Kawai, A. Suda and B. Jeffriess. Finally, we wish to acknowledge Albert Caton's review and useful comments on an earlier draft of this manuscript.

6. REFERENCES CITED

Anonymous. 1993. Tuna News. Report from the 1993 Southern Bluefin Tuna Industry-Research-Management Workshop, November 1993, Hobart, Australia. CSIRO Fisheries. 8 p.

Anonymous. 1994. Report of the Southern Bluefin Tuna Trilateral Workshop. 17 January - 4 February 1994. CSIRO, Hobart, Australia. 41 pages plus appendices.

Bayliff, W.H. 1993. An indexed bibliography of papers on tagging of tunas and billfishes. Spec. Rep. I-ATTC 8: 91 p.

Caton, A.E. (ed). 1991. Review of aspects of southern bluefin tuna: biology, population and fisheries. In: Deriso, R.B., and W.H. Bayliff (eds.). World Meeting on Stock Assessment of Bluefin Tunas: Strengths and Weaknesses. Spec. Rep. I-ATTC 7: 181-350.

Caton, A.E. 1994. Commercial and recreational components of the southern bluefin tuna (Thunnus maccoyii) fishery. In: Shomura, R.S., J. Majkowski and S. Langi (eds.). Interactions of Pacific tuna fisheries. Proceedings of the First FAO Expert Consultation on Interactions of Pacific Tuna Fisheries, 3-11 December 1991, Noumea, New Caledonia. Vol. 2: Papers on biology and fisheries. FAO Fish. Tech. Pap. (336/2): 344-369.

Caton, A.E, P. Ward and C. Colgan. 1995. The Australian 1989-90 to 1994-95 southern bluefin tuna season. Commission for the Conservation of Southern Bluefin Tuna Scientific Meeting, 10-19 July 1995, Shimizu, Japan. NRIFSF. SBFWS/95/1: 23 p.

Davis, T. 1993. Application of ultrasonic tracking to determine the surfacing behaviour in SBT: Report 1992/93. Fifth Workshop on Southern Bluefin Tuna Recruitment Monitoring and Tagging programs. 28 June-2 July, 1993, Hobart, Australia. CSIRO Fisheries RMWS/93/9: 7 p.

Davis, T. 1994a. Vertical and horizontal movements of southern bluefin tuna in the Great Australian Bight determined by ultrasonic telemetry. Abstract. In: Kleiber, P., and R. Rasmussen (eds.). Proceedings of the 45th Annual Tuna Conference, 23-26 My 1994, Lake Arrowhead, California, page 67.

Davis, T. 1994b. Behaviour of southern bluefin tuna determined by ultrasonic telemetry. Sixth Workshop on Southern Bluefin Tuna Recruitment Monitoring and Tagging programs, 18-21 July 1994, Hobart, Tasmania. CSIRO Fisheries. RMWS/94/9: 15 p.

Fonteneau, A. 1987. Surface versus longline tropical fisheries in the Atlantic Ocean - What interaction? Expert Consultation on Stock Assessment of Tunas in the Indian Ocean, 4-8 December 1986, Colombo, Sri Lanka. Collect. Vol. Indo-Pac. Tuna Dev. Mgt. Programme: 191-216.

Fonteneau, A. 1988. Interaction between tuna fisheries: A critical review based on some Atlantic examples. Collect. Vol. Sci. Pap. ICCAT 28: 109-129.

Gunn, J., T. Carter and N. Clear. 1994. SBT strontium chloride marking experiment: Progress report for 1993/94. Sixth Workshop on Southern Bluefin Tuna Recruitment Monitoring and Tagging programs, 18-21 July 1994, Hobart, Tasmania. CSIRO Fisheries. RMWS/94/4: 13 p.

Gunn, J., T. Polacheck, T. Davis, M. Sherlock and A. Betlehem. 1994. The development and use of archival tags for studying the migration, behaviour and physiology of southern bluefin tuna, with an assessment of the potential for transfer of the technology to groundfish research. ICES CM 1994. Mini-Symposium on Fish Migration 21: 14 p.

Hampton, J. 1989. Population dynamics, stock assessment and fishery management of Southern Bluefin Tuna (Thunnus maccoyii). Ph.D Dissertation. University of New South Wales, Kensington, Australia. 273 p.

Hampton, J. 1991. Estimation of Southern bluefin tuna Thunnus maccoyii natural mortality and movement rates from tagging experiments. Fish. Bull. NOAA-NMFS 89: 591-610.

Hampton, J. 1994. A review of tuna fishery interaction issues in the western and central Pacific Ocean. In: Shomura, R.S., J. Majkowski and S. Langi (eds.). Interactions of Pacific tuna fisheries. Proceedings of the First FAO Expert Consultation on Interactions of Pacific Tuna Fisheries, 3-11 December 1991, Noumea, New Caledonia. Vol. 1: Summary report and papers on interaction. FAO Fish. Tech. Pap. (336/1): 138-157.

Hearn, W.S. 1994. Models for estimating SBT age at length during the transition period. 13th Trilateral Scientific Meeting on SBT, 20-29 April 1994, Wellington, New Zealand. MAF Fisheries. SBFWS/94/13: 5 p.

Hearn, W.S., G.M. Leigh and R.J.H. Beverton. 1991. An examination of tag-shedding assumptions, with application to southern bluefin tuna. ICES J. Mar. Sci. 48: 41-51.

Hearn, W., A.W. Whitelaw, T. Polacheck and K. Williams. 1994. The 1990-93 Southern Bluefin Tuna Tagging program off Southern Australia. In: Whitelaw, W. (ed.) Proceedings of the East Coast Tuna Science-Industry-Management Workshop, 23-24 March 1993, Eden, Australia, CSIRO Fisheries: 123-127.

Klaer, N., and T. Polacheck. 1993. A method for calculating alternative catch-at-age matrices from fine-scale Japanese longline catch and size data. 12th Trilateral Scientific Meeting on SBT, 13-19 October 1993, Hobart, Australia. CSIRO Fisheries. SBFWS/93/15: 17 p.

Kleiber, P., and B. Baker. 1987. Assessment of the interaction between North Pacific albacore, Thunnus alalunga, fisheries by use of a simulation model. Fish. Bull. NOAA-NMFS 85: 703-711.

Lenarz, W.H., and J.R. Zweifel. 1979. A theoretical examination of some aspects of the interaction between longline and surface fisheries for yellowfin tuna, Thunnus albacares. Fish. Bull. NOAA-NMFS 76: 807-825.

Majkowski, J., and G.I. Murphy. 1983. CSIRO southern bluefin tuna tagging: scientific achievements and future objectives. Aust. Fish. 42: 26-27.

Majkowski, J., W.S. Hearn and R.L. Sandland. 1988. A tag-release/recovery method for predicting the effect of changing the catch of one component of a fishery upon the remaining components. Can. J. Fish. Aquat. Sci. 45: 675-684.

Miyabe, N. 1994. A brief analysis of fishery interaction for bigeye tuna in the Pacific Ocean. In: Shomura, R.S., J. Majkowski and S. Langi (eds.). Interactions of Pacific tuna fisheries. Proceedings of the First FAO Expert Consultation on Interactions of Pacific Tuna Fisheries, 3-11 December 1991, Noumea, New Caledonia. Vol. 1: Summary report and papers on interaction. FAO Fish. Tech. Pap. (336/1): 239-248.

Nishida, T. 1994. Japanese longline fisheries of southern bluefin tuna (SBT: Thunnus maccoyii) in 1990s. 13th Trilateral Scientific Meeting on SBT, 20-29 April 1994, Wellington, New Zealand. MAF Fisheries. SBFWS/94/3: 48 p.

Polacheck, T. 1988. Yellowfin tuna, Thunnus albacares, catch rates in the western Pacific. Fish. Bull. NOAA-NMFS 87: 123-144.

Polacheck, T. 1994. An overview of interaction issues among the fisheries for southern bluefin tuna. In: Shomura, R.S., J. Majkowski and S. Langi (eds.). Interactions of Pacific tuna fisheries. Proceedings of the First FAO Expert Consultation on Interactions of Pacific Tuna Fisheries, 3-11 December 1991, Noumea, New Caledonia. Vol. 1: Summary report and papers on interaction. FAO Fish. Tech. Pap. (336/1): 264-298.

Polacheck, T., K. Sainsbury and N. Klaer. 1994. Assessment of the status of the southern bluefin tuna stock using virtual population analysis - 1994. 13th Trilateral Scientific Meeting on SBT, 20-29 April 1994, Wellington, New Zealand. MAF Fisheries. SBFWS/94/14: 40 p.

SPC (South Pacific Commission). 1994. Oceanic Fisheries Programme: Work Programme Review 1993-94 and work plan 1994-95. Seventh Standing Committee on Tuna and Billfish, 5-8 August 1984, Koror, Palau. Tech. Rep. Tuna Billfish Assess. Programme, S. Pac. Comm. 65 p.

Shingu, C. 1978. Ecology and stock of southern bluefin tuna. Jap. Assoc. of Fishery Resources Protection. Fisheries Study 31. [English translation available by M.A. Hintze as CSIRO Division of Fisheries and Oceanography Report 131 (1981) 79 p.]

Suzuki, Z. 1986. A study of interaction on yellowfin tuna between longline and purse seine fisheries in the eastern Atlantic. Coll. Vol. Sci. Pap. ICCAT, SCRS/86/38: 15-26.

Suzuki, Z. 1994. A review of interactions between purse seine and longline on yellowfin (Thunnus albacares) in the western and central Pacific. In: Shomura, R.S., J. Majkowski and S. Langi (eds.). Interactions of Pacific tuna fisheries. Proceedings of the First FAO Expert Consultation on Interactions of Pacific Tuna Fisheries, 3-11 December 1991, Noumea, New Caledonia. Vol. 1: Summary report and papers on interaction. FAO Fish. Tech. Pap. (336/1): 158-151.

Whitelaw, W. 1992a. Turn Tuna into T-shirts. Professional Fisherman (March): 23-24.

Whitelaw, W. 1992b. Tuna Tagging. Western Fisheries (Winter): 40-44.

Whitelaw, W., and B. Hearn. 1991. CSIRO goes SBT tagging. Professional Fisherman (June): 15-16.

Whitelaw, W., and V. Mawson. (eds.). 1993. Inaugural Southern Bluefin Tuna Science-Industry-Management Workshop (Proceedings); (date, location, org. sponsoring meeting), 80 p.

Whitelaw, W., B. Hearn, T. Polacheck and K. Sainsbury. 1992. CSIRO and Japan come together for a new Southern Bluefin Tagging program. Aust. Fish. 51: 10-12.

Williams, K. 1982. Tagging method. In: Majkowski, J. (ed.). CSIRO data base for southern bluefin tuna (Thunnus maccoyii (Castlenau). Rep. CSIRO Mar. Lab. 142: 7-8.

Williams, K. 1983. Big numbers of SBF tuna tagged off Esperance. Aust. Fish. 42: 26-27.

Williams, K. 1992. The technique of tagging. Aust. Fish. 51(6): 15-17.

Williams, K. 1993. SBT Longline tagging - Tasmania 1993: Trip report. Working paper. Fifth Workshop on Southern Bluefin Tuna Recruitment Monitoring and Tagging programs, 28 June-2 July 1993, Hobart, Australia. CSIRO Fisheries. RMWS/93/11: 10 p.