2. THE FISHERIES AND THEIR CATCHES (Continued)

2.1.4 Other driftnet fisheries in the North Pacific

Although the three fisheries described above are the major ‘large-scale’ fisheries, and also apparently the only ones operating outside EEZ's, there are numerous other driftnet fisheries in the region. From the perspective of their impact on non-target species it is worth examining some of these briefly.

US driftnet fisheries are largely directed towards salmon in coastal or estuarine waters during the summer. Douglas (1989) and Barlow et al (1990) provide summaries of US driftnet fisheries. In Alaska, some 550 vessels are licensed to use driftnets for salmon in Prince William Sound and the Copper River delta; these nets are up to 150 fathoms (274m), with meshes sizes of 124– 178mm and a total depth of up to 27m. There are about 160 vessels licensed to use driftnets in the South Unimak region, and a further 158 permits for set or driftnet fishing along the Alaskan Peninsula. In these two areas mesh sizes must exceed 5.25 inches (133mm) and nets may not exceed 200 fathoms (366m). In Southeast Alaska there are 468 driftnet licences (for nets of up to 300 fathoms or 550m), and there are 460 licences for driftnets in Cook Inlet, where nets must be less than 150 fathoms (274m) in length and have meshes of less than 6 inches (152 mm). In Bristol Bay there are 1746 driftnetters permitted to use nets of up to 274 m in length with meshes of 4.5 to 6.75 inches (114–171 mm). Throughout Alaska, salmon nets are attached to vessels while fishing and are generally soaked for 2 or 3 hours or less.

Marine mammals known to be caught in these nets include Steller's sea lions, harbour seals, harbour porpoises, Dall's porpoises, killer whales, white whales and even grey and humpback whales. Levels of mortality are generally poorly known, but in the late 1970's about 500 harbour seals and 400 Steller's sea lions were being caught annually in the Prince William Sound and Copper River areas of Alaska, together with smaller numbers of harbour and Dall's porpoises, and sea otters.

Entanglement of birds has not been studied in detail in these fisheries, but Sealy and Carter (1984) suggest that mortality of marbled murrelets in driftnet fisheries may be a significant problem.

In addition to Alaskan salmon driftnet fisheries, around 900 boats are reported to use driftnets for salmon in Washington and Oregon, around the Columbia River, Grays Harbour and Willapa Bay, and nearly 4000 more boats operate set and drift gill nets in and around the Puget Sound region (Douglas 1989). A few hundred more vessels use driftnets for shads, barracuda and bass in Oregon and California (Uchida 1985). Salmon nets are strictly controlled. Monofilament is illegal in Washington, Oregon and Alaska. In Puget Sound driftnets may be 549m long with mesh sizes of from 114–120mm and in Oregon they are 457m long with a 127mm mesh. Entanglement of harbour porpoises, especially, has been noted in some of these fisheries, but entanglement rates are not known (Barlow et al 1990).

The California driftnet fishery for sharks and swordfish should also be mentioned again among the coastal driftnet fisheries, although fishing occurs up to the limits of the EEZ. Fishing is prohibited within 75 miles of shore during the grey whale migratory period. This fishery has been described by Barlow et al (1990) and Miller et al (1983). Nets of up to 1000 fathoms (1829 m) are employed with mesh sizes of 18–24 inches (457–610 mm), and a mean depth of around 40 m. The nets are usually suspended 5–27m below the surface. Catches in 1989 included 1360 tonnes of swordfish and around 3000 tonnes of shark. Marine mammals entangled in this fishery have included grey whales, common dolphins, minke whales, northern right whale dolphins, short-finned pilot whales, Pacific white-sided dolphins. In 54 observed sets in 1990, 5 common dolphins, 2 Pacific white-sided dolphins and 1 short-finned pilot whale were reported (Barlow et al 1990). Assuming a net length of 1829m, this would be equivalent to around 80 cetaceans entangled per 1000 km of netting set.

In the Soviet Union's far east, several types of coastal driftnets are, or were, reportedly used. An anchovy driftnet 45m long, 200 meshes deep, a hanging ratio of 0.6 and a mesh size of 14mm is described by Andreev (1966, cited in Uchida 1985). Driftnets were also reportedly used for Pacific saury, herring, mackerel and salmon. Herring nets are 30m long and 6 to 15.2m in depth, salmon driftnets are reportedly 30m long and 3.3m deep, mackerel nets employ a 40mm mesh are 6.4m deep and 30m long, while saury nets are 36m long, 5.1m deep and have a mesh size of 16mm. No recent information nor effort data could be located for these fisheries, nor any information on entanglement of non-target species.

Japanese coastal driftnet fisheries are diverse, and many vessels are involved, with a large array of net specifications. Uchida (1985) states that 38000 Japanese vessels use gill nets to the exclusion of other gears. Table 21 gives catches by species in gill nets in coastal waters. Drift and set gillnets are not separated in official statistics, so it is assumed here that all pelagic species may be taken in driftnets. It should be noted however, that in Japan, as well as some other areas, bottom driftnets and fixed surface nets may be used (Yamaha 1986). Driftnets used for Spanish mackerel in the Nagasaki prefecture (Uki district) are around 1 km in length, with a mesh size of 115cm, constructed with nylon monofilament, and are operated from September to May (Yamaha 1986). Other Spanish mackerel nets may have smaller meshes (Uchida 1985).

For sardines a 43mm mesh is used, and 40 units of netting 30–48m in length may be set at one time (1.2 km – 1.9 km). Similar gears are used for mackerels but larger meshes (70–85mm) may be used. Information on non-target catches is not available for any of these fisheries, but some accidentally caught cetaceans are reported by Japan to the IWC each year. Chinese driftnet fisheries are similarly diverse. Species caught include Spanish mackerel, herring (Ilisha), anchovy, and even swimming crabs. Several thousand vessels are involved. A summary of the Chinese driftnet fisheries is shown in Table 22. Chinese fishermen are being encouraged to adopt driftnetting as a more selective method of fishing than trap fisheries. Quantitative information on non-target catches is not available, but Zhou and Wang (1990) report finless porpoises, in particular, being caught in coastal driftnet fisheries.

The Republic of Korea has a fishing fleet which includes many thousands of gill net vessels (13890 in 1982) of which over 10% (1458 in 1982) were classified as offshore gillnet vessels; the remainder being coastal gill net vessels (Anon. 1983). The proportions of these types of vessel using driftnets are unknown, and there is apparently no information on non-target catches.

[Pelagic fish are assumed to be taken in drift gillnets, demersal fish in fixed gillnets]

N.A. indicates data are not available;
a = approximation;

2.2 Central and South Pacific

2.2.1 Large mesh gill net fishery

The most widely known driftnet fishery in the Southern Pacific is the large mesh (albacore) driftnet fishery in the Tasman Sea and in waters to the east of New Zealand, between 30° and 45°S in the Subtropical Convergence Zone.

This fishery was apparently initiated as a result of Japanese exploratory driftnet fishing for slender tuna between 1982 and 1987 (SPC 1989). Between 10 and 20 Japanese vessels participated between 1983 and 1988. In the 1988/89 season, however, 64 Japanese and possibly as many as 130 Taiwanese vessels participated in this fishery. After protests from coastal states in the region, the Japanese and Taiwanese agreed to cut their fleets to 20 and 24 vessels respectively, and the fishery is due to be phased out by 1991. In fact, only 19 Japanese vessels participated in the 1989/90 season. It is worth noting, in passing, that Roper et al (1984) state that Japanese vessels at that time were operating a driftnet fishery for flying squid in the Tasman Sea. This was apparently only a trial fishery.

The dramatic increase in the number of Japanese vessels participating in the 1988/89 season was said to be the result of salmon driftnetting vessels re-locating to the South Pacific after exclusion from the US EEZ. The albacore surface fishery is conducted during the austral summer. The season is dependant upon weather conditions, but typically may extend from November to March, whereas the salmon fishery is confined to June to August. The concerns of coastal states in the region gave rise to the Tarawa Declaration, adopted by the South Pacific Forum in July 1989, which determined to develop a convention to give effect to a zone in the South Pacific free from driftnetting and resolved that Forum member nations would take all possible measures to prevent and discourage driftnetting in the region. In November 1989 a convention was opened for signature in Wellington which would prohibit fishing with long driftnets in the South Pacific region.

One of the main causes for concern over the driftnet fishery was the lack of adequate information on its catches or on effort. Preliminary Japanese driftnet CPUE data were presented to the second South Pacific Albacore Research (SPAR) Workshop in 1989, and these data showed an increasing trend in CPUE. This was thought to be due to a shift in fishing effort towards the peak of the season. Watanabe (1990b) provided more detailed CPUE data for the Japanese fleet in three fishing areas; Watanabe's data are reproduced in Table 23 together with Japanese catch data and some limited information on Taiwanese catches presented to the 2nd SPAR workshop. Again, there is no evidence of a decline in CPUE, but changes in fishing practices and season may hamper the interpretation of such CPUE series. The absence of reliable data on Taiwanese catches or effort is also a severe limitation. Other surface fisheries accounted for around 9000 tonnes of albacore in 1989, while the longline fishery took 29,000 tonnes.

The gill nets employed in the albacore driftnet fishery are reported to vary between 20 and 55 km in length, with mesh sizes of from 160 to 200mm. The nets are 10 to 15m in depth. Coffey and Grace (1990) reported that Japanese driftnets used in the Tasman Sea were on average 40 km in length and consisted of eight individual 5 km nets laid end to end, with a 180mm multi-filament mesh, 10m in depth. Results of Japanese research cruises in the South Pacific have shown that a 180mm mesh is most suitable for capturing 65 to 79cm albacore (SPC 1989). Taiwanese vessels observed by Coffey and Grace (1990) were reported to use 200mm mesh with nets of 15m in depth set as 5 individual 9 km nets, totalling 40 km on average. Japanese vessels were observed setting nets in parallel, 3 km apart, working in a group of eight vessels.

[Note Taiwanese figures are preliminary estimates: 1989 figure is based on a catch rate of 300 tonnes per vesselassuming 60–130 vessels fished; Korean catches were made by a single research vessel in 1988/89.]

Non-target catches

Information on the catch composition of driftnetting vessels in the South Pacific is restricted to two reports. Coffey and Grace (1990) identified more than 15 species, two of which were mammals. They quantified catches in 126 km of driftnet. Sharples et al (1990) recorded a minimum of 46 species including 5 cetacean species in two cruises of the Japanese research vessel Shinhoyo Maru. Their list is reproduced below, in Table 24. They quantified catches in sets amounting to more than 1100 km of netting.

Common dolphins were the most frequently recorded mammals in both reports. A mean entanglement rate of 56 per 1000 km of netting set was observed for this species by Coffey and Grace in the Tasman Sea. On the assumption that 20 vessels were operating in the Tasman Sea in the 1989-90 season, each setting 40 km of netting per day for 3 months, Coffey and Grace estimated a total catch of some 4600 dolphins, 6300 billfish, 3500 sharks, 2700 sunfish and 13800 ray's breams, as well as 900,000 tunas.

Sharples et al made observations on board Japanese cruises in 1989 and 1990 in the Tasman Sea and to the east of New Zealand. A total of 37 sets were observed, involving a total of more than 1100 km of netting. Observed marine mammal entanglements included 53 common dolphins, 10 striped dolphins, 1 Risso's dolphin, 1 pilot whale and 1 southern bottlenose whale. The entanglement rate for common dolphins amounted to around 48 animals per 1000 km overall, and that of striped dolphins 9 per 1000 km. Entanglement rates were much higher in the Tasman Sea than in the STCZ. Other non-target species caught included 3 leatherback turtles. The remaining species are listed in Table 24. Skipjack and pomfrets were both taken more frequently than albacore in the Tasman Sea.

It should be noted that the observations made in the Tasman Sea cannot be assumed to be representative of those in the STCZ. Nevertheless, Coffey and Grace's estimates of total captures were based upon the assumption of only 20 vessels operating in the Tasman Sea, whereas during the 1988/89 season, as many as 195 vessels may have operated in the South Pacific as a whole, suggesting far higher total catches in that season than those estimated by Coffey and Grace.

2.2.2 Other driftnet fisheries of the South and Central Pacific

Other driftnet fisheries in this region are poorly documented. A trial driftnet fishery was operated in Yap waters in 1989 in a joint Yap-Japanese venture (Goldblat 1989). 7 or 8 nets were set, each measuring approximately 2 miles in length (i.e. 26-30 km overall) and 7.6m deep. A three inch (76mm) mesh was used. A total of 24 sets were made. The total netting set was therefore 624-720 km.

The main species caught was skipjack tuna (909), but sharks, mahi-mahi and a variety of other tunas were also taken. In addition a total of 97 dolphins, 10 turtles, 21 manta rays and 11 whales were also caught, 3 of which were filleted and frozen.

The species of whales and dolphins caught were not recorded but catch rates were evidently very high, around 126 to 146 dolphins per 1000 km and 15 to 18 whales per 1000 km. 20 turtles caught included 2 adult leatherbacks (released alive but injured) and possibly one ridley, while the remainder were green and hawksbills. This fishery has not been developed.

Driftnet fisheries are widespread all along the south and central American coasts, and are pursued by a wide range of vessels for numerous species. Statistics on levels of effort or catch are not widely available.

In Chile a swordfish driftnet fishery operates from May to August or September along the entire coastline from Peru to Valdivia (39° 50S), from 15 miles to 150 miles from the coast, and occasionally as far as 200 miles from shore. Vessels are generally around 15 or 16 meters in length, 16GT. About 500 of these vessels operate from San Antonio, Quintero, Valparaiso and Constitucion. The nets employed are nylon multifilament with a 200 to 220mm mesh loosely hung as tangle nets (E₁ = 0.3 to 0.4), and are generally 1.5 km in length, very rarely up to 3 km. Nets are set 10 m below the surface. Landings of swordfish are shown in Figure 16. There has clearly been an increase in effort over the past decade. (J.L. Brito, pers. comm).

Non-target species caught include porbeagles and blue sharks, as well as Mola spp. Leatherback turtles are also taken relatively frequently. Brito collected 15 leatherbacks opportunistically from fisherman in 1989, and 10 in 1988 (Frazier and Brito 1990). Other species reportedly entangled include sperm whales, balaenopterids, killer whales, southern right whale dolphins and Burmeister's porpoises.

Coastal driftnet fisheries in Peru are also known to take a wide variety of small cetaceans in a fishery which is at least in part directed at small cetaceans (Read et al 1988, van Waerebeek and Reyes 1990). Catches include dusky and common dolphins, Burmeister's porpoise and bottlenose dolphins. Other target species include a variety of sharks, dorado, and eagle rays. Over 2500 vessels are involved in set and drift gill net fisheries which operate as far as 100 nautical miles from shore, using nets up to 8 km in length (Reyes and Oporto 1990). Van Waerebeek and Reyes recorded 1100 small cetaceans taken at one port, Pucusana in 1987. Total catches of small cetaceans have been estimated at around 10000 animals per year, mainly dusky dolphins (Read et al 1988). Duffy et al (1984) also note the capture of unknown numbers of Humboldt penguins (Spheniscus humboldti) and red-legged cormorants (Phalacrocorax gaimardi) in “drifting gill nets set for turtles, porpoises and large fish” in Peru.

Artisanal fisheries in Ecuador are thought to catch small cetaceans in nets including driftnets (M. Prieto pers. comm.), but no quantitative data on driftnet fisheries are available. Vidal et al (1990) also report driftnet fisheries in Costa Rica (operating up to 54 km from shore) for a variety of pelagic species, and a shark driftnet fishery on the Pacific coast of Panama. Data on non-target catches in these fisheries are unavailable.

Documentation of further driftnet fishing in the Pacific is lacking. However, it is worth noting that Japanese research cruises in the South Pacific have operated in the Southeastern Pacific as well as in the Southwest around New Zealand. It is reported, for example, that JAMARC found albacore concentrations east of New Zealand, in the central South Pacific and off Chile (SPC 1989). It is not known whether Japanese or Taiwanese vessels have taken advantage of this latter resource yet.

Fig.16. Swordfish landings in the Chilean swordfish driftnet fishery

2.3 Indo-Pacific and Indian Ocean

Driftnets are widely used throughout the Indo-Pacific and in the northern Indian Ocean by coastal fishermen, targeting tuna species and other scombrids. The main driftnet fishery in International waters is the Taiwanese fishery which operates in at least two areas in the Indian Ocean, fishing for tunas.

2.3.1 Large-mesh driftnet fishery in the Indian Ocean

Little information is available on the Taiwanese high seas gill net fishery in the Indian Ocean. According to published statistics (IPTP 1989), substantial tuna catches by Taiwanese vessels using driftnets in the Indian Ocean first started in 1985, when 721 tonnes of albacore were caught from FAO statistical areas 51 and 57 (Indian Ocean). Catches of albacore increased quickly over the next three years, and other species also became increasingly evident in the statistics (see Table 25).

Taiwanese driftnetting for tunas in the Indian Ocean was initiated in 1983 by a single vessel, and has expanded rapidly subsequently. Landings are made primarily at Koahsiung, but also at Bangkok, and, to a lesser extent, Port Louis, Penang and Capetown (Hsu and Liu 1990). Table 26 reproduces information on the numbers of Taiwanese vessels fishing in the Indian Ocean with driftnets from 1983 to 1989 (data from Hsu and Liu 1990).

Figures in parenthesis are provisional. Landings includes some catches from other areas.

The fishery is reported to operate primarily in two areas. These areas are shown in Figure 17 for 1987-8 and include a large part of the Arabian Sea, as well as a large area in the southern Indian Ocean between 25°S and 45°S, and from 35° to 115°E. These fishing areas, however, are not consistent. Data for the 1986-7 season are also shown, in terms of albacore CPUE, in Figure 18. (Data from Hsu and Liu 1990: CPUE units not stated.)

It seems likely, therefore, that fishing areas may shift from year to year, and this may be reflected in the catches, as proportions of different species have changed markedly over the few years for which there are data.

Statistics on effort, other than those presented in Figures 17 and 18 below, are not available. There does not appear to be any regulation of the fishery, although log books are completed by skippers and sent via their fishing companies to the Tuna Resources Research Centre of the Institute of Oceanography of the National Taiwan University. The TRRC compiles statistics for submission to IPTP in Colombo.

The type of gear employed by Taiwanese vessels in the Indian Ocean is similar to that used in the South Pacific. Mesh sizes of 200-220mm are used, but nets are reported to be 45-47 m in depth. About 700-900 nets are set at a time with a total length of 37-47 km, suggesting tans of just over 50m.

Information on non-target species is not available. Hsu and Liu (1990) indicated that sharks were a major component of the catch in the 1986-87 season, 24% of the catch (by number) was sharks, but in 1987-88 only 0.5% was reported as sharks. Presumably this may be due to differences in what was discarded between the two seasons. No data on marine mammal, bird or turtle catches are available.

Fig.17. Taiwanese driftnet fishing effort in the Indian Ocean - albacore CPUE 1986/87
(Maps from Hsu and Liu - units not given)

Fig.18. Taiwanese driftnet fishing effort in the Indian Ocean - albacore CPUE 1987/88.
(Maps from Hsu and Liu - units not given)

2.3.2 Other driftnet fisheries of the Indo-Pacific

An important Taiwanese driftnet fishery was operating in the Arafura Sea, north of Australia between 1974 and 1986, and was studied by Harwood et al (1984), Harwood and Hembree (1987) and Hembree and Harwood (1987). In 1976, 67 Taiwanese vessels were operating in the Australia FCZ, but this was reduced to 30 in 1979, increased to 38 in 1984, but was fixed at 18 after August 1985, prior to the termination of the fishery in 1986. The fishery was mainly targeting sharks, mackerel and tuna. Vessels involved were from 160 to 380 tonnes, using 145–190mm (average 170mm) multifilament mesh net with a depth of 15m. The average length of nets deployed increased from 8.2 km in 1979 to 16.0 km in 1985.

Over a five year period (1981-85), 17,467 sets were made, and 407 of these were observed. A total of 319 cetaceans was recorded caught in these observed sets. Using the figures given by Harwood and Hembree (1987) for the average lengths of nets deployed by season, and the numbers of observed sets and observed cetacean captures, approximate entanglement rates can be found. There are, for June to October 1981: 833 cetaceans per 1000 km (25 per 30 km observed); October 1981 to October 1986: 55 cetaceans per 1000 km; 1982/83:33 cetaceans per 1000 km; 1983/84: 50 cetaceans per 1000 km, and 1984/85: 88 cetaceans per 1000 km. Harwood and Hembree warn against the assessment of catch trends in this fishery due to substantial changes in fishing method and areas of operation during the period of observation. Nevertheless, the entanglement rates, described above (overall rate = 63.32), although high, are comparable with some others examined previously. The species involved in these captures were bottlenose dolphins (60%), spinner dolphins (35%), and spotted dolphins (4%) with a single false killer whale and a single Indo-Pacific humpbacked dolphin also observed. Total catches of all species over a four and half year period were estimated at 14000 cetaceans (3100 per year), and these catches were an important factor in the closure of the fishery.

After prohibition from fishing in Australian waters the fleet appears to have moved to Indonesian waters. In 1987 a total of 17427 fishing days were reported fished by Taiwanese driftnet vessels fishing in and around the Arafura Sea (Anon 1988c). The number of vessels involved was not stated, but must have exceeded 48 even if year round fishing was possible. The total catch amounted to 20 000 tonnes in 1987, and consisted of small and large sharks (25% and 15% by weight respectively), Spanish mackerel (20%), longtail tuna (19%), and other fish. The distribution of effort is shown in Figure 19. No information on non-target catches was recorded, but oceanographic conditions are similar to those in northern Australian waters, so that similar cetacean catch rates might be expected.

Driftnets are widely used by most of the countries in the Indo-Pacific region, and Southern Asia, and these fisheries are reviewed briefly below.

Catch statistics from the Southeast Asian Fisheries Development Centre (SEAFDEC) indicate that 18% of the Philippines catch of fish in 1986 (243000 MT) was taken using gill nets, not all of which would be in driftnets. In 1987 19706 tonnes of tunas were taken by gill net in the Philippines, or about 7% of the total Philippine tuna landings for that year (IPTP 1989), and these may be assumed to be caught in driftnets. The main areas for driftnetting appear to be in Archipelagic waters and to a lesser extent on the Pacific coast, but driftnets do not appear to be a particularly important gear in the Philippines (IPTP 1988). Effort statistics are not available for Philippine fisheries.

Fig.19. Distribution of estimated total effort (in 1000 pcs.hr) operated by Taiwanese gillnetters (100–500 tons) in the Arafura Sea (From Anon. 1988c).

Dolar (1990) reported on a clupeid and needlefish driftnet fishery at Pamilacan, Philippines, involving 30 vessels using about 1 km of netting each; she gave an estimate of around 20 dolphins entangled per year in this one fishery. Species were thought to include Fraser's, spinner and spotted dolphins.

Indonesian driftnet fisheries are not well documented, except for one or two specific fisheries. Driftnets are widely used on the east coast of Sumatra, where over 12000 small fishing vessels use a variety of gears, but mainly driftnets to catch tuna (IPTP 1988). Over 22000 small vessels employ tuna gill nets, among other gears, in coastal waters, in northern Java, and over 1200 in coastal waters of south Java. In Kalimantan over 4800 vessels use gill nets among other gears, in inshore waters for tuna. In Sulawesi tuna are caught mainly by trolling and gill nets do not feature in IPTP statistics. Similarly, Irian Jaya and Maluku have few gill nets recorded. However, the Islands of Nusu Tengarra from Bali to Timor use gill nets and trolls, and over 7700 vessels are recorded as using gill nets. Overall then, nearly 48000 vessels are reported to use tuna gill nets in Indonesia. SEAFDEC statistics indicate that over 148000 vessels use gill nets in Indonesia; presumably some of these may also be driftnets for non tuna species. The size distribution of vessels is not available. There is very little information on the specifications of the gear used.

Stequert and Marsac (1989) describe the drift gill net fishery at Pelabuhan Ratu in south Java, where 169 vessels were operating in 1983, mainly in the 10–12m class (2.4–4 GT). Driftnets are generally 60–65m in length, and 18–20m deep with an 80–110 mm mesh. Up to 20 nets may be set together, equivalent to a total deployment of up to 1.3 km. Boats usually go out daily, but may sometimes stay out for 2 or 3 days at a time, and nets are set at night. According to Gafa (1987) the number of operational days averaged 5491 annually between 1981 and 1985. Another important driftnet fishery operates out of Prigi, where about 50 vessels operated in the early 1980's (Gafa 1987); boats and gear are similar to those at Pelabuhan Ratu, but Gafa states that meshes of 127 to 165mm are used at Pelabuhan Ratu, whereas, meshes of 89mm to 114mm are used at Prigi. It seems likely that most of the remaining tuna gill net vessels in Indonesia are similar to or smaller than these vessels. There is no information on non-target catches.

Coastal driftnet fisheries are widespread in Malaysia. A total of 11419 vessels are reported to use gill/driftnets in Peninsular Malaysia, and a further 6818 in East Malaysia. Most of those in East Malaysia are fishing for tuna (6053 tuna vessels use gill nets in Sabah and Sarawak-IPTP 1988), but only around 2900 vessels in Peninsular Malaysia target tuna with driftnets; most of these are around 10 tonnes GT, and target small tunas. Additional target species include scads, sardines and a variety of other species including swimming crabs and prawns.

Numbers of fishing vessels using driftnets for a variety of species in Peninsular Malaysia are listed in Table 27 for 1970–1987. Fishing is generally confined to within 12 miles of the shore, and mostly within 5 miles. Nets are constructed in 30 fathom sections (=55m). Small vessels may use only 4–5 sections, while some of the larger ones may set up to 60 sections (<3 km). Mesh sizes range from about 1“to 3” (25mm to 76mm), and sometimes, until recently, up to 250mm (P. Chee pers. comm.). Large meshed nets have recently been banned due to their potential impact on turtles (R. Noordin pers. comm.).

In Brunei some 1182 vessels are reported to use gill nets, but no further information on their use could be found. In Hong Kong, likewise, there are reported to be 716 gill netting vessels in 1988 which land around 20000 tonnes of fish per annum (Anon 1988b), but no further details were found. Information of Vietnamese and Cambodian fishing vessels was not found. Information on Chinese driftnet fisheries was reviewed above in Section 2.1. Singapore has 74 vessels registered as gill netters.

The number of Taiwanese vessels gill netting in Southeast Asian waters is not known, but catches by species and by gear type are given in SEAFDEC statistics. Taiwanese gill net vessels landed 87244 tonnes of fish in 1987, 19636 tonnes of which were sharks, and 17000 tonnes of which were tunas, billfish and king mackerels. A further 9129 tonnes were sardines. Presumably much of the catches of these species may have come from driftnets.

Driftnet vessels in the Thai fleet fishing in the Gulf of Thailand include 327 vessels driftnetting for Spanish mackerel, 219 using encircling mackerel gill nets and 17 using pomfret gill nets. In addition 1693 use ‘other gill nets’ and 2641 shrimp gill nets. On the Indian Ocean side of Thailand the equivalent figures are Spanish mackerel driftnet: 38, mackerel encircling gill net:4, other gill nets:123 and shrimp gill net: 653 (SEAFDEC 1989). IPTP (1988) also indicate that some 256 Thai vessels may driftnet for tunas in the Gulf of Thailand. Most of these vessels are around 20 GT, and the major target species is longtailed tuna. Around 30 vessels are reported to fish for tuna on the western coast; catching anywhere between 20 and 900 tonnes of tuna per year in this area. These tuna vessels are apparently included in the Spanish mackerel driftnet fleet, as fishing for tuna with gill nets evolved from the Spanish mackerel fishery (Boonragsa 1987). The nets used are 3 to 5 km in length, and 30–50m deep, at least on the western side and presumably in the Gulf of Thailand too (Boonragsa 1987).

Table 28 is a summary of catches by gill net in the SEAFDEC area by nation, and numbers of gill net vessels where recorded (SEAFDEC 1989). Outside the SEAFDEC area (Southeast Asia), important driftnet fisheries exist in Bangladesh, Sri Lanka, India, Pakistan, Iran and, to a lesser extent, Oman and some of the other Gulf States. These are reviewed briefly below.

Bangladesh catches almost no tuna (16 tonnes in 1988), but has important driftnet fisheries for hilsa, pomfrets, and a wide variety of other species. Driftnets account for around 30% of Bangladeshi marine fish catches. There are two main driftnet fisheries, a small mesh fishery (70– 100mm) mainly targeted at hilsa, and a large mesh fishery (180–200mm) targeted at jewfish, catfish, tripletails, pomfret, snapper, spanish mackerel and sharks (Pajot 1980a). The small meshed nets are the most common and widely used in estuarine waters during the southwest monsoon (July-September) and during the northeast monsoon (October-April). Large mesh vessels operate largely during the winter (October to March) on account of calmer weather, and fish up to 50 km from shore. The number of vessels engaged in these fisheries is not known, but Bangladesh government statistics indicate that there may be in excess of 3300 mechanised vessels using gill nets and/or seine nets. Bergstrom (1982) surveyed driftnetters at the three major landing sites, Patherghata, Bridgeghata and Cox's Bazaar, and estimated that around 125 large mesh vessels operated from Patherghata, and about 340–410 small mesh vessels from all three ports.

In both fisheries, crews of 8–9 persons are employed and fishing trips of 7–12 days, with about 2 sets a day are made. Nets are mostly nylon monofilament and average lengths at the three ports ranged from 1.3 km to 2.2 km. Some vessels carried up to 3.8 km of netting. Trials of lighter twined netting have been carried out by the FAO's Bay of Bengal Programme to improve the economic return in these fisheries (Pajot and Das 1984). No mention of marine mammal or other non-target catches were made in the reports.

The drift gill net fisheries of India are described as one of the mainstays of the artisanal as well as the small mechanised sectors of the fishing industry in India. Catches include tunas, seerfishes, sharks, mackerel, sardines, pomfrets and a variety of other pelagic species. A majority of the 150,000 artisanal vessels as well as possibly 6000 to 8000 small mechanised vessels employ drift gill nets. Nets are usually 50–70m in length and seldom over 1.5 km, and are operated on both coasts generally within the 50m isobath, with a variety of mesh sizes (IPTP 1990). The majority of the 3250 vessels reported to use gill nets for tuna by the IPTP (1988) fish in the western provinces of Gujerat (1225) and Maharashtra (1139). Marine mammals are taken incidentally, but generally these catches have yet to be quantified. Marine mammal species taken in set and drift gill net fisheries were identified by Mohan (1990). These included spinner dolphins, bottlenose dolphins, common dolphins, humpback dolphins, and finless porpoises. Mohan suggests that the number of small cetaceans landed at Calicut market has declined from a peak of 53 in 1980 to less than 5 per year more recently. Largest catches were reported to be along the southwestern coast.

The Sri Lankan drift gill net fishery is perhaps the best known of the Indian Ocean artisanal and small-scale driftnet fisheries. Pajot (1980b) estimated that around 2000 mechanised boats regularly use large mesh driftnets in Sri Lanka, with a further 1500 using them on occasion. Nets of 3 to 4.5 km are commonly used and meshes of from 90 to 180mm may be employed, though most are usually 140–152mm. The fishing season in the south and west, for tuna, is primarily from June to September (IPTP 1988), but may continue at a lower level year round.

According to Leatherwood and Reeves (1989), in the south and west most gill netters use 25 to 35 pieces of net with meshes of 3.75 to 7 inches (95mm–178mm), generally setting them at dusk. Nets are reportedly rigged to fish 12–18 feet (3.7–5.5m) below the surface; around 1800 vessels may be employed on this coast. In the north and east the fishing seasons are primarily January to May, and September to December. The preferred total net length is reported to be around 750–900m, and as many as 570 3.5 tonne vessels and an unassessed number of 17.5ft vessels may be involved. Pajot (1980b) states that the 3.5 tonne vessels fish up to 40 km offshore.

There is also a small fleet of around 80 (in 1986) larger vessels (11m) fishing outside coastal waters with driftnets. These are designed to fish from 60–160 km from shore. Catches include small cetaceans which are consumed locally.

Catch rates for cetaceans (and tunas) are seasonally and locally variable. Reported estimates of cetacean catch rates range from 0.044 to 0.167 small cetaceans per boat per day (average 0.063). Depending upon which of the reported average total net lengths is used (750 to 900m or 3 to 4.5 km), the mean catch rate per boat per day may be equivalent to between 14 and 84 small cetaceans per 1000 km of netting set.

Total catches in Sri Lanka have been estimated differently, at around 25000 to 45000 cetaceans per year by Leatherwood and Reeves (1989) after considering a variety of different survey results, and at around 13000 per year (IPTP 1990), but the basis for these estimates is unclear. The species composition is shown in Table 29, from Dayaratne and de Silva (1990).

In Pakistan, it is reported (IPTP 1990) that driftnets of over 10 km in length, and vessels of around 50 GRT, are used to catch skipjack, yellowfin and billfish. Vessels operate trips for up to 2 months, to the coasts of Oman and Somalia. Presumably some fishing may occur in international waters which might classify this as a ‘large scale’ fishery, but more detailed data are lacking. Pakistan is reported to have landed 32000 tonnes of tuna in 1988, all by gill nets (IPTP 1989). Details on effort and non-target catches have not been located. Around 290 vessels are reported to fish for tuna with gill nets (IPTP 1988).

Iranian fisheries are not well documented, in English anyway, but driftnets are clearly an important fishing method. Tuna are reported to account for some 15 to 16% of the total national catch (Nikouyan, 1988). Fishing seasons are from April to September in the Sea of Oman, and from November to February in the Persian Gulf. Most tunas are caught by artisanal fishermen using driftnets. There are reported to be some 2596 multipurpose vessels engaged in the fishery, from 5 to 100 GT (Nikouyan 1990). Gillnets range from 500m to 8 km, with nets of 120 meshes deep, and a mesh size of 140 to 160 mm for tunas and 110 or 130 mm for seerfish. Iranian catches of tuna in 1988 amounted to 21087 MT, mostly longtailed tuna, kawakawa and seerfish, all of which were reported taken by gill netting (IPTP, 1989).

Throughout the Persian Gulf, driftnets are widely used in coastal tuna fisheries. IOFC (1989) reviewed the large pelagic resources of the Arabian Sea and Gulf area. Landings are given below in Table 30 by species group. Gillnets are stated to be the main gear used, but some may also be taken in handlines and troll lines. In Oman at least 200 dhows use driftnets, particularly in the region of Sur, to catch tuna. Nets consist of 12 100m sectors, and are about 6m deep with an 8 inch (200mm) mesh. The fishery is pursued for around nine months of the year and there is an overlapping kingfish (Scomberomorus) fishery too. Vessels fish out to about 10 miles from shore. (D. Evans, pers. comm.).

The fisheries of the Southwestern Indian Ocean (including Somalia) are reviewed by Saunders, Sparre and Venema (eds, 1988). Driftnet fishing becomes increasingly less common on the east coast of Africa and the islands of the Southwest Indian Ocean compared with the Southern Asian countries.

In the Comores there are about 35 “vedettes”, 9 metre vessels, and some 3000 pirogues (4 – 7 m) almost all of which use handlines; there are said to be 15 locally made gillnets (140 mm mesh, 100 m long and 1.5 m deep) in use, as well as 100 nets provided through Japanese aid (Williams James 1988). The way in which these nets are used is not stated, and no information on non-target catches is available. Driftnets appear to be very little used, if at all, in any of the other Indian Ocean island states.

Some 1234 vessels are reported to constitute the artisanal fishing sector in Somalia, and ‘mesh nets' are among the gears used. The proportion using driftnets is not known, but total catches were only 4,477 tonnes in 1985 (van Zalinge 1988).

In Kenya there are slightly less than 2,000 artisanal vessels, 400 of which are recorded as gill nets vessels. Total catches by gill nets amounted to 2,688 tonnes in 1984. Much of this was evidently demersal fish, and so, presumably, derived from set nets, but a minimum 16 % consisted of tunas and seerfish, presumably from driftnets. The industrial sector in Kenya does not include gill netters (De Sousa, 1988).

In Tanzania there were 3,690 artisanal vessels recorded for 1986 of which 277 were motorised. The most popular gear is the handline, but 8842 gill nets and 3590 shark gill nets were recorded in 1986, although the number used as driftnets is not stated. Only about 25 % of Tanzania's catches are pelagic fish, and of this (12,000 tonnes in 1986) at least 15 % (1800 tonnes) is thought to be taken by purse seines (Nhwani 1988). Driftnets evidently could only account for a small proportion of the total catch.

In Mozambique there is little collection of statistical data, so that information on vessels, gear and catches is limited. There is a driftnet fishery for Kelee shad (Hilsa kelee) in Mozambique, in Maputo Bay. Four motorised boats (6.5 – 8.5 m long) and 434 artisanal boats are involved. The motorised boats use 3 200 m driftnets, 100 meshes deep, and a mesh of 2 inches (51mm). Artisanal boats use exactly the same gear, but some fishermen use up to 5 nets instead of 3. Catches have averaged 3115 T annually from 1984 to 1986. The total amount of netting which could be deployed amounts to around 263 km. No information on non-target catches is available.

Madagascar's fisheries consist of around 60 motorised craft used for shrimp fishing, and somewhere around 7000 unmotorised canoes, which use handlines and gillnets. Gillnets, however, are reportedly used for demersal species, and pelagic resources in the area are rather poor (Ralison 1988).

Before leaving the Indian Ocean, it is worth noting that a 1982 report by the Fisheries Agency of Japan (FAJ 1982b) charts fishing areas for marlins and other taken in drift gillnets. Four areas are shown in the Indian Ocean (as well as the major areas on Pacific). These are in the Bay of Bengal, South of the Seychelles, to the east of the Chagos archipelago, and to the north of Mauritius. No other information on Japanese fishing with driftnets in these areas has been found.

2.4 Atlantic and Mediterranean

2.4.1 Large scale fisheries in the Atlantic and Mediterranean

The only important large scale pelagic driftnet fisheries in the Atlantic and Mediterranean which have been reported are the Italian swordfish driftnet fishery (recently terminated) and the French driftnet fishery for albacore. There are, however, persistent indications that Taiwanese vessels may be employing driftnets in some areas.

In the South Atlantic driftnetting is illegal in South African waters, but in 1990, 166 vessels from Taiwan (Province of China), Korea and Japan applied for, and were given, permits to enter South African ports with gill nets on board. A further seven vessels are known to have entered South African waters illegally with gill nets on board. One of these vessels ran aground and is reported to have had penguins and seals in its freezer, and over 100 km of driftnets stowed in board (R. Krohn, pers. comm.). Another was found to be carrying driftnets (and penguins) upon docking in Cape Town for repairs. A third (Korean) vessel was arrested for using driftnets inside South African territorial waters. These incidents were all reported in the local press at the time.

The presence of large numbers of (mainly) Taiwanese vessels in South African waters with driftnets on board, together with two which had penguins in their freezers, does not necessarily imply any driftnet fishing in the southeast Atlantic. Papers presented to the IPTP expert consultation on tuna in July 1990 indicated that Cape Town was among the ports used by vessels fishing in the Indian Ocean to land tuna. Driftnetting is known to occur in waters of the southern Indian Ocean to the east of South Africa, from where Cape Town is the nearest suitable port.

Several of the vessels involved, however, also took on board South African deckhands, and indicated that they were going to jig for squid in the Falklands (only 4 of the 166 vessels had licences to jig for squid in the Falklands (J. Barton pers. comm.), but there is also a squid fishery outside the Falkland Island FCZ). A South African deckhand reported that in fact, after leaving Cape Town they sailed southwest for a week towards the Falklands to jig for squid, but on the way stopped for a month to driftnet for tuna near a mountainous island which could be seen from the deck. Tuna transhipment occurred at sea, and around 100 other vessels were fishing there. A South African research vessel also came upon 5 vessels fishing around Tristan da Cunha, and caught its propeller in a driftnet. Penguins were observed dead in the water close to one of the fishing vessels. The deckhand indicated that apart from tuna, the driftnets also caught “many” penguins, 15–20 dolphins, 3–4 small whales, sunfish, and “lots” of sharks and other fish in the month of fishing. The limited information on this fishery, which occurs in part at least within the Tristan da Cunha 200 mile EEZ, is summarised by Ryan and Cooper (in press).

A second controversy arose in August 1990, when two Taiwanese vessels, the Her Sheng No 1 and the Her Hsing No 1 were photographed in Port of Spain harbour, Trinidad, with driftnets stacked on their rear decks. The photographs were published in local and US newspapers. Some 15 such vessels were reported to be present. The Taiwanese fishing company involved denied that driftnets were being used to catch tuna in the area, but claimed the nets were for catching squid (The Trinidad Guardian 17.08.90). There are several species of flying squid, including Ommastrephes bartrami, present in the Atlantic, but so far there have been no reports of squid driftnetting.

Spanish fishermen have reported encountering Taiwanese driftnet vessels in the Atlantic, and have cut their nets. Three Taiwanese driftnet vessels were observed transhipping in the Azores in August 1990. They had originally been rigged as squid jiggers (at least one was known to have been jigging for squid in the South Atlantic). They were currently equipped with driftnets, targeting tuna. The crew of one vessel indicated that there were a total of 21 such vessels driftnetting in the North Atlantic, that they were harassed by Spanish fishermen, and that they frequently entangled cetaceans and turtles (J. Gordon pers. comm.)

The French driftnet fishery for albacore was started experimentally in 1986 to augment catches of the existing French troll fishery for albacore; trollers operate driftnets by night and troll by day. The fishery operates from June to the end of September, mainly more than 200 miles from shore in the Biscay area. The fishery starts in June near to the Azores and follows the albacore schools as they move north and east. The number of trollers deploying driftnets rose from 16 in 1988 to 37 in 1989. Catches by gillnets amounted to some 750 MT in 1988, which compares with a total catch by all gears for the North Atlantic stock of albacore of 60300 T in 1988. Overall albacore catches have declined in recent years, due to a decrease in effort, while CPUE indices have shown an increasing trend in the baitboat fleet, and no trend in the trolling fleet (ICCAT 1990).

A variety of mesh sizes from 80–120mm have been used in this gill net fishery, with meshes of around 90mm producing the best results. Net panels about 50m in length and 20 to 36m in depth are used, rigged with a hanging ratio (E₁) of 0.6. Nets vary in total length from between 2500 to 6000m, although the Greenpeace vessel Sirius reported observing nets of 20 km in this fishery.

French observers of driftnetting operations in 1989 reported that 90.5% of the catch was albacore, 3.5% bluefin tuna, and the remaining 6% was made up of swordfish and sharks. Eight out of about 130–150 trips were observed, and two unidentified small cetaceans were taken during this period. Greenpeace observations included 84 albacore, 10 blue sharks, 2 Ray's breams and one common dolphin in the sections of net examined by the vessel Sirius. Observations in 1990 have included both striped and common dolphins, and total catches of more than 400 dolphins at current levels of fishing effort have been suggested (Antoine 1990).

The Italian swordfish driftnet fishery was banned in 1990, and restarted at least temporarily during 1991, after much litigation, but with restrictions on the amounts of netting allowed per boat. Until the early summer of 1990 this was one of the largest driftnet fisheries in the world.

Driftnets for tuna and swordfish have been used in the Mediterranean since classical times at least (Di Natale 1990a), but use of such nets expanded rapidly in the 1980's as a result of government encouragement of Italian fishermen to use driftnets, which were considered to have more desirable conservation properties than some other gear types. The expansion of the fleet, to more than 700 vessels by 1989, was accompanied by an expansion of the range of the fleet, away from Sicilian and Calabrian waters in the south, to the Ligurian Sea in the north. About 90% of the fleet was using nets of 12–13 km in length, and around 28–32 m in depth. Meshes of 180 to 400mm were used, for the capture of albacore and swordfish. Some of the fleet were using 6 km or less of netting, while a few vessels set 20 km or more.

Catches of swordfish in gill nets by Italian vessels increased from 1300 tonnes in 1984 to 1650 MT in 1988. The impact of the fishery on the stock, however, is not clear due to an inadequate time series of CPUE and a lack of reliable age estimates. Nevertheless, the mean length of swordfish landed at southern Italian ports is reported to have declined from 136cm in 1985 to 124cm in 1989 (Di Natale 1990a). The modal size class is 115cm, (all gear types), which compares with modal sizes of 170cm in longline fisheries in other parts of the world (Palko et al 1981).

A great number of non-target species has been recorded in Mediterranean driftnet fisheries. Table 31 lists the species of fish which have so far been identified (Di Natale pers. comm.) At least 44 non-target fish species have been recorded, but catch rates are unknown. In addition to these, both turtles and marine mammals are also caught.

Captures of marine mammals in this fishery have been recorded over a fairly long period of time. Harmer (1917) noted the capture of a sperm whale in ‘tunny nets’ at Lipari in 1902, while further records have been supplied by Duguy et al (1983) and Di Natale and Mangano (1983) (reviewed in Northridge 1984).

In the past few years increasing numbers of cetaceans have been reported entangled. Firstly, it was noted that increasing numbers of strandings of dead cetaceans were being recorded, especially in the Ligurian Sea. It was also noted that an increasing proportion of stranded cetaceans bore evidence of entanglement in netting (net fragments or scars, or tails missing). The trends are reviewed by Notarbartolo di Sciara (1990). The species involved were sperm whales, (23 out of 131, or 17.5%, of entangled animals), Cuvier's beaked whales (1.5%), pilot whales (6%), Risso's dolphins (3%), bottlenose dolphins (8.4%), and striped dolphins (44%). These animals clearly represent only a small fraction of the total actually entangled. From personal interviews with fishermen involved in the fishery, Di Natale (1989) postulated that somewhere around 3000–5000 cetaceans might be caught in all.

The relatively high proportion of sperm whales among these recorded entanglements should be noted. Sperm whales, perhaps due to their size or feeding habits (they are deep divers) are rarely reported in driftnets, so that the Mediterranean fishery appears to be unusual in this respect.

Catches of both leatherback and loggerhead turtles have also been recorded in this fishery, but catch rates for turtles, as for cetaceans, are not available. The Italian fishery was discontinued in July 1990 after it was prohibited by the Italian Government.

2.4.2 Other driftnet fisheries in the Atlantic and Mediterranean

There are, in addition to the Italian driftnet fishery, driftnet fisheries on a much smaller scale in several other Mediterranean countries. These were summarised at a recent GFCM/ICCAT expert consultation on the evaluation of stocks of large pelagic fisheries in the Mediterranean area, in Bari, June 1990.

In Algeria a 3 km long driftnet is now being fished experimentally off the western coasts of the country, and so far no marine mammal entanglement has been recorded. In Morocco a fleet of 30–40 artisanal vessels is reported to be fishing for small tunas with driftnets of 3 to 4 km in length. No information on entanglements of non-target species is available (GFCM/ICCAT 1990).

A few (less than 10) artisanal driftnetters are reported to fish from Corsica, and in Greek waters there is reported to be one vessel which recently started using a 10 km driftnet in the Aegean, and there are also 13 smaller vessels using 3 km nets in the same area. In a review of data on fishing vessels and gear used in the Mediterranean, Dremiere and Nédélec (1977) describe a Romanian drifting trammel net, but no information on its use was provided. Similarly a Turkish sardine driftnet was described, as being 365m in length 33–36.50m deep and with a 32mm mesh. The number of vessels employing this gear is not known. Smaller numbers of artisanal vessels employ similar gears in Italian waters (Di Natale 1990b), and doubtless elsewhere in the Mediterranean.

Two French vessels are reported to be using driftnets of about 3 km in length for tuna in the Gulf of Lions, in the south of France. Up to 70 Spanish boats may also use driftnets to catch tuna in the south of Spain, around the straits of Gibraltar and notably in Moroccan Atlantic waters (Serna and Alot 1990). Nets of from 1.5 to 3.5 km are reportedly used, and no marine mammal entanglements have been recorded (to June 1990), although Spanish observers have recorded and released three turtles.

Driftnet fisheries are not well documented on most of the West African coast. Driftnets are not, apparently, used in Mauritania (Josse 1989). In several of the other West African countries (Sierra Leone, Cote d'Ivoire, Ghana) encircling gill nets are widely used (R. Basimi pers. comm., COPACE 1989). Drift gill nets are important in Ghana and Nigeria at least. Doyi (1984) describes the types of driftnets used in this region. Nets are from 180m up to 650m in length, and from 2.5 to 50m in depth. Mesh sizes may vary from 45mm up to 200mm. The numbers of driftnets deployed in the West African region is unknown, but CECAF estimates suggest around 40000 artisanal fishing vessels operate in the coastal waters of West Africa from Mauritania to the Congo (Lawson and Robinson 1983).

Maigret (1990) reported the introduction of a new driftnet fishery for tunas, swordfish and sharks in the Cote d'Ivoire and Ghana since 1983. There are about 30 pirogues making about 1500 trips per year in all from two ports. Dolphins are apparently caught ‘regularly’. Such catches are illegal and are not officially reported. Statistics on the Ghanaian driftnet fisheries are limited. ICCAT statistics indicate that Ghana reported catches of some 1193 tonnes of bigeye tuna, billfish and swordfish by gill net operations in 1988. Maigret (1990) reports a well developed artisanal pelagic fishery of some 8000 pirogues, but the proportion using driftnets has not been reported. Maigret also reports the use of driftnets to target tunas by an unspecified proportion of 1500 artisanal vessels operating from Sao Tome and Principe.

European Atlantic coastal driftnet fisheries are not documented in much more detail. In southern Europe there is some coastal driftnetting for pilchards and sardines, and in northern Europe driftnets are used for salmon. In Portugal over 11000 vessels are licensed to use gill nets, but these are most usually set gill nets; driftnets are used to catch both small and large pelagic fish (Sequeira and Ferreira 1990). Ireland supports one of the largest driftnet fisheries in northern or western Europe. This is a salmon fishery and much of it is illegal. Around 900 licences were issued in 1988 for salmon fishing with driftnets in Irish tidal waters (Moriarty pers. comm.). It is widely known in Ireland that driftnetting for salmon, especially at night, is one of the major occupations in certain districts, and it is clear that the actual number of vessels using driftnets is higher than the 900 licences might suggest. The total amount of netting deployed is unknown. Several species of small cetaceans, including common and striped dolphins, and harbour porpoises are known to be entangled in these nets.

Driftnetting for salmon also occurs under licence in England and Wales but not Scotland. The most important area is the northeast of England where somewhat over 100 licences are granted annually for driftnetting for salmon between April and September. Harbour porpoises are regularly caught in this fishery, which operates out to about 4 miles from shore using 400 or 600m nets with a mesh size of around 120mm. Entanglement rates are only crudely known and appear to average about 6 porpoises per boat per year; most of these are released alive as the nets have to be tended continuously by law and fishing is only allowed by day (Northridge 1988).

A driftnet fishery for salmon was recently closed in Norway after it was revealed that catches of harbour porpoises were high. In two months of 1988, 580 licensed fishermen caught and reported 96 harbour porpoises. Although the salmon driftnet fishery has now been banned, there is still reported to be a small mackerel driftnet fishery operating in the northern North Sea or Skagerrak (A. Bjorge pers. comm.), but enquiries on this fishery have proved fruitless.

Driftnet fisheries, mainly for salmon, but also for herring and mackerel, operate in the Baltic and Kattegat. Danish, Swedish, German, Polish, Finnish and Soviet vessels all participate. The Baltic salmon fisheries are regulated by the International Baltic Sea Fishery Commission, who have set catch quotas and an annual closed season from 15th June to 15th September in the driftnet fishery. Nets are reported to range up to 21km in length, with a mesh size of 160mm. Annual catches of salmon have averaged 3673 tonnes in the last five years, and a TAC of 3750 was set in 1991. About 75% of the total catch is taken in offshore fisheries, and about 80% of this offshore total is taken in driftnets. An unknown proportion of the inshore catch is also taken in driftnets. Present catch levels are thought to be high enough to risk the future of wild stocks of salmon (ICES 1990). A variety of fish, including cod (Gadus morhua), herring (Clupea harengus), plaice (Pleuronectes platessa) flounder (Platichthys flesus), turbot (Psetta maxima), lumpsucker (Cyclopterus lumpus), garpike (Belone belone) and mackerel (Scomber scombrus) have been reported taken in this driftnet fishery. Birds, including guillemots (Uria aalge), razorbills (Alca torda), eiderduck (Somateria mollissima) and loons (Gavia sp.), and porpoises (Phocoena phocoena) have also been reported (Christensen and Larsson 1979). Estimates of the catch rates of these species are lacking.

A US fishery has recently developed in the Northwest Atlantic for swordfish using gill nets. Some 15 vessels were reported to have taken part in 1989 and 30 in 1990. The fishery operates along the edge of the North American continental shelf, mainly off the New England coast. Approximately 1.5 miles (c. 2.5 km) of netting are deployed by each vessel, and a mesh size typically of around 20 inch (500mm) is used (T. Smith, pers. comm.) The nets are set very loosely to tangle fish as well as gill them. A brief inspection of 5.6 km of netting by the MV Greenpeace in 1989 revealed 26 swordfish, 2 bigeye tuna, and 4 dolphins. Read (1990) presented a table of cetacean catches by this fishery, indicating that common dolphins are the most frequently taken species (39%), followed by bottlenose and Risso's dolphins (15% and 14%). Beaked whales and long-finned pilot whales were also taken (12% and 20%), as well as some spotted and striped dolphins.

A gill net fishery for salmon has existed at west Greenland since 1959, but it was not until 1965 that firstly Faroese and Norwegian fishermen, and later fishermen from Denmark and Greenland, began driftnetting offshore, up to 40 miles (73 km) from the coast. Nets are generally monofilament nylon, through multifilament is also used, have a 140mm mesh size and are used in fleets of up to 100, each net being around 25–30m long, and 5m deep. The fishery is controlled by a TAC and by control of the start and end of the season (Lear and Christensen 1975, ICES 1988).

In 1987 the fishery was opened on 25 August and ended on October 7_th. The TAC was 935 MT, and a nominal catch of 966 MT was recorded, slightly higher than the previous year, and up considerably from a low of 297 tonnes in 1984. A peak catch of 2689 tonnes was recorded in 1971, but from 1972 to 1975, non-Greenlandic vessels were phased out of the fishery and catches and the TAC both declined until the mid 1980's. The TAC is divided into 2 components, a free quota for all fishermen with a licence, and a small boat quota only available to boats smaller than 30 feet (9 metres) (ICES 1988).

The fishery is reported (in 1987) to be dominated by small boats, and to be concentrated now in inshore waters. The small boats used on average only 40 nets, whereas bigger boats used on average 99 nets. The nets are patrolled while they are in the water, and in most cases are cleared prior to hauling. Up to 350 boats may participate in the fishery (ICES 1988).

Non-target catches include small cetaceans and seabirds. Lear and Christensen (1975) examined catches of harbour porpoises in non-Greenlandic vessels during 1972. They estimated that around 1500 porpoises were caught by this part of the fishery, and suggested that catch rates were probably higher in the non-Greenlandic sector of this fishery, which fished further offshore, than in the Greenlandic sector. Catch rates increased in terms of animals caught per 100 mile-hours of netting ranged from 0 up to 21, with considerable variability between areas and weeks of the fishery. Catches per 100 nets set averaged 0.29 overall, which, if a 33m net length is assumed (“by far the majority were about 33m”), amounts to a mean catch rate of 88 per 1000 km of netting set.

The foreign fishery is now closed, but catches of porpoises continue in the Greenlandic fishery. Estimating catch rates is rather difficult in this fishery, as porpoises are to some extent targeted as a food source, and may also be shot (C. Kinze pers. comm.)

Lear and Christensen (1975) also mention two whales, one a pilot whale, which were taken in their sampling period; doubtless other cetaceans continue to be taken, but presumably at a low rate. Birds are also taken, notably Brunnich's guillemot (Uria lomvia) which was at one time considered threatened by this fishery (Tull et al 1972). Other species affected are Razorbills (Alca torda), little auks (Alle alle), black guillemots (Cepphus grylle) and Atlantic puffins (Fratercula arctica). Total catches during 1976–80 were estimated at around 50000–100000, considerably less than in previous years after the foreign (offshore) fishery was closed (Evans 1984).

Salmon gill netting in Canada is now restricted to the use of fixed gill nets (Read 1990). Through the remainder of the East Coast of North America there are a variety of small scale driftnet fisheries targeting species such as shads, some of which are known to entangle small numbers of marine mammals as well as turtles and birds (O'Hara et al 1986).

Drift gill net fisheries in Delaware target weakfish, croaker and shad. This is an onshore or estuarine fishery, in which the length of a net is limited to 3000 feet (914m) from May to September (but not at other times), and in which fishermen must tend the net continuously. Only around 30 nets were being used in 1977 (O'Hara et al 1986). Loggerhead turtles have been recorded caught in small numbers. In and around Chesapeake Bay over 6000 fishermen use inshore gill nets, including driftnets. Unknown numbers of bottlenose dolphins, as well as an occasional whale, and numerous water fowl, are taken by these coastal fisheries (O'Hara et al 1986)

A large onshore fishery, mainly for shads, in which over 4500 vessels are involved (Douglas 1989) employs small drift and set nets in North and South Carolina and Georgia. Regulations controlling mesh sizes and lengths vary between regions. About 15 bottlenose dolphins per year may become entangled, as well as an unknown number of turtles and loons and other birds. Harbour porpoises and humpback whales have also been reported in this fishery infrequently (Read 1990).

In recent years two new driftnet fisheries have developed off Florida, for sharks and for Spanish mackerel. The shark fishery is reported to involve 24 vessels (Douglas 1989) but no other details have been located. The king mackerel fishery started in the early 1980's but only involved a few boats until 1986. In 1986 the number of vessels increased to 7 then 13 in 1987. Nets average 3000 yards (2742 metres) or more, and one vessel was deploying 7000 yards (6398 metres) in 1987. The boats work outside 3 miles from shore off the Florida coast from April to September, using a 5 inch (127mm) mesh and a No. 9 nylon twine, nets being about 150 meshes (15.24m) deep. Driftnetting is done at night, and soak times of up to 12 hours are reported. In 1987 National Marine Fishery Service observers observed 38 operations, 5.2% of the total recorded effort (73) trips and recorded catches, landing and discards. Their results (from Schaefer et al 1989) are reproduced below in Table 32. No marine mammals or turtles were caught, but one leatherback turtle became entangled but escaped as the net was being hauled.

Driftnetting in the Caribbean and South America is not well documented at present. Driftnets are widely used in parts of the Mexican east coast, for large pelagic species of fish (J. Gonzalez-Cano pers. comm.), but no quantitative data are available on effort or catches. In much of the insular Caribbean fish traps and various lines are the most commonly used gear. In some islands, however, gill nets are used. It is reported that in Jamaica, for example, there is “a continuing slight shift to gill netting” away from traps (WCAFC 1986). In Barbados flying fish are reportedly caught in driftnets with a 41mm mesh size (Storey 1984, cited in Mahon et al 1986). Nylon gill nets are reportedly used in the French Antilles for catching jacks, balahoo, flying fish and mullets, and the Barbados driftnet fishery has also been introduced to Dominica and Grenada (Hess 1961 cited in Mahon et al 1986). Vidal et al (1990) report some gill netting, including driftnetting, for sharks, spanish mackerels, jacks and dorado in Costa Rica, for grey mullets in the Dominican Republic, where these authors report on the capture of a young humpback whale, and for unspecified fish in Puerto Rico.

Van Waerebeek (1990) reported on dolphin catches in a gill net fishery for pelagic fish (including carangids and sharks) in French Guiana. Around 20–25 Brazilian vessels operate off French Guiana and one estimate was that 4–5 dolphins of two species were taken per trip. Each trip lasts up to a week. Van Waerebeek speculated that the species were most likely bottlenose dolphins and tucuxi (Sotalia fluviatilis). The method of use of the gill nets, however, was not stated.

In Trinidad there is a small scale fishery for king mackerel involving over 100 vessels of around 10m (crew of 2 to 3). The gear used is 4 to 41/2 inch (102–114mm) mesh driftnets, constructed mainly with multifilament (nos. 12–15) twine. Panels are 100 – 150m in length and about 10m in depth. Up to 5 panels of netting are set at a time (500 – 750m). Fishing is from dusk for around 6 hours, nets being hauled around midnight. In 1989 5325 boat trips were made. The main target species are Scomberomorus brasiliensis and S. cavalla (Sturm pers. comm.) The only detailed record of any marine mammal entanglement in this fishery involved an 8 to 10 year old female killer whale (Orcinus orca), part of a pod of about 15, which was entangled in a driftnet which was being hauled in, on June 10th 1987 in the Gulf of Paria. Other members of the pod attempted to release the whale by bumping the boat and pulling the net. Dolphins (of unknown species) are also reportedly sometimes taken in nets (T. Ottley, pers. comm.)

Driftnet fisheries in Brazil are extensive along much of the coast, but few statistical data are available. In the State of Maranhao, for example, an estimated 5000 artisanal boats may use large mesh driftnets, often only a hundred metres, or a few hundred metres, in length (V. Batista, pers. comm.). Driftnets for catfish are also used in the Amazon and are reported to catch tucuxi.

Along the coasts of Uruguay and Argentina there are extensive gill net fisheries, including a large mesh fishery for sharks in Uruguay, which is estimated to have killed over 3000 franciscana dolphins (Pontoporia blainvillei) between 1974 and 1989 (Praderi 1990). Similar shark fisheries exist in southern Brazil and in Argentina, and dusky dolphins (Lagenorhynchus obscurus) and Burmeister's porpoises (Phocoena spinipinnis) are known to be taken too. Most or all of these gill net fisheries, however, appear to be set net fisheries rather than driftnet fisheries (Crespo and Corcuera 1990).