Bottom trawling fleets predominate in many Mediterranean fisheries, being responsible for a high share of total catches and, in many cases, yielding the highest earnings among all the fishing sub-sectors. The high profitability of this fishing practice is largely due to its low selectivity with respect to sizes and species caught, and to the high harvests generated. Trawlers have dramatic effects on the ecosystem including physical damage to the seabed and the degradation of associated communities, the overfishing of demersal resources, and the changes in the structure and functioning of marine ecosystems derived from the depletion of populations and the huge amount of bycatches and associated discards.
Whilst the problems related to the impact of bottom trawling on Posidonia beds and soft bottoms have been dealt with elsewhere in this report, the present section focuses on the ecosystem effects of trawling derived from its low selectivity and the issues relating to the capture of undersized individuals and discarding. Bycatches (and subsequent discards) of particularly vulnerable species or groups are covered in other parts of this report, as are the effects of trawl discards on marine seabird populations. The ecosystem effects of discards reported below refer to demersal communities.
Bottom trawling fisheries in the Mediterranean are essentially multispecies. Monospecific fisheries are very rare and are largely limited to deep shrimp fisheries on muddy slope bottoms. The high marketability of small fish in many countries encourages the targeting of the juvenile fraction of some species, often in violation of laws regarding minimum sizes. Demersal populations are consequently overfished, shallow areas (within the three-mile coastal limit or on bottoms less than 50 m deep, depending on the country) are illegally trawled and small, illegal mesh sizes are used. Examples are widespread throughout the Mediterranean and are not detailed here since they mostly concern recurring issues related to classical fisheries management. The well-known massive seasonal harvest of undersized red mullet, which are caught on shallow grounds when they settle, is though worth mentioning. The paradigmatic case of the hake fishery using bottom longlines and otter trawling gears in the Gulf of Lions also deserves highlighting. Data from the late 1980s clearly showed that the trawling fishery exploited the juvenile fraction of the population since the mean size of catches was only 17.9 cm, which strongly contrasted with the 48.2 cm corresponding to longline catches (Lleonart, 1990).
Information on discards in Mediterranean trawl fisheries confirm the magnitude of the problem, though they vary considerably in amount and composition depending on region, boat size, season, bottom type and depth of the exploited ground. The first regional study addressing the magnitude of discards in the western Mediterranean involved the monitoring of fishing fleets in seven ports (six Spanish and one Italian). Combined data gave discard estimations ranging from 23-67% of total catch in bottoms less than 150 m deep, 13-62% in bottoms 150 to 350 m deep and 14-43% in slope bottoms deeper than 350 m (Carbonell, 1997; Carbonell, Martin and de Ranieri, 1998). Data from a single locality, the Catalan port of Vilanova i la Geltru (north-west Mediterranean), illustrate this high quantitative variability. Monitoring of the fleet there revealed that the annual average of discards ranged between 13% and 39% of the total catch for small boats (< 150 hp) and between 17% and 48% for larger boats (> 150 hp), depending on the depths exploited. The amount discarded, however, peaked at 75.4% and 66.6%, respectively, in the case of larger boats operating in spring and smaller ones operating in the summer on shelf bottoms (< 150-m depth).
Similar high discard levels have been reported for other Mediterranean trawl fisheries. Total annual discards in Sicily during the 1980s were estimated at around 70 000 t, accounting for an average of 44-72% of catches (Charbonnier, 1990). The monitoring of fleets operating in three major Greek fishing grounds (Ionian Sea, Cyclades Islands and Thracian Sea) during 1988-1997 yielded discard estimations of 40%, 55% and 25% of the total catch of fish, crustaceans and cephalopods, respectively (Machias et al., 1999). Field studies carried out in 1995 showed that the fraction discarded by the trawl fleet operating in the Cyclades area, in the Aegean Sea, amounted to 59% of the total catch in bottoms less than 150 m deep, 63% in bottoms 150-200 m deep, and 37% in grounds deeper than 300 m (Vassilopoulou and Papaconstantinou, 1998). On the whole, discards in the Hellenic commercial trawl fishery are estimated to account for 45% of total catch (Stergiou et al., 1998). The rapido beam trawler fleet (56 units) based in Chioggia in the Adriatic Sea produces qualitatively heterogeneous discards depending on the species target. Whilst pectinid fishing involves the exploitation of sandy bottoms offshore and discards consist of echinoderms (32% in weight), crustaceans (26%), molluscs (23%) and porifers (15%), flatfish fishing is carried out on muddy coastal areas, where molluscs and crustaceans account for the bulk of discards (60% and 30%, respectively).
High discard levels are also common in the case of Mediterranean deep sea trawling fisheries. Discards by the trawling fleet operating on the upper slope (230-611 m) off Alacant (south-east Spain) have been estimated at 34.6% of the total catch (Soriano and Sánchez-Lizaso, 2000). The low selectivity of trawling is highlighted by data from this fishery showing that up to 95 species are taken; only 12 of these account for nearly 89% of the total, and 89 of them are discarded. The analysis of discards in the Norway lobster (Nephrops norvegicus) and red shimp (Aristeus antennatus) fisheries at 280-720 m in the Balearic Islands (western Mediterranean), estimated at an average of 42% of the total catch, led the authors to conclude that an important fraction of the catch of the two deep-sea decapod crustacean fisheries of the Western Mediterranean is discarded (Moranta, Massutí and Morales-Nin, 2000). Longer tows, to compensate for the reduced biomass, seem to result in lower selectivity by the mesh and higher discard rates.
Discarding can also involve important commercial species, especially smallest size classes. Discards of commercial species in Greek waters are reported to range from 0% for red mullet (Mullus surmuletus) to 10% for hake (Merluccius merluccius) and shrimp Parapenaeus longirostris (Machias et al., 1999). The bulk of discards (66%) in the Balearic deep sea crustacean fisheries at a depth of 300 m referred to above correspond to undersized marketable species. The study of hake discards (Merluccius merluccius), forkbeard (Phycis blenoides) and poor cod (Trisopterus minutus capelanus) in the trawl fishery of the northern Tyrrhenian Sea revealed that they can reach high levels, depending on the species, the season and the depth exploited (Sartor et al., 1999). Maximum estimates of discards were 34.1% of total catch (in weight) for hake, 41% for forkbeard and 39% for poor cod, whereas total annual mean discards in the traditional trawl fishery amounted to 39%, 65% and 57% respectively in numbers of individuals. All individuals under 10 cm are discarded in all three species.
Although a proportion of discards in Mediterranean trawling fisheries may survive, few helpful data on which to base quantitative estimates exist. Observations derived from experiments on aquaria carried out on board point to the low mortality of crustaceans caught as a bycatch in Catalan trawl fisheries, whereas survival rates of fish are highly heterogeneous and vary strongly according to the species (i.e. 0% for Trachurus spp. and 100% in Scyliorhinus canicula) (Sánchez, 2000). Another study of bycatch survival in the rapido fleet operating in the northern Adriatic showed low mortality in all taxa examined during the three to four hours following capture (Pranovi et al., 1999).
The impact of discards goes far beyond single-species demographic effects, since discarded biomass can alter ecosystem structure by favouring scavengers (Moranta, Massutí and Morales-Nin, 2000). The consequences of the fishing-driven increase in food supply stemming from have seldom been addressed by specific studies.
The only work dealing with this issue in the Mediterranean is based on photographic surveys carried out off the Catalan coast in the northwestern Mediterranean, and focuses on the estimation of the consumption rate of fishery discards by scavengers (Bozzano and Sardà, 2002). The study used a baited camera, which was set on the sea floor at a depth of 100 and 300 m in two areas subjected to trawling with continual discards. Eight fish and nine crustacean species were recorded feeding on the baits, and the benthic snake eel Ophichthus rufus was the main scavenger species, followed by isopods (i.e. Cyrolana borealis) and amphipods (i.e. Schopelocheirus hopei). Sporadic scavenging behaviour was even reported for common fish species such as Spicara spp. and Trachurus spp. Discarded material seems to enter demersal food webs quite quickly, as suggested by the high consumption rates recorded. In all cases baits were fully consumed within 24 hours, and consumption rates reached maximum levels in deep bottoms at night. The authors concluded that the prevalence of O. rufus indicated an environment dominated by a monospecific scavenger guild, whose competitors and predators have probably been eliminated by fishing activity. This conclusion is particularly interesting since it highlights the multiple effects of fishing on complex systems such are communities and ecosystems: fishing can favour a single species within the demersal ecosystem by both removing its competitors and independently increasing its food availability through discards.
There is compelling evidence that discards by Mediterranean unselective trawling fleets are significant. The effect on marine communities is twofold: at a single-species level, the population dynamics of a species are altered, and at the ecosystem level profound changes occur because of the disruption of food webs. Ecosystem modifications are triggered by the change in the biomass and demographic structure of the different species as well as by the increasing food supply for scavenger and opportunistic species. It is worth noting that the latter can result in the trophic connection of separate sub-systems (i.e. pelagic and benthic), making ecosystem consequences even more dramatic.
Although bottom trawling is inherently rather unselective, bycatches and discards can be minimized. Trawling can be limited and technical measures can be introduced to improve selectivity. Trawl selectivity within an area depends on many factors, ranging from the depth exploited or the kind of bottom, to the season. Most impacting scenarios could be avoided by restricting trawling both spatially and temporally. In this context, current provisions banning trawling in coastal waters less than 50 m deep or three miles offshore should be enforced effectively. Trawling gears could be made more selective by using higher mesh sizes or incorporating special excluding devices, such as those based on rigid grids. The former solution may be difficult to apply in Mediterranean waters for social and political reasons, but the development and compulsory use of excluding devices increasing selectivity (such as those in use in some North Atlantic waters) deserve attention. Alternatively, the use of a square mesh can also improve selectivity. It is convenient to mention here that shorter trawling hauls are known to reduce discard rates (Stergiou et al., 1998, Moranta, Massutí and Morales-Nin, 2000).
Partial solutions and technical improvements notwithstanding, the banning of bottom trawling in large marine protected areas throughout the Mediterranean Basin appears to be the only way of maintaining a sample set of demersal ecosystems free of the damage caused by this widespread fishing practice. These areas would moreover be very useful as a basic reference guide to healthy bottom communities in the context of a future ecosystem-based management of Mediterranean fisheries.
Pelagic longlining in Mediterranean waters inflicts considerable mortality on elasmobranchs, marine turtles and seabirds taken as bycatch or even (in the case of the former) target species. It is obvious, however, that large pelagics, the objective of this fishery, is the group most impacted by this gear. The main species targeted in the Mediterranean are swordfish (Xiphias gladius), bluefin tuna (Thunnus thynnus) and to a lesser extent, albacore (Thunnus alalunga). Bluefin tuna and swordfish are exceptional in the Mediterranean context for being the only species whose populations are subjected to an international TAC-based management regime. The overall issue of the sustainable management of their populations is beyond the scope of this report, and the discussion below focuses instead on the selectivity of surface longline fisheries operating in Mediterranean waters as it affects the immature, small-sized fraction of their dwindling populations and the degree of compliance with current international legislation.
A variety of medium-scale and industrial pelagic longlining fleets operate in Mediterranean waters, ranging from local coastal state fleets to large industrial foreign fleets, whether Japanese, flag of convenience (FoC), or even unflagged pirate fleets. FoC and pirate fleets are estimated at about 100 units (GFCM, 1997). Surface longline gears, including those used by local Mediterranean fleets, are deployed in large areas since line lengths of 50-60 km (bearing several thousand hooks) are not rare. Longline fleets in quest of their highly migratory target fish species, even local ones, are highly mobile, covering virtually the whole Mediterranean basin. A significant share of catches is taken in international waters, more than 12 miles offshore.
The Spanish longline fleet operates from the Strait of Gibraltar (5ºW) to 7ºE near Sardinia, and from 42ºN to the Algerian coast (Camiñas and De la Serna, 1995). In the early 1990s a Spanish fleet of 30 longlines operated throughout the year in the southwestern Mediterranean. In the summer months, when the swordfish fishery peaks, the number of Spanish boats rose to 60-80. This local fishing effort was complemented by about 30 Japanese and 30 FoG longliners (Aguilar, Mas and Pastor, 1992). Overall, some 145 Spanish longliners target swordfish in Mediterranean waters and a further 100 artisanal boats operate in coastal waters during the summer. Seventy percent of total yearly effort in this fishery is concentrated in the summer and autumn. Bycatch, excluding turtles, accounts for 10% of total landings in weight (Camiñas and De la Serna, 1995).
Italian longlining fleets targeting swordfish and albacore are based mostly in Sicily, Puglia, Sardinia, Campania and Liguria, and comprise more than 1 500 boats operating mainly in the Gulf of Taranto, the south Adriatic and the Aegean Sea (Camiñas and De la Serna, 1995). About 27 longline units operated in 1997 in the vicinity of the Santuario dei Cetaceii, in the Western Central Ligurian Sea, where driftnets have been banned since 1992. These fleets, however, are able to reach much more distant grounds. In 1992, the Sicilian fleet operated from Crete and Cyprus in waters close to Egypt and the rest of the north African coast (Cavallaro and Luca, 1996). Italian longline fleets are also known to reach Iberian waters during the autumn. In the central southern Tyrrhenian Sea, swordfish have historically been fished with driftnets (spadara) but an important longline fishery has recently been established at Mazzara del Vallo in the Strait of Sicily (Di Natale et al., 1996).
The Greek National Statistic Service includes longlining in the broad category of coastal fisheries and although no specific figures are available, it is estimated that the swordfish fishery accounts for more than 50% of the total professional fishing effort by Greek fleets in western Greece (Panou et al., 1999). A total of 47 longline boats were known to be based in the Ionian Islands and the Epirus coastal region alone in the mid-1980s. Camiñas and De la Serna (1995) gave a total figure of 400 boats from 70 ports being involved in the Greek swordfish fishery in 1991. The main fleets, concentrating 50% of total Greek production, are based in Kalymos (south-east Aegean) and Chania (Crete). Of the total annual catch, 70% is taken at the peak of the season, from May to September, in an area covering the Aegean Sea, the Ionian Sea and even the Levant Sea. About 180 vessels are involved in albacore fishing in the central and northern Adriatic.
Seasonal differences in the size of swordfish caught by the Spanish longline fleet operating in the Mediterranean have been reported, suggesting that different age groups are targeted in different seasons (Camiñas and De la Serna, 1995). Smaller specimens are caught during the autumn months, when fishing is carried out in more coastal areas, peninsular and insular (for the Balearic Is.). The Italian longline fleet is also known to operate near the coast in the Strait of Sicily during the autumn (Di Natale et al., 1996). In Greek waters, however, the fishing of swordfish is prohibited by law from October to January (Panou et al., 1999).
The selectivity of longline fishing in the Mediterranean with respect to ICCATs minimum legal sizes for swordfish and bluefin tuna are a matter of concern. The percentage of legally undersized swordfish with respect to current EU legislation (< 120 cm LJFL) caught by Spanish longliners in the Mediterranean was 81-83% in 1992-1994 (Anonymous, 1995, cited in Raymakers and Lynham, 1999). A recent study commissioned by TRAFFIC and WWF confirmed the previous figures, and demonstrated the Spanish longline fleets' non-compliance with its international and EU legal obligations (Raymakers and Lynham, 1999). The study, based on observers at the main Mediterranean Spanish ports from June to September 1998, showed that 86% of a sample of 2 097 swordfish landed from 171 vessels had been illegally fished (<120 cm., and probably <25 kg). This sample represented about 7.5% of the 1991-1995 annual average of swordfish caught by Spanish fleets in the Mediterranean. As for bluefin tuna, 210 out of a sample of 254 individuals (or 83% of the total) landed by ten longline vessels were below the minimum legal size of 6.4 kg. The monitoring of swordfish catches by the Italian longlining fleet operating in the central and southern Tyrrhenian Sea and the Strait of Sicily also pointed to the predominance of immature, small-sized individuals in this fishery (Di Natale et al., 1996). The mean weights of individuals caught in the southern and central Tyrrhenian Sea were 16.8 kg and 12 kg respectively, and 17.5 kg in Sicily. These values contrast sharply with the current minimum weight of 25 kg recommended by ICCAT.
Albacore longlining also has negative consequences on swordfish and bluefin tuna populations. Di Natale et al. (1996) report small-hooked surface longlines targeting albacore in western Italian waters catching very small swordfish, weighing less than 3 kg. De Metrio et al. (1997) investigated the catches of the albacore longline fleets operating in 1995 in the Gulf of Taranto (north Ionian Sea), the eastern coast of Sicily (south Ionian) and the north Sicilian coast (south Tyrrhenian), an area fished by a fleet of nearly 150 vessels. Comparisons of landings at ports and catches on board revealed that most swordfish catches were not reported at the ports. Catches of young (class 0) swordfish and bluefin tuna were estimated at 53.2% and 10.1%, respectively, of the total catch in number of individuals, which point to high absolute catches.
Apart from harming important groups taken as bycatch, pelagic longlining in Mediterranean waters is clearly unselective with respect to non-target undersized fractions of the populations that are the object of the fishery. Some data even point to immature large pelagic fish being the bulk of surface longline fisheries. This applies mainly to swordfish and, to a lesser extent, bluefin tuna. Regardless of whether small specimens are caught because of the intrinsic action of the gears or merely reflect overfishing of populations, known to be at low levels, action could be undertaken to minimize the negative impact of present longline practices. The creation of no-fishing zones in strategic areas and seasons, for example spawning and nursery grounds or coastal areas in autumn, could be considered as recommended also by the authors of the TRAFFIC-WWF study. The extension of the Spanish fisheries jurisdiction to a vast region in the western Mediterranean (Royal Decree 1315/1997) provides an opportunity to enforce EU Regulations (derived from ICCAT Recommendations) and implement other new measures in these former international waters.
This section does not set out to deal with the issue of the monospecific management of large pelagic populations, but it is clear that pelagic longlining in the Mediterranean induces high levels of mortality in several ecologically valuable and biologically vulnerable species as well as in non-target, legally protected fractions of swordfish and tuna populations, to the extent that the fishery might just as well be targeting this latter group. Large pelagic species are apex predators and key players in Mediterranean pelagic ecosystems and their conservation appears to be essential to keep ecosystems healthy. Overfishing of pelagic apex predators (bonito and mackerel) in the Black Sea may have triggered a trophic cascade effect working down to lower trophic levels, making the system less resilient to external changes (Daskalov, 1999). The well-known Mnemiopsis invasion led to the collapse of fisheries in the late 1980s. All the evidence strongly suggests that current policies should be revised in favour of an ecosystem-based management of large pelagic fisheries and the related surface longlining fishing practices.
The diversity and economic importance of artisanal gears in small-scale fisheries are essential features of Mediterranean fishing. Stergiou, Petrakis and Politou (1996) consider that small-scale fishing is socioeconomically more important than trawling and purse seining in Greece since it occupies 87.5% of all boats, 57.5% of total fishing power (in HP) and produces near half of the total wholesale value of catch. The heterogeneity of gears and target species makes it difficult to reach any general conclusions as to the impact of these small-scale practices on the ecosystem. Factors such as the season of the year, the characteristics of the area exploited (depth, type of bottom, etc.) further complicate the picture. Some trends emerge nonetheless, such as the higher selectivity of some gears and the negative effects of other artisanal practices. Ghost fishing by abandoned or discarded small-scale gears is another issue of potential importance in the Mediterranean. These points are discussed below, excluding the specific effect of some small-scale fisheries on the populations of endangered species such as monk seals or turtles addressed in previous sections.
Static nets are usually highly selective, catching larger fish than, in most cases, trained nets. Different types of nets can, in turn, also differ deeply as to intra- and interspecific selectivity. A comparative study of catches in eight types of net gear (both beach seines and static gill and trammel nets) in the Aegean Sea revealed that large meshed trammel nets yielded the biggest commercial catches as a proportion of total catches (Stergiou, Petrakis and Politou, 1996). In another study, the relative selectivity of trawlnets, bottom longlines and gillnets operating on slope bottoms (between 200-700 m) in the Southern Adriatic Sea was analysed with respect to three demersal species: blackmouth catshark (Galeus melastomus), rockfish (Helicolenus dactylopterus) and blue whiting (Micromesistius poutassou) (Ungaro et al., 1999). The results showed that gillnets (rete ad imbrocco) were always the most positive selective gear for size of individuals caught. The modal length of blackmouth catshark caught by gillnets, for instance, was 54 cm, contrasting sharply with only 16 cm reported for trawlnets. Sbrana et al. (1999) carried out a comparative study of interspecific selectivity with three kinds of static nets: monofilament gillnets, trammel nets with a monofilament inner panel and entirely multifilament trammel nets, and also tested the effect of different mesh sizes. The study concluded that whereas the number of species caught was negatively correlated with mesh size of a given gear type, inter-specific selectivity decreased from gillnets to trammel nets; the trammel nets with three multifilament nets were the least selective of all. However, the target species in the Sardinian cuttlefish (Seppia officinalis) fishery using trammel nets constituted up to 78% of the total catch weight (Cuccu et al., 1999).
Beach seines, deployed in very shallow grounds to catch small fish, are common in some Mediterranean waters and are relatively unselective. They are used in Italy, where they are known as sciabica, and other countries to catch small sardine fry (biancheto), transparent goby (Aphia minuta) (rosetto) and sand eel (Gymnammodytes cicerellus) (cicerello). The beach seine fleet in Crotone (northwestern Ionian Sea) operates on bottoms less than 9 m deep (Carbonara et al., 1999). A different kind of beach seine (sonsera) is used to catch sand eel in a limited area off the coast of northern Catalonia (northwestern Mediterranean). Fine-meshed trawling gears are employed in Adriatic waters mainly in the Gulf of Manfredonia (Casavola, De Ruggieri and Lo Caputo, 1999). The transparent goby fishery is allowed to operate there from January to March, and catches mainly Aphia minuta (53.7%) and small sardine (39.7%), together with other fish (including juvenile anchovy) and benthic invertebrates (Casavola et al., 1999). The Sardinian squid (Loligo vulgaris) fishery works with beach seines on shallow bottoms ranging from 25 m deep almost to the shore. Whereas squid catches consist of adult individuals, contrasting with the local trawlnet catches, Loligo vulgaris accounts for only 20% of the total catch in weight and salema (Sarpa salpa) makes up the bulk of the catch (72%) (Cuccu et al., 1999). A comparative study of small-scale gears used in the south Euboikos Gulf (Greece) during 1992 and 1993 revealed that beach seines were very effective in catching younger and smaller specimens, leading the authors to conclude that banning of beach seines is essential for the conservation of demersal and inshore diversity (Stergiou, Petrakis and Politou, 1996). It is worth mentioning that the minimum legal beach seine cod-end mesh size in Greek waters is only 8-mm bar length.
Game fishing is a growing leisure activity in many Mediterranean waters, and probably has a significant impact on some species, for example bluefin tuna and swordfish, whose low age classes suffer particularly. As many as 380 000 juvenile swordfish are estimated to be caught annually by non-commercial fishermen in Calabria (De Metrio et al., 1997). The impact of this activity on marine populations and ecosystems in the Mediterranean remains to be adequately addressed.
The massive use of fixed nets (and other artisanal gears such as traps) in many small-scale Mediterranean fisheries, makes ghost fishing by abandoned or discarded gears a potentially important problem in Mediterranean waters but has attracted scant attention. Erzini et al. (1997) carried out an experimental study of gillnet and trammel net ghost fishing in shallow (15-18 m) rocky bottoms in the Atlantic waters off the coast of the Algarve in southern Portugal. The results of the study indicated that abandoned gillnets yielded more catches than trammel nets as measured by the mean number of fish caught by 100 m-length pieces of nets after 120 days of deployment on the bottom (gillnets: 344 fish specimens entangled; trammel nets: 221 fishes entrapped). Whilst catches decreased gradually over time, nets continued to catch fish four months after the experiments had started. Osteichthyes were the most numerous group among the 39 species recorded, accounting for 88.8% of the total specimens in number. The other groups included molluscs, gastropods and crustaceans. Sparidae species, however, made up about 33% of total catches in numbers. There is evidence suggesting that nets lost in deep water may have an even longer effective fishing life span, running to years. This is matter of concern since some deep gillnet fisheries (such as the Italian rete ad imbrocco in the southern Adriatic) operate in Mediterranean waters.
The results of the study mentioned above also implicated ghost fishing in disturbing demersal food-webs in a similar way to that reported for trawl discards. The authors described considerable scavenging pressure on entrapped fish by octopuses, cuttlefish, conger eels, moray eels and wrasses (Coris julis), which could have led to an underestimate for the actual fishing capacity of discarded nets.
The high diversity of artisanal gears (and species targeted) and the importance of small-scale fisheries in many Mediterranean coastal waters introduce considerable additional complexity to the overall issue of the ecosystem-based management of Mediterranean fisheries. In this context, Stergiou, Petrakis and Politou (1996), referring to Greek small-scale fisheries, stated that management strategies based on single species calculations will be of limited value, and advocated the promising alternative approach of a management regime based on marine harvest refuges. This holistic approach overcomes, in part, the problems related to the variable intraspecific and interspecific selectivity of different gears and other variable factors such as bottom type, depth of setting, seasons and the phenomenon of ghost fishing.
There is enough scientific consensus to support the total banning of some artisanal gears in Mediterranean waters. Beach seines should be eradicated from EU Mediterranean waters from January 2002. All fishing with coastal seines will also be prohibited by 2001 in Turkish Aegean waters, as long demanded by many local artisanal fishermen (Anonymous, 1999b). Game fishing is a superfluous non-commercial practice and must be prevented from inflicting any additional damage on vulnerable species such as swordfish and bluefin tuna.
In general terms, and leaving managerial issues aside, many artisanal fisheries (such as static or bottom longlines) are probably more selective than trawling practices, and therefore a preferable, much less ecosystem-impacting alternative, provided that discarding gears at sea can be stopped.
The outstanding impact of bycatches by surface swordfish driftnets fleets on many vulnerable groups inhabiting Mediterranean waters, as reported in some detail in previous sections, makes a summary of the present status of these fleets desirable. Details of technical aspects of gears and specific fleets are not included here, since the controversial issue of driftnets has been extensively discussed and an extensive literature is already available (see Paul, 1994, for a global, world-wide account of this issue).
The Italian Mediterranean driftnet fleet of at least 650 vessels in 1996, using nets measuring on average 10-12 km long, has long been at the centre of the debate, though it is not the only one operating in Mediterranean waters. Driftnet fleets continue their activities despite successive international initiatives banning or limiting this low selective fishing practice (swordfish represented only 18% of the Italian driftnet catch in numbers, but nearly 50% in weight; Di Natale, 1996). Resolutions 44/225 and 46/215 adopted in 1989 and 1991 by the General Assembly of the United Nations recommended the imposition of a moratorium on all large-scale pelagic driftnet fishing by 30 June 1992. European Regulation (EC) No 345/92 prohibited driftnet fishing in the Mediterranean with nets more than 2.5 km in length, as did the General Fisheries Commission for the Mediterranean (GFCM) in 1997 under Resolution 97/1, a binding recommendation. Effective moves to restructure the Italian driftnet fleet have been made since the adoption of European Regulation (EC) No 1239/98 and later regulations totally banning the use of driftnets by Community fishing vessels within and outside Community waters from 1 January 2002. Finally, in November 2003 ICCAT laid down a binding Recommendation (03-04) completely banning the use of driftnets for fisheries of large pelagics in the Mediterranean.
Some fleets indeed limited driftnet fishing in Mediterranean waters during this long political process, whilst others grew rapidly. The Spanish Mediterranean swordfish driftnet fleet is an example of the former. In 1993-1994, 27 boats illegally deployed nets 3-5 km long on the Mediterranean side of the Gibraltar Straits (Silvani, Gazo and Aguilar, 1999). This fishery was particularly unselective, with swordfish catches accounting for only 5-7% of total catch in numbers, which was mostly sunfish (Mola mola) (71-93%) and other species such as dolphins and turtles (see the respective sections above). After 1994, these boats stopped using large-scale driftnets and changed target species. Other fleets, on the contrary, have continued to expand, in some cases taking advantage of gears supplied from reconverted fleets. This is the case of North African countries and Turkey, despite national legislation banning large-scale swordfish driftnetting in most of them. Italian and Greek fishermen are known to sell their equipment to Turkish fishermen (A.C. Gücü, pers. comm.). According to Tudela et al. (2005) at least 177 Moroccan vessels carry out large-scale driftnet activities in the Alboran Sea and Straits of Gibraltar areas (357 according to Moroccan sources; document ICCAT SCRS/2002/139). In addition to this important North African fleet, the other major fleets involved are Italian (about 90-100 vessels still exist), Turkish (45-110 vessels; Akyol et al., 2003, SCRS/ICCAT, 2001) and French (46-75; SGFEN/STECF, 2001). Many evidences point to other countries like Algeria as being also likely driftnetters, though confirmed official information is not available.
Solid legal instruments already exist to tackle the issue of driftnet fishing in the Mediterranean, especially after the recent total ban issued by ICCAT. Their enforcement should be a priority for the different coastal states and the concerned Regional Fisheries Organizations (GFCM and ICCAT).