3. DISCUSSION

3.1 The current status of driftnet fisheries.

It is clear from the preceding section that driftnet fisheries are both diverse and very widespread. With the exception of a few island areas, such as the southwest Indian ocean, parts of the Caribbean and central Pacific, where driftnets are not much used, driftnet fisheries have extended to, or remained in, most areas of the world's seas, and make a significant contribution to fishery yields.

In Table 33 the driftnet fisheries discussed above are listed by geographical area. This is not intended to be an inventory of all the world's driftnet fisheries, as many of these remain little documented, but it serves instead to make several points about the nature of present day driftnet fisheries.

Firstly, the number of boats involved is enormous, but most of these are small coastal craft deploying perhaps a few hundred metres of netting each. In order to obtain some idea of the scale of the various fisheries, a crude estimation has been made, where possible and based on the information reviewed above, of the amounts of netting which might be deployed during a fishing operation by a typical vessel in each of these fisheries. This figure, multiplied by the number of vessels involved, gives a very crude impression of the amount of netting (in thousands of kilometres) available for deployment in each of these fisheries.

Clearly this is not an ideal way to compare such fisheries, especially as the number of operations per unit time is not included, nor is there any way of judging the areal extent of most of these fisheries, but there is a paucity of relevant data with which to make a more detailed comparison.

Table 33 therefore, also demonstrates the approximate scale of the various fisheries described above. While it is clear that a large proportion of the total available fishing effort is deployed by industrial driftnetting vessels, notably in high seas fisheries for squid and tunas, there are areas of the world where comparable amounts of netting may be deployed by artisanal fishermen, for example in Sri Lanka and neighbouring countries. The deployment of these large amounts of driftnetting by industrial fleets in areas of open ocean, however, is largely a new development since the 1970's, and this represents a major change in the way that driftnets are being used.

In the days when driftnets were extensively used in industrial fisheries for the capture of schooling pelagic species such as herring, nets were generally set in areas of high fish density, where schools of the target species were known to be concentrated. Today such areas are generally the domain of other gear types, while driftnets may be used in areas where the target species occurs at relatively lower densities.

Only selected non-target species, known to occur relatively frequently are listed, as a guide to the species involved overall.

As an example, during the early years of the 20th century, British driftnet vessels might have expected to catch around 1000 herring per haul in a 32m net (11m deep, 600,000 meshes), and as many as 2000–3000 in a good haul (Butcher 1979). Inspection of recent catch records from Japanese salmon research cruises (FAJ 1989b) indicates that a catch of more than a thousand fish in 2.5 km of netting is unusual, and the mean catch during 1988 research cruises was only 5 fish per net (45m in length). Similarly, in the North Pacific squid driftnet fishery, catches amount to only 45–50 squid per km of netting in the commercial fleet. Even allowing for differences in the sizes of the mesh and the target species involved, it is clear that a much more dispersed resource is being targeted in these latter fisheries. In such situations, catch rates necessary for making a profit are only reached by using a commensurately larger amount of netting. The ready availability of cheap plastic twines since the 1950's has made this possible.

This change in the way in which driftnets are now being used appears to be the main reason for the controversy which surrounds them, at least with regard to their perceived competition with other fisheries. The criticisms of driftnet fisheries summarised in the Introduction are considered gain below, in relation to possible management measures.

3.2 The criticisms of driftnet fisheries and possible management procedures

3.2.1 Competition with other fisheries.

3.2.1.1 Increased efficiency

Existing fisheries targeting relatively dispersed stocks have found that driftnets may be able to catch more fish at a lower cost, or that driftnet vessels may be able to target the fish in areas where they are otherwise uneconomic. In this way, driftnet fisheries may out-compete other fisheries, just as in the earlier decade of this century trawlers out-competed driftnet vessels for more densely schooling fish.

In fact, there is very little information on the relative economic performances of driftnetting vessels compared with other vessels targeting the same stock, although the assertion that the former are at an advantage is fairly frequently made. Clearly the relative economic advantages will depend upon numerous factors such as the price of fuel, and not least the density of the fish resource. The more fish aggregate, the more economically they may be caught by purse seine or trawl. The depth zone at which the fish swim may also influence the relative economic efficiencies of the competing gears too.

Competition between vessels using different gears is a common feature of many fisheries around the world, and has usually been resolved by legislation to ensure separation of the competing gears (as with Japanese driftnet and longline vessels) or by banning one or other of the gears. In such situations it is not necessarily driftnets which are banned. In English coastal waters, trawlers are prohibited from targeting herring which spawn in parts of the Thames estuary, in order to allow a traditional small scale driftnet fishery for herring to continue. Where such measures are not taken, the less economic fishery will ultimately disappear.

Within waters of national jurisdiction, such decisions are relatively easily made. The situation becomes more problematic where fleets of different nations are fishing for the same resource with different gears in international waters. In this case there are only two possible solutions to the problem. It may be resolved if there is a competent regional fishery authority which is able to regulate fishing practices, or, failing that, by bilateral agreements between the nations involved. The regulation of high seas driftnetting for salmon in the North Pacific provides an example.

3.2.1.2 Resource depletion

The relative economic efficiency of driftnets when used to fish dispersed resources has also been taken to suggest that resource over-exploitation may occur more easily. Thus it may be claimed that as a fish stock is depleted, if it becomes more dispersed, while catch rates of other gears fall off rapidly, driftnet catch rates may continue at a level which is still economical, thereby enabling driftnet fisheries to deplete a resource more quickly.

In fact, such an argument can be used exactly in reverse for gears such as purse seines, which rely on fish aggregating. Certain fish such as pilchards and tunas, continue to aggregate even while their overall numbers are depleted, keeping local densities high even though the overall biomass might be dangerously low. Purse seiners could therefore maintain good catch rates, even at very low stock levels, provided fish can be located economically.

On the other hand, because the cost of deployment of drift nets is relatively low, economical catch rates may be possible at dangerously high levels of fishing effort. This may contrast with gears which are more expensive to deploy, and which may become uneconomical beyond relatively low levels of fishing effort, thereby enabling some degree of ‘self-regulation’. Such considerations, however, are only relevant where there is an absence of any effective fishery management, where fisheries are left to self-regulation.

Where effective management does occur, there is no great difference between driftnets and other gears in their ability to deplete fish stocks. The level of depletion is determined by the level of fishing effort, and it is the job of fisheries management bodies to control the level of fishing effort. Once again, this is theoretically straightforward in national waters, but is not always so easy on the high seas, where international fishery regulatory bodies are required and enforcement is usually difficult. Few such bodies are available at present, although under the provisions of the United Nations Convention on the Law of the Sea (UNCLOS), nations fishing on the high seas have a responsibility to co-operate in ensuring appropriate conservation measures are taken, and to establish regional fisheries organisations to this end.

3.2.1.3 Resource wastage

Although there are clear responsibilities stated in the UNCLOS for the sustainable management of high seas resources, there is no provision for the rational use of those resources. A frequent criticism of driftnetting is that the practice is responsible for great wastage of fishery resources. The primary cause for concern is due to wastage during the fishing operation.

Wastage comes at four levels. Firstly, it is argued, that as fish which are caught in driftnets are often left in the water for several hours prior to retrieval on deck, a relatively high proportion of the catch is spoiled and has to be discarded. This loss might be expected to be most important in warmer waters.

The proportion of fish which is discarded due to spoilage in driftnet fisheries has not often been reported on, but observations on Japanese vessels in the South Pacific led to estimates of a high-grading loss of approximately 2% of the catch. The experimental fishery in Yap waters was reported to have had a high proportion of spoiled fish, but that proportion was not estimated.

Secondly, fish species which might otherwise be considered useful, and which are included in the catch, may be discarded. Driftnets are in this way held to be unselective. The low species selectively of this method of fishing can be observed by referring to Tables 8, 16, 24, 31, and 32, where large numbers of species are observed in the catch. The results of Japanese research cruises reinforce this point. Murata et al (1989) reported on a series of four research cruises aimed at flying squid in the north Pacific during 1988. Three of these cruises used driftnets and one employed squid jigs. More than half the total squid were taken in the jigging cruise, but whereas during the jigging cruise only four species were caught, all of which were squids, and 97% of which by number were the target species, in the driftnet cruises more than 36 species of fish, squid, bird and mammal were taken and only 35% by number were the target species, although it should be noted that a range of several mesh sizes was used in the driftnet cruises.

High levels of non utilised catch are not necessarily found in all driftnet fisheries. French observers in the albacore fishery operated in the north Atlantic, for example, have reported catches of 90.5% albacore by number, and a further 3.5% bluefin tuna. It should also be remembered that, as was noted in the introduction, some other gears such as demersal trawls may often yield very low proportions of a target species in their catch.

The capture of a wide range of species, in whatever gear, can only be categorised as wasteful where many of the species are discarded. It should be remembered that in many small scale driftnet fisheries a large proportion of species caught may be retained for sale or consumption; with the larger scale or industrial driftnet fisheries, however, this is less likely to be true. In such fisheries there is a narrower definition of what constitutes a target species, and in general only the higher value species such as squid and tunas are likely to be retained. This is therefore likely to lead to a higher discard rate in these fisheries.

The third form of wastage in driftnet fisheries comes in the dropout of fish from the net during hauling. An exact definition of dropout is hard to find, but it is generally taken to be the proportion of fish observed to fall from the net as it is being hauled. This quantity has been measured by observers working on a number of vessels. In the south Pacific, a mean rate of 8.7% (range 0–20.8%) was observed in one cruise, and 3.7% (0–8.1%) in another. In the North Pacific squid driftnet fishery, dropout rates of 3–10% have been recorded.

The dropout rate as measured from the deck is clearly a minimum estimate of total fish loss. Fish may fall from the net as it is being hauled but before leaving the water, and fish may fall from the net, or be removed from the net by sharks or marine mammals, during the period in which the net is soaking. It is not clear whether or not such losses should be included in the dropout category. The fourth category of wastage is therefore the loss due to fish which fall from the net before leaving the water.

Estimates of such losses are very hard to obtain. One experiment in England suggested that about 5% of salmon were removed from driftnets by seals before the net was hauled. Removal by sharks is also commonplace in warmer waters, and Goldblat (1989) noted the higher proportion of sharks in the nets the longer they had been in the water, suggesting that they had been attracted to the nets. Within this category, further losses might be expected from those fish which encounter the net and escape alive with lost scales, or skin or muscular damage. Some of these may then die later. Again, estimating such mortality is very difficult. In some areas, however, it is clear that a high proportion of fish encounter driftnets: in the South Pacific sub-tropical convergence zone gill net marks were observed on up to 19% of troll caught fish in 1988–89 (Hampton et al 1989).

The wastage of fishery resources has been considered a problem in several other fisheries around the world, including tropical shrimp trawl fisheries and temperate industrial fisheries for fishmeal. In some of these, management measures have been adopted to limit the waste. However, there have been very few management initiatives which address these types of wastage in driftnet fisheries. Net damage to albacore in the South Pacific has certainly been a concern of the South Pacific nations, and in this context has been partly responsible for the elimination of driftnetting in the area. Wastage of fishery resources in other driftnet fisheries has not been addressed, but this may be an issue which nations fishing in international waters have an obligation to address at present.

3.2.1.4 By-catch of valuable fishery resource

The relatively low species selectivity which often characterises driftnets has also led to problems where driftnet fisheries have a by-catch which is the main target of other fisheries. North American origin salmon have been taken by squid and other driftnet fisheries in the North Pacific, for example, while southern bluefin tuna are also taken in Taiwanese driftnets in the southern Pacific and Indian Ocean, which may thereby reduce the possible value of the bluefin which could have been caught and sold by other methods for the lucrative sashimi market. In both situations the by-catch has been seen as a useful addition to the catch. The situation is therefore different to the problem of resource wastage, but causes considerable antagonism between the fisheries involved. In the North Pacific management measures have been adopted based on the observation that the flying squid and the salmon are distributed in water masses of differing temperature. By ensuring that vessels are kept out of the salmon zone, high catches of squid can be maintained, while minimising the salmon by-catch.

Where a target and a by-catch species share the same distribution, the problem is less easy to resolve. Nevertheless, this is a common problem in fishery management, and one for which a variety of management measures are available. Areas with the most vulnerable concentrations of the by-catch species may be closed to the offending fishery, closed seasons may be implemented to ensure that spawning or feeding aggregations of the by-catch species are protected, total effort or total catch by the offending fishery may be limited to ensure a limited catch of the non-target species, or in the extreme, where a particularly vulnerable by-catch species is involved, the offending fishery may be banned entirely.

Management measures such as these are readily adopted in waters under national jurisdiction, and have also been adopted within areas controlled by regional fishery management bodies. The problem is less easy to resolve on the high seas, and in areas without effective regional management regimes.

3.2.1.5 Physical impediment

The physical impediment that driftnets may pose to other fisheries and other vessels in general, has been little documented. There are numerous anecdotes concerning driftnets ensnaring the propellers or keels of other vessels, but there does not appear to have been any detailed study of the frequency with which this occurs, nor of the economic consequences. On the other hand, because of the cost in terms of the loss of netting, driftnets would be expected to be set away from shipping lanes and away from areas where there are other major fisheries. In coastal waters there are often zones to minimize such gear conflicts. The systematic collection of information on vessel entanglement with driftnets would go some way to ascertaining the scale of this problem.

3.2.1.6 Synthesis

Most of the issues raised above involve competition between different fisheries, and most such types of competition have been addressed by fishery management bodies for a considerable length of time. Driftnet fisheries, operating in international waters and outside the control of regional fishery bodies, have simply extended the problems of competition between gears and between fisheries into new areas. Most of the problems of competition between driftnet and other fisheries operating on the high seas could be resolved by the extension of effective fishery management regimes to cover those areas, provided the appropriate databases can be developed reliably and enforcement schemes can be put in place.

3.2.2 Environmental impacts

Although many of the controversies surrounding driftnets may be familiar issues concerning competition between fisheries using different types of gear, in one major respect this is not the case. The large number of species taken in driftnets has already been mentioned in relation to the possible wastage of fishery resources, but this issue is also of concern from other perspectives. The removal of large numbers of animals which are not generally considered as fishery resources, has given rise to environmental concerns in several areas.

Although the capture of large numbers of non-target species is not a unique feature of driftnets, there is a concern that, because these nets fish near the surface, air breathing animals (mammals, birds and reptiles) feature in the non-target catches in relatively larger numbers, compared with the non-target catches in a demersal trawl, for example.

There is a further concern that high seas driftnet fisheries, operating in the relatively unproductive open ocean environment, may threaten the stability of this ecosystem, by catching unduly large proportions of rare animals with low resilience to exploitation.

Whether or not the open ocean ecosystem is any more susceptible to irreversible damage from indiscriminate fishing than coastal ecosystems is difficult to judge, and would be even more difficult to measure. Coastal or shelf water communities tend to be much more diverse (and therefore possibly more stable) than open ocean communities, but may still contain species or groups of species which are highly vulnerable to non-specific fishing methods. Such species are also often dependent upon very localised critical habitats for their feeding or reproduction and are easily threatened by the degradation of such habitats through coastal fishing or pollution. Coastal marine mammals such as the sirenians, and communities resident in seagrass beds are two such examples.

Open ocean habitats, however, are in general less productive than coastal waters, and so may be considered more likely to contain longer-lived and slower-growing species with low fecundity. Many open ocean species also appear to be quite rare(sunfish or beaked whales for example). On the other hand, open ocean species may also have a wider distribution than some of the coastal species, and so, at a species level at least, may be less vulnerable to depletion from locally high mortalities, although local populations may be as vulnerable in either location.

Whether any of the relationships or links between species or communities in the open ocean environment are more vulnerable to disruption or destruction than those in coastal environments is not clear either. At present then, it is impossible to make a general statement about the relative vulnerabilities of the two environments. Indeed assessing the overall impact of driftnet fisheries on either coastal areas or open oceans, at an ecosystem level, is not practicable at present.

The impact of driftnets on individual species might be somewhat easier to address, in theory if not in practice. Of particular concern are the air breathing animals. These are characterised by relatively low fecundity, slow population growth and low sustainable yields, especially when compared with fast growing fishery resource species like squid or tunas. They are therefore considered more vulnerable to increased mortality rates than most of the exploited species. Some of the fish species which are not generally considered as fishery resources may also be of concern.

The difficulties involved in examining the effects of driftnet fisheries on entire ecosystems therefore forces the issue to be examined through individual species. Some of those species which were found to be impacted by driftnet fisheries are reviewed below, but in most cases, it will be seen, data on catch rates or population sizes are inadequately known.

3.2.2.1 Individual species impacts

Dall's porpoise. Phocoenoides dalli

The major non-target casualty of the Japanese North Pacific salmon fisheries is the Dall's porpoise. Considerable attention has been payed to the affect of the driftnet fisheries on this species. Dall's porpoises are distributed throughout much of the temperate North Pacific, and through both the Bering and Okhotsk Seas, while their southerly limit is described by the 17° or 18° isotherm. At least two forms are recognised, the truei form, which occurs mainly along the Japanese Pacific coast, and the dalli form which is found throughout most of the range. Upto six discrete stocks of the dalli type may exist as well as separate truei stock.

Population estimates of Dall's porpoise are problematic, because of the species' habit of approaching vessels, which makes population surveys based on transects difficult to interpret. Total population estimates have ranged from 1.5 to 2.1 million animals. Reported catch rates have varied from as low as 0.79 per 1000 km of netting per set in the landbased fishery to as high as 58 per 1000 km of netting per set in the high seas fishery in the US EEZ. Total catches have been estimated by US scientists to range from 12000 (1982) to 3100 (1986) for both salmon fisheries. Reported catches are somewhat lower (Table 4). No trend in population size was observed between 1980 and 1984 (Jones et al 1986). The Bering Sea stock is estimated to be between 78% and 94% of its pre-exploitation size, whereas the stock in the Western Pacific is thought to be between 66% to 91% of its original size.

The impact of the squid driftnet fishery has yet to assessed, but if the catch rates observed in the 1989 Joint Observer programme are representative of most years, then around 6000 more Dall's porpoises might be taken in this fishery. This would yield a total catch of up to about 8400 animals based on US estimates of catches in the salmon fishery in recent years.

Previously observed catch rates in the squid fishery have been higher, however, (Table 17) so that a higher total catch may be possible. The impact of the large meshed gill net fishery is not known, but due to its generally more southerly distribution would be expected to be less than that of the squid gill net fishery.

Although the impact of the salmon gill net fishery on the Dall's porpoise has been particularly well studied over a number of years, the results of these studies do not provide a conclusive answer as to how the driftnet fisheries have affected this species.

Northern fur seal Callorhinus ursinus

Northern fur seals are distributed across the North Pacific ocean generally in water with a temperature of more than 15°. The population size underwent an approximate 50% reduction between 1972 and 1980, the causes of which are still unclear; more recently the population appears to be more or less stable, at around 1.2 million.

Northern fur seals are also impacted by the salmon driftnet fisheries, but numbers involved have been reported to be low, around 200 per year based on observations in the US EEZ (Cranmore 1988). It should be noted, however, that catch rates of northern fur seals in Japanese salmon research cruises have run at a long term average of slightly less than half those of Dall's porpoises (Table 5) so somewhat higher catches may be possible, although it should be remembered that research cruises results are not directly comparable to the commercial fishery.

Catch rates (fatalities) in the squid fishery were reported to be around 0.73 to 1.6 per thousand km of netting in 1988 and 1989 (Table 17), suggesting a possible total kill of somewhere about 2200 to 4800. Further catches may be expected in the large mesh fishery, again possibly at a lower level than the squid fishery.

Northern right whale dolphin Lissodelphis peronii

This species is confined to offshore temperate waters of the North Pacific, apparently with two centres of population density, off California and off Japan. Total population sizes are unknown, but Cranmore (1988) suggested a total population off California of around 80,000 based on estimates by Dohl et al (1980 and 1983). The population size on the western side of the Pacific and in the Central Pacific are not known.

This species is evidently considerably affected by the squid driftnet fishery. Based on catch rates observed in the 1989 joint observer programme, a total catch of slightly less than 20000 was proposed (Table 17). Observed catch rates in previous surveys have been both lower and higher. This species may also be taken in the Pacific large mesh fishery, but catch rates are unknown. The order of magnitude of the possible total catch clearly gives some cause for concern, but the overall impact cannot be assessed without further information on total population sizes and better estimates of the total catch.

Pacific white-sided dolphin Lagenorhynchus obliquidens

This species is also confined to the temperate North Pacific, from Japan and Baja California northwards. A peak population estimate off the US coast during the autumn was around 86000. Population distribution is not understood, and there appear to be no further population estimates elsewhere in the Pacific. Catch rates in the 1989 squid driftnet observer program were relatively high at 3.56 per 1000 km of net set, suggesting a total catch of perhaps slightly more than 10000. Observed catch rates in 1988 were slightly lower, (2.89) but would still yield a total catch estimate of more than 8000. Again there is clearly cause for some concern over the scale of these catches, despite inadequate population estimates.

Common dolphin Delphinus delphis

Common dolphins are a cosmopolitan species found throughout tropical and temperate waters of the world. They are also reported in a number of driftnet fishery catches, but few estimates of total catches have been made. Population sizes and stock distributions are not well known, but as its name implies it is generally a numerous species of cetacean.

Catches of common dolphins were reported in the Pacific squid driftnet fishery and an approximate catch of some hundreds was estimated (Table 17); this seems unlikely to pose a threat to any Pacific population, although the possibility of vulnerable local populations should not be overlooked. In the South Pacific, Coffey and Grace (1990) estimated a catch of some 4600 dolphins (mainly common dolphins) in the Tasman Sea in 1989/90. The impact of this on any local populations is also, of course, unknown. Common dolphins would be expected in most of the other driftnet fisheries reported in section 2. They have been reported in Irish, French, U.S. (Atlantic and Pacific), Japanese large mesh and Peruvian driftnet fisheries but estimates of catch rates are lacking.

Sperm whales Physeter macrocephalus

Sperm whales are also a cosmopolitan species, occurring from tropical to polar waters. Population sizes are poorly known. The large size of individual sperm whales might make them seem unlikely victims of driftnet fisheries. There are in general very few records of sperm whale entanglements in most of the worlds seas. Sperm whales have been recorded in the Chilean swordfish driftnet fishery, but in unknown numbers. In the Mediterranean, however, this species is one of the most frequently recorded of the species which became entangled in the swordfish fishery. At least twenty three sperm whales have been reported entangled in Italian driftnet fisheries in the past three years (Notarbartolo di Sciara 1990), and the total is presumed to be more.

Di Natale (1990b) thought that driftnets in the Mediterranean must be a significant source of mortality in that population. The population size or relationship to other populations is unknown.

Striped dolphin Stenella coeruleoalba

Distributed in warm temperate and tropical waters worldwide, striped dolphins were taken in the Italian driftnet fishery more frequently than any other species (44% of known cetacean entanglements). Based on the information reviewed above, catches seem likely to have run into the thousands per year, but the effect on the Mediterranean population is unknown.

In the Sri Lankan fishery about 6–11% of 13 000 to 40 000 cetaceans entangled a year may be striped dolphins (i.e. 780 to 4400 in total: Table 29). Again the effect of this on any local population cannot be guessed.

Other marine mammals

In general, for remaining marine mammal species, there is too little information available to give much idea of the impact of any specific fishery. In the Sri Lankan fishery, spinner dolphins are most frequently taken (33–47%) which implies a catch of possibly as many as 18,800. Although this figure seems large, there is no information on abundance with which to compare it. Other potentially large catches are evident for pygmy and dwarf sperm whales, which are generally regarded as rare species but which occur as 2–6% of the Sri Lankan catches.

Mention should also be made of the beaked whales. Beaked whales are in general not commonly seen, deepwater mammals, and some species are known from only a very few stranded animals. Unidentified beaked whales have been recorded in Sri Lankan driftnets, and two or three southern bottlenose whales have been reported in South Pacific albacore driftnets. Beaked whales are one of the more commonly found cetaceans in the Atlantic US swordfish driftnet fishery too. Although the numbers of animals reported entangled is low, population sizes may also be low, so that any mortalities should be a cause for concern.

A striking feature of Table 33 is the lack of information on non-target catches in so many of the coastal driftnet fisheries. This may be considered a worrying situation as there are a number of small cetaceans species with restricted distributions, confined to coastal waters about which there is very little information. Where such species overlap in their distribution with intense driftnet fisheries, there may well be a cause for concern. Examples of such a situation include the finless porpoise (Neophocaena phocaenoides) which is confined to coastal waters of the Indo-Pacific region, which in turn is an area of intense driftnetting activity. Clearly it would useful to have a clearer picture of the effects of coastal driftnet fisheries too.

Birds

Captures of birds in driftnet fisheries are not well recorded, except in the North Pacific salmon and squid fisheries. There the major species involved are the dark (sooty and short-tailed) shearwaters, (Puffinus griseus, P. tenuirostris) tufted and horned puffins (Lunda cirrhata, Fratercula corniculata), Laysan albatross, (Diomedea immutabilis) and the black-footed albatross (Diomedea nigripes).

Using the crudely estimated total catches for the squid fishery in Table 18, and estimates from Jones and DeGange (1988) (Tables 6 and 7) estimates of catches of dark shearwaters exceed 40,000 in the two fisheries, but the total world populations of these two species are about 50 million birds (sooty shearwaters) and 40 million birds (short-tailed shearwaters) (Anon 1988a). The tufted and horned puffin populations are also numbered in the millions (6 and 2 million respectively, Anon 1988a) so that again, catches of 38 000 and 9 000 respectively seem to be unlikely to affect total population size. Catches of black-footed albatrosses may amount to perhaps 5000 or so, which compares with a population estimate of some 198,000 birds. The Laysan albatross is thought to number around 2.5 million birds in total (Fefer et al 1984), whereas estimates of catches extrapolated from observed catch rates may approach 14000, or 1.8% of the population. Population structure and dynamics are too poorly known for the effects of these catches to be assessed.

Rockhopper penguins (Eudyptes moseleyi) are reported to have been taken in apparently large numbers in the recently developed fishery around Tristan da Cunha. Enquiries to the Island revealed that at least one penguin colony has declined in recent years, but it is thought that this may be in part at least due to illegal taking as bait for lobster pots (B. Fines pers. comm.). The impact of Peruvian driftnet fisheries on Humboldt penguins (Spheniscus humboldti) is also unknown.

Turtles

Turtles do not appear in large numbers in many of the driftnet fishery observations, and of those which are, many are not identified to species level. Catch rates observed in the Yap trial fishery were certainly very high. Systematic information on other tropical driftnet fisheries is lacking, and catch rates have only been recorded in the more temperate waters of the North and South Pacific. The leatherback turtle (Dermochelys coriacea) is the species most often identified, although a green turtle (Chelonia mydas), an olive ridley (Lepidochelys olivacea) and loggerheads (Caretta caretta) have also bee recorded (Gjernes et al 1990).

Leatherback turtles are global in their distribution and have documented range extending from 47°S to 71°N (Pritchard and Trebbau 1984). Major rookeries in the Pacific are confined to Mexico where Pritchard (1982) estimates that some 75 000 females (more than half the known world population) nest on the beaches from Michoacan to Oaxaca; another 12000 may nest elsewhere in the eastern Pacific. A further 13000 leatherback nests have been reported on a 17.8 km beach in Irian Jaya (Bhaskar 1984). The numbers of females nesting at Terangganu in Malaysia has fallen from around 1800 females nesting per year in the 1950's to fewer than 100 in 1988. Females typically average 4–5 nests per year and return to the same nesting grounds on 2–3 year cycles (Marquez 1990). Important nesting and feeding grounds also exist in the eastern and western Pacific for other sea turtle species, but none are predictably found north of 40°N.'

The observed catch (kill) rates of leatherback turtles in the squid driftnet fishery might indicate at least 250 animals per year are taken (not all turtles were identified to species level, so this is a conservative interpretation). Balazs (1982) reported 5 dead leatherbacks wrapped up in sections of squid driftnet, floating at the surface, apparently between 35°-45°N in the Pacific. Additional mortalities in the Chilean driftnet fishery may amount to several hundred per year, and in Malaysia, Chan et al (1988) estimated catches in gillnets of 77 and 33 leatherbacks in 1984 and 1985 respectively. Large mesh gillnets have subsequently been banned in Malaysia. Preliminary observer data from the gillnet fishery off the coast of California show one leatherback killed during four months in 1990 (Scott Eckert, US NMFS pers. comm.). Clearly, when additional mortalities in the Pacific large mesh driftnet fishery and the potential impact of other coastal driftnet fisheries are considered, significant mortality due to driftnet fisheries cannot be discounted.’

Fishes

Among the fishes, blue sharks (Prionace glauca) are apparently the most widely caught species, but despite the very large numbers taken there is simply not enough information on the life-history parameters to assess the likely scale of the impact of such mortalities. Clearly they could be a cause for some concern. Other species, such as the sunfishes (Molidae) are also too poorly known to assess the affect of accidental captures in gill nets, but impacts on rarer fish have so far received very little attention.

Conclusion

In conclusion it may be said that for most of the gill net fisheries of the world, information on catch rates is too poor to make any reasonable estimate of total catches of non-target species. Where information does exist, however, as in the North Pacific salmon and squid fisheries, and in Sri Lanka for cetaceans, the estimation of total catches is hampered by very variable catch rates for most non-target species between seasons and from area to area. It is clear, however, that the crude indications of possible total catches of some species do not encourage any idea that impacts may be minor.

3.2.3 Environmental management considerations

The management of fisheries to minimise environmental impacts has been widely addressed in many areas. Management measures adopted have included banning certain gears from specified areas to preserve species or habitat types, restricting the extent to which gear types may be used, which has included the setting of quotas for non-target species, and the development of gear modifications to minimise the impact on non-target species.

Management measures, however, need to be framed within management objectives. In the case of the environmental impacts of driftnets, several possible environmental objectives may be relevant. For example, the maintenance of stability in the open ocean ecosystem might be considered an appropriate environmental objective, or the elimination of catches of a particular turtle species might be considered.

Although impact on the open ocean ecosystem was listed above as a possible criticism of high seas driftnet fisheries, it was pointed out that forecasting the long-term impact of a fishery on such a system would be extremely difficult. The functional relationships between the various components of the open ocean ecosystem are still so poorly understood, that attempting to model or manage the effects of any open ocean fishery would be impracticable. A coastal fishery would be even more problematic, as coastal areas are in general so heavily fished by numerous different fishing gears, as well as being influenced by other anthropogenic factors, that the specific effects of a driftnet fishery on the coastal ecosystem would be impossible to forecast.

An alternative objective might therefore be to manage a driftnet fishery in such a way as to control the impact on individual species or populations. This was the approach taken in the Mediterranean where the Italian large mesh driftnet fishery was banned due to its non-selective impact on cetacean populations. Even this approach, however, is problematic. There are at least three possible approaches to the problem. These are (1) suppressing the impact, (2) limiting the impact to acceptable levels, and (3) disregarding the impact.

The first of these three strategies, in the extreme, could be taken to suggest management measures prohibiting any capture of the species involved. Where a marine mammal species which is distributed in the same time and area as the fishery is concerned, a complete prohibition of the fishery would almost certainly be required. Such an approach would be very difficult to justify, not least because, to be consistent, almost all other fisheries would have to be prohibited. Marine mammals are occasionally killed in almost all fisheries, and the impact on the food and income of coastal human populations would be unacceptable to most governments.

A more appropriate interpretation of the “no-.impact” strategy might be to adopt management measures which would reduce the probability of any mortalities in a given time period to a specified low level, with the aim of maintaining the highest possible levels of abundance for the species concerned. This is the ultimate objective of the US Marine Mammal Protection Act, where captures of, or injuries to, marine mammals of any species must be reduced to “insignificant levels approaching a zero mortality and serious injury rate”. Management measures might then include time area closures, or regulations on net length and number of vessels. Again, however, it should be pointed out that where the species of concern is a marine mammal, then the management measures are likely to include very severe restrictions, if not total prohibition of driftnet fishing.

The second approach is to manage the fishery in such a way as to maintain indefinitely a certain population size of the non-target species. This requires additional information before management measures can be devised. It is necessary first to institute an independent research programme to investigate the population numbers and dynamics of the species concerned. The effects of increased mortality rates need to be ascertained, the ‘original’ population size needs to be estimated, and the population structure (in terms of stock distribution and genetic inter-relationships) must be well understood.

There are of course considerable difficulties in performing these tasks. Estimating catch rates, for example, can be extremely difficult as such estimates are often widely different from one survey to another. As the non-target species is not being actively sought by the fishery, and where there is little correlation between the distribution of the target species and that of the non-target species, there is clearly likely to be much variation between years due to changing distribution patterns of the fishery and the non-target species. Estimates of northern fur seal catch rates in the North Pacific squid driftnet fisheries provide a good example of this (see Table 17). Neither can it be assumed necessarily that there is any real equilibrium population size to be used as a yardstick with which to compare current population sizes. It will also be necessary to decide what level of population depletion is ‘acceptable’, to determine criteria for establishing ‘acceptable’ levels of depletion, and to implement monitoring programmes to monitor both the fishery and the non-target species. Management measures might include time/area restrictions, gear modifications, effort limitations or even closure of the fishery if no other effective means of achieving the objective is found.

It is possible to attempt to manage a fishery in this way, as has been done for the US yellowfin tuna purse seine fishery, for example, but it should be remembered that the cost of such an exercise, if fully implemented, may eventually outweigh the value of the fishery, and if the exercise is to be financed by the fishery, then it could in some instances make that fishery uneconomic.

With either of the above two strategies, it is also necessary to determine which of the species or populations being impacted by the fishery are to be studied. If the most vulnerable populations are to be used as indicators of environmental damage, then it is also necessary to have a certain amount of background information before the choice of species or populations is made. This is an important preliminary process because it might be easy to overlook the depletion of a little known or rare species, or to overlook the existence of discrete populations. When dealing with existing fisheries it is also necessary to bear in mind the possible cumulative impact which the fishery may have already had on non-target populations, and indeed the possibility that existing fisheries may have altered the ecological equilibrium of a system by the removal of predatory fish, and thus allowed marine mammal numbers to increase beyond their ‘pristine’ state.

The third ‘laissez faire’ approach has simple management consequences. Driftnet fisheries are allowed to continue without consideration of their impact on non-target species. Clearly this is what has happened until recently in most fisheries. The present concerns over driftnet fisheries, including the United Nations resolution (44/225) on large-scale pelagic driftnet fisheries, would suggest that, on the high seas at least, such an approach is no longer acceptable.

A further concern regarding driftnet fisheries is that lost or discarded driftnets may continue to ‘ghost fish’ after their loss, continuing to ensnare animals for some time. Where large enough amounts of netting are being deployed, even a relatively small discard or loss rate may result in large amounts of lost netting, and concern has been expressed that such netting may continue to ensnare animals for some time, as the nylon is resistant to decay. Eisenbud (1985) estimated that more than 17 km of netting might be lost every night in the North Pacific, amounting to several thousand kilometres of netting per year.

Most derelict fishing gear recovered on beaches in the Gulf of Alaska consisted of fragments of trawl webbing, while gill net fragments formed only 1% of entanglement debris in 1989 (Johnson 1990). The relatively low proportion of drift netting may simply be due to the relative proximity of the two types of fishery. A Canadian volunteer driftnet survey programme has monitored derelict driftnets along the Canadian west coast since 1986. There were 25 reports in 1987, 6 in 1988 and 89 in 1989; the increase in 1989 is thought to have been due to increased observer effort. No entangled animals were found in 1989, while one dead bird was found in 1988. These lost driftnets are thought to have originated in the squid driftnet fishery. (Hargreaves and Carter 1988, 1989). Japanese experiments on free floating driftnets have indicated that discarded netting is wound up by the action of water and wind, and that 2 km net segments are wound into a complete ball after about 20 days, making continued fishing impossible (Mio et al 1988).

Although there is as yet little direct evidence that ghost netting is a major problem, the pollution of the marine environment with net fragments is clearly something to be avoided. Under Annex V of the International Convention for the Prevention of Pollution from Ships (MARPOL), which came into effect on December 31st 1988, disposal of netting at sea is illegal, and the 41 contracting governments are required to provide reception facilities in ports for the disposal of such garbage. Accidental loss rather than dumping of nets at sea remains a more difficult problem to resolve.

3.3 Prospects for driftnet fisheries and their management

The United Nations resolution (44/225) on large scale pelagic driftnets has highlighted some important problems concerning these fisheries. The resolution states that large scale pelagic driftnet fisheries (i.e. those operating on the high seas) should be terminated by 1992 unless “… effective conservation and management measures be taken based upon statistically sound analysis to be jointly made by concerned parties of the international community with an interest in the fishery resources of the region…”. Although this phrase is no doubt open to numerous interpretations, it is clear that such “analysis” will need to be carried out within the framework of existing or new fisheries management bodies and international conventions. This is also stipulated in Articles 118 and 119 of UNCLOS.

The UN resolution, however, also encouraged coastal countries to take appropriate measures and to co-operate in collection and submission of scientific information on driftnet fishing in their own exclusive economic zones. The environmental problems posed by high seas driftnet fisheries are essentially no different from those posted by coastal driftnet fisheries, so that a joint approach to the problem would seem to be appropriate if coastal nations are to fulfil their duties and responsibilities for management of the resources allocated to them under UNCLOS.

Fisheries management regimes for the high seas areas are needed, however, for more reasons than those covered by the UN resolution. The expansion of driftnet fisheries onto the high seas has largely been driven by the exclusion of distant water fleets from coastal EEZ's, and the increasing world demand for fish products. It therefore seems inevitable that high seas resources will become increasingly under pressure, and not just from driftnet fisheries. The rational and sustained use of the resources of these areas with a minimum of conflict between competing fisheries, can only be achieved by the implementation of management regimes with adequate scientific and enforcement capabilities.