Australia: State of the Environment Second Technical Paper Series (Coasts and Oceans), Series 2
David Barratt, John Garvey and Jean Chesson
Bureau of Resource Sciences, Australia
Department of the Environment and Heritage, 2001
ISBN 0 6425 4745 9
"Total bycatch (ie non-target species, discards, incidentals, etc) of fishing relative to catch of target species"
(see indicator 8.4; Saunders et al. 1998, p25)
Bycatch in trawl fisheries is identified as one of four indicators of harvesting pressure in Saunders et al. (1998). The other components are: the extent of legal and illegal export of wildlife; the number of permits requested and issued for the harvest of wildlife; and the proportion of each major species' reproductive biomass removed by fishing activity.
The tender document for this project states that the removal of indigenous organisms from the wild is a measure of the pressure of such activities on biological diversity. Capture of bycatch species in fishing reduces population sizes and may change the demographic characteristics of populations, affecting fecundity and rates of recruitment.
For this study, the indicator is defined as the total bycatch of fishing relative to catch of target species, or more specifically, the weight of non-target species captured in a fishery compared to the weight of target species taken.
To develop this indicator, it is first necessary to determine what is meant by "bycatch". The "catch" of a fishery can be divided into six components as shown in Table 4.1. (Chesson and Whitworth 1999). The "target species" are defined as the main species caught by the fishery. There is an important distinction between the target species of a fishery and species targeted (deliberately sought) by an individual fisher on a particular fishing trip. For many fisheries, the target species are easy to specify. For other fisheries, the classification of some species is ambiguous. The ambiguity is not a problem in the context of this project provided each species is assigned to one and only one category.
|Components of "catch"||Target species||Non-target species|
|Other- Organisms affected by fishing gear which do not reach the deck.||E||F|
Referring to the table, commonly used terms can be defined as follows:
- total catch: A + B + C +D
- retained catch: A + B
- discarded catch: C+D
- bycatch: B + D + F
- byproduct: B.
The National Bycatch Policy uses the above definition for bycatch, but restricts its attention to D and F. Some definitions of bycatch also include discarded target catch (C), others exclude E and F (Alverson et al 1994).
Chesson and Whitworth (1999) show that there are at least four issues of concern related to bycatch:
- effects on operating costs
- catch utilisation
- direct effects on non-target species (ie mortality, injury)
- indirect effects such as
- effects on community interactions (predation, competition)
- food provisioning (for crabs, seabirds etc) from discards.
The last two issues are relevant to this study. In terms of direct effects on non-target species, catch components B, D and F are relevant. In terms of indirect effects, there are two aspects. The total removal from the system (A, B, C, D, E and F) can be regarded as a measure of overall pressure. The total discards (C and D) can be regarded as a measure of food provisioning. In each case, the fate of the individuals caught must also be taken into account. All retained catch is killed. Depending on the fishery and species, most or some of the discarded catch is killed. Mortality for the "other" catch component is variable.
Since both direct and indirect effects are likely to be of concern, we assess data availability in terms of each of the six components identified in Table 4.1.
Harris and Ward (1999) provides a comprehensive review of bycatch in Commonwealth fisheries and is the source of much of the information in this chapter.
In Commonwealth fisheries, catch information is recorded in compulsory logbooks and fisheries returns (Table 4.2). Some of these data collection programs have been running for more than two decades. However, in most cases only the retained catch (including byproduct) is reliably recorded by fishers. When compulsory fishers' logbooks have included fields for recording discards, compliance has been poor. This problem is exacerbated when discards are counted against quota if reported. Even when there is good cooperation from fishers, detailed recording of bycatch is time-consuming and technically difficult. A high level of taxonomic skill is needed to reliably identify many of the organisms.
From examination of State reports it appears that some fisheries record some non-target species, mainly byproduct and discarded vulnerable species (eg turtles and marine mammals), but discards of the majority of non-target species are not recorded. In some fisheries, there is no regular reporting of even the target species. In Queensland, NSW and WA, recording of bycatch is currently being considered for compulsory inclusion in management plans.
Other sources of bycatch information include observer programs set up specifically to quantify non-target catch (for example, Liggins (1996) and Garvey (1998a) in the South East Fishery) and specific research programs (for example Munro (1984), and Poiner and Harris (1996) for the Northern Prawn Fishery). As with any data collection program, it is important that observer programs have adequate quality assurance. The Commonwealth Sub-Antarctic fisheries have perhaps the most extensive observer coverage with two full-time scientific observers carried on every cruise. More diffuse data exist from fish surveys and bycatch information obtained incidentally in other research programs.
For most fisheries in Australia, there is insufficient information to implement the bycatch indicator. Exceptions are the main prawn (Northern Prawn Fishery, Torres Strait Fishery) and fish trawl (South East Fishery) fisheries, and the longline tuna fisheries (Eastern Tuna and Billfish and Southern Bluefin Tuna). Where information is available, it is usually for only one or a few points in time.
The amount of discarded catch (C + D) is highly dependent on the method of fishing. The highest proportion of catch discarded have been recorded in the major prawn trawl fisheries (for example, 95% in the Northern Prawn Fishery) and fish trawl fisheries (for example 50 -86% in the South East Trawl Fishery trawl fisheries). Fisheries such as the squid jig and purse seine fisheries are highly selective and have little "non-target" catch and almost no discards. Even within a trawl fishery, however, the level of discarding can vary greatly by region, season and fishing operation. In the tropical prawn fisheries, discards are rapidly eaten by birds, dolphins, sharks and bottom scavenger fishes. The fate of discards in other fisheries is unknown.
Based on published and unpublished information and a general knowledge of the fisheries concerned, taking particular account of the fishing method, we have estimated the proportion of total catch discarded in each of the 144 managed fisheries in Australia. The results are summarised in Table 4.4. In the absence of actual discard data, this approach could be combined with information on retained catch to obtain a rough national estimate of total removals (ignoring differential survival). Repeating this in subsequent years would provide a trend over time, but the trend would reflect only changes in the use of different methods. It would not reflect use of bycatch reduction devices or other bycatch management measures.
|Retained Target||Retained Non-target||Discarded target||Discarded Non-target||Other Target||Other Non-target||Number rare animals|
|Bass Strait central zone scallop||H||recent years||y||?||y||n||n||n||n|
|Christmas Island||B||After 1992||y?||?||n?||n?||n?||n?||n?||Traditional and developmental fishery|
|Cocos (Keeling) Islands||B||After 1998||y?||?||n?||n?||n?||n?||n?||Traditional and developmental fishery|
|Coral sea||B||After 1995||y||?||n||n||n||n||n||Small industry, little known, was under State jurisdiction before 1995|
|East coast deep water trawl||B||After 1993||y||?||n||n||n||n||n||Developmental fishery|
|Eastern tuna and billfish||H||After 1989?||y||y||?||y?||n||y?||y||Observer data and logbooks|
|Great Australian Bight trawl||H||After 1988||y||y||n?||n||n||n||n||Discards may now be included in the logbook since 1997|
|Heard Island and McDonald Islands||H||After 1997||y||y||y||y||y||y||y||Small fishery- 2 boats, 2 observers on every trip|
|Jack mackerel||H||After 1985||y||y||y||none||n||?||none||Targeted fishery, little bycatch < 0.1%. Logbooks from Tasmania|
|Macquarie Island||H||After 1995||y||y||y||y||y||y||y||Small fishery- 1 boat, 2 observers on every trip|
|Norfolk Island||B||After 1996||y?||?||n?||n?||n?||n?||n?||Traditional fishery|
|Northern prawn||H||After 1970||y||y?||n||n||n||n||y?||Data can be used from earlier research surveys|
|North west slope trawl||H||After 1986||y||y||n||n||n||n||n||Logbooks- Only retained catch is known|
|South east trawl||H||After 1992||y||y||n||y||n||n||y||Data estimated from observer samples|
|South east non-trawl||H||Up to 1997||y||y||y||n||n||n||n||Commonwealth logbook introduced in 1997|
|Southern bluefin tuna||H||After 1985||y||y||n||y||n||y||y||Early data not reliable, Australian data less reliable, Japanese data good after 1989, very good after 1995|
|Southern shark||H||After 1960||y||y||y||n||?||n||n||Shark fallout from nets may be estimated|
|Southern squid jig||H||After 1996||y||y||none||?||?||n||?||Bycatch very low|
|South Tasman Rise||B||After 1998||y||y||y||y||n||n||y||Data estimated from observer samples|
|Torres Strait lobster||H||After 1997||y||y?||y||none||y?||n||y?|
|Torres Strait prawn||H||After 1989?||y||n||n||n||n||n||y?||Data can be used from other ecological studies|
|Western deep water trawl||H||After 1992||y||y||N||n||n||n||n||Logbooks- Only retained catch is known|
|Western and southern tuna and billfish||H||After 1979||y||y||?||?||n||?||?||Some observer data- annual surveys|
Definitions: y= Information is recorded, y?= Information probably recorded, ?= Unknown if information exists , n?= Information probably not recorded , n= Information not recorded
References (Refs): H= Harris and Ward 1999, B= BRS Status report
|Fishery||Landed catch||Non-target catch||Species of particular concern|
|Amount (t)||Value (A$m)||Amount (t)||Composition(No. of species)||Discards(%)|
|Northern Prawn||10 300||138||30-60 000||~200||95||Turtles, snakes, sawfish|
|Southern Bluefin||5 000||86||1460 (1995)||~50||83||Seabirds (albatross), sharks|
|South East Trawl||24 500||51||~ 25 000||~300||50-86||Seals, sharks|
|Bass Strait Central Scallop||1 000||21 (1995)||?||?||high||None(?)|
|Torres Strait||1 500||25||4-8000 (1980s)||~200||99 (1980s)||Turtles, snakes, dugong (dive)|
|Southern Shark||3 200||16||500-1000 (1970s)||?||?||White shark, seals|
|Eastern Tuna and Billfish||8 000||25||1520 (1995)||~60||60 (longline)||Seabirds, sharks|
|South East Non Trawl||1 400||7||?||? >67||?||Seals|
|Great Australian Bight||2 650t||6||?||? >52||?||None(?)|
|Southern Squid||1 800||2 (1995)||low, <18 ?||Few||-||Seals|
|Western Tuna and Billfish||350||2.1||?||? ~60||? >50||Seabirds (esp. 2 spp), shark|
|Jack Mackerel||9 900||n.a.||low, <10 ?||Few||-||Seals (low concern)|
|Western Trawl||360||1.6||?||? ~ 400||?||None(?)|
|Macquarie||n.a.||low? (tentative)||? > 23||little||Albatrosses, seals, penguins (#)|
(#) Precautionary concern because of closeness to a unique island reserve.
n.a. Not available
|Discarded catch as a percentage of total catch||No. of fisheries||Proportion of all fisheries|
|1-9% (Very low)||52||36%|
Fishing method: Fish trawling, Danish seining
Landings: 27 900 t 
Amount: ~ 25 000 t
Species: 16 quota species/species groups
Composition: ~300 species
Value: A$56m 
Fate: 50-86% discard.
Species of public concern: Seals, deepwater sharks
The South East Trawl Fishery is one of the oldest commercial fisheries in Australia and supplies fresh fish to Sydney and Melbourne markets. It also supplies fish for export and is an important source of employment for many New South Wales, Victorian and Tasmanian coastal communities. In 1997, it supplied about 28 000 t of marine product worth A$56 million (Caton et al. 1998).
The current South East Trawl Fishery is a complex fishery extending from Sydney southwards around Tasmania to Kangaroo Island in South Australia. It encompasses a range of demersal habitats and 16 subfisheries in 6 statistical zones (Tilzey 1994). These zones are Eastern Zones A and B, Eastern Tasmania, Western Tasmania, Western Zone and Bass Strait (Figure 4.1).
In 1997 there were 108 active vessels operating in the fishery. Over 80 commercial species are caught. Sixteen of these species or species groups (column 2, Table 4.5) together account for about 80% of the retained catch and are managed under a system of total allowable catches (TACs). The TACs are allocated to fishers as individual transferable quotas. Some of the remaining commercial species are caught in quantities that may warrant their eventual inclusion as quota species, but most occur as a small incidental catch.
Target and Non-target Species
The distinction of target and non-target species is particularly difficult in this fishery. The fishery has quota species that are regularly assessed and are the focus of fishery management. For most purposes, the quota species are designated as the target species of the fishery, and all other species as non-target species. This categorisation should not be confused with the concept of targeting as applied to the behaviour of individual fishers on a particular fishing trip. It is possible for a fisher to actively seek non-target species. Non-target species that are frequently sought by individual fishers are listed in Table 4.5. Conversely, target species may be caught incidentally while fishing for other species and would not be considered a target species by that fisher on that occasion.
Figure 4.1 Location of South East Fishery statistical zones
Source: (modified from Tilzey 1994)
Sources of Information
Trawl catch records extending as far back as 1918 were unearthed in 1993 in a project to process and computerise historical catch data collected by CSIRO before the 1960s (Klaer 1996, Klaer and Tilzey 1996). That project also obtained records from the owner of a trawler company that operated from 1938 to 1961. However, only retained catches were recorded and these data provide an indication of what are now mostly the quota species and also of some "non-target" species such as Chinese leatherjackets, barracouta, latchets and skates/rays. Since commercial species could have been high-graded, records of the less commercial species such as sharks and dogfishes depended greatly on catches of the more valuable species or on market conditions (Klaer 1996).
A compulsory fishing logbook system was introduced for the South East Trawl Fishery in 1985. The logbook recorded fishing operations and the retained catch of quota species and some other commercial species. However, the logbook records provide no information on discarded catch.
|Family group||Target species (quota)||"non-target" species (non-quota)
Species considered targets by fishers
|Number of other "non-target" species in family|
|Hakes||Blue grenadier||Southern hake||0|
|Dories||John Dory, mirror dory||King dory||3|
|Scorpion fishes||Ocean perches||8|
|Flatheads||Tiger flathead, flathead||Deepwater flathead||0|
|Whitings||Eastern school whiting||King George whiting||0|
|Trevallies||Silver trevally||Yellowtail scad||2|
|Trevallas||Blue-eye trevalla, blue and spotted warehou||4|
|Prawns||Royal red prawns||3|
|Hound sharks||Gummy and school shark||0|
|Dogfishes||Brier shark, piked dogfish, Centrophorus sp., Daenia sp., dogfish (unspecified)||10|
|Saw sharks||Saw sharks||0|
|Angel sharks||Australian angel shark||1|
|Oreos||Spiky oreo, smooth oreo||4|
|Squids||Arrow squid, squid (unspecified)||0|
Tilzey (1994) compared the catches recorded in logbooks with verified landings and found that the logbook underestimated the landings from 1986 to 1989. The extent of under-recording was low for high-volume, desirable species but increased for the low-volume, retained "non-target" species.
The best information on "non-target" species comes from onboard observers. In the SEF, an observer program has been running since the early 1990s collecting data on fishing operations, total catch of all species and discards for a statistical sample of SEF fishing trips.
The Scientific Monitoring Program (SMP) was set up as a pilot program in 1992 in response to concerns that ITQs would reduce the quality of the information collected by the logbook program and would also cause an increase in levels of discarding (Garvey 1996). The SMP was coordinated by BRS and included fisheries agencies of New South Wales, Tasmania and Victoria. New South Wales had a collaborative monitoring program that also collected catch and discard data from the Eastern A region and central northern New South Wales (Liggins 1996).
In 1996, AFMA assumed responsibility for funding the SMP and its role was expanded by combining on-board monitoring with port-based length data collection. During 1996 and 1997 this program was run by BRS on an interim basis while AFMA funded a study to design a more comprehensive sampling strategy (Smith et al., 1997). It is important to note that the sampling design made much use of the SMP pilot data.
In 1998 the sampling strategy was implemented through the Integrated Scientific Monitoring Program (ISMP). This program is currently run by the Marine and Freshwater Research Institute in Queenscliff. This program is one of a number of measures contained within the South East Trawl Fishery Management Plan 1998 (section 4 (c)). The management plan also makes it obligatory for vessels to carry observers if requested. Details of this plan relating to the monitoring of bycatch can be found in Appendix V.
During 1996 and 1997 the New South Wales Fisheries Research Institute research vessel FRV Kapala carried out replicate surveys of fishing areas around Sydney, Ulladulla and Eden to quantify the changes in relative abundance and composition that have taken place since surveys undertaken between 1976 and 1977 (Graham et al., 1997). By using the same gear, in the same locations and seasons, changes that have taken place in target and "non-target" species over 20 years could be examined. Using the results of that survey, Andrew et al. (1997) found marked changes in the comparative abundance of upper continental slope fishes between 1976-7 and 1996-7, including a 62 percent decline in commercial fish species, and strongly inferred that sustained fishing (predominantly trawling) over this period was the major cause.
Characteristics of the Catch
Figure 4.2 illustrates one type of summary information that can be obtained from the SMP. Total catch (retained and discarded) can be estimated and tracked over time. As the SMP also records information by species, it is also possible to examine individual species. The SEF is an exception in this regard. Few fisheries have this level of information on discards and non-target catch.
There are large differences in the proportion and catch rate of "non-target" species between regions ranging from 74% (1 043 kg per shot) in the western region, to 14% (236 kg per shot) in the predominantly orange roughy trawl fishery off eastern Tasmania (Table 4.6).
Composition of non-target catch
There is a wide range of "non-target" fauna in the catch of the South East Trawl Fishery. The SMP recorded 339 taxa or taxa groups in the catch. "Non-target" species made up 36% of the bony fishes, 97% of the crabs and shrimps by weight, and 100% of sharks and rays, squids and calamari, starfishes and sea cucumbers, and sponges and invertebrates. The "non-target" species nominated as targets by fishers made up 92% of the catch of squids, calamari and shells; 27% of the catch of sharks and rays, and only a small proportion (about 5%) of the catch of bony fishes and crabs and shrimps.
Figure 4.2 Total catch (discarded plus retained) by quarter for the Western Zone of the South East Fishery estimated from discard data obtained by the Scientific Monitoring Program. The error bar is the standard deviation of the estimate.
|SMP monitoring results|
|Total catch rates (kg/shot)||1390||1640||1750||1420||455|
|"Non-target" species (kg / shot)||767||236||378||1043||277|
|Percentage of catch|
|"Non-target" species ("targets")||10%||3.||11%||13%||5%|
|"Non-target" species (other)||45%||11%||11%||61%||56%|
|"Non-target" species (all)||55%||14%||22%||74%||61%|
|Average shot time (h)||3||2||3||3||-|
|Annual fishery effort (shots)||17 725||5381||3168||6683||7056|
|Estimated "non-target" catch (tonnes)||13 595||1270||1198||6970||1954|
1Includes annual trawl effort of 1825 shots from Bass Strait
Generally, less than 1 kg per shot, or less than 1% of the catch of observed vessels consisted of starfishes and sea cucumbers, sponges and other invertebrates (Table 4.7). Vessels observed from the Eastern A and B zone had the highest average catch rates of sharks and rays (231 kg per shot), crabs and shrimps, and squids, calamari and shells. The western zone also had high catch rates of squids, calamari and shells, crabs and shrimps, and sharks and rays. East Tasmania had a high catch rate of bony fishes, and low catch rates of sharks and rays, crabs and shrimp, and squids, calamari and shells, similar to West Tasmania. Catch rates of the Danish seiners were very different to the trawlers and had the highest proportion of sharks and rays (20%).
Because the catch rates of some "non-target" species are so low, they are concealed in an examination of the overall catch of the fishery. This can be partly overcome by examining the information by families of fish.
|Major animal group||% composition||CPUE (kg/shot)|
|Target||"Non-target" (T)|| A&B
|Sharks and rays||0%||27%||73%||231||60||60||130||88|
|Crabs, shrimps etc.||3%||6%||91%||16||1||1||6||1|
|Squids, calamari, shells etc.||0%||92%||8%||57||6||5||56||12|
|Starfish, sea cucumbers etc.||0%||0%||100%||4||1||<1||<1||<1|
Fate of non-target species
Most "non-target" species are discarded. SMP results indicated that the discarded proportion of the "non-target" catch varied from 50% to 86% between sectors. There were 157 "non-target" species that were always discarded. Since fishers may also discard target (quota) species in response to market forces or ITQ limits, the amount of discarding is a poor indicator of the amount of the "non-target" catch. This is in contrast to the prawn fisheries where discards and "non-target" catch are highly correlated.
There are sizeable components of the "non-target" catch that are retained and make an important contribution to the commercial operation of the fishery.
There are no studies of the mortality resulting from trawling, exposure on the fishing vessel and subsequent discarding in the fishery. Studies undertaken in northern Australia indicate that most of the fish (bony fishes and sharks and rays) die, but that some of the other animal groups, such as crabs and other crustacea, may have a higher rate of survival.
Species of particular concern
A total of 46 fur seals were recorded in the monitored catch of 1886 trawl shots fished between 1993 and 1995. Of 19 instances where the condition of the seals was also recorded, 15 seals were caught dead by trawlers in Eastern A and B, and four were caught alive by Eastern Victorian Danish seiners. Fur seals are caught across almost all sectors of the South East Trawl Fishery except perhaps the northern part in New South Wales. Many of the catch records were located between Eden and Lakes Entrance (Figure 4.3) possibly because this is also where there was a large number of vessels monitored.
Figure 4.3. The location of fur seal catches in the South East Trawl Fishery for zones other than New South Wales. Location information of New South Wales catches not accessible.
Source: SMP database. Extracted by John Garvey, BRS
Using underwater video cameras, (Shaughnessy and Davenport 1996) observed numerous instances of Australian fur seals in the vicinity of both trawl nets and benthic dredges as they were towed on the bottom or during retrieval. During this research three fur seals were caught dead out of 111 trawls of 30-minute duration fished between Wilsons Promontory and Bermagui at depths of 16-602 m. The mortality rate (0.027/trawl) was comparable to the figure for similar research cruises in South Africa (Shaughnessy and Davenport 1996). However in two subsequent research cruises, no seals were caught from 100 similar trawl shots fished in the same area (S. Davenport, CSIRO, February 1998, pers comm).
These results are preliminary indications of the incidence and mortality of seals in trawling operations and should be interpreted with caution. Their localised and chance nature indicates that larger numbers of trawl shots would have to be recorded across the entire fishery in order to reliably estimate the catch and mortality of seals in this fishery.
Fishing debris in the form of net pieces sourced from the South East Trawl Fishery has been responsible for the entanglement of fur seals (Pemberton et al. 1992, Pemberton 1993). Surveys of fur seal populations in Bass Strait and off Tasmania between 1989 and 1993 found that between about 1.5-2% of the population were entangled - a third of the entanglements were identified as trawl nets. It is possible that this "ghost fishing" mortality has, at times, been as large or greater than the mortality associated with the visible catch. It is also possible that some seals are cut free of the net and released alive but still tangled in netting (P. Shaughnessy, CSIRO, February 1998, pers comm). Live seals taken during fishing operations can be aggressive and difficult to release without endangering the safety of the crew.
Entanglement of seals in netting has been associated with the orange roughy sub-fishery. In recent years, the extent and fishing effort of this sub-fishery has decreased possibly reducing the level of seal entanglement.
Stevens (1993) considered deepwater dogfish to be some of the shark species most at risk as a result of commercial fishing operations. These deepwater species are a "non-target" species of trawlers fishing for orange roughy. In view of what is known of their biology, the species' ability to withstand fishing pressure is considered low.
Effect of fishing on non-target species
An ongoing CSIRO study is describing features of the fishery ecosystem of South Eastern Australia and modelling the relationship of several of the fishing methods used with fish populations of a range of habitat types. The study is also investigating the feeding relationships of fish on and off reefs to determine the importance of habitat. Its results will assist other ecological assessments of the effects of fishing on "non-target" species in the South East Trawl Fishery.
|Fishing method: Prawn trawling|
|Landings: 9 202 t ||Amount: 30-60 000 t|
|Species: Penaeid prawns||Composition: ~200 species|
|Value: A$105m [1996/97]||Fate: ~95% discard. Scavenged by dolphins, birds, sharks and fish|
|Species of public concern: Turtles, snakes, sawfish, seahorses and pipefish|
The Northern Prawn Fishery (NPF) is a trawl fishery for penaeid prawns extending across northern Australia from Cape Londonderry in Western Australia to Cape York in Queensland. The fishery has banana prawn and tiger prawn sub-fisheries.
The banana prawn sub-fishery lasts for a few weeks in April and May and uses trawls with a high headrope, fished mostly during the day in shots of short duration (5-30 min) that target dense schools of banana prawns. Tiger prawns do not form dense aggregations, and are fished at night using trawls of low headrope in shots of long duration (180 min).
The fishery is managed through input controls in the form of limited entry, gear restrictions, and time and area closures. The fishing fleet has been restructured in recent years to reduce fishing capacity and improve profitability. There are approximately half as many vessels in the fleet now as in the mid 1980s. In 1997, 129 boats fished and landed a total of 9 202 t of prawns. This consisted of 4 587 t of banana prawns and 2 694 t of tiger prawns (Caton et al. 1998). In 1996/97 the annual value of the fishery was about $105 million, making it the most valuable fishery managed by the Commonwealth.
Although the size of the fleet has decreased substantially, the adoption of new technology such as global positioning systems continues to increase the effective fishing effort on the targeted prawns.
Target and non-target species
Penaeid prawns (such as banana, tiger and endeavour prawns) are the target species of this fishery and are regularly assessed. All other animals caught are considered "non-target" species. For this fishery the term "bycatch" is synonymous with the "non-target" catch. The term "byproduct" is also used to identify the retained "non-target" catch.
Sources of Information
Commercial and exploratory surveys
The Northern Prawn Fishery developed as a result of exploratory surveys undertaken jointly by CSIRO, the Queensland Department of Harbours and Marine and Craig Mostyn and Co between 1963 and 1965. The southern Gulf of Carpentaria around Karumba and Mornington Island was trawl-surveyed monthly with several types of banana and tiger prawn nets at night and during the day (Munro 1984, Rainer and Munro 1982, Rainer 1984). The surveys were directed by the noted fish taxonomist from CSIRO, Ian Munro, whose influence resulted in detailed information on the identity and numbers of all animals caught while fishing for prawns. Consequently, detailed pre-fishery information on the abundance and distribution of "non-target" species is available for some areas of the Northern Prawn Fishery.
In 1983, the tiger prawn sub-fishery of southern Gulf of Carpentaria was surveyed for a year with prawn trawls used commercially in the fishery. The weights and numbers of "non-target" animals were also collected in 10 of the 12 monthly fishing trips (Poiner and Harris 1986).
In 1985 and 1986, surveys examined the changes in the "non-target" fish communities that had taken place since the 1963 exploratory period. Areas where there had been little trawling were compared to those that had been intensively trawled, to determine the impact of trawling for prawns on "non-target" fish species (Harris and Poiner 1991).
Between 1986 and 1991, a series of investigations mostly using on-board observers and commercial samples were undertaken over a wide range of fishing grounds from the western most part of the Northern Prawn Fishery (Joseph Bonaparte Gulf) to the Edward Pellew Islands in the south-western Gulf of Carpentaria. These investigations estimated the composition, amount and commercial potential of the "non-target" catch (Pender and Willing 1989, Pender and Willing 1990, Ramm et al. 1990, Pender et al. 1992.)
The catch of turtles between 1980 and 1988 was estimated from some of the above investigations (Poiner et al. 1990) and from selected and trained fisher volunteers in 1989 and 1990 (Poiner and Harris 1994, Poiner and Harris 1996).
The Northern Prawn Fishery was one of the first Commonwealth fisheries to introduce compulsory logbooks that were completed by fishers and cross-checked with onshore processor returns. These have been a reliable source of information on the target prawn catch and the fishery effort since 1970. For many years logbooks allowed a voluntary recording of "non-target" species. In 1995, it became compulsory for fishers to record the retained catch of "non-target" species and turtles.
The abundance and diet of "non-target" species that feed on prawns on the trawl grounds were investigated to estimate their effect upon prawn mortality (Blaber et al. 1990, Brewer et al. 1991, Salini et al. 1990, Salini et al. 1994).
Characteristics of the Non-target Catch
Banana prawn sub-fishery
The amount of "non-target" catch in the banana prawn sub-fishery is not known. Though it has sometimes been assumed to be of the same order as that of the tiger prawn sub-fishery (Blaber et al. 1990, Andrew and Pepperell 1992) it is likely to be considerably less.
It is difficult to get realistic estimates of "non-target" catch in this sub-fishery. The fishery is based on closely kept information on the location of schools found through aerial surveys or from characteristic echo sounder traces. It is therefore difficult to duplicate the catch with a scientific survey, and extensive monitoring would be required by observers to estimate the incidence of shots mistakenly made on schools of fishes instead of prawns.
Tiger prawn sub-fishery
From catch rates around Groote Eylandt, Grey (1984) estimated that the "non-target" catch of the tiger prawn sub-fishery was about 35 000 - 45 000 t. Using 293 trawl shots observed from 20 commercial vessels that fished across the Northern Territory, from the Limmen Bight to Joseph Bonaparte Gulf in 1988, Pender and Willing (1989) estimated a mean catch rate of "non-target" species of 144 kg per hour per vessel. From logbooks they estimated that 25 700 t were taken in the Northern Territory, and revised the estimate for the entire Northern Prawn Fishery to 47 000 t. Prawn trawlers fishing Albatross Bay in the Queensland region caught approximately 6 600 t of fish annually in 1987 and 1988, based on a prawn to fish bycatch ratio of 1:10 (Blaber et al. 1990).
The catch for the southern Gulf of Carpentaria in the Queensland region, was calculated from catch rates collected in a 1983 survey (Poiner and Harris 1986). Survey catch rates of "non-target" species for a tiger prawn trawl of 49 m headrope length, as used in the sub-fishery at the time, scaled up by the logbook fishing effort indicated that about 16 500 t of "non-target" species would have been taken in the southern Gulf of Carpentaria in 1983.
These results highlight the difficulty of estimating the "non-target" catch of a fishery which extends over such a large area. The target prawn catch and the fishery effort varies between years and between regions and "non-target" catch is also likely to be variable. The separate estimates available are fragmented in time and across regions. Overall they suggest that the "non-target" catch is large and has probably ranged from approximately 30 000 to 60 000 t per year depending on total effort in the fishery.
Since these estimates were made, the number of boat days fished have decreased. If the fishery had a negligible impact on the "non-target" species populations at the time the estimates were made, the decrease in effort would have resulted in a decrease in "non-target" catch. However, if "non-target" species populations were over-exploited at the time, the reduction in fishing effort would not have reduced the "non-target" catch but could have increased it. Blaber et al. (1990) considered the area around Albatross Bay in the Queensland region was lightly fished by prawn trawlers in 1988 because the "non-target" fish catch was less than 10% of the estimated biomass of fish present. However, Harris and Poiner (1991), found that although the overall exploitation of "non-target" fish species may have been low, populations of some benthic species had been significantly reduced as a result of trawling. Consequently, reduced effort from a restructure of the fishery could have changed the composition as well as reduced the total "non-target" catch.
Composition of the Non-target Catch
The composition of "non-target" catch is examined separately for the banana and tiger prawn fisheries.
Banana prawn sub-fishery
The early exploratory surveys of the Gulf of Carpentaria (GOC) in the mid 1960s, encountered banana prawn schools in a "balled" up condition, containing many tonnes of banana prawns and very few other species or trash fish (Munro 1984). However, at other times, schools of prawns were accompanied by schools of fish that positioned themselves directly above or slightly to the side of the prawn mass. Numerous anecdotal reports by fishers since then suggest that schools of dollarfish, leatherjackets, grunts, or catfish (Taylor 1995) are occasionally mistaken for prawns. At such times large hauls of those "non-target" species may be taken and quickly discarded. Advances in echo sounder technology and communication over the last decade are likely to have resulted in better targeting of banana prawns and lower catches of "non-target" species.
Tiger prawn sub-fishery
Bony fishes made up of 68% of the "non-target" catch in the Northern Territory region (Pender et al. 1992); sharks and rays 9%; crustaceans (crabs, bugs, mantis shrimps, non-penaeid prawns) 3%; molluscs 4%; echinoderms 3%; sponges 2%; other invertebrates 1% and reptiles 1%. The target penaeid prawns constituted 11% of the catch. The ratio of "non-target" species to penaeid prawns was 9.6:1, and the ratio of "non-target" fish to penaeid prawns was 6.5:1. There were 234 species from 86 bony fish and shark families identified. Grinners dominated the catch in weight, and other abundant families included threadfin bream, dollarfishes, shovel-nosed rays, trevallies, goatfishes and leatherjackets (Table 4.8).
The composition of the "non-target" catch in the southern GOC in 1983 was different. Grunts dominated the catch in weight, and other abundant families included threadfin bream, whiting, dollarfishes, leatherjackets, trumpeters and grinners (Table 4.9). Stingrays were the only cartilaginous fish in the most abundant fish families, unlike in the Northern Territory sector where shovel-nose rays, stingrays and sharks were included. Dollarfish, grunts, sardines and anchovies, croakers and rays were the most abundant fish families at Albatross Bay in the eastern GOC between 1986 and 1988 (Blaber et al. 1990).
These separate studies indicate important local and regional differences in the composition of the "non-target" catch. Large-scale differences exist in the patterns of abundance for "non-target" species between the eastern and western parts of the Northern Territory region (Ramm et al. 1990). To be effective, measures aimed at reducing the "non-target" catch would have to consider these regional and local differences in species composition.
Fate of non-target species
Most of the "non-target" catch is discarded. Observations of the fishing activities aboard vessels indicate about 3% (700-1 000 t) of the "non-target" catch was retained in 1988 (Pender et al. 1992), consisting mostly of bugs, squid, sharks and fish. This was mostly sold through the markets at ports such as Darwin and Cairns associated with the fishery. Small quantities of mantis shrimp, black coral, shark jaws, sawfish snouts, and bailer and other shells were collected as curios. Snakes were occasionally retained and sold to tanneries in the past (Ward 1996).
In 1995, it became compulsory for retained catches of "non-target" species to be recorded in logbooks. This amounted to 84 t in 1995, and increased to 424 t in 1996 when 292 t of squid was recorded from the Northern Territory sector. Bugs were the major retained catch in the Queensland, squid in Northern Territory, and coral prawns in Western Australia (Table 4.10).
These records are not cross-checked through processor returns as is the prawn catch and would not account for all the "non-target" species retained. The total retained catch from logbooks is less, and has fewer species than the estimates made from on-board observations in 1988, even though the utilisation of "non-target" species has probably increased since then. For example, it is highly likely in 1996 that squid catches exceeded the reported 292 t. However, the logbook records may accurately reflect the relative importance of these "non-target" species in the retained catch in each Northern Prawn Fishery sector. For example, whiting are prominent in the retained catch of the Queensland region, and sharks in Northern Territory region; consistent with the studies that have examined the abundance of these "non-target" species in the catch from these regions.
Most of the discarded catch is dead on return to the sea as a result the trawling operation and exposure on the deck of the fishing vessel (Hill and Wassenberg 1990).
Species of particular concern
The catch of some "non-target" species is of concern because they may be vulnerable to localised extinction even though they may be rarely caught. These include long-lived animals that are susceptible to the increased mortality due to fishing, or susceptible because they have a restricted distribution overlapping with the fishing ground. Some may also have a favoured status with the public.
|Fish family||% Wt||Catch (t)|
|Fish family||% Wt||Catch (t)|
Prawn and turtle habitats overlap in the shallow waters of the Northern Prawn Fishery (Limpus and Miller 1993). As a result, trawls dragged along the seabed to catch prawns also sometimes capture turtles. Turtles caught are mostly flatbacks (59%), but also include loggerheads (10%), olive ridleys (12%), greens (8%) and hawksbills (5%) (Poiner and Harris 1996). Studies of the Northern Prawn Fishery turtle catch have not recorded leatherback turtles (Poiner and Harris 1996), and anecdotal reports indicate it is an extremely rare catch. Turtles are taken primarily by the tiger prawn sub-fishery; the banana prawn sub-fishery catches are likely to be much less because the fishing season and the trawl tow duration are shorter. However, data is yet to be analysed to quantify the turtle catch of the banana sub-fishery.
All marine turtles are listed in the Convention on the International Trade in Endangered Species (CITES), and the International Union for the Conservation of Nature (IUCN) as either endangered or vulnerable species. All marine turtles are nationally protected in Australia by the National Parks and Wildlife Act (1975). The loggerhead turtle was declared endangered and the green, leatherback, hawksbill and olive ridley were declared vulnerable under the Endangered Species Protection Act 1992 (ESPA). In recent years, there have been concerns about the impacts of prawn trawling throughout northern Australia (including the Queensland East Coast) on populations of marine turtles. For example, there is particular concern about the effect of trawling on the eastern loggerhead population where there has been a 50-80% decline in nesting turtles over the last 10-15 years (Limpus and Reimer 1994), and on the green turtle population nesting on the southern Great Barrier Reef. These concerns resulted in the nomination of prawn trawling as a key threatening process to turtles under the ESPA in 1995. In 1998 it was decided that prawn trawling should not be listed as an endangering process under the ESPA. It is likely that prawn trawling will be nominated as a key threatening process to turtles under the new Environment Protection and Biodiversity Conservation Act 1999.
Poiner et al. (1990) estimated that, as a result of prawn trawling in the Northern Prawn Fishery between 1980 and 1988, about 5 700 turtles were caught annually of which about 340 died. The study identified that better information on the mortality of turtles was needed. Subsequently, trained volunteer fishers recorded catches of turtles, and their condition on capture. Using those results, the catch was estimated at about 5 500 in 1989 and about 5 200 in 1990 (Poiner and Harris 1996). About 14% of turtles caught drowned in trawl nets. Another 25% were injured or comatose when brought on board, of which half recovered when left for more than thirty minutes on the deck. Turtles are noted for their hardiness and ability to survive amputations, but the mortality rate of healthy, injured or resuscitated turtles released from trawling is unknown. Poiner and Harris (1996) estimated that in the unlikely event that all injured and resuscitated turtles died after being released, about 2100 turtles would have died each year in 1989 and 1990 as a result of trawling in the Northern Prawn Fishery. A turtle monitoring project has been running in the NPF since 1998. The estimates of turtle captures are similar to those reported by Poiner and Harris (1996). However, the mortality rate has reduced due to improved resuscitation techniques.
|Bugs - tails||13271||9092||223||22586|
|Bugs - whole||34133||15057||22||49212|
*No data for 1996
Sea snakes are found throughout the warm shallow waters of northern Australia. Based on tanning factory records and specimens sent to the Conservation Commission of Northern Territory, at least 14 species are taken in the Northern Prawn Fishery (Ward 1996). The spine-bellied seasnake and the elegant seasnake were most common in the catch (Wassenberg et al. 1994).
Heatwole and Burns (1987) estimated a mean catch rate of 0.26 snakes per hour of trawling for prawns, of which about 10% died in the banana prawn sub-fishery, and 42% died in the tiger prawn sub-fishery based on the different durations of trawl tows. Pooled results of research and commercial trawls indicated that about 120 000 snakes were taken in the Gulf of Carpentaria alone in 1991, of which between 29 800 and 67 000 died (Wassenberg et al. 1994). By far the largest catches of snakes (70%) were taken in shallow coastal waters less than 15 m deep. Fisher returns of dead snakes from trawlers in the NPF indicated significant differences in the species taken in the tiger and banana prawn fisheries (Ward 1996).
Little is known of sea snake population biology, or the level of mortality that they can sustain. In 1995, nomination of prawn trawling as a threatening process to sea snake populations was later withdrawn. Though sea snake products have been traded internationally since 1930s, sea snakes are not considered threatened and are not currently listed by CITES (Marsh et al. 1993).
In 1990, licences issued under Queensland fishery management regulations allowed collection of up to 10 000 sea-snake skins per year under licence (Marsh et al. 1993). Some of the snake skins processed at the time came from the Northern Prawn Fishery. In 1994, responsibility for the conservation of sea snakes was passed to the Queensland Department of Environment and the licences were not renewed.
Sea-snake products can only be exported from Australia under a licence issued by the Commonwealth Department of the Environment. This licence requires that the product is regulated by a management plan specifically for seasnakes, and is not the result of a catch taken in waters under Commonwealth jurisdiction. Currently, there are no licences issued for the export of seasnakes.
At least three marine species of sawfish occur in the coastal waters of the Northern Prawn Fishery (Last and Stevens 1994) and are caught by trawlers (Stevens 1993). The narrow sawfish was recorded from prawn trawl catches in the early exploratory surveys of the mid 1960s, and is still moderately common in the Gulf of Carpentaria from inshore areas to 40 m depth (Last and Stevens 1994). Sawfishes are more rarely caught than turtles. Nothing is known of the effect of trawling on their populations.
Sawfishes were recently proposed for listing in CITES because populations worldwide, particularly of the freshwater and estuarine species, are under threat. The proposal was not accepted. Australia still has reasonably intact isolated populations of sawfishes, but their restricted habitats and life histories may render them susceptible to fishing.
The freshwater sawfish was listed as vulnerable under the ESPA (1992) legislation, in October 1997. The nominations for the estuarine and marine sawfish species were not successful.
Seahorses, pipefishes and seamoths
Thirty-six species of seahorses and pipefishes were listed as vulnerable worldwide under IUCN listing criteria for marine species (Hudson and Mace, 1996). These included the spotted and the spiny seahorse which have been recorded in catches of prawn trawls in the Northern Prawn Fishery. Seahorses and pipefishes are vulnerable because of their restricted and often inshore habitats which are under increasing human pressure throughout the Indo-Pacific.
Seahorses rarely occur in prawn trawls, and then usually on occasions when trawlers accidentally foul their nets close inshore in the vicinity of sea grass beds. Seamoths, particularly the slender seamoth, are commonly taken in trawls, but this species is not considered vulnerable. In more recent years, seahorses, pipefishes and seamoths have been collected and dried for the Asian medicine market. Because of high prices (>US $1,000 per kg) in this market the component of the catch destined for Asian medicines could increase.
As of January 1998, seahorses and pipefishes can only be exported from Australia under a permit issued by the Commonwealth Government environmental agency (Environment Australia). Permits are only granted for captive bred specimens or those taken under a management regime approved under the Wildlife Protection Act 1982 to ensure that the harvest is not detrimental to the species or population.
Effect of fishing on non-target species
After 20 years of fishing for prawns in the south-east Gulf of Carpentaria, the same "non-target" fish communities that had existed before the fishery developed, were still present (Poiner and Harris 1986). There had not been any significant change in the abundance of 65% of the "non-target" species examined; whereas 22% of the "non-target" species had decreased, and 12% had increased. "non-target" species that decreased were mostly benthic, whereas those that increased mostly were mid-water species. No fish species had disappeared, but leatherjackets had decreased dramatically and a grinner (the short-finned lizardfish) had substantially increased in abundance. The diel behaviour of the "non-target" fish community had also changed - many species were more abundant in the catch during the day, after the fishery had developed. These changes were considered to have been caused by the fishing effort of prawn trawlers, the discarding of bycatch, and natural environmental changes to the sea bottom. (Harris and Poiner 1991).
It is thought that marine turtle populations cannot withstand significant mortality above natural levels because they grow so slowly and take so long to mature (> 30 years for loggerheads, Limpus and Reimer 1994). The effect on the population is usually assessed through modelling the survival through each life history stage from eggs, to adults in response to fishing mortality.
For loggerheads, the turtle species under greatest threat, about three per cent of the population that forages in the Northern Prawn Fishery were killed annually as a result of trawling for prawns. There is considerable scientific debate about whether this mortality represents a major threat to the population nesting on the Queensland east coast. Models used to study this effect were mostly static and did not include the possibility that as the turtle population decreased, its intrinsic rate of growth could increase through greater availability of food and habitat resources. There are also numerous other sources of mortality in northern Australia such as harvesting by neighbouring countries and by Australian indigenous peoples, destruction of habitat, and predation of eggs by introduced animals and goannas. Apart from prawn trawling, few of these sources of mortality have been quantified. A recent modelling study of an Australian population of loggerhead turtles indicated that fox predation of turtle eggs could have had a greater impact on the population than trawling (Chaloupka and Limpus 1997). Such studies are important in ensuring that the appropriate emphasis is placed in addressing the different risk factors affecting the population. Trawling is one of those risk factors and there is general agreement that the fishery should reduce its contribution to total mortality by using fishing practices that avoid catching turtles. Turtle exclusion devices, extensively trialed in this fishery in recent years, are very successful at reducing the turtle catch (Brewer et al. 1996) and have been promoted through a program that encourages the fishers to trial, adopt and improve the devices (SFC 1996, Eayrs et al. 1997). A recent NPF bycatch action plan sets a date for the compulsory use of turtle excluder devices and requires that, by the end of the year 2000, the fishery catches only 268 turtles annually (5% of the estimated catch of turtles ten years earlier).
The review of data quality and availability and the two case studies illustrate the complexity of the "bycatch" issue. Removal of animals from the sea is likely to have both direct and indirect ecological impacts, but our knowledge and understanding of these is very limited.
The two case studies are atypical in that they are examples of fisheries with unusually large amounts of information on catch other than retained target species. Even in these cases, the link between catch and ecological impact is poorly understood. For all species, target and non-target, the ability to determine population status is greatly enhanced by fishery independent information, but even then it is difficult to distinguish effects of fishing from other environmental effects. Short and long term changes in the abundance of non-target species in the Northern Prawn Fishery have been attributed to the fishing activities of prawn trawlers as well as environmental changes in local conditions.
What can be validly and reliably measured and what changes can be detected? When changes are detected or inferred, we have limited ability to judge their significance. Fishing will invariably affect ecological relationships between species. What changes are acceptable?
In the absence of general answers to this question, there is a need for pragmatic approaches tailored to the specific situations of individual fisheries. For example, where a fishery involves interactions with species of particular concern (marine mammals, turtles, seabirds or other species that are threatened or endangered) it may be appropriate to determine a conservative upper limit to mortality caused by fishing. If this limit is reached, the fishery is closed or equivalent management action is taken. This type of approach is taken in the United States with respect to marine mammals. Where the main concern in a fishery is the unnecessary removal and discarding of animals, the management objective may be to reduce the discard rate to below an agreed target. Where particular communities of organisms are of concern, it may be appropriate to close certain areas to fishing either completely or on a seasonal basis. In each case, the objective and hence the appropriate indicator is specific to the particular fishery.
In its present form, the proposed indicator is not a useful indicator of marine disturbance. It is poorly defined and it is unclear what aspect of marine disturbance it is attempting to measure. An increase in the amount of bycatch could reflect a failure of bycatch reduction measures or an increase in the abundance of bycatch species. Conversely, a decrease in bycatch may reflect successful implementation of bycatch reduction measures or a decline in the abundance of bycatch species. For this reason, we recommend against the use of catch of any particular species as an indicator of the state of that species. Catch is also a poor indicator of pressure on a particular species because the impact of a particular catch depends on the population size and life cycle characteristics of the species. Catch is an important piece of information in determining the impact of fishing on a species, but is not sufficient and should not be used as an indicator in its own right.
There does appear to be merit in using total removals (retained catch plus discarded catch) as a gross measure of pressure on marine ecosystems that could be compared against estimates of total productivity. For example, it has been estimated that world wide, approximately 110 million tonnes per year of fish and invertebrates are killed by ocean fishing (National Academy of Sciences 1999). Comparison with various estimates of the total productivity of ocean ecosystems at around 100 million tonnes per year, suggests that global removals are near the maximum sustainable level. Total seafood catch is listed as one of the ANZECC core environmental indicators (E+S 3, ANZECC 1999).
Estimates of total removals require an estimate of the total retained catch and the total discarded catch taking into account that some proportion of the discarded catch might survive. Total retained catch is generally recorded, although it has not been compiled on a national scale by fishery. Very little data on discarded catch is collected routinely in Australian fisheries. It is generally agreed that on-board scientific observers are required to collect detailed data, especially at the species level, and this has been implemented in only a handful of fisheries. The cost would be prohibitive in most fisheries. An estimate of total removals for a fishery, however, requires only an estimate of discard rate which can be based on knowledge of the fishing method (as demonstrated in Table 4.4), observer data, logbook records, research studies or a combination of all four. Where adequate information is not available, there would be a benefit to a fishery to collect more reliable data to demonstrate the effectiveness of a bycatch reduction measure or other factors that reduced its discard rate below the rate otherwise assumed.
- Based on the fishing methods involved, it is estimated that 114 (79%) of Australia's 144 commercial fisheries have low to very low discard rates (less than 25% of the total catch discarded). There are 19 fisheries with moderate discard rates (25 to 50% of total catch discarded) and 11 fisheries with high discard rates (greater than 50% discarded).
- The South-East Fishery is one of the few examples where bycatch information is available over time. Between 1994 and 1997, the amount of discarded catch was variable but showed no obvious trend.
- "Bycatch" is an ambiguous term. The indicator needs to be more clearly defined to specify precisely which bycatch issue it is attempting to address.
Based on our investigation of this indicator, we recommend that:
- Environment Australia pursue the total removal indicator as listed in the ANZECC core indicator set, but only as an aggregate across species. We recommend against catch (retained, discarded or total) as an indicator of either state of, or pressure on individual species. Catch may be an essential ingredient of an assessment of state or pressure, but is insufficient as an indicator in its own right.