S12.1: Seabird mortality on longlines in Australian waters: A case study of progress and policy

¹Tasmanian Parks and Wildlife Service, PO Box 44A, Hobart, 7001, Tasmania; ²Tasmanian Museum and Art Gallery, PO Box 1164M, Hobart, 7000, Tasmania, fax 61 03 62333 477, e-mail rgales@delm.tas.gov.au;. ³Australian Bird & Bat Banding Scheme, Environment Australia, PO Box 8, Canberra, ACT 2601, Australia

Each year tens of thousands of seabirds are accidentally killed on longline hooks set in the world’s oceans. Seabird bycatch during longline fishing occurs when birds are attracted to fishing vessels by discards and baits and ingest baited hooks during the setting or, less commonly, hauling of the longline. The hooked birds are subsequently pulled under the water by the weight of the line and drown. The extent of this mortality is such that longline fishing is acknowledged today as representing the most pervasive threat to seabirds, causing widespread declines in populations across the world (Alexander et al. 1997; Birdlife International 1995; Gales in press; Croxall in press).

Longline fishing techniques are used to target pelagic and demersal finfish and shark species, and longline fisheries span the world’s oceans. Longline fishing commenced in the southern oceans in the 1950’s. Whilst bycatch of seabirds by longlines was first reported from band returns in the early 1980ís (Morant 1983), the potential magnitude of the problem was not fully realised until Weimerskirch and Jouventin (1987) documented a dramatic decline in Wandering Albatross Diomedea exulans populations on the Crozet Islands. These authors proposed that the decline was due to longline and trawl fisheries impacting on albatrosses in the marine environment. Studies by Brothers (1991) and Murray et al. (1993) based on direct observation of bycatch rates aboard Japanese longline vessels targeting Southern Bluefin tuna (SBT) Thunnus maccoyii in Australian and New Zealand waters, respectively, confirmed the threat to seabirds from longline fishing operations

The extent of seabird mortality is poorly known for most of the world’s longline fisheries (Cooper in press). However, broad comparisons of the limited data available show that catch rates in the order of 0.4 birds per 1000 hooks are typical, although rates an order of magnitude higher are also documented (summarised in Alexander et al. 1997 and Cooper and Wanless in press; Barnes et al. 1997). Information on bird bycatch in the Southern Oceans is best known for the Japanese longline fishery targeting Southern Bluefin tuna in the Australian and New Zealand exclusive economic zones (EEZ). Since 1988, fisheries observers in these countries have routinely recorded the number of seabirds observed caught during Japanese fishing operations. Over the last ten years the hauling of more than 9 million tuna longline hooks have been observed in the Australian EEZ (5% of all hooks set), and over 5 million in New Zealand (8% of all hooks set).

Within the Australian Fishing Zone (AFZ, Fig. 1), Tuna longline fishing has been carried out by Japanese fishing vessels since 1979, and by Australian vessels since the late 1980s. Japanese fishing effort in the AFZ reached a peak in 1988, but since then the annual fishing effort by the Japanese fleet has been declining. Conversely, the domestic fishing effort has been increasing and in 1997 over 7 million hooks were deployed. There is however no observer scheme for this fishery and so information on domestic seabird catch rates is limited.

Analyses of the trends of seabird catch rates in the AFZ by Japanese longliners since show an apparent fall from the 1988 bycatch figure of 0.4 birds/1000 hooks to levels of between 0.1-0.2 birds/1000 hooks (Gales et al. in press; Brothers et al. 1998b). These catch rates translate to a total mortality of between 1000 and 3500 seabirds per year in the AFZ, depending on the level, area and season of deployment of effort. Most of the birds killed are albatrosses, including species recently listed as Threatened and Endangered (Croxall and Gales 1998). Of concern is that, despite awareness of the seabird bycatch issue having increased, and the adoption of mitigation measures advocated, seabird mortality rates in 1997 in the AFZ Japanese tuna longline fishery were the highest since 1988. The reasons for this apparent trend of increased catch rates are not clear (see Gales et al. in press).

So, a decade after the severity of bird bycatch in longline fisheries was first documented, during which time the issue has escalated in the international marine conservation agenda, the problem still persists at levels which are considered to be unacceptably high (Environment Australia 1998). Failure to resolve this problem led the Australian Government in 1995 to list the ‘incidental catch (or bycatch) of seabirds during oceanic longline fishing operations’ as a Key Threatening Process under the Federal Endangered Species Protection Act 1992. This listing requires that a Threat Abatement Plan be prepared to identify the actions necessary to abate the threat. The subsequent plan, which was finalised in 1998, takes the approach of prescribing an overall reduction in seabird bycatch to a level below 0.05 seabirds/1000 hooks. The plan specifies that this reduction must be achieved within five years. Failure to do so would indicate that the measures prescribed to achieve this figure, or their level of implementation, are inadequate, or insufficient in reducing bycatch. Such a failure would also imply that the future for albatrosses, and other species of seabirds, is indeed, bleak.

The ten year evolution of the seabird bycatch issue in the AFZ, where the magnitude of the problem was first documented, is presented here as a case study. The implementation and efficacy of the existing mitigation measures are discussed and the nuances of the limitations of observer derived data are explored. Hopefully this chronology and synthesis of earlier work (Brothers 1991, Gales et al. in press, Brothers et al. 1998 a, b; Brothers et al. in press) will be useful for other nations facing the problem of high levels of seabird bycatch caused by longline fisheries operating in their EEZs.

A longline consists of a main line with numerous baited hooks attached on branchlines. The configuration of the longline can vary considerably depending on the target species, fishing area and the size of the fishing vessel. The line can be set parallel to the surface on floats or anchored vertically. The number and type of hooks and the length of the branchlines also depend on the target species (Chapman 1990).

Longlines are typically mid-water set (pelagic) or bottom set (demersal), depending on the target species. Demersal longline operations in other areas are known to have high seabird catch rates (see Cooper and Wanless in press) but within the AFZ it is pelagic operations which have posed the highest risk to seabirds (Environment Australia 1998). The interactions between seabirds and the Japanese and Australian domestic pelagic longline fleets are the focus of this paper.

For pelagic longline fishing a single line (mainline) to which many hooks are attached (via branch lines) is deployed. The Japanese longline fleet targeting Southern Bluefin tuna consists of vessels of 45-60 m in length which set between 2400 and 3000 barbed, steel hooks per day on 40 m branch lines suspended from a 100 km mainline. Baits are cast approximately every 6 seconds for the 5 hour period it takes to set the line. The line is recovered after a 5 h soak period at depths between 60-150 m, and this typically takes 12h. Details of Japanese longline operations in the AFZ are documented in Ward (1996).

Australian domestic pelagic vessels range from 18-45 m in length. The fleet largely operates within 100 nm of shore and set lines of 20-40 km with 600-1500 hooks. Setting usually takes 2 - 4 hours, after which the baited hooks soak at depths of 30 - 200 m for 3 - 10 hours. Hauling the line takes 6 - 10 hours. Further details of domestic longline fishing operations in the AFZ are presented in Brothers and Foster (in press).

The Japanese longline tuna fishery started in 1955 and effort increased to maximum levels of 100 million hooks set in the Southern Oceans each year through the 1980s. Whilst the total global effort decreased during the 1990s, the proportion of effort expended in the AFZ increased to over 30% by 1993/94, equivalent to 21 million hooks. Since then, Japanese effort in the AFZ has contracted, with most of the effort in the Austral winter. In the 1997 fishing season 7.7 million hooks were set, all in the winter season. There has been no Japanese tuna fishing within the AFZ since then, following the failure of the members of CCSBT (Australia, New Zealand and Japan) to reach agreement over tuna quota limits for 1998.

The Australian domestic tuna longline effort has increased each year. Since 1989 the annual effort has risen from 1.4 million hooks to 7.8 million hooks set in 1997 (Fig. 2).

The Australian Fisheries Management Authority (AFMA) is responsible for the management of the tuna fishery in the AFZ. AFMA personnel have observed the activity of the Japanese longline vessels fishing in the AFZ since the declaration of the zone in 1979 (Caton and Ward 1979). Until 1988 the information obtained from this fishery observer scheme was relatively informal and focused on liaison and validation of logbooks. During the 1990s the nature of reporting evolved to be quantitative, and aimed to examine a representative, random sample of fishing effort throughout the AFZ at a level of approximately 10% of all hooks set. In the 1997 fishing year, a total of 720 300 hooks were observed, 9.4% of all hooks set. Over the last ten years more than 9 million tuna longline hooks have been observed hauled in the AFZ (Fig. 2).

In addition to AFMA fisheries observers, Tasmanian Parks and Wildlife Service (TASPAWS) biologists have also observed fishing operations on Japanese longliners since 1988. This involvement allowed the identification of parameters which were relevant to the quantification and understanding of seabird bycatch, a necessary precursor to efforts to solve the problem. As a result of this program, collection of additional records detailing fishing gear and operations became standard practice for fisheries observers. Seabirds retrieved from hooks were also retained for subsequent examination by TASPAWS biologists. Whilst there was resistance to retaining birds in the early years, increasing co-operation over time has resulted in a greater proportion being retained.

Unfortunately, there has not been a dedicated observer program established for the Australian domestic longline tuna fishery, even though effort in this fishery now exceeds that for the Japanese industry. However, between 1994 and 1997 opportunistic observations of seabird bycatch on domestic vessels were carried out by TASPAWS scientists who observed 39000 hooks set over 49 days (Brothers and Foster in press, Brothers et al. 1998a. b). Limited data on bird bycatch was also collected by fisheries scientists conducting tuna research (Whitelaw 1997). Therefore, much of what has been learnt on the seabird bycatch issue in the AFZ stems from the Japanese tuna longline fishery.

Data for the analyses of the trends in seabird bycatch in the Japanese SBT fishery in Australian waters was derived from both the AFMA fisheries database and TASPAWS seabird bycatch database. Analyses of the data for the Japanese fishery between 1988-1996 have been recently synthesised by Klaer and Polacheck (1995), Gales et al. (in press), Brothers et al. (1998 a, b; in press ). Seabird bycatch rates in the domestic tuna fishery in the AFZ are documented in Brothers and Foster (in press), Whitelaw et al. (1997), and Brothers et al. (in press b).

For analyses of seabird bycatch, the AFZ was divided into seven regions (Fig. 1), and each fishing year into two seasons, summer (October to March) and winter (April to September). Timing of the setting of longlines was partitioned into day, night and twilight. The total number of birds killed in each area and season were estimated from the observed catch rates (see Klaer and Polacheck 1995, Gales et al. in press for details).

Seabird carcasses hauled aboard, when retained by observers, were frozen and returned to TASPAWS. Parameters examined included identification of species, sex, age class, breeding status, morphometrics, stomach contents, and material collections for determination of DNA analyses and heavy metal concentrations. Records of banded birds were documented and sent to the Australian Bird and Bat Banding Scheme.

Analyses of the Japanese longline seabird bycatch data show that seabird catch rates vary with year, fishing area and fishing season (Brothers 1991, Klaer and Polacheck 1995, Gales et al in press). Despite some inadequacies in the data, catch rates show that over 30 000 seabirds are likely to have been killed on Japanese tuna longlines in the AFZ between 1988 and 1997.

In the AFZ most birds are killed in waters around southern Australia during the summer fishing season, even though most fishing effort occurs during winter. The distribution of fishing effort and the locations and seasons where seabirds were hooked and retrieved by observers illustrate this pattern (Fig. 1). The propensity for higher catch rates in the southern regions during summer is probably the result of several features including: (1) seasonal and spatial changes in the numbers and species of seabirds attending the vessels. An increasing proportion of ‘diving’ birds may result in more birds being killed because they may be able to retrieve the baits and so get hooked themselves, and also make more hooks available to kill other birds, (2) an increased proportion of line setting during daylight hours in summer (most birds are caught when line setting occurs in daylight).

Efforts to interpret trends in the seabird catches by the Japanese fleet since 1988 are constrained by the interactions of factors which influence seabird catch rates, such as fishing season and area, weather patterns, moon phase, fishing practice and gear, and use of mitigation devices (Klaer and Polacheck 1995, Gales et al. in press; Brothers et al. in press). Nonetheless, it appears that following the quantification of the initial 1988 bycatch estimate of 0.41 birds per 1000 hooks (Brothers 1991), the overall catch rate of birds quickly fell, to a level of ca. 0.1-0.2 birds per 1000 hooks (Fig. 3). Despite the expectation that catch rates would continue to decline as knowledge on threat mitigation increased, it is surprising to note that this level has been subsequently sustained. Any reduction in the total numbers of birds caught during the 1990’s in the AFZ therefore reflects decreased fishing effort rather than a decline in catch rates. Recently recorded catch rates of 0.1-0.2 birds per 1000 hooks equate to 1000-3000 birds being killed in the Japanese AFZ fishery each year. There is insufficient data to estimate the number of birds killed by the domestic fishery each year.

Closer inspection of spatially and seasonally adjusted figures showed that catch rates may have actually increased during the 1990’s (Fig. 3), (Brothers et al. in press b). This increase is most noticeable for the winter fishing season, when most fishing now occurs, and in particular for the Tasmanian region, one of characteristically high levels of fishing effort and bird catch rate. The reasons for both the lack of a continuing decline in catch rate, and the apparent rise in the winter catch rate are not clear, but may involve factors such as limited or incorrect use of mitigation measures, or changes in fishing practices and equipment.

Whilst this review concentrates on our experiences with longline fisheries operating within the AFZ, it is necessary to consider the impact elsewhere of other longline fisheries on seabird populations. Most albatrosses and petrels traverse the oceans, and so encounter the fleets of several different fisheries. For example, those same populations of seabirds interacting with tuna longlines set in the AFZ may also encounter the toothfish fishery in more southerly regions of the Southern Ocean. Thus there is a need to consider the collective global impacts of different fisheries on seabirds if they are to be effectively managed and conserved.

There are no reliable estimates of the number of seabirds killed on longline hooks around the world each year. More than 20 000 longline vessels are thought to operate throughout the world (FAO 1998). This figure includes only licensed operators, and the level of illegal fishing is thought to be substantial. The size and extent of illegal operations are not known, but it is estimated that unregulated fishing may increase estimates of the numbers of birds killed in the regulated toothfish fishery off Prince Edward Island by 5-20 times (Ryan et al. 1997). or by an order of magnitude for the CCAMLR toothfish fishery (CCAMLR 1997).

Brothers (1991) estimated that 44 000 seabirds were killed annually in the Japanese southern bluefin tuna fishery in the southern oceans. Whilst Japanese effort in this fishery has declined since that estimate was prepared (>100 million hooks each year 1980-87 to 54 million hooks in 1994) the effort expended by other nations in the region has increased. It is estimated that Taiwanese effort south of 30°S, for example, has increased to 150 million hooks each year (Tuck and Polacheck 1997). The combined effort by the Japanese and Taiwanese fleets then now doubles that used by Brothers in his calculations. The implications to seabird bycatch are obvious.

Other fleets also catch seabirds (reviews by Alexander et al. 1997; Cooper and Wanless in press). Preliminary estimates for the Hawaiian pelagic longline fishery are of more than 10 000 albatrosses killed each year between 1990 and 1994 (Anon 1996). In the Southern Ocean, demersal longline fleets targeting Patagonian toothfish kill substantial numbers of seabirds. In addition to the legal fishing here, there is increasing illegal activity and based on information presented to CCAMLR (synthesised by Cooper and Wanless in press) up to 145000 birds may have been killed on longlines targeting Patagonian toothfish in the 1996/97 season. This impact on species is considered by CCAMLR (1997) to be unsustainable for the most impacted populations of albatrosses and petrels in the southern Indian Ocean.

Typically, seabird bycatch data are rarely robust because of how they are collected and assessed. There are two principal problems: (1) the event of a bird being caught is, statistically, a rare event. On only 15% of the 3500 hauls observed between 1991 and 1995, have seabirds been observed caught. It is partly this ‘infrequency’ that delayed the realisation of the seriousness of the problem and (2) not all seabirds killed on longlines are seen by observers. The seabird mortality estimates therefore are minimums as they are based on the number of birds seen hauled by the observer, and do not represent the total number of birds hooked.

From watching entire sets and hauls, Brothers (1991) estimated that about 30% of birds killed on longline hooks during setting come off the hooks and are not hauled aboard. The missing birds were presumed to have come off the line during the soak or during line hauling, or been scavenged by sharks. Of the carcasses hauled aboard, one third are damaged and partly eaten (Brothers et al. 1998b). Consequently most data being collected for seabird bycatch is assessed from observations of line hauling (consistent with focus of fisheries observers), rather than line setting, and it is during line setting that 97% of the birds are hooked (Gales et al. in press).

In addition to the accidental dislodgement of the birds off hooks is the potential for crew to deliberately cut birds away before the line is hauled aboard under the scrutiny of the observer. Estimates of the frequency of line cutting have been made each year since 1995 (Gales et al. in press). Whilst the number of hooks observed in these trials is relatively small, it appears that this source of ‘error’ is significant, with potentially only half the birds that were killed being hauled aboard in some years.

To increase the reliability of bycatch estimates requires: (1) extensive observations of line setting AND hauling; (2)  a requirement that all hooks are landed, with observer verification of species identity before discard of non-bird bycatch; and (3) increased monitoring which focus on seabird bycatch, rather than observations which are collected incidentally to fish monitoring programs.

Seabirds which are hooked on longlines are typically those whose foraging activities include surface scavenging, a propensity to follow boats and to feed on baits and discards, and are large enough to swallow hooked bait.

Whilst initial concerns were focused on the albatrosses, affected species also include smaller petrels, shearwaters, gulls, skuas and gannets. More than 60 seabird species have been recorded as killed on longlines (Cooper and Wanless in press), and in the AFZ at least 21 species are known to be affected (Table 1). Except for two species, all are Procellariiformes, nine of which are albatrosses. Seventy six percent of the 750 carcasses retained from Japanese SBT hooks between 1988 and 1997 were albatrosses. It was the emergence of these type of data which lead to longlining being recognised as the most pervasive threat facing this group of birds.

The threats and impact of anthropogenic influences, such as longlining, on albatrosses are especially serious given the limited capacity of albatross populations to cope with increased levels of mortality. Albatross species are very long lived. They produce a single chick every one or two years (depending on the species) and the chicks can take as long as 11 months to fledge. Breeding does not occur until 5 -12 years of age. This reproductive strategy results in low levels of recruitment. Even slight additional mortality affecting young or adult birds will significantly reduce the viability of most species.

In the Australian zone, examination of the seabirds retained has allowed assessments of the vulnerability of species and populations (see Gales et al. in press). Determinations of sex and age cohorts is also important given the differential impact of elevated adult and juvenile mortality on demographic processes. For the seven most commonly caught species in the AFZ the age and sex classes of birds killed varies with respect to fishing grounds and season (Table 2). Yellow-nosed albatrosses, for example, are most frequently caught in the waters off South Western Australia, and whilst no age bias was evident in the total samples, seasonally adjusted assessment shows that adults (94%) are killed predominantly during the winter months, whereas immatures (80%) are killed almost exclusively in summer. For Wandering albatrosses, it is typically males which are killed in both the summer and winter fishing seasons, and during winter, most of these birds are immatures.

To understand which species are affected by a longline fishery, it is necessary to collect substantial amounts of data. For example, observation of 108 662 hooks in 1988 identified 45 seabirds caught, of which 33 were examined to species. Five albatross species and a giant petrel species were identified (Brothers, 1991). Over the next 8 years, during which a further 8 million hooks were observed, a further 11 species were added to the list. Since 1992 there has been one additional species identified each year (Fig. 4). The sequence of identification of these species however, is not simply a reflection of the sequence of encounter. Analyses of the data suggests that until 1992, when fisheries observers were becoming aware of the importance of the issue, there was a tendency to preferentially retain the large albatrosses (Gales et al., in press). This tendency seemingly decreased with education and corresponding increases in the proportion of dead seabirds retained by observers for accurate species identification and processing onshore (8-45% retained each year to 1992; after which 70-100% retained each year). Since 1992, when more representative species collections have been retained by observers, the continuing number of species being identified shows how long it takes to accrue reliable information on the species composition of the seabirds being killed.

Lists of species impacted by longline fisheries cannot be transferred between different fleets, even in the same area. The species of seabirds killed in the AFZ by the Japanese longline fishery differ from the species identified as being killed by domestic boats. For the domestic fishery, of the six seabird species so far identified from the domestic boats (Table 1) only two have been recorded from the Japanese fleet. Differences in fishing equipment, as well as the time and area of fishing (domestic operations closer inshore and more often setting at night) are thought to account for this difference.

The retention by observers of all seabirds caught is then very important, particularly when the focus of observing is fish management. This is because, unless specifically and intensively trained, fisheries observers may not correctly identify birds. Our experience indicates that even trained fisheries observers correctly identify only ca. 50% of all corpses they retain. Appropriate extensive examination and sampling of the carcasses also allows identification of sex and age cohorts, sampling for genetic and morphometric characteristics and identification where possible to subspecies.

The importance of this requirement has been highlighted by the recent revision of albatross systematics, which resulted in the recognition of 24 species, an increase from the previously recognised 14 species (Robertson and Nunn in press). Collection of the full suite of morphometric measurements, as well as genetic sampling since 1992 has enabled us to review the initial species identifications (Table 3). Consequently, we now know that 13 of the currently recognised species are impacted by longlining in the AFZ. Importantly, this has also allowed us to then examine these species in light of recent assessment of their conservation status based upon IUCN criteria (Croxall and Gales in press). Such recognition is critical to our understanding of the impacts of longline fishing on albatross taxa. The 13 species of albatrosses known to be killed in the AFZ include one endangered species, 10 vulnerable species, one near threatened species and one species for which the data are inadequate to assign a status. This recasting and re-listing of the taxonomic and conservation status of albatrosses indicates the critical threat posed by longline fishing to many of the seabird species which occur within the AFZ.

Generally it is the larger and more aggressive species (giant petrels and the large albatrosses) which compete for baits on the surface, whilst the smaller more proficiently diving species (petrels and smaller albatrosses) retrieve submerged baits. Once retrieved, these baits may then be contested for on the surface by individuals of several species, the larger birds generally being the more successful competitors (Brothers 1991). The size of hook influences the size of birds hooked, as smaller species generally only pick at baits, and so may or may not be hooked, whereas the larger, more aggressive birds swallow the baits whole, along with the embedded hook. In Japanese pelagic longline operations most birds (97%) are dead when hauled aboard, having drowned after being hooked as the line is being set. Only few (3%) are alive when hauled, these birds having become tangled or hooked on lines during the line haul (Gales et al. in press).

In the Australian domestic tuna longline fishery, a greater proportion of birds caught are hauled aboard alive than in the Japanese tuna fishery, because the gear used within this fishery is lighter (more buoyant), which permits some birds hooked during the set to remain on the surface and not be dragged under by the weight of the line. Furthermore, opportunities for birds to be caught during the haul are considerably greater on domestic vessels (Brothers and Foster in press).

In the AFZ, at latitudes between 16°S and 45°S, most seabirds get hooked during daylight hours. This characteristic reflects the capacity of seabirds to locate their food by sight (although smell may assist in detection) and consequently their feeding patterns are mainly diurnal (Harper, 1987; Weimerskirch and Wilson 1992; Nevitt et al. 1995; Hedd et al. 1997). Hooks set at night are 85% less likely to catch birds than those set during the day (Brothers et al. in press). This divergence is especially marked during winter. During summer the effect is diminished during periods of bright and full moonlight. These tendencies for increased daytime seabird catch rates have been also found in other longline fishing areas (Murray et al. 1993, Ashford et al. 1995; Barnes et al. 1997).

In addition to light levels, depth also dictates the accessibility of food sources to seabirds. Most observations of albatrosses feeding at sea document surface seizing, or birds making shallow dives or surface plunges (Harper et al. 1985; Thompson 1992). Recent work has shown that, as well as short, shallow plunge dives (mean duration = 2 secs; max depth = 3 m), Shy Albatrosses also undertake swimming dives which are both longer (mean duration = 6 secs) and deeper (up to 7.4 m) (Hedd et al. 1997). The smaller light-mantled Sooty Albatrosses have been recorded diving to depths over 12 m (Prince et al. 1994). The diving capacity of petrels exceeds that of albatrosses, and petrels also may forage under lower light regimes (Huin 1994, Barnes et al. 1997). Sooty Shearwaters, for example average dives to 39 m and attain maximum depths of 67 m (Weimerskirch and Sagar 1996). A capacity for diving enables these birds to retrieve baited hooks for some time and distance behind the vessels.

The time during which the baited hooks are accessible to the birds before sinking out of their reach varies with fishing gear and practice. Typically in pelagic longline operations most baits are taken within 20 seconds of being deployed, and whilst within 100 m of the stern of the ship (Brothers 1991). Synthesis of the data derived from the diving depths and speeds of Shy Albatrosses and sink rates of longline hooks (Pemberton et al. 1995) shows that the birds have about 5 seconds after the bait is not protected by the streamers of bird scaring lines (>100 m astern) and before the bait descends out of their reach. These data, coupled with observations of birds feeding behind boats and the operation and performance of fishing gear, are useful in design and use of effective mitigation measures.

A mitigation measure, in the present context, is best described as a modification to fishing practices (operational changes) or fishing equipment (gear changes) which reduces the likelihood of seabird bycatch. There have been several reviews of mitigation options and measures including Alexander et al.  (1997), Bergin (1997), Environment Australia (1998) and most recently the FAO commissioned review by Brothers (in press).

For pelagic longlining targeting tuna, there are currently choices of combinations of six mitigation options, three involving gear changes and three requiring operational changes (Table 4). Five of these options have moderate to high potential to reduce seabird bycatch, preventing access and detection of the baited hook to the birds through increasing the sinking rate of the hooks, deterring the birds whilst the hooks are accessible, or through fishing at night. In contrast, the option of offal dumping, which simply distracts the birds during setting, has a low potential of reducing bycatch.

In 1988, Brothers (1988, 1991) prescribed four measures to reduce bird mortality and increase fishing efficiency within the southern ocean Japanese tuna longline fishery: (1) the use of bird scaring lines; (2) mechanised bait throwing; (3) adding weight to increase hook sink rates; and (4) confining line setting to hours of darkness. These same measures, together with the use of thawed bait, are still the most effective mitigation measures available today - a decade later. Whilst awareness of the issue of seabird bycatch has risen rapidly over this time, it would seem that acceptance and implementation of mitigation measures has not occurred at a similar pace. The apparent reluctance by fishers to broadscale adoption of mitigation measures perhaps reflects the persisting inadequacy in the effectve use of the measures, or perhaps it reflects a cultural intransigence to change in long held fishing practices.

Interpretation of changes in seabird catch rates resulting from the use of mitigation measures is complex and constrained by the inter relationships of factors affecting catch rate, the ongoing changes to fishing practices and equipment, and inadequate level of data collection of factors pertinent to seabird catch rate (Brothers et al. in press). In statistical assessments of the efficacy of mitigation measures these factors can combine to precipitate misleading and confusing results and interpretations. Consequently, whilst the need for statistical verification of the effect of mitigation measures on catch rates of seabirds, as well on target species and other non target species remains, there is also a strong need for pragmatism.

Five options for operational and gear changes which are known to have moderate to high potential for the reduction of capture of seabirds in the pelagic tuna longline fishery are briefly described below:

Most seabirds caught on longlines are active during the day. Consequently, fewer seabirds are caught on longlines which are set during hours of darkness. Night setting is widely advocated as the single most effective and reliable measure to reduce seabird bycatch.

Reductions in seabird catch rate in the order of 60-96% have been reported for night setting (Murray et al. 1993; Klaer and Polacheck 1995; Cherel et al. 1996; Brothers et al. in press). In the AFZ, where an overall reduction in catch likelihood of 85% has been reported for lines which are set at night, compared to day set lines, the seabird catch likelihood decreases by 2% for every 1% of hooks set at night (Brothers et al. in press). The efficacy of night setting in reducing seabird catch rate in the AFZ has been found to be highest in winter (99%), the effect being diminished in summer due to significant interactions with moonlight levels (Brothers et al. in press)..

Despite the consistent evidence that night setting is effective in reducing bird bycatch, there has not been widespread adoption of night fishing as standard practice in the AFZ. In recent years the proportion of hooks set at night by the Japanese fleet has remained at less than 20% (Gales et al. in press).

To accelerate the adoption of night setting as a tactic to reduce bird bycatch the Threat Abatement Plan (Environment Australia 1997) prescribes night setting of hooks as one of the three options for pelagic longline fishing in the AFZ. Night setting is also encouraged and promoted in other fishing grounds e.g. New Zealand and Alaska, and required in others, such as CCAMLR waters.

Frozen bait floats and so is accessible to seabirds for longer periods. This problem is exacerbated when the fish baits also have inflated swim bladders.

In pelagic longlining in New Zealand waters, decreases of 69% in seabird catch rates have been reported when thawed baits were used (Duckworth 1995). In the AFZ reduced catch rates have also been observed when thawed rather than frozen baits have been used, but this difference diminishes in winter (Brothers et al. in press). The reasons for the seasonal influence include the differences in ambient temperature but also data deficiencies preclude rigorous comparisons of this feature in different weather regimes.

In experimental trials fish with inflated swim bladders floated, even when thawed (Brothers et al. 1995). These trials highlighted the importance of careful bait selection (to avoid species with swim bladders), or the option of adding weights near each hook to negate the buoyancy effect of swim bladders in baits.

The use of thawed bait, on lines which are sufficiently weighted to be inaccessible to seabirds immediately after they are set, has been prescribed as one of the three options in the Threat Abatement Plan (Environment Australia 1997). The use of thawed bait is prescribed in two demersal fisheries (CCAMLR and NMFS/Alaska), although the transfer of this measure to demersal fisheries is of little or no benefit. (Brothers in press).

A bird line consists of a main cord suspended from a pole above the stern of the vessel with a number of streamer cords attached. The line and streamers hang over the area where the hooks enter the water and moves in an unpredictable way, making it difficult for the birds to gauge how to obtain a bait without colliding with a streamer.

To be effective, a bird line must be deployed from adequate height, be towed astern directly above the area where baits enter the water, and be constructed of the appropriate materials (Brothers 1996). Efficacy of birds lines is also influenced by weather, and species composition of aggregations of seabirds (different species respond to bird lines differently). For maximum protection some vessels use two lines deployed either side of the mainline.

Reductions of 30-70% in seabird catch due to bird line use have been reported for pelagic longline operations (Brothers 1991, Klaer and Polacheck 1995), although small sample sizes precluded reliable statistical analyses of these data. Analyses of larger datasets have failed to detect statistical differences in catch rates due to bird line use (Duckworth 1995; Brothers et al. in press). This is believed to reflect the poor quality of the data and also the need for effective deployment on bird lines (Brothers et al. in press). Case studies of bird line use show that, with correct use and refinement, bird lines can assist in reducing seabird catch rates by 85% (Brothers et al. in press).

Use of bird lines has been made mandatory for all longline vessels operating in CCAMLR waters since 1991, in the New Zealand EEZ since 1993, and in Australian waters south of 30°S since 1995. Despite this, in a survey of 14 sets by domestic vessels in 1997, a bird line was used during only 1 set. As a part of the Threat Abatement Plan, it is recommended that bird lines are required for all longline vessels in the AFZ. In 1997 Japan undertook to make use of bird lines mandatory on all their pelagic longline vessels.

When baited hooks are cast manually they often land in the area of propeller turbulence, and so their sinking is delayed. Bait casting machines are designed to throw the bait accurately under the bird line and clear of propeller turbulence and thereby increase the sink rate of baits. Not coincidentally, use of bait casting machines also relieves the crew of the onerous task of manually casting hooks, a fact which has assisted with industry adoption of the machines.

When used in conjunction with an appropriate bird line, it has been estimated that reductions in seabird catch rates of 60-80% are achievable (Brothers 1993, Klaer and Polacheck 1995). Statistical comparisons of seabird catch rates however have failed to show significant reductions in catch rates when bait casting machines are in use (Brothers et al. in press). Further, the casting capacity is not consistent between different brands, this difference confounding the efforts to assess their efficacy. Improper use of bait casting machines is thought to contribute to the unexpected results of the statistical analyses. Bait casting machines are often operated at maximum throwing capacity, and so may cast baits away from propeller turbulence but also outside the area protected by bird lines.

Bait casting machines were first developed in 1992/93 and their use has steadily increased since then to the current level where over half the Japanese fleet observed fishing in the AFZ are fitted with bait casting machines (58% of 31 vessels in 1997). Compact units, suitable for use on smaller domestic boats have also recently been developed. Voluntary codes of practice include adoption of bait casting machines in both New Zealand and Australian waters.

Weights added to branch lines in pelagic longline operations accelerate the sink rate of hooks and so reduce the time that bait is available to birds. Weighting regimes appropriate for the reduction of seabird catch rates are influenced by the use of other protection devices (bird lines), the setting speed of the vessel, and the diving capabilities of the attendant seabirds.

Sink rates of baited hooks alone have been measured experimentally and operationally (Brothers et al. 1995; Pemberton et al. 1995) but there are few data available to statistically quantify the efficacy of weighted lines on seabird catch rates. Comparisons of seabird catch rates in the AFZ between light lines (nylon) and heavy lines showed that the lighter lines which sink more slowly caught 3-7 times as many birds as lines constructed of heavier materials (Brothers et al. in press). Recent assessments of seabird catch rates on lines with different weight regimes showed that fewer birds were caught on weighted lines (Brothers et al. 1998b)

Using archival hook monitors Pemberton et al. (1995) also showed that efforts to fish at a predetermined depth are hampered by the effects of currents which cause the lines to ascend prior to hauling. If fish catch rates are improved by targeting specific depth and temperature regimes, addition of weights to retain hooks at desired depths should improve catch rates of targeted fish.

The only regulations imposed regarding weighting regimes on longlines are those in place since 1997 for demersal operations in CCAMLR waters. For pelagic operations in Australian waters, the Threat Abatement Plan (Environment Australia 1998), offers ‘use of lines which are sufficiently weighted to cause the baits to sink out of reach of diving birds immediately after they are set’ as part of one of three mandatory options to be selected by fishers in 1998 and beyond.

The above accounts briefly relate the details of five currently used/available mitigation options. The second generation of mitigation measures, more conceptual options, are described by Brothers (in press) and the TAP (Environment Australia 1998) and include underwater setting, ‘smart’ protected hooks, and ‘smart’ artificial baits and lures.

A consideration of the currently available measures would suggest that, to immediately assist in reducing bird bycatch a combination of several of the following measures should be implemented: (1) setting of hooks should be restricted to night; (2) appropriate bird lines and bait casting machines should be used at all times (including night set lines), such that baits land outside the area of propeller turbulence and the bird line effectively protects the baits whilst close to the surface; (3) baits are thawed and cast with, not against, the prevailing wind; (4) lines are weighted appropriately to optimise hook sink rates and (5) appropriate and innovative education of fishers regarding the correct use of mitigation measures must be accelerated.

Examination of the assessments which have been undertaken for the above options indicate that: (1) the impact of mitigation measures on target and other non-target species has yet to be resolved. Integrated use of time depth recorders and hook monitors are required to further understand implications of different line setting strategies; (2) for any fishery, among the current options, a suite of measures, rather than any single measure is required (see above); (3) results of assessments using inadequate data pertaining to efficacy of mitigation measures can be misleading and precipitate highly inaccurate interpretations; (4) most aggregated datasets being used to assess efficacy of mitigation measures on seabird catch rates are inadequate for that purpose (see Brothers et al. in press); (5) not all mitigation measures can be simply transferred between different longline fisheries (pelagic and demersal), nor between different vessel types or fishing practices (large and small pelagic longline vessels which operate at different setting speeds) and (6) a logical, pragmatic approach to reduction of bird bycatch must prevail.

Ultimately, mitigation measures that are adopted across the board will be those which are economically and operationally advantageous to fishers, and so will not rely on extensive monitoring for compliance. The best solution is one where there is no opportunity for birds to access baits and/or effective complete removal of the source of attraction. This may be achieved by engineering and vessel design modifications which allow for effective underwater setting and hauling, effective and exclusive bait protection, or use of ‘baits’ to which birds are not attracted. Only time will tell, but time may well be running out for a number of seabird populations.

During the 1990s Australia has been very active in the issue of seabird conservation (Fig. 5), and has taken a three pronged approach to the problem: (1) assessment of the conservation status of seabirds, and subsequent listing of threatened species under the Commonwealth Endangered Species Protection Act 1992 (ESP Act); (2) identification of key threatening processes affecting the conservation of seabirds, and the subsequent listing of those processes under the Commonwealth ESP Act; and (3) international action to complement domestic conservation actions for species with a global distribution.

In 1992 Environment Australia (then Australian Nature Conservation Agency) funded the preparation of a global review of albatross population and threats (Gales 1993). This review highlighted the paucity of information on status of albatross populations and also concluded, after an assessment of all threats known to be impacting albatross populations, that direct mortality in association with commercial fishing operations was the most serious threat facing albatross populations.

In 1995 the Macquarie Island population of Wandering Albatrosses was also listed as Endangered under the Endangered Species Protection Act.

The threatened status of albatrosses worldwide, and their vulnerability to longline related mortality was recognised at the First International Albatross Conference held in Hobart in 1995. The workshop associated with this conference agreed there was a need to review the global conservation status listings in light of the taxonomic revision of the group (Alexander et al. 1997). This re-evaluation revealed that albatrosses have the highest proportion of threatened species in any bird family (that comprises more than a single species, Croxall and Gales in press). This global evaluation was then widely consulted in the 1997 assessment of albatross status by the Australian Government, an assessment which culminated in the listing of four albatross species as Endangered and thirteen species as Vulnerable. Three species were judged as not meeting the criteria for listing. Recovery plans for all listed species are scheduled to be implemented by 2000. These plans must provide for the research and management actions necessary to stop the decline of, and support the recovery of, the species listed so that its chances of long-term survival in nature are maximised.

The Australian Government recognised the threat and listed the incidental catch (or by-catch) of seabirds during oceanic longline fishing operations as a key threatening process on Schedule 3 of the Endangered Species Protection Act 1992 on 24 July 1995. The Act requires the preparation and implementation of a Threat Abatement Plan within three years for nationally coordinated action to alleviate the impact of longline fishing activities on seabirds in Australian waters. The plan was subsequently developed in consultation with the pelagic longline fishing industry, non-government conservation groups, scientists and government authorities responsible for conservation and fisheries management.

The objective of the Threat Abatement Plan is to reduce seabird by-catch in all fishing areas, seasons or fisheries to below 0.05 seabirds per thousand hooks, based on current fishing levels. This represents a reduction of up to 90% of seabird by-catch within the Australian Fishing Zone, and should be achievable within the five year life of the Plan.

The plan aims to significantly reduce the by-catch of seabirds by: (1) prescribing the appropriate modifications to fishing practices or equipment (mitigation measures); (2) providing for development of new mitigation measures; (3) educating fishers and the public; and (4) collecting information necessary to improve knowledge of seabird-longline fishery interactions.

Specific actions in the plan prescribe the mitigation measures which must be used by domestic and foreign longline vessels in longline fisheries, fishing areas and fishing seasons in the AFZ. These include the following measures for pelagic longline fishing in the AFZ: (1) night setting of hooks as one of three mandatory options available for selection by fishers; (2) the use of lines which are sufficiently weighted to cause the baits to sink out of reach of diving birds immediately after they are set, as part of one of three mandatory options to be selected by fishers; (3) the use of thawed bait, as part of one of three mandatory options to be selected by fishers; and (4) a requirement that from 1998 all vessels operating in the AFZ will carry bird lines and use them when appropriate. Use of bird lines below 30 ⁰S will remain mandatory.

The progress of the plan will be evaluated each year. If the adoption rates of mitigation measures are inadequate and if seabird by-catch rates do not decrease throughout the initial five year life of the Plan, it is intended that other mitigation measures, such as area or seasonal closures, will be investigated.

The Threat Abatement Plan and Recovery Plans represent Australia’s domestic contribution to the global conservation of seabirds. However, conservation of many of these migratory seabird species will rely on more than Australian action. For example, mitigation strategies such as those outlined in the Threat Abatement Plan need be pursued in international waters and the Exclusive Economic Zones of other Southern Hemisphere nations. The Australian Government is pursuing such action through international fora such as the Convention for Conservation of Migratory Species of Wild Animals (CMS), the Ecologically Related Species Working Group of the Commission for the Conservation of Southern Bluefin Tuna (ERS of CCSBT), the Convention for the Conservation of Antarctic Marine Living Resources (CCAMLR), and the Food and Agriculture Organisation of the United Nations and its Committee on Fisheries (COFI).

The Convention on the Conservation of Migratory Species of Wild Animals (CMS or Bonn Convention) provides a framework for enhancing the conservation status of migratory species through the co-operative efforts of the countries along the migratory path of those species. The CMS provides for listing of migratory species under two appendices:

Appendix I lists ‘critical’ species which are on the brink of extinction such as the Blue Whale and Short-tailed Albatross. Taking of animals listed under Appendix I is strictly prohibited.

Appendix II provides for listing of species that require or would benefit from international co-operative agreements for conservation. The appendix does not prohibit taking of listed species. It does provide for conservation agreements to be made which focus international research and management efforts on conservation of the listed species.

Australia proposed the listing of eleven species of albatrosses to the appendices of the CMS because it believed that international cooperation was required to reduce the threats, specifically that presented by longline fishing, to these species. This proposal was considered and accepted by CMS in April 1997 at the 5th Conference of Parties to the Convention.

Following that listing, Australia is encouraging the development of a Regional Agreement to be developed under CMS with southern hemisphere range states to ensure cooperation to achieve effective conservation of albatrosses globally. A key focus of this cooperation will be the need to ensure that conservation measures of the type that Australia is pursuing within the Australian EEZ are applied to other countries EEZ's and the high seas. If conservation measures are not applied to areas outside the Australian EEZ then the status of Australian albatross populations cannot be secured.

International agencies, of which Australia is a member state, which have been active in the area of seabird and longline interactions, include the Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR). CCAMLR has canvassed the issue of seabird bycatch on longlines since the 1980s , particularly since the report documenting the observations by Brothers was tabled in 1990. In 1994, the year in which CCAMLR established a working group to focus on the issue (Incidental Mortality Arising from Longline Fishing Working Group), CCAMLR also endorsed a conservation measure aimed at reducing bycatch (Conservation measure 29/XIII). The initiatives of CCAMLR in the seabird bycatch issue has been well documented by Haward et al. (in press) who conclude that the CCAMLR initiatives are important in that they offer a ‘mitigation model’ for use by other international fisheries management organisations.

Other international forums, to which Australia is a party, which have adopted seabird bycatch on longlines as an issue (see Haward et al. in press) include the World Conservation Union (IUCN) which called for actions to reduce incidental mortality of seabirds to ‘insignificant levels for affected species’ (IUCN 1997). More recently, the Food and Agriculture Agency of the United Nations (FAO) established in 1997 a Seabird Technical Working Group which is charged with provision of guidelines to reduce the incidental catches of seabirds in longline fisheries (to be adopted in 1999 by FAO’s Committee of Fisheries).

How many seabirds are killed on longlines is a complex and difficult question to answer. How many seabirds can be killed, without compromising the conservation status of any population, is also a difficult matter to resolve, although one approach being considered is e.g.: New Zealand’s MALFIRMs (maximum allowable levels of fishing related mortality, similarly referred to as a TAC). Ensuring that these levels are not exceeded, both in legal and illegal fishing operations, is logistically, near impossible.

Seabirds have most likely been killed on longline hooks since longlining became a popular fishing technique in the early 19th century. In these early days the impacts on non target (and target) species was limited by the lack of capacity to range far off shore and preserve the catch. In the late 19th Century and early 20th Century it was the land based actions of sealers and feather hunters who threatened the survival of some albatross populations (Medway in press). These threats then abated but were replaced by threats at sea made possible by two technological innovations - refrigeration and long range boats - these innovations being driven by an increasing global demand for fish.

Pelagic longline fishing, made possible by long range ships with large capacity freezers, has expanded since the 1950s and now represents the latest - though potentially most serious threats faced by albatrosses. Some fifty years later, just as technological innovation to fishing practices created the opportunity for widespread adverse impacts of longline fisheries on birds, it seems that it is again going to be technological innovation which will ultimately be the only way to effectively reduce the impact to acceptable levels over the worlds oceans. Operational and legislated changes may be effective within the EEZs, but on the high seas, the domain of these seabirds, it will take a new generation of longline fishing practices to secure the future of these birds. The other event which will reduce the impact of the fishery may be the commercial extinction of the target species.

The challenge which lies ahead is threefold. We must ensure that sufficient effort and resources are expended to ensure the successful and timely development and implementation of changes to fishing practice which alleviates the threat of longline fishing to seabirds. We must also ensure that we are sufficiently alert to identify the next threat which will impact on these birds, as history shows that anthropogenic impacts are serial and successive. Thirdly, we must ensure that the lessons learned from the process of efforts to mitigate the threat of longline fishing on seabirds are applied to efforts to ameliorate other threats so that the pace of effective change is accelerated.

The Australian Fisheries Management Authority have provided the observer datasets and the cooperation of the observers has played a vital role in the data collection for more accurate assessments of seabird bycatch. Environment Australia has provided continued funding for this work and significant input has been given by Andrew McNee, Jason Ferris and Narelle Montgomery. TASPAWS continues to support this work, does the Australian Research Council through the provision of a QEII Fellowship to RG.

Alexander, K., Robertson, G. & Gales, R. 1997. The incidental mortality of albatrosses associated with longline fishing. Australian Antarctic Division, Hobart.

Anon. 1996. Hawai’i longliners hook, kill thousands of albatrosses annually. Environment Hawai’i, January 1996: 1, 4-5.

Ashford, J.R., Croxall, J.P., Rubilar, P.S. & Moreno, C.A. 1995. Seabird interactions with longlining operations for Dissostichus eleginoides around South Georgia, April to May 1994. CCAMLR Science 2: 111-121.

Barnes, K.N., Ryan, P.G. & Boix-Hinzen, C. 1997. The impact of the hake Merluccius spp. longline fishery off South Africa on procellariiform seabirds. Biological Conservation 82: 227-234.

Bergin, A. 1997. Albatross and longlining - managing seabird bycatch. Marine Policy 21: 63-72.

BirdLife-International 1995. Global impacts of fisheries on seabirds. Birdlife International, Cambridge, UK.

Brothers, N. 1988. Approaches to reducing albatross mortality and associated bait loss in the Japanese longline fishery. Unpublished report of TASPAWS.

Brothers, N. 1991. Albatross mortality and associated bait loss in the Japanese longline fishery in the Southern Ocean. Biological Conservation 55: 255-268.

Brothers, N. 1993. A mechanised bait throwing device for longline fisheries - performance assessment of a test machine. Department of Parks, Wildlife and Heritage,

Brothers, N. 1996. Longline fishing dollars and sense: Catching fish not birds using bottom set or mid-water set longlines. Tasmanian Parks and Wildlife Service, Hobart.

Brothers, N. in press. A review of longline fisheries seabird bycatch mitigation measures and their effect on other marine species. Background paper for Food and Agricultural Organisation of the United Nations, Rome, 1998.

Brothers, N. & Foster, A. in press. Seabird catch rates: an assessment of causes and solutions in Australia's domestic tuna longline fishery. Marine Ornithology :

Brothers, N., Foster, A. & Robertson, G. 1995. The influence of bait quality on the sink rate of bait used in the Japanese longline tuna fishing industry: An experimental approach. CCAMLR Science 2: 123-129.

Brothers, N., Gales, R. & Reid, T. 1998a. Seabird interactions with longline fishing in the AFZ: 1996 seabird mortality estimates and 1988-1996 trends. Wildlife Report 98/1, Parks and Wildlife Service, Tasmania.

Brothers, N., Gales, R. & Reid, T. 1998b. Seabird interactions with longline fishing in the AFZ: 1997 seabird mortality estimates and 1988-1997 trends. Wildlife Report 98/2, Parks and Wildlife Service, Tasmania.

Brothers, N., Gales, R. & Reid, T. in press. Interpretation of effects of mitigation measures on seabird catch rates by analysis of Japanese longline fishery observer data within the Australian Fishing Zone, 1988-1995. Biological Conservation

CCAMLR 1996. Fish the sea not the sky. How to avoid by-catch of seabirds when fishing with bottom longlines. Commission for the Conservation of Antarctic Marine Living Resources, Hobart.

CCAMLR 1997. Report of the Working Group on Fish Stock Assessment. Hobart: Commission for the Conservation of Antarctic Marine Living Resources.

Cherel, Y., Weimerskirch, H. & Duhamel, G. 1996. Interactions between longline vessels and seabirds in Kerguelen waters and a method to reduce seabird mortality. Biological Conservation 75: 63-70.

Cochran, W.G. 1977. Sampling techniques. John Wiley and Sons.

Cooper, J. & Wanless, R.M. in press. The incidental bycatch of seabirds in specific longline fisheries: a worldwide review. Background paper for Food and Agricultureal Organisation of the United Nations, Rome, 1998.

Croxall, J.P. in press. Research and conservation: a future for albatrosses? In: G. Robertson & Gales, R. (eds) The Albatross: Their biology and conservation. Surrey Beatty and Sons, Chipping Norton, Australia.

Croxall, J.P. & Gales, R.P. in press. Albatross populations: status and threats. In: G. Robertson & Gales, R. (eds) The Albatross: Their biology and conservation. Surrey Beatty and Sons, Chipping Norton, Australia.

Croxall, J.P., Rothery, P., Pickering, S.P.C. & Prince, P.A. 1990. Reproductive performance, recruitment and survival of Wandering Albatrosses Diomedea exulans at Bird Island, South Georgia. Journal of Animal Ecology 59: 779-796.

Duckworth, K. 1995. Analysis of factors which influence seabird bycatch in the Japanese Southern Bluefin Tuna longline fishery in New Zealand waters; 1989 to 1993. New Zealand Ministry of Agriculture and Fisheries.

Duckworth, K. & Wells, M. 1995. Reduce your catch - advice on how not to catch seabirds. Seafood New Zealand, November 1995: 48-49.

Environment Australia 1998. Draft Threat Abatement Plan for the Incidental catch (or bycatch) of seabirds during oceanic longline fishing operations. Environment Australia Biodiversity Group, Canberra.

FAO 1998. Draft Plan of Action For Reducing Incidental Catch of Seabirds In Longline Fisheries. Rome.

Gales, R., Brothers, N. & Reid, T. in press. Nature and magnitude of seabird mortality in the Japanese longline fishery in the Australian Fishing Zone 1988-1995. Biological Conservation

Harper, P.C. 1987. Feeding behaviour and other notes on 20 species of procellariiformes at sea. Notornis 34: 169-192.

Harper, P.C., Croxall, J.C. & Cooper, J. 1985. A guide to foraging methods used by marine birds in Antarctic and sub-Antarctic seas. Biomass Handbook 24: 1-22.

Haward, M., Bergin, A., & Hall, H.R. 1998. International legal and political bases to the management of the incidental catch of seabirds. In: G. Robertson & Gales, R. (eds) The Albatross: Their biology and conservation. Surrey Beatty and Sons, Chipping Norton, Australia.

Hedd, A., Gales, R., Brothers, N. & Robertson, G. 1997. Diving behaviour of Shy Albatrosses Diomedea cauta in Tasmania: initial findings and dive recorder assessment. Ibis 139: 452-460.

Huin, N. 1994. Diving depths of White-chinned Petrels. The Condor 96: 1111-1113.

IUCN 1997. Resolutions and recommendations. World Conservation Congress Montreal, Canada 13-23 October 1996. Gland: International Union for Conservation of Nature and Natural Resources. 95 pp.

Klaer, N. & Polacheck, T. 1995. Japanese longline seabird bycatch in the Australian Fishing Zone April 1991 - March 1994: catch and catch rates by area and season and an evaluation of the effectiveness of mitigation measures. CSIRO Division of Fisheries, Hobart.

Medway, D.G. 1998. Human-induced mortality of Southern Ocean Albatrosses at sea in the 19th century: a brief historical review. In: G. Robertson & Gales, R. (eds) The Albatross: Their biology and conservation. Surrey Beatty and Sons, Chipping Norton, Australia.

Murray, T.E., Bartle, J.A., Kalish, S.R. & Taylor, P.R. 1993. Incidental capture of seabirds by Japanese Southern Bluefin tuna longline vessels in New Zealand waters, 1988-1992. Bird Conservation International 3: 181-210.

Nevitt, G.A., Veit, R.R. & Kareiva, P. 1995. Dimethyl sulphide as a foraging cue for Antarctic procellariiform seabirds. Nature 376: 680-682.

Pemberton, D., Brothers, N.P. & Gales, R. 1995. The significance of investigating hook sinking rates to seabird mortality and fishing efficiency in the longline fisheries for tuna. Tasmanian Parks and Wildlife Service.CCSBT/ERS/WG 1995.

Robertson, C.J.R. and Nunn, G.B. in press. Towards a new taxonomy for albatrosses. In: G. Robertson & Gales, R. (eds) The Albatross: Their biology and conservation. Surrey Beatty and Sons, Chipping Norton, Australia.

Ryan, P.G., Boix-Hinzen, C., Enticott, J. W., Nel, D.C., Wanless, R. & Purves, M. 1997. Seabird mortality in the longline fishery for Patagonian Toothfish at the Prince Edward Islands: 1996-97. CCMALR WG-FSA-97/51. 15 pp.

Thompson, K.R. 1992. Quantative analysis of the use of discards from squid trawlers by Black-browed Albatrosses Diomedea melanophris in the vicinity of the Falkland Islands. Ibis 134: 11-21.

Tuck, G. & Polacheck, T. 1997. Trends in tuna longline fisheries in the Southern Ocean and implications for seabird bycatch: 1997 update. CCSBT/ ERS/WG/1997.

Ward, P.J. 1996. Japanese longlining in Eastern Australian waters. Bureau of Resource Sciences, Canberra.

Weimerskirch, H. & Jouventin, P. 1987. Population dynamics of the Wandering Albatross, Diomedea exulans, of the Crozet Islands: causes and consequences of the population decline. Oikos 49: 315-322

Weimerskirch, H. & Wilson, R.P. 1992. When do Wandering Albatrosses Diomedea exulans forage? Marine Ecology Progress Series 86: 297-300.

Weimerskirch, H. & Sagar, P. 1996. Diving depths of Sooty Shearwaters Puffinus griseus. Ibis 138: 786-794.

Whitelaw, W. 1997. Some observation on seabird by-catch from Australian longline fishing vessels: 1994-1996. CCSBT-ERS/97/

Table 1. Summary of incidence of seabirds recorded as bycatch in Japanese and domestic tuna fishery in AFZ

Table 3. Status of species caught on longlines in AFZ with population information summarised from band returns.

Fig 1. Distribution of Japanese pelagic longline fishing effort within the AFZ since fishing commenced in 1979. Distribution of carcase returns from Japanese pelagic longline fishing vessels operating in the AFZ since 1988.

Fig 2. Numbers of hooks set and observed by Japanese and domestic pelagic longline fishing vessels within the AFZ.

Fig 3. (a) Observed seabird catch rate on Japanese pelagic longline fishing vessels throughout the AFZ. (b) Observed seabird catch rate on Japanese pelagic longline fishing vessels, divided by season (Summer = October to March; Winter = April to September). (c) Observed seabird catch rate on Japanese pelagic longline fishing vessels during winter in the area around Tasmania.