Global agrofood chains, retail and catering: the case of the fish sector

by

John Wilkinson
Programa de Pós-graduação em Desenvolvimento, Agricultura e Sociedade
Universidade Federal Rural do Rio de Janeiro (CPDA/UFRRJ)

1. Introduction

Global agrofood chains orchestrated by an equally globalized supermarket sector have been seen to be the decisive factors currently transforming food production and consumption not only in the developed western economies, but also throughout the world and especially in the developing countries. Fresh produce, and in particular the fruit and vegetable sector, has been identified as the wedge for imposing supermarket hegemony back through the food chain. This sector has received special attention because it represents the key new dynamic component of food diets in developed economies and global middle-classes in the long-term evolution from carbohydrates to animal proteins to vitamins. At the same time, as a new sector, there are few long-standing major players for the supermarkets to negotiate. In addition, it is a sector where the advantages of logistics are foremost and where processing is rudimentary, with low technical or know-how barriers to entry (Reardon, Timmer, Barret and Berdegue, 2003; Dolan, Humphreys and Harris Castell, 2001 ).

The fruit and vegetable sector has also attracted attention for the way in which complex forms of coordination and contracting are emerging as an alternative to both the traditional markets and vertical integration. This has been further stimulated by more general economic tendencies to de-verticalisation, which have placed a premium on new forms of economic coordination, whose analysis has found a powerful tool in transaction cost approaches (Reardon, Codron and Busch, 2001; Green and Schaller, 1998).

As a complement to this focus on supermarket hegemony, there has been a call to pay more attention to the internal transformations in the food systems of developing countries, consequent on supermarket globalisation via FDI. This serves as a corrective to the almost exclusive attention which has been given to the new role of non-traditional food exports from developing countries as the major feature of the globalised food system (Athukorala and Sen, 1998). In Latin America, for instance, it has been calculated that marketing through the domestic(ated) supermarket system is more than twice as important as exports (Reardon and Berdegue, 2003).

While the fruit and vegetable sector provides a flagship for the demonstration of this thesis, it is also shown that supermarkets are more generally bypassing traditional distribution systems, turning to specialized suppliers, setting up their own distribution centres (often with levels of regional and global integration) and imposing qualitatively new demands on all would-be suppliers as regards scale, variety, regularity, hygiene and specific quality requirements. Although the dramatic substitution effect on traditional retail outlets is noted, the focus of the analysis is on the retroactive redesigning of the different food chains through criteria imposed by the supermarket system and legitimized largely in terms of their being responses to consumer demand.

The analysis of the globalization of the fish sector, which follows, will be developed in dialogue with the above framework. In an industry which until recently has been largely based on hunting practices it would not be surprising to find the greater influence of sector specific features. On the other hand, generic bottling and canning technology, which was developed in France and England at the turn of the nineteenth century for a wide range of foods, was also applied to fish. A sardine, canning factory was set up as early as 1824 in France, popularising a product which still remains an important component of the urban diet (Shephard, 2000)^[24]. The drama of fish-stock depletion and the varied responses which it has provoked, however, have played a major role in defining not only the current profile of capture fisheries (especially the drive to vertical integration) and the regulatory regime of which it is an inseparable part, but also the new patterns of coordination involved in certified sustainable fishing systems. In addition, the shift to different forms of fish farming were in large part justified as (complementary) solutions to the same problem, although the issue of food security in the context of projected demographic trends in Asia played a key motivating and legitimating role.

Nevertheless, as the House of Lords enquiry in England makes clear, the supermarkets are seen to be playing a leader role in the reorganization of the British fisheries both through their global sourcing practices and their upstream coordination which now extends to the promotion of "own brand" flagships (House of Lords, 1999). This polemical move has been justified precisely through the need to ensure the criteria indicated earlier as the defining characteristics of supermarket intervention in distribution systems - quality and programmable regularity.

Supermarkets are similarly committed to the promotion of fish farming, to such an extent indeed that they have shown themselves resistant to EU labelling requirements imposed in the wake of critical evaluations on the health and environmental impacts of these fishing practices (Blythman, 2003). In this case, supermarkets appear to be in danger of falling into the trap of production "lock-in", freedom from which has traditionally marked them off from other economic actors in the food chain, and constituted one of their key competitive advantages.

On the other hand, the fish chain also highlights the increasing importance of potential competitors to the supermarkets for the consumer dollar in the form of catering where traditional fish outlets either remain strong (fish and chips in England being the limit case) or expand on the basis of new globally sourced supplies (shrimps in the US), resorting to distinct distribution systems and forms of coordination (Jakle and Scutlle, 1999).

The peculiar origins of fish procurement in capture systems have decisively shaped the patterns of industrial organization in this food chain, whose trajectories appear to be dominated by the compulsive problem-solving logic identified by Rosenberg (1975). Challenges at the point of consumption, particularly those relating to smell and bones, have similarly oriented the drive to new products and processes. This would explain the special appeal of the "ready to heat" fillets of salmon, whose pleasant appearance, minimal smell and bonelessness, would seem to fit the bill, if only it were not "fish farming dependent".

In developing countries, supermarkets are much more restricted by the nature of existing fish supply systems, which are less malleable than fruit and vegetable supplies or even fresh meat. As a result, they appear to be mostly at the stage of replacing, or competing with, traditional retail outlets but still largely based on traditional distribution channels. In Latin America, supermarkets are responsible for a rapidly increasing percentage of fresh fish sales, but these are sourced from the traditional wholesalers (Wiefels, 2003). Few developing countries have their own fleets and their Exclusive Economic Zones (EEZ) are usually rented out to foreign trawler fleets, which fish for their markets of origin (Thorpe, Ibarra and Reid, 2003).

Much domestic supply, therefore will be based on the uneven quality of artisan fishing, itself under heavy pressure from the industrial fishing fleets, and also, in many countries, from the emergence of fish farming. The challenges, therefore, to supermarkets in developing countries, would appear to be particularly acute in the case of reorganising supplies based on new levels of quality and regularity. In the case of new fish products, on the other hand, the supermarkets must compete with already established export or catering oriented distribution channels. As supermarkets consolidate their position in relation to traditional retail outlets we are likely to see an acceleration of global sourcing for the domestic market further curtailing the options open to the artisan fishing sector.

In the following sections we will characterize the unique features, which have informed the specific patterns of globalization in the fish chain. First, we will consider patterns of world production and the unique role of world trade in the characterization of the fisheries sector. This will be followed by a review of the features leading to the current crisis of capture fishing. The third section evaluates the emergence of global patterns of fish farming. In each of these latter two sections the role of large-scale retail and catering in the restructuring of the global fish chain is analysed.

2. Fish and global patterns of production, consumption and trade

World trade in processed, semi-processed or prepared food products as a proportion of total world production for core diet foods such as milk, meats, fruit and vegetables ranges from as low as 5 percent up to 15 percent. In 1995 the value of total trade in horticulture products was US$15 billion, with meats at US$22 billion and tropical beverages reaching US$29 billion (UNCTAD, 1997).

Global trade in fish products in the same year reached US$46 billion and by 2001 that figure was over US$56 billion. Almost 40 percent of total fish production in live weight equivalent is incorporated into world trade, with 50 percent of this coming from developing countries. The net revenue from fish exports by developing countries reached US$18 billion in 2001, more than the combined net value of their rice, coffee, sugar, tea, banana and meat exports (Ruckes, 2003). On the other hand, Japan, the USA and the EU are responsible for almost 80 percent of total imports. Over the last twenty years, trade flows have completely reversed, with the developed world shifting from being a net exporter of 818 000 tonnes of food fish in 1973 to a net importer to the tune of 4 045 000 tonnes by 1997 (Delgado et al, 2003).

The trends in production are no less marked. Fish production based on capture had its boom period in the 1950s and 1960s with an average 6 percent annual growth rate. From there on, however growth slowed down dramatically to around 2 percent in the next twenty years and went into absolute decline in the 1990s. Aquaculture (inland and marine), on the other hand, led by salmon and especially shrimp farming, has had an astonishing record of annual growth: 5 percent in the 1950s and 1960s, 8 percent in the 1970s and 1980s, and 10 percent in the 1990s (Sundar, 2002).

From being a basically subsistence activity combined with rice production, aquaculture was responsible for 37.5 million tonnes of production in 2001, equivalent to 40 percent of total capture fishery in that year (97.6 million tonnes). By 2020 it is projected that aquaculture will account for 40 percent of total fish production (Delgado, 2003). Initially, overwhelmingly based on shrimp production in Asia and salmon in Chile, Canada, Norway and Scotland, fish farming is now extending to white fish species and is being expanded in the Mediterranean (sea bass, bream) and Oceania (tuna, kingfish) with experimental development of off-shore aquaculture (halibut, haddock) in the USA (Staniford, 2003).

Given this dramatic shift-around in trade, the global fish chain could well be analysed within the parameters of the out-sourcing of production phases to developing countries, which has been a common practice in other activities, exploiting advantages of cheap labour, abundant and cheap resources, and lax regulatory regimes (Gereffi, 1994). There is much truth in this perspective, but it should be noted that salmon fish farming from the Northern hemisphere is now the prime source of supermarket sales for this product in the developed countries. Regulation of fish farming in the developed world has been equally lax, primarily because it has been seen as the only viable solution to the intractable problem of marine fish stock depletion. If the experiments in offshore fish farming advance it may well signal a return of production to the waters of the developed countries.

The shift from capture to aquaculture brings with it all the characteristics of intensive farming systems, thoroughly analysed in the agrofood literature, particularly in the case of feed-lots, pigs and poultry, and involving similar, if not even greater, problems related to the environment, waste-product pollution, chemicals, veterinary products (antibiotics, hormones) and intensive feeding systems. These latter have been singled out as the Achilles heel of fish farming to the extent that there is no ready substitute for fishmeal based on pelagic, oily fish species which constitutes a crucial component of capture fishing. It is estimated that 35 percent of fishmeal and 70 percent of fish oil is already dedicated to aquaculture, figures which are projected to rise to 56 percent and 98 percent respectively by 2010 (Staniford, 2003). It comes as no surprise that soybeans are already being experimented as an alternative feed supply, although the effects on final food taste may be a limiting factor. This shift from capture to aquaculture has been legitimized by analogy with the evolution from extractivism to the intensive production paradigm of the green revolution, and, to push home the point, has been baptized as the "blue revolution" (Economist, 2002).

The progressive convergence of animal protein production systems has led to an ever greater involvement in the fish sector by leading players from other food chains situated at different stages along the production process. The first move in this direction, as we have seen, was the result of the application of canning technology, but in this case, while the technology was generic, firms tended to remain specialized in accordance with the origin of the raw material in question.

With the advent of frozen and prepared foods, pioneered by Birds Eye^[25], fish products became one option within a portfolio of "main meal" dishes, leading to the direct involvement of firms from different food-chain backgrounds - Unilever, Tyson and Conagra. Fish farming, in its turn, has attracted global firms in animal feed, such as Ralston Purina, together with upstream inputs and genetics firms, among whom Monsanto is very much in evidence. Nutreco, the largest global fish farm firm is equally involved in pig, poultry and animal feed production.

Global fish consumption patterns are as striking as trade flows, showing symmetrically inverse tendencies. In 1973, developed countries were responsible for 55 percent of global fish consumption, declining to 31 percent by 1997. This increase in developing country share has been largely due to China (11 percent to 36 percent in the same period), with India and Southeast Asia doubling their share, whereas in sub-Saharan Africa consumption stagnated and declined in the 1990s.

Fish is a much more decisive component of the protein diet in developing countries, representing an average of 20 percent in less developed countries, but rising to as much as 50 percent in some countries (Bangladesh), whereas in developed countries, also with sharp variations, it represents an average of 13 percent. Per capita consumption in the developing world (excluding China), however, was calculated at 9.2 kg per capita, as against 21.7 kg in the developed world, (Delgado et al., 2003). While developing countries, therefore, are more dependent on fish for their protein needs, there is a net flow of fish from these countries for consumption in the developed world (Sundar, 2000).

The depletion of capture fish resources in Northern waters and the extraordinary surge in fish capture and especially fish farming in the South, almost exclusively, (with the decisive exception of China), devoted to exports for northern consumers, in addition to acute social and environmental issues, starkly poses the dilemmas of global food security.

3. Capture fishing: from a hunting to a sustainable harvesting industrial trajectory?

Transatlantic trade in cod was already well established in the 16th century replacing the Icelandic routes. French boats, better provisioned with salt could each carry back a 25 000 catch of processed and prepared cod, while the British, with salt in short supply, had to dry their cod on land before setting back. The Newfoundland settlements, which emerged, were based on the cod economy, and this, with time, would become an important motive in the events leading to the Boston Tea Party (Kurlansky, 1999). On the Pacific coast, already in the early 19th century, the Hudson Bay Company promoted trade in salmon, which was exported to England (Dawson, 2003).

The icing of fish, a technology learned by the British from the Chinese, began to be used as early as the 1780s to open up trade in fresh salmon from Scotland to London. This technology, initially used for the marketing of landed fish, was rapidly applied to preserving fish at sea, allowing for larger fleets, longer fishing times and deep sea fishing. In its turn, these changes provoked adaptations in the ships and the gear (Shephard, 2000).

As early as the 1870s the USA Congress, concerned with the decline in New England fish, created the Commission of Fish and Fisheries. Before the end of the century, regulation controlling fishing in its own waters began to be put into place, including net restrictions, closed seasons and spawning escapement requirements. Technologies developed during the Second World War and applied to trawler fleets transformed the nature of international fishing in the 1950s and 1960s. These included the application of radar and sonar to locate and follow migratory stocks. The use of diesel engines permitted cheap fuel allowing fleets to stay out longer and fish deeper (Safina, 2003).

The factory trawlers made their appearance in the early 1950s as an adaptation from whale fishing and by the 1970s they dominated the world's fishing with the Soviet Union sporting 400, Japan, 125, Spain 75, France and England 40 each (Rogers, 2003). These trawlers could scoop up 400 tonnes of fish in a single haul, could stay at sea for months and process the fish into frozen filleted slabs. This mass, primary processing, production system promoted the development of the quick frozen convenience fish food sector, which in its turn was transformed by the consolidation of large-scale retail.

The supermarket format itself owed much to the revolutionary implications of the cold food chain. Such a development was already predicted in a Birds Eye advert of 1930 which argued that frozen foods would change the way people shopped and that there would be a new kind of store in the future "not a grocery store or a meat market or a fish market or a delicatessen, but all four rolled into one. It is a food store in the broadest sense of the word." (Shephard, 2000). Today, Birds Eye is one of Unilever's three leading frozen fish brands.

Six per cent annual growth rates of capture during the first two decades after World War II were the result of this revolution in fishing practices, leading to the serious depletion of world fish stocks on the high seas. The factory trawler fleets also threatened stocks in coastal waters, and with them domestic supplies and the livelihoods of artisan fishing communities, which were the key units of many national fishing economies. In particular they threatened the traditional combination of fishing and on-land primary fish processing which together had guaranteed the viability of these fishing communities (Alcock, 2002).

Territorial disputes became common (of which perhaps the most famous were the British-Iceland cod wars), not now within the three nautical mile range, which was seen to be part of national sovereignty, but over rights to fishing on a nation's continental shelf and to migratory fish whose origins were in the bays and rivers of a particular country. A series of "Law of the Sea Conferences" attempted to impose limits on catches and these were followed by the enactment of national 200 mile Exclusive Economic Zones (EEZ) which were to change the rules of the game but not necessarily slow down the rate of fish stock depletion.

The USA Magnuson Law of 1976, instituting the EEZ, although it had been preceded by Iceland´s unilateral action in a similar sense, was the landmark for worldwide national legislations. While rich in fish, the USA when compared to Japan, the Nordic countries and the then Soviet bloc, was essentially a meat (initially red and then increasing white) consuming country and it had no fleet to compete with the factory trawlers (Buck, 1995).

The Magnuson Legislation led to replacing the open sea regime with national and regional fishing management schemes but it did not change the fishing model. Rather it allowed the USA to become a competitor and by 1999 it also had 50 factory trawlers. Each trawler cost now anything from US$40-65 million and the largest could net and process one million pounds of fish a day, every day.

Four or five firms are responsible for 20-30 percent of the world fish catch, with the largest of them, the Norwegian, American Seafoods/KRG1, controlling 10 percent of the global white fish market and 40 percent of the pollock quota in the Baring Straits (Buck, op.cit.). With the decline in cod, pollock became a key new product, which was transformed into surimi for the Japanese market. As artificial crab it also became an important input for the burgeoning seafood restaurant chains in the USA By 1996, the top eight seafood chains in the USA had a turnover in excess of US$3 billion (Jakle and Sculle, 1999).

In addition to accelerating the depletion of the principal fish stock, the trawlers were responsible for enormous waste in the form of bycatch, calculated in 1994 in the case of New England as 145 percent of the total ground fish landings. It is also thought that they may be provoking irrevocable damage to the ocean's ecosystems as they drag up the biomass from the seabed.

The combined pressures of depletion and threat to long-term sustainability led to restrictive legislation on catch and gear in Northern waters, involving bans and quotas. One important consequence was a shift to the waters of the developing world. Unlike the USA, these countries may have acquired EEZ rights, but their ships were unable to take advantage of them and the financing of new fleets was for most countries prohibitive. Rights to fish were either rented out or foreign fleets allowed to fish under national flags. For the EU, this was an ideal alternative to selling off fleet capacity, and Spanish trawlers took over the Argentine/Uruguay hake grounds with the incentive to these countries of tariff exemption for the import of the catch to Europe (Thorpe, op.cit, 2003).

In many zones, a Total Allowable Catch (TAC) was imposed, but without individual quotas this led to a "race-for-fish" based on ever larger boats and shorter fishing seasons, favouring the international fleets and further threatening the ice chilling fleets and local fish processors. By the 1990s the management systems in most developing countries were in disarray, which, particularly in a climate of liberalization policies, pointed the way to privatizing strategies.

The introduction of Individual Fishing Quotas (IFQ), often transferable (ITQ), might in principle have mitigated the pressures of the "race-to-fish" syndrome. By then, however, the fleets were dominated by large firms, allied to foreign capital, and most countries exhibited a high degree of "regulatory capture", with powerful interest groups influencing fishing policy. In the Nordic countries, particularly Iceland, and the USA, some success seems to have been achieved in defending local fishing systems through the design of the quota system (Pacific Whiting Conservation Cooperation, Sylvia and Munroe, 2003), but this is less feasible where management systems are weaker.

With 70 percent of capture stocks either fully or over-fished, the pressure for change was pursued along two lines - from quotas to sustainable management systems, and from the capture fishing to aquaculture. For some, these two options were complementary, for others aquaculture was seen as a substitute for capture fishing. Yet others saw in aquaculture a radicalization of the industrialization of fishing, with even more pronounced social and environmental consequences.

On the international regulatory front, the Oceans chapter of the Rio Summit's Agenda 21 was followed by the FAO Code of Conduct for Responsible Fisheries, the UN Agreement on Straddling Stocks and Highly Migratory Fish Stocks, and the United Nation Convention on the Law of the Sea (UNCLOS). Within the fishing industry by far the most significant initiative was the establishment of the Marine Stewardship Council (MSC), which emerged from a joint initiative between Unilever and the World Wide Fund for Nature (WWF). The MSC, which is now independent and financed by charitable trusts, has adopted three basic principles as the criteria for its certification:

"A fishery must be conducted in a manner that does not lead to over-fishing or depletion of the exploited populations and, for those populations that are depleted, the fisheries must be conducted in a manner that demonstrably leads to their recovery. Fishing operations should allow for the maintenance of the structure, productivity, function and diversity of the ecosystem (including habitat and associated dependent and ecologically related species) on which the fishery depend. The fishery is subject to an effective management system that respects local, national and international laws and standards and incorporates institutional and operational frameworks that require use of the resource to be responsible and sustainable." (Heap, 2000)

Unilever is the world leader in frozen final food products, and its three brands, Findus, Birds Eye and Igloo, account for some 25 percent of the frozen fish market in Europe and the USA While aquaculture is beginning to move into these species, it offers no short-term perspective for a firm such as Unilever, which has therefore become a proponent of sustainable capture fishing, planning to source all its fish from such certified systems by 2005. To date, it claims to have 25 percent of its fish from sustainable certified sources and MSC certificates have now been attributed to the Salmon Fish from Alaska, Hoki from New Zealand and Rock Lobster from Australia (Unilever's site)

Another major development has been the adoption of flagged vessel schemes by the Asda/Walmart and Tesco supermarkets in Britain. While the contracts are triangulated through suppliers or processors, the supermarkets commit themselves to taking the whole catch. The principal motivation in this case seems to be improvements in supply chain management, particularly to ensure fresh supplies, seven days a week. The quality of the supplied fish is also guaranteed through the stipulation of fishing practices for fresh fish - maximum time at sea, storage and icing practices. In the House of Lords inquiry the question was raised whether quality in this sense might not in fact increase discards. Whether or not this is the case, it is the issue of consumer oriented quality rather than sustainability which seems to be to the fore. Along a similar consumer driven logic "locally caught" supplies are being promoted in certain regions (House of Lords, 1999).

As with other food products, scares are leading to the implementation of traceability systems and the EU is financing a programme, Tracefish, for the elaboration of voluntary industry standards on quality and traceability, which will probably begin to be implemented as from 2004. Initiatives are already in place which involve the use of electronic weighing scales, enabling catches to be bar-coded for weight, time and place of capture, together with water temperature (MacDubhghaill, 2003). As in the flagship examples, however, it is not clear that the quality and traceability take on board the issues of sustainability, which are the basis of the MSC certification schemes.

4. Fish farming

Perceptions of the tragedy of ocean fish-stock depletion have to a great extent been mitigated by the promotion of a sentiment of inevitability, with capture fishing being viewed as the counterpart to hunting and extractivism in agriculture, both of which would now be replaced by increasingly convergent intensive production practices. The notion of the "blue revolution" captures this vision perfectly, presenting fish farming as an evolution to more efficient practices, the more so as these systems extend beyond salmon, trout and shrimps to white fish species. In addition, a conjunction of powerful economic and political interests has reinforced and accelerated this shift.

As we saw earlier, fish, increasingly in the form of fish farming, is now the single most important food export from developing countries, and aquaculture has received decisive support from international financing, most notably the World Bank and the Asian Development Bank. In addition, this sector has been seen to provide opportunities for less developed countries such as Bangladesh, and has led to the emergence of developing country transnationals, of which the Thai Charoen Pokphand (CP) Group is the most notable. More generally, fish-farming provides decisive new markets for the global food industry players in the genetics, inputs and feeds sectors - Nutreco, Monsanto, Ralston Purina. Lastly, and perhaps most importantly, large-scale retail and catering have seized on salmon and shrimps as the ideal fresh fish alternatives and have committed themselves heavily to the promotion of these markets, based largely on fish farming both in the North and the South.

Shrimp and salmon have been the most spectacular examples of the growing importance of fish farming which is now almost exclusively responsible for the increase in world trade of these products. Shrimps provide a paradigm case for critiques of North-South trade. With the shift to aquaculture, a traditional extensive, subsistence, mixed cropping regime of shrimps and rice in the South was replaced by intensive, export monoculture for luxury consumption in the North.

In the US, the second largest importer after Japan, 70 percent of shrimps are consumed outside the home, particularly in the seafood restaurant chains dominated by the Red Lobster group, which alone consumes 5 percent of world exports. According to the President of ICEC Seafood Corporation: "Red Lobster's advertising of shrimp has stimulated all consumption and in part made possible the 250 percent increase in US shrimp consumption since the advent of aquaculture" (Greenpeace, 1997). Asian developing countries have provided the major supply base, followed by Central America and more recently Brazil. With the development of aquaculture, labour is shifted off the land to the processing factories in much the same way as other out-sourced, labour intensive, stages of global production chains. In 1980, eight Asian countries were responsible for 76 000 tonnes, a figure which had risen to 448 000 by 1997. Ecuador, the leading Central American producer, increased its production from 9 000 to 130 000 tonnes in the same period (Goss, Burch and Rickson, 2000).

This explosive growth has been accompanied by sharp price oscillations, far-reaching environmental impacts (elimination of mangroves, water pollution, diminishing biodiversity), conflicts over landuse and threats to local communities, together with devastating outbreaks of disease. A combination of untried technology and the perspective of huge short-term gains led to the emergence of a "boom and bust" profile to the industry.

Taiwan, initially the leading shrimp-farming exporter, suffered a crippling disease outbreak in the late 1980s from which it only began to recover in 2000. In compensation, it exported its technological know-how to Thailand, which quickly transformed itself into Asia's leading exporter, giving rise to the first, developing country transnational in the sector, the Charoen Pokphand (CP) Group. Strong support from the World Bank and the Asian Development Bank, together with the growth of the CP Group made shrimp production an irresistible option for most Asian countries - Indonesia, Bangladesh, India, Vietnam. Disease, the search for ever cheaper labour costs, and the Asian financial crisis, accentuated this nomadic tendency, which extended beyond Asia to Central and South America and, in the most recent period, also to Africa.

On the other hand, as with the case of fruit and vegetables, the transnationalization of retail has also stimulated domestic demand, with battered fried shrimps, together with other fish specialities, now experiencing 10-20 percent annual growth rates in Thai and Singapore supermarkets. A more complex globalization is replacing the North-South paradigm, with the frontiers between export and domestic markets becoming more fluid (UNCTAD,1997).

While global players have stimulated the development of shrimp farming exports from developing countries, protectionist measures in the forms of tariff escalation, which has been as high as 30 percent for some products, and more rigid quality standards (implementation of HACCP systems) imposed formidable barriers to access, particularly to the US and EU markets. Tariff escalation has tended to decline, however, and less developed countries have benefited from tariff exemption. Countries, such as Bangladesh, have shown the viability and benefits of adjusting to more stringent quality controls, although these can be activated for purely protectionist motives, as would seem to be the case in the current measures being taken against Brazil, which has emerged as the latest developing country exporter of shrimps to the US market (Cato and Subasinge, 2003; Josupeit, Lem and Lupin, 2003).

New quality standards may lead to greater control over contaminants (antibiotics) detectable in the final product, but they do not necessarily address the broader issues of environmental (and social sustainability). There are some indications however that the more opportunistic features of the industry are ceding ground under the joint impact of local struggles (particularly in the case of India), the "busts" which have followed the "booms", and the Asian financial crisis which led to the crisis of big players such as the Thailand CP Group. As a result, there is now a greater presence of small shrimp farming operations, a greater concern with hatchery operations to ensure shrimp stocks and a shift to lower density production systems (Cyriac, 2003). This would not seem to be the case, however, with new entrants such as Brazil, whose drive to become a world leader in shrimp exports is based on very high per hectare production, which may reproduce the disease and pollution syndrome of earlier Asian shrimp farming, although Brazil is now in a position to benefit from the collective learning of this global industry, together with the new technologies and know-how now available (Rocha, 2003).

If catering in Japan and the USA has been primarily responsible for the growth of the shrimp economy, supermarkets have played the decisive role in the dramatic popularisation of salmon consumption, which, from being a high priced luxury in the 1970s, is now an everyday fresh fish option and a staple of the booming, ready-made sandwich sector. Some 90 percent of salmon supplied to British supermarkets now comes from salmon farms (Edwards, 2002).

While Chile is fast on its way to becoming the leading exporter of salmon, the majority of farmed salmon still comes from the North - Norway and Scotland. Chile's exponential expansion of salmon production in the 1990s, however, (involving a twenty-fold increase) has led to rapidly falling prices and bankruptcy for many Norwegian producers (Milliken, 2003). More recently, salmon farming has become a major focus of health concern, as research has pointed to high levels of contamination, leading the EU to require labelling distinguishing farmed from wild salmon. Supermarkets have resisted this move, both calling in question their "consumer sovereignty" orientation and suggesting that they have increasingly become stakeholders in the move to fish farming.

If salmon farming were under attack solely for environmental factors we might anticipate a decisive shift to the South, especially as Chile plans to double its production over the next ten years. It should be said, however, that environmental opposition is also gaining ground in Chile as salmon farming moves further south into regions of the country valued particularly for their tourism (Centro Oceanos, 2001).

In Scotland, the environmental critique has been particularly forceful, focussing on pollution levels (sewage equivalent of 9.4 million people), the threat to shellfish sites, algal blooms and escapes of farm fish into the wild. It has also been calculated that the economic value of salmon angling (235 million pounds), which is particularly threatened by escapes and pollution, is on a par with salmon farming. Its beneficiaries, however, are more diffuse (tackle, clothing, hotelry, travel), and have less clout than the highly concentrated salmon farming sector where two firms - Norsk Hydro (Norway) and Nutreco (via Marine Harvest) - control two thirds of the 340 farms in Scotland. Globally, the top seven companies control 40 percent of world farmed salmon production (Friends of the Earth, 1999).

The challenge to salmon farming goes beyond environmental issues and reaches to the central vulnerability of all fish farming - dependence on concentrated fishmeal and oil pellets made from capture-fish pellagic species, which are both highly contaminated and rapidly becoming depleted, as " fishing down the food web" increases with the exhaustion of large white-fish stocks.

High levels of contamination in farmed salmon have been officially recognized by the European Commission, British Government advisors declared salmon to be the most contaminated food sold by British supermarkets, and USA research on Scottish farmed salmon has concluded that it is a serious cancer health risk based on detected levels of dioxins and polychlorinated biphenyls (PCB). Although the British Food Agency maintains that the health benefits of fish consumption outweigh any risks, the European Union has determined that farmed fish must be labelled. This measure has been resisted by supermarkets which calls in question their self-proclaimed responsiveness to consumer interests and exemplifies the degree to which they have now become "locked-in" to certain supply chain solutions.

Threatened by competition from Chile, Scottish fish farming is now moving into the production of cod and haddock with the direct support of supermarkets, such as Marks and Spencer. From being a highly specialized activity, therefore, fish farming is increasingly being adopted as a generalized solution to the dwindling supplies of large white fish species from capture-fishing, attracted also by the high prices which this decline provokes.

It is estimated that by 2006, most large fillets of cod consumed in Britain (total consumption 170 000 tonnes) will be from fish farming. Cod, it should be added, consume twice the feed of salmon with an equivalent expected increase in pollution. Sea cage farming of carnivorous finfish has now become a global business, from Norway to New Zealand, from the Mediterranean to Mexico from the Americas to Australia and includes sea bass, bream, barramundi, halibut, tuna and other more exotic species. In spite of the setbacks provoked by disease and environmental regulation and the uncertainties of the market, this move of fish farming into the principal white fish species is being promoted by leading multinationals Monsanto, Nutreco, Stolt, Cermaq and Ralston Purina (Standiford, 2003).

In a devastating indictment of this tendency, Standiford, points to five fundamental flaws in sea cage fishing: untreated wastes, mass escapes, diseases and parasites, toxic chemicals, and fish feed/food. He argues that in theory new practices and technologies may palliate the first four flaws, although only inland, closed containment systems, such as those used by Future SEA Farms and Fish Protech in Australia offer guaranteed protection, so much so that Fish Protech has received authorization to locate in drinking water collection areas. The fifth flaw, however, fish meal and oil, has no "acceptable" solution.

By 2010 it is estimated that at current rates of growth, 56 percent of fish meal and 98 percent of fish oil will be consumed in fish farming. Rather than being an alternative to capture fishing, the oceans are being drained of their oily, pellagic species to service the expansion of fish farming. Already many species are disappearing and ocean biomass collection is developing to compensate the depletion of fish meal stocks.

Within the trajectory of the fish farming alternative many solution are being explored to deal with the first four flaws identified by Staniford, involving off-shore farming in deep water to deal with pollution and disease with a consequent reduction in chemicals, together with genetic control of reproduction to deal with escapes (an adaptation of "Terminator" technology).

Until these options are operational or financially attractive alternatives, however, the dominant tendency is to move from overpolluted areas into pristine waters, imitating the frontier mining/farming of yesterday, in spite of increasing opposition from environmentalists, local communities and tourist interests. As for the fifth flaw, the solution currently being explored is soybean pellets, closing the loop with white meat production and unifying the production matrix of the animal protein complex.

Transforming carnivorous fish into vegetarians is the obverse of feeding animal meats to cows, and although it may be less of a health risk, its effect on taste has proved to date an important obstacle. But for how long? The evolution of chicken taste is a ready reminder of consumer adaptability. Staniford's optimal solution would be the elimination of finfish carnivorous sea farming and the promotion of inland shell fish farming. Barring that, he argues, the only solution is closed containment systems or "close-down".

5. Conclusion

Global value chain analysis, with its distinction between supply and demand oriented production systems, and parallel work on the globalization of food retailing, have opened up important insights into the current dynamics of agrofood system globalization. The one has highlighted the way in which demand driven considerations have redesigned food chains almost from scratch, while the second approach has also showed the need to go beyond a simple North-South divide as large-scale retail breaks down the frontiers between export and domestic markets even in developing countries.

The paradigm for analysing these tendencies has been the fruit and vegetables sector where backward linkages have been recreated and redesigned in accordance with the evolution of logistics and quality demand criteria. The fish food chain can be usefully analysed within this perspective. The shift of supermarkets towards direct contracting with trawler ships for capture fishing and their leading role in the creation of a popular salmon, fresh fish market are clear confirmation of the way retail is currently piloting changes in the global organization of agrofood systems.

At the same time, the fish sector demonstrates a much more complex interplay of interests even at the demand end of the chain where catering has played a more autonomous role, particularly in the promotion of shrimps and similar seafood products. More decisively, the fish sector displays enormous inertial characteristics based on the historical accumulation and consolidation of interests around different segments of the global food chain, which have erupted in conflicts and even wars and have only been partially contained by increasing regulation. Faced with this out of control juggernaut where advanced technologies are harnessed to short term but long consolidated interests, the supermarket involvement in capture fishing has been limited to the fine-tuning of logistics and the quality (freshness) of supplies.

The most farsighted initiatives to date have come from the fish-processing sector, dependent on globally sourced white fish supplies, in the form of the Management Stewardship Council (MSC), jointly promoted by Unilever and the World Wide Fund for Nature (WWF).

On the other hand, the retail sector has played a decisive role in the promotion of the fresh salmon market based on fish farming. Unlike the fruit and vegetable sector, which is more readily subject to surveillance and consumer friendly practices, fish farming is dominated by powerful global upstream players and by a complex still untried technological base, with enormous social and environmental externalities. The evidence to date would suggest that the retail sector itself has now become "locked-in" to a production system, initially seen as the guarantor of idealized consumer demand for quality (organoleptic and nutritional) and practicity but now transformed into an object of suspicion and confusion. The supermarkets refusal to label salmon as "fish farmed", spells perhaps the end of its innocent courtship of the consumer, based on demand satisfaction.

As retail assumes a more and more direct role in the organization of the value chain it also loses the flexibility to respond in an uncommitted way to fluctuations in consumer perception. At the same time, its commitment to final product quality falls far short of consumer demand once this takes on board the more conflict-ridden issues of sustainability. A satisfactory solution to the dilemmas of the global fish chain cannot be left to the supermarket´s putative capacity to internalize consumer demand, but requires the direct involvement of the consumer-citizen in a complex mix of direct action, public regulation and new consumer practices.

REFERENCES

Alcock, F. 2002. "Scale Conflicts and Sectoral Crisis: the Fisheries Development Dilemma" in F. Biermann et al Proceedings of the 2001 Berlin Conference on the Human Dimension of Global Environmental Change, Potsdam

Athukorala, P.C. & Sen, K. 1998. Processed Food exports from Developing Countries: Patterns and Determinants

Balsevich, F., Berdegue J.A., Flores, L., Mainville, D., Reardon, T., Busch, L. & Unnevehr, L. 2003 (forthcoming). "Supermarkets and Produce Quality and Safety Standards in Latina America" in American Journal of Agricultural Economics

Blythman, J. Supermarkets Mislead Public over Seafood (n/d)

Buck, E.H. 1995. Individual Transferable Quotas in Fishery Management, CRS Report for Congress

Cato, J.C. & Subasinge, S. 2003. Food Safety in Food Security and Food Trade, Focus 10

Cyriac, K.J. 2003. The Marine Products Export Development Authority, India, FAO, Rio de Janeiro

Dawson, R. 2003. Vertical Integration in Commercial Fisheries, PhD Thesis, Virginia

Delgado, C. et al. 2003. Outlook for Fish to 2020, IFPRI

Dolan, C., Humphrey, J. & Harris-Pascal, C., Horticulture Commodity Chains: The Impact of the UK Market on the African Fresh Vegetable Industry, IDS, Working Paper, 96

Economist, 2002. The Blue Revolution

Friends of the Earth, Scotland, 1999. The One that Got Away

Gereffi, G. 1994. "The Organisation of Buyer Driven Global Commodity Chains" in Gereffi, G. & M. Korzeniewicz (eds) Commodity Chains and Global Capitalism, Westport

Goss, J., Burch, D. & Rickson, R.E. Agri-Food Restructuring and Third World Transnationals: Thailand, the CP Group and the Global Shrimp Industry

Green, R. & Schaller, B. 1998. " La Place du Contrat de Référencement dans l ´Évolution des Rapports d´Échange entre Fournisseurs et Distributeurs, INRA-ESR, Paris

Greenpeace. 1997. Shrimp: the Devastating delicacy

Edwards, R. 2002. Farm Salmon is now the most Contaminated Food on the Shelf

House of Lords. 1999. Examination of Witnesses

Jakle, J.A & Sculle, K.A. 1999. Fast Food, John Hopkins University Press, Baltimore & London

Josupeit, H., Lem, A. & Lupin, H. 2000. Aquaculture Products: Quality, Safety, Marketing and Trade in R.P. Subasinghe et al Aquaculture in the Third Millennium, Bangkok

Kurlansky, M. 1999. Cod, Vintage

MacDubhghaill, U. 2003. The Outlook for the UK Fishing Industry, IntraFish.com Industry Report

Milliken, M. 2003. Salmon Farms Spawn Fortunes and Critics in Chile, Santiago

Reardon, T. & Berdegue, J.A. 2002. "The Rapid Rise of Supermarkets in Latin America", Development Policy Review

Reardon, T., Timmer, Barret & Berdegue, J.A. 2003. "The Rise of Supermarkets in Africa, Asia and Latin America", American Journal of Agricultural Economics, 85, 2003

Reardon, T., Codron & Busch. "Agrifood Grades and Standards in Developing Countries: Issues of Inclusion and Exclusion", in International Food and Agribusiness Management Review

Rocha, I de Paiva. 2003. Advantages and Constraints of the Brazilian Fishery Industry, FAO, Rio de Janeiro

Rodgers, R.A. 2003. Mesh Flesh: Fish farms and Environment/Equity Issues, York

Ruckes, E. 2003. World Fish Trade, Demand Forecasts and Regulatory Framework, FAO, Rome

Safina, C. 2003. Fishing off the Deep end - and Back, Blue Ocean Institute

Shephard, S. 2000. Pickled, Potted and Canned, Headline

Staniford, D. 2003. Closing the Net on Sea cage Fish Farming, (Paper: Queensland Conservation Council and the Australian marine Conservation Society), Queensland

Sundar, A. 2000. "Marine Resources", in Marketing the Earth: The World Bank and Sustainable Development, www.foe.org

Thorpe, A., Ibarra, A.A & Reid, C. 2003. The New Economic Model and Fisheries Development in Latin America, (paper: Marine Environmental Politics in the 21st Century), Berkeley

UNCTAD. 1997. Opportunities for Vertical Diversification in the Food processing Sector in Developing Countries, UN

Wiefels, R. 2003. Changes in Fish Distribution Channels in Latin America: the Increasing Role of Supermarkets, FAO, Rome

Wilkinson, J. 2003. Globalisation, Food processing and Developing Countries: Driving Forces and Impact on Small Farms and Firms, FAO, Rome

WWF. 2001. Scotland, Bitter Harvest