SESSION 2
Aquaculture

Food safety controls from farm to table with emphasis on aquaculture shrimp in Thailand

Wanwipa Suwannarak
Department of Fisheries, Bangkok, Thailand

ABSTRACT

The Department of Fisheries in Thailand has adopted an Action Plan on Food Safety from Farm to Fork to ensure that shrimp and shrimp products are safe and comply with national and international standards. In addition to that, procedures (including good aquaculture practices, good manufacturing practices and hazard analysis critical control point systems) have been improved in all sectors of the food supply chain. This includes:

- Good aquaculture practices at hatcheries and nurseries

- Good aquaculture practices at farm level

- Good practices for feed/drug and chemical suppliers

- Good practices for harvesting, marketing and processing

- Good manufacturing practices and hazard analysis critical control point systems at processing level

- Import and export controls

Main activities involve registration, training, technical assistance, follow up, monitoring and control of production standards, sampling, approval and certification procedures and controls for imports and exports.

INTRODUCTION

Thailand has emerged as one of the largest seafood suppliers to the global market. The volume and value of exported products are increasing and the quality is accepted worldwide. Exported seafood products are important foreign income sources for the country. The total value of exports for 2002 was US$ 3.7 billion. Black Tiger shrimp (Penaeus monodon), mostly from aquaculture production, is a major export commodity. Currently over 25 000 shrimp farms with a culture area of approximately 73 000 ha are in operation. In 1999, total shrimp production from culture and wild catch reached 250 000 tonnes and more than 95 percent was exported as frozen and processed products. Frozen shrimp products accounted for 136 000 tonnes, earning US$ 1.3 billion and have since increased steadily. By 2002, frozen shrimp export value had reached US$ 1.8 billion. This clearly indicates the potential of the Thai seafood industry.

FOOD SAFETY CONTROL

In order to sustain the Thai food industry (including the shrimp industry) and to maintain consumer confidence in product safety, 2004 has been declared as the "Food Safety Year" by the Thai government. The general aims are to apply standard practices to all sectors of the value chain from primary production to the market place and to apply a single standard for products whether destined for domestic or overseas markets.

As a result, the Department of Fisheries (DOF), an office under the Ministry of Agriculture and Cooperatives responsible for seafood products, has adopted the "Action Plan on Food Safety from Farm to Table". This is being implemented through many legislative acts - the Fisheries Act, Food Act, Animal Diseases Act and Import and Export Control Act.

The goal of the action plan is to ensure that shrimp and shrimp products are safe and reach or exceed national and international standards. The stakeholders include hatcheries, farms, feed suppliers, drug and chemical agent suppliers, handlers, brokers and fish markets, primary processors, processors, importers and exporters. In addition, process improvement has been emphasized using good aquaculture practice (GAP), good manufacturing practice (GMP) and hazard analysis critical control point (HACCP) systems in all sectors of the food supply chains.

Following the plan mentioned above, the strategies are described as follows:

1. Good aquaculture practices for hatchery and farm

All aquaculture hatcheries and farms are required to register with the DOF and gain approval to begin operations. This ensures that guidelines on site selection, farm design, operational standards and chemical and drug usage are followed.

Farm inspections are carried out to ensure that approved operational procedures and standards are properly implemented and also to ensure that chemical and drug residues are not present. DOF has operated 22 mobile units in major fish culture areas. Each mobile unit is assigned to an area and is required to inspect approximately 8 farms per week. The mobile unit then moves to other areas until all areas are covered. By this inspection procedure, each farm will receive at least five visits annually.

2. Good practices for feed, drug and chemical supplier

All feed, drug and chemical usage in aquaculture must be registered with the DOF. The DOF does not allow any feed factory to manufacture medicated feeds. Feed wholesalers and retailers also have to acquire a license from the department. In addition, feed formulae and labels must be approved before sale is permitted. Feeds are sampled regularly and a full composition analysis undertaken, including testing for drug contamination.

To strengthen drug control, use of chloramphenicol and nitrofurans in animal feed were banned in Thailand in 1998 and 1999 respectively. However, both are still allowed for human medical treatment. Nevertheless, the Thai Food and Drug Administration and Customs Department have strong controls on imported drugs, including inspection and monitoring programmes.

At hatchery and farm levels, information is disseminated widely and training on drug and chemical usage is conducted. Frequent inspections are performed to ensure compliance with standards.

3. Good practices for harvesting and marketing

All holders, brokers and fish markets must be registered with the DOF. To maintain quality of fish and prevent contamination during handling, a quality control and inspection supervisory programme is applied. The programme is conducted through supervision of handling techniques, monitoring equipment utilization, quality maintenance programmes, standardized transportation techniques and training and extension on processing techniques.

4. Good manufacturing practices and HACCP controls for processing plants

All processing plants must be registered with the DOF. Currently, the fish products safety and quality control programme is based on GMP and general principals of food hygiene, which are implemented as the prerequisite programmes. All processors under DOF approval must implement HACCP systems.

Now, Thailand is one of the few countries in the world that enforces compulsory HACCP implementation in fishery establishments. In 1991, for the first time, DOF implemented a voluntary HACCP fish inspection programme which was subsequently strengthened and became mandatory in 1996. To maintain their approval status, DOF monitors the establishments and their products regularly.

5. Import and export control

To control the quality of exported products especially with regard to drug residues, DOF has acquired and installed Liquid Chromatography Mass Spectrophotometer (LCMSMS) equipment in the laboratory to determine metabolites of nitrofurans in shrimp products. Currently, all shipments of cultured shrimp are subjected to full inspection prior to export. Pre-shipment inspections are conducted to identify possible drug residues in finished products. This is to ensure compliance with DOF and importing country requirements. When contamination exceeds established limits, the lot is rejected.

When the drug residue test is positive, the license for farms that supply raw material to processors will be withdrawn until the proper corrective measure is taken and subsequently test results meet the requirement. In addition, drug and chemical vendors that supply to the farm will also be blacklisted. The processors are required to verify their HACCP plan, which identifies possible problems and details action plans to cope with such problems, to the DOF. The DOF then will perform a follow-up inspection to ensure compliance. These procedures are conducted to guarantee that all potential problems are addressed and that reoccurrence is prevented.

Beside the export control, DOF also conducts the surveillance programme for drug residues and shrimp diseases in imported products. The imported lots will be rejected and then returned back to the country of origin when the residue levels exceed the established limits and DOF will take the legal actions against the importer including fines and charges.

The main activities in the action plan are as follows.

1. Registration of all involved sectors and improving standards in all sectors

2. Training in inspection and auditing for official inspectors

3. Training in GAP/Code of Conduct (CoC) for farmers

4. Training in good hygiene practice (GHP)/GMP/HACCP for handlers and processors

5. Technical assistance for all sectors

6. Follow up, monitor and audit the implementation of GAP, CoC, GHP, GMP and HACCP for all sectors

7. Test samples of feed, drugs, chemicals, shrimp fries, shrimp and shrimp products

8. Control the usage of shrimp movement document

9. Control production standards

10. Certify hatcheries, farms, feed/drug and chemical suppliers, handlers, suppliers and processors

11. Issue health certificate for exported products and provide information of health certificate status issued by the DOF for EU member countries through web based notification

12. Control imported products to ensure that it meet the SPS requirements.

THE CURRENT ISSUES

DOF controls and monitors many aspects of the shrimp industry. Special attention is given to the prevention of drug and chemical residues as well as microbiological contamination. Prevention of environmental pollution and maintenance of water quality in the production areas are also vitally important. Implementation of the plans has led to Thailand being able to produce high quality shrimp products. At present, the number of cultured marine shrimp shipments in which drug residues are detected is less than 10 percent. Meanwhile, the concentration level of contaminated substances in detected products is decreasing. Ninety percent of contaminated samples have very low levels of nitrofuran residues (0.3 to less than 1.0 ppb). Nevertheless, contamination of semi-carbazide in wild caught shrimp and in ingredients is still found as well as nitrofuran contamination in imported shrimp.

Potential hazards in aquaculture fish

Hector M. Lupin
FAO Fishery Industries Division, Rome, Italy

INTRODUCTION

Fish production from aquaculture reached 29.1% of total fish production by weight in 2001 (FAO, 2003). Worldwide aquaculture has been growing at an average compounded rate of 9.2% since 1970, exceeding the growth rate of capture fisheries (1.4%) and farmed terrestrial animals (2.8%) in the same period. The most impressive growth in aquaculture production has occurred in Asia, and in particular in China; per capita supply of aquaculture fish outside China has also increased from 0.6 kg in 1970 to 2.3 kg in 2000. Different forecast scenarios analysed by FAO agree on the continued increasing contribution of aquaculture to total fish supply in the coming decades, with probably not less than a doubling of the current volume of production by 2030 (FAO, 2003).

There is no doubt that fish aquaculture is a booming industry. However, development of aquaculture can be hindered in practice by a number of factors such as water and area availability, environmental impact, limitation on feed production (e.g. fish meal availability) and lack of safety or perceived lack of safety in the context of a "risk society". In particular, with respect to fish safety, the knowledge of potential and relevant hazards that may affect aquaculture fish and fish products is essential in order to develop risk analysis schemes that could ensure consumer confidence and allow for the responsible and sustainable development of aquaculture.

A Joint FAO/NACA/WHO Study Group which met in Bangkok, Thailand, in July 1997 produced a report on the food safety issues associated with aquaculture products (WHO, 1999). This report puts aquaculture hazards under a more international perspective and identifies a number of areas where hazards in aquaculture are of importance for consumers in developing countries, such as in the case of parasites in freshwater fish. The report proposes a strategy to achieve safety assurance in aquaculture fish, based on risk assessment and application of the HACCP system to production. The report also identifies a number of knowledge gaps and research needs; part of the objective of this paper is therefore to attempt to identify research needs related to new or emerging hazards in aquaculture. A further purpose of this paper is to contribute to the discussion of a strategy to perform a hazard analysis for aquaculture products at production level.

GENERAL CHARACTERISTICS OF AQUACULTURE THAT COULD AFFECT RISK MANAGEMENT

The number of cultured fish species reaching the market is increasing. In addition to classic species such as carp, tilapia, trout, salmon and shrimp, additional species have been added, such as sea-bream, sea-bass and mussels in Europe and catfish in the United States. Many other species, ranging from turbot to octopus, are in different stages of development to obtain a market share. A recent study indicates that in Europe alone there are more than 50 different species produced at commercial level or are in a position to pass to commercial production (Basurco, 2003). It has been estimated that by the year 2000, there were 210 different farmed fish and plant species already in production worldwide. Of these farmed species, 131 were finfish species, 42 mollusc species, 27 crustacean species, 8 plant species and 2 amphibian and reptile species (Subasinghe, 2000). In addition, it has been estimated that about 35 fish species have so far been genetically modified (OECD). Table 1 presents the increase in the number of cultured fish species in the different regions between 1970 and 2000.

TABLE 1
Increase in the number of species cultured in the different regions (1970-2000)

Source: Subasinghe, 2003

TABLE 2
Production and economic importance of the different groups of cultured fish and aquatic plants

Source: FAO, 2002

The number of cultured species is important because they may present different types of possible hazards (as an extreme example we could mention the production of puffer fish in China for both internal consumption and export). The number of cultured fish species will surely increase, since there are markets for many different species of fish, rather than one single fish market. The point here is that once a specie is sold on the market, it should have its own hazard analysis, regardless of its trade volume and economic importance. This is a point of practical importance, though very often neglected in literature.

Important aspects to understanding certain safety developments in aquaculture fish are production volume and economic importance. The current volume of production and economic importance of the different groups of cultured fish and aquatic plants can be seen in Table 2.

Freshwater fish, mainly consumed in the country of production, incur relatively little interest in terms of hazard analysis research due to its small impact on international markets. Nevertheless, since, in large areas of the world, aquaculture fish is consumed raw, particularly in SE Asia, the risk and severity associated with the hazard of trematode infections is paramount (WHO, 1995). Crustaceans (mainly shrimp), which constitute only a fraction of the volume of production (3.6%), and in terms of absolute value represent 16.6% of the total, have a comparatively large visibility in terms of hazard research. The guiding criteria for this seems to be the unitary price rather than the absolute value and the fact that most of the farmed shrimp produced in developing countries are exported and consumed in developed countries.

An important aspect which can be noted from Table 2 is the importance of cultured molluscs, both in weight terms and in absolute value. International trade of live bivalves, particularly from developing countries, is limited when compared with other fish trade. However, molluscs, and in particular bivalve production, is increasing at a rate similar to that of seaweed and aquatic plant production and is second to fish (FAO, 2003). Again, this importance is not totally reflected in hazard analysis literature; for instance, the public health problems with V. parahaemolyticus and Hepatitis A recorded in some Chinese regions that might be linked to bivalve consumption (INFOYU, 2003). In analyzing possible hazards in aquaculture, it is necessary to avoid the temptation to concentrate only on hazards that may affect current international fish trade.

The existence of many different production systems (micro-environments) that are utilized in practice (e.g. extensive, intensive, semi-intensive, re-circulating systems, etc.), which in turn may be related to different macro-environments of sea water, fresh water and brackish waters, are a field open to many possible hazards. There is no standardized production, and the same fish specie could be produced following different production methods; for example, intensive and extensive fish farming, integrated or not, in floating (or bottom) cages at sea, in ponds on land (but with flowing sea water) and in re-circulating water systems, and so on. It is therefore possible to expect hazards to originate both from the macro-environment (e.g. water sources) and from the micro-environment (e.g. water management, chemicals utilized, etc.), and one cannot generalize the occurrence and relevance of these hazards.

Fish feeds are of utmost importance in many aquaculture systems. In the case of intensive farming, different feeds might be utilized as well as different veterinary drugs. However, drug residues might also appear, for instance, in "extensive" farming products if the fish or shrimp ponds are "integrated" to intensive chicken or pig production farms. Feeds could also introduce hazards other than drug residues in cultured fish through contamination of fish or plant meals. Some hazards may occur naturally while others are man made. The hazards that may be introduced by feeds are manifold.

There are no internationally agreed aquaculture definitions even for basic terms like "intensive" or "extensive" aquaculture, which in turn makes it additionally difficult for the fish safety expert to systematize possible hazards.

A SUGGESTED APPROACH TO HAZARD ANALYSIS IN AQUACULTURE

A large number of potential and relevant hazards are already identified in regulations and/or guidelines. Those hazards are, in general, either post-harvest hazards, and/or hazards specifically linked to aquaculture conditions, e.g. hazards related to water utilized in aquaculture. Two examples of these regulations and guidelines are the FDA Fish and Fisheries Products Hazards and Control Guidance (FDA, 2001b) and the Council Directive 79/923/EEC of the European Union on the quality required for shellfish waters (EU, 1979). Some examples of these hazards are given in Table 3.

Some cultured fish species such as salmon are already well established in terms of production methods, thereby making it easier to in turn develop a specific hazard analysis starting from production. However, the same is not true for most farmed species. Well established procedures for a given specie in a given situation cannot necessarily be extended to different situations for the same specie, not to mention for different species in different situations. This paper will not review the hazards presented in Table 3, but will attempt to analyse whether current regulations and guidelines, though initially for processing conditions, are useful to identify the hazards when related to production conditions. A possible way to classify and study potential hazards in aquaculture, related primarily to production conditions, could be to group them as follows:

(i) Common (known) post-harvest hazards. These are the hazards resulting from handling, processing and packaging (post-harvest in general) and are usually already included in the HACCP hazard analysis step.

(ii) Hazards due to macro- and micro-environmental contamination and/or conditions. The most important are certainly those linked to water sources, including contamination during normal operation or under exceptional situations (floods, storms). Potential hazards that may build up within the same system (micro-environment) should be included in this category.

(iii) Hazards due to the contamination of fish feeds. There is sufficient experimental evidence that feeds may be the conveyor, in particular of chemical hazards, which later accumulate in fish.

(iv) Hazards due to the use of veterinary drugs. Direct hazards as residues in fish, and indirect as development of microbial resistance to antibiotics.

TABLE 3
A classic view of possible and relevant hazards in aquaculture

Notes:

(1) Difference between "natural" and "contaminant" could be misleading, particularly for hazards that may come from the environment (because are naturally there) or because their levels are particularly high at certain locations due to man made contamination.

(2) It is better to split the hazards associated with veterinary drugs, since they are of a different nature.

COMMON (KNOWN) POST-HARVEST HAZARDS

Common (known) hazards appear at first sight as the most easy to identify and deal with. A typical concern in aquaculture is, for instance, the possibility of localized contamination (e.g. heavy metals, chemicals) due to water contamination, floods, etc. There may, however, be situations that are not due to "accidents", such as harvesting methods and practices which could increase the prevalence and concentration of pathogens.

The existence of biological contamination from a natural aquatic environment (exposure to open air, soil, winds, etc.) cannot be ruled out, even when appropriate safe aquaculture practices are in place. In this case, there is a need to recognize a hazardous step, i.e. the passage from the live animal to animal food. When the live fish (including shrimp and bivalves) is removed from the water it is exposed to a different environment (lack of water, increase of temperature^[4]), making it subject to stress. Even if the fish (or shrimp) continues to live, bacterial counts change rather quickly, particularly in warm climates.

Live fish gives the perception of safety, but this can be a wrong assumption, particularly when related to fish outside its natural environment in conditions of increased temperature, and/or packed in stressful conditions. Delayed icing increases pathogen counts during shrimp harvest under tropical conditions (MIP, 2000), even if the shrimp continues to live. It has also been demonstrated that a delay in icing increases the level of V. parahaemolyticus in oysters which, in turn, noticeably increases the risk for consumers of the raw oysters (FDA, 2001a).

Whereas there is no generalized data as yet, existing references make the passage of live fish to fish as food a possible candidate for a CCP (Critical Control Point) for pathogen contamination in aquaculture products, in particular small fish, like bivalves and shrimp. This CCP would be in the pond at the time of harvesting, and from a general FSO (Food Safety Approach), it would be more effective than the CCP at plant reception for the same purpose.

HAZARDS DUE TO MACRO AND MICRO-ENVIRONMENTAL CONTAMINATION AND/OR CONDITIONS

Water is perhaps the first source of concern for any specific type of aquaculture production method and fish specie, and as such is addressed in most regulations. Chemical and biological hazards that might affect fish through water are manifold, due to contamination (e.g. pesticides) or endemic (e.g. V. parahaemolyticus in sea water); under steady state conditions (e.g. pathogenic E. coli strains due to the proximity to inhabited areas) or due to seasonal or unusual events (e.g. floods, biotoxins).

Biological hazards

Currently the control of most of these macro-environmental hazards is accomplished through water monitoring by means of indicators, but there is evidence that this risk management procedure may not always be effective in practice to control some hazards, such as the potential risk of enteroviruses and hepatitis A virus in mussels - and likely in other bivalves - as discussed by Romalde et al. (2002) and Croci et al. (2003).

Some biological hazards could build up in water within the aquaculture systems (micro-environment). The most common could be the formation of biofilms that may protect pathogenic bacteria (eventually increasing prevalence or concentration) and/or act as a place for spreading microbial resistance to antibiotics. Whereas this type of hazard is addressed by the periodic cleaning of ponds and equipment, done for fish health reasons, there may be a need to analyze such a procedure with a view to ensuring human public health as well.

Parasites

The problem of parasites associated with aquaculture that might represent a hazard to humans, in particular trematodiases, nematodiases and cestodiases, has been discussed in some detail (WHO, 1999). Another interesting example of parasitic hazards is Cryptosporidium spp. In particular, Cryptosporidium parvum, an oocyst-forming apicomplexan protozoan, is an obligate intracellular parasite that infects the epithelium in the gastrointestinal tract of humans and various animal hosts. Cryptosporidiosis is generally a self-limiting disease, with a high degree of morbidity and a low rate of mortality (the latter related, in general, to vulnerable populations). Children in developing countries are the most exposed to this parasite. Cryptosporidium is currently reputed as an emerging hazard in a number of foods, in particular raw or improperly cooked molluscs (Millar et al., 2002).

Relatively recently, it has been found that sewage treatment plants are not effective to eliminate Cryptosporidium oocysts (Cataldo et al., 2001). In addition, it has been found that regulatory indicators of contamination used in the EU were not effective in detecting the oocysts, and current waste-water treatments were ineffective in preventing their passage into the water environment (Bonadonna et al., 2002). Up until 2002 no epidemiological record of human cryptosporidiosis associated with molluscs had been recorded (Millar et al., 2002). However, a number of positive detections of Cryptosporidium oocysts have been recorded between 1997 and 2003 in clams, mussels, oysters and cockles in different EU countries, as well as in Canada and the USA (Millar et al., 2002; Gomez-Couso et al., 2003).

Cryptosporidium oocysts are good candidates to become a relevant hazard in aquaculture of molluscs, both fresh and seawater species, throughout the world for the following reasons: (i) an infective dose is relatively small, as few as 10 viable oocysts (Okyusen et al., 1999); (ii) the prolonged survival up to one year of oocysts in seawater, retaining their infectivity when filtered by benthic mussels (Tamburrini and Pozio, 1999); (iii) chlorination alone has not been successful in eliminating waterborne Cryptosporidium oocysts (as they are resistant to the biocide activity of free chlorine); and (iv) there is no effective antimicrobial treatment to eradicate this parasite from the gastrointestinal tract of symptomatic individuals (Millar et al., 2002).

Chemical hazards

Biotoxins

In scientific literature, in addition to those "classic" biotoxins listed in Table 2, a large number of biotoxins associated with molluscs (e.g. conotoxins) appear; however, their importance from the point of view of food outbreaks in humans may be rather limited, even though they could have local importance. As a matter of fact, there are epidemiological indications that some toxins associated with molluscs are of public health importance at a national and/or regional level, such as the case of venuripin shellfish poisoning in Japan (usually associated with oyster and clams) that produces fulminant hepatic failure (33% of deaths and the callistin shellfish poisoning associated with the Japanese callista clam.

With the growing importance of freshwater bivalve culture, in addition to the classic hazard represented by human pathogens, there is also the need to search for the possibility of biotoxin production in freshwater environments. The production of PSP by freshwater cyanobacteria has been already ascertained (Lagos et al., 1999; Pereira et al., 2000; Saker et al., 2003), as well as the accumulation of paralytic shellfish poisoning (PSP) from cyanobacteria in freshwater mussels (Negri and Jones, 1995). The PSP hazard is, therefore, not only related to marine species.

In addition to the possibility of biotoxin hazards due to macro-environmental conditions as discussed in the previous paragraphs, there is a less known potential source of micro-environment production of biotoxins at pond level due to phytoplankton and harmful algal blooms, originated in turn to faulty fertilization and feed composition and rate. A number of harmful dinoflagellates and cyanobacteria species have by now been identified, as well as toxic substances, such as the saxitoxin family (with PSP and PSP-like symptoms), venerupin and cyanobacteria toxins (Alonso-Rodriguez et al., 2003) formed from these species.

Whereas a number of these toxins are also toxic for fish species which act, in practice, as an indicator of the hazard, they accumulate in fish and may constitute a hazard for humans in general or to certain populations at risk. The response of the cultured fish to the toxin may not necessarily be a good indicator of the effects of the same toxin on humans.

Toxic substances

Toxic chemical hazards that might appear due to water, soil and even air contamination are a classic field of study and do not need to be reviewed here. However, it is necessary to point out that there may also be a possibility that the toxic substance is formed inside the fish farm. An example of this type is the possibility of bromate formation in sea water recirculating systems during the oxidation of naturally-occurring bromine by ozone (Tango and Gagnon, 2003).

HAZARDS DUE TO THE CONTAMINATION OF FISH FEEDS

Hazards due to the contamination of fish feeds constitute a large chapter of potential hazards in aquaculture. Feeds tend to concentrate hazards, in particular chemical hazards (mainly POPs - persistent organic pollutants - and heavy metals), which in turn accumulate in fish. For example, research on farmed salmon has consistently found higher levels of PCBs (polychlorinated biphenyls), PBDEs (polybrominated diphenyleters) and OPs (organochlorine pesticides) than in wild counterparts. This in turn has been linked to the elevated contamination of such chemicals in salmon feed (Easton et al., 2002).

Feed should be considered in hazard analysis in very broad conceptual terms. For instance, in SE Asia, integrated fish farming combines intensive husbandry (chicken and pigs) with extensive aquaculture. In such cases, animal manure is excreted into the ponds and subsequently fertilizes the water; this supports the growth of photosynthetic organisms which in turn support fish growth. Whereas the procedure of utilizing animal manure in this manner does not actually represent fish "feed" in classic aquaculture terms, it is in effect fish feed, and unexpected hazards for extensive fish farming such as antimicrobial resistance could appear as a result of antibiotics fed to the chicken and pigs (Petersen and Dalsgaard, 2003).

Changes in the composition of fish feeds, in particular with the increased use of groundnut meals and cereals or other sources of vegetal protein and lipids, may present hazards of natural toxins of vegetable origin like aflatoxins, T-2 toxins, vomitoxin and some toxic natural feed components (e.g. oxalic acid, anti-vitamins, etc.). Again, some producers argue that cultured fish is in itself the control of contamination; however, this cannot be generalized mutatis mutandis or extrapolated to the possible effect on humans.

Even if in many countries there are already regulations related to animal feeds, and in particular fish feeds (see for instance CFIA, 2003), there is a growing consensus that feed production (for food animals, and in particular aquaculture fish) should be produced under HACCP plans (den Hartog, 2003).

HAZARDS DUE TO THE USE OF VETERINARY DRUGS

Drugs are utilized for different purposes in aquaculture. For instance, antibiotics could be used as therapeutants, as prophylaxis and as a growth factor. The regulatory and production tendency is to reserve the use of approved antibiotics only as therapeutants, avoiding their use in prophylaxis and as a growth factor. This tendency is compounded with a move to develop and utilize vaccines rather than antibiotics, as is already occurring in some countries with salmon production, and in general to reduce as much as possible the use of all veterinary drugs. There are basically three types of hazards due to veterinary drugs;

(i) Residues in fish of authorized drugs above the allowed limits;

(ii) Use of banned drugs for aquaculture (or not specifically authorized for its use in aquaculture); and

(iii) Development of pathogen strains resistant to antibiotics (permitted or not).

Whereas some hazards, e.g. residues of authorized veterinary drugs below MRLs (maximum residue limits), could be reputed to be very low, this does not automatically mean the same can be said for all drugs, in particular to the use of non-approved or prohibited veterinary drugs (Lupin et al., 2003).

In many countries, a non-approved veterinary drug for aquaculture is not exactly a banned drug, since it can eventually be utilized legally under an "off label" (or "cascade") scheme. Even if regulations are changing to limit that possibility, and single professionals are hardly in a position to assess the risk to humans embodied in most "off label" uses, the problem remains in countries where the dispensing of drugs in aquaculture is not done through professionals. The possibility of selling free, uncontrolled drugs, regardless of whether they are banned or permitted, is usually known as "over the counter" and is practiced in some countries. This was certainly at the root of problems experienced by some developing countries with residues of banned antibiotics (e.g. chloramphenicol) in exported shrimp from 2001 to 2003.

The regulatory scenario is also complex in other senses. For example, in some countries a veterinary drug can be legally banned only if it has been a previously approved drug. Some veterinary drugs utilized in aquaculture for many years, such as malachite green (that together with its metabolite leucomalachite green, are reputed to be hazards to humans), have never been a formally "approved drug" for aquaculture in many countries. This creates a rather confusing situation in legal and practical terms which, among other things, is not a good example for developing countries' aquaculture. Developed countries are introducing changes in regulations to prevent this type of situation, but this is not yet a worldwide tendency.

Antibiotics can be banned for use in food animals for two basic reasons. Firstly, residues might be toxic for humans (e.g. chloramphenicol) and, secondly, public health authorities have decided to keep such antibiotics exclusively for use in humans.

The increasing hazard of antimicrobial resistance in general (Harrison and Leederberg, 1998) and in relation to animal foods (NRC, 1999) is well known. The epidemiological impact of microbial resistance on human has, by now, been widely documented. The main problem continues to be to find a balance between responsible animal production and human health aspects; aquaculture is no stranger to this dilemma.

Whereas to assign responsibility for specific antimicrobial resistance to the use of a given antibiotic in a specific context could, in most cases, be elusive, resistance in pathogens associated with aquaculture has been identified in a number of countries and fish products (Radu et al., 2003; Petersen and Dalsgaard, 2003; Miranda and Zemelman, 2002; Castro-Escarpulli et al., 2003).

In particular, Aureli et al. (2003) reported on the resistance of Listeria monocytogenes isolated from food, including smoked salmon, in Italy. Zhao et al. (2003), authors associated with the FDA, reported on the antimicrobial-resistant Salmonella serovars isolated from imported foods to the USA, indicating that:

"This study indicates that antimicrobial-resistant Salmonella are present in imported foods, primarily of seafood origin, and stresses the need for continued surveillance of food-borne zoonotic bacteria from imported foods entering the United States."

All this seems to indicate that regardless of the difficulties in defining a neat chain of events (cause-effect) to perform quantitative risk assessments (e.g. in the way it can be done for a pathogen), antimicrobial resistance would have a regulatory impact on fish trade.

A lack of harmonization and transparency in the use of veterinary drugs in aquaculture is common place. This is, for instance, reflected in the fact that only one of the many antibiotics in use (tetracycline) has a Codex Alimentarius MRL, or from the fact that a different number of antibiotics are approved at country level by each EU member state. A large part of the current problem with the use of veterinary drugs, including pathogen strains resistant to antibiotics in fish (and in other food animals as well), is due to arguable practices and situations associated with the lack of appropriate regulations dealing with the sale and use of veterinary drugs and/or the possibility or impossibility of enforcing existing regulations. It could be said that the true hazard is not the use of veterinary drugs per se but from our incapacity as human beings to deal with veterinary drugs in a responsible and rational manner.

DISCUSSION AND CONCLUSIONS

Potential hazards of aquaculture fish are not necessarily always relevant hazards. However, to decide when a potential hazard is a relevant one continues to be a difficult task. Risk assessment, in principle, could allow us in each case to decide on the relevance of each hazard in a given situation, but neither all the necessary risk assessments are available nor the appropiate data to make them possible. The existence of epidemiological data could be another way of identifying a relevant hazard from a potential one, but again other aspects like severity or regionalism may play a role in practice.

The HACCP focus on hazards and relevance (risk/severity) is defined by regulations even if there is not always a risk assessment available. In the particular case of implementing HACCP in aquaculture production, it seems necessary to review the procedure to perform a hazard analysis. On one hand, there could be additional, new or emerging hazards that may appear; on the other hand, known hazards may become relevant or more relevant in a given aquaculture context. This approach seems necessary, in turn, to develop suitable strategies at risk management level. A general conclusion could be that, as it was said at the beginning of HACCP implementation, and perhaps even more applicable in the aquaculture situation, "each case is (or could be) a case".

Introduction of the HACCP system in fish aquaculture and feed production, as already recommended by a number of authors, seems a necessary step for the continued development of commercially based aquaculture. But more information and studies are necessary to determine, for instance, what could be controlled through GHP and GMP applied to aquaculture production, what would require of a full CCP approach, and how to improve the effectiveness of environmental monitoring methods.

The present paper indicates that a number of potential hazards in aquaculture products have been reported in literature since the publication of the Joint FAO/NACA/WHO Report (WHO, 1999), even if their relevance has not been ascertained in all cases. Existing aquaculture information and data should be revisited and systematically analyzed with a hazard analysis approach. In particular, it is necessary to keep in mind that useful information and data could come from very different sources and academic fields, as is reflected in the references of this paper. This multi-disciplinary effort most certainly has a role in the continued successful development of aquaculture and is the only way to obtain proper risk assessments.

REFERENCES

Alonso-Rodriguez, R. & Paez-Osuna, F. 2003. Nutrients, phytoplankton and harmful algal blooms in shrimp ponds: a review with special reference in the Gulf of California. Aquaculture, 219(2003):317-336.

Aureli, P., Ferrini, A.M., Mannoni, V., Hozic, S., Wedell-Weergaard, C. & Oliva, B. 2003. Susceptibility of Listeria monocytogenes isolated from food in Italy to antibiotics. Int. J. of Food Microbiology, 83(2003):325-330.

Basurco, B. 2003. International Centre for Mediterranean Agronomic Studies (CIHEAM). Mediterranean Agronomic Institute of Zaragoza (IAMZ); Zaragoza, Spain, personal communication.

Bonadonna, L., Briancesco, R., Cataldo, C., Divizia, M., Donia, D. & Pana, A. 2002. Fate of bacterial indicators, viruses and protozoan parasites in a wastewater multi-component treatment system. Microbiologica, 25: 413-420.

Castro-Escarpulli, G., Figueras, M.J., Aguilera-Arreola, G., Soler, L., Fernández-Rendón, E., Aparicio, G.O., Guarro, J. & Chacón, M.R. 2003. "Characterization of Aeromonas spp. isolated from frozen fish intended for human consumption in Mexico"; Int. J. of Food Microbiology, 84(2003): 41-49.

Cataldo, C., Briancesco, R. & Bonadonna, L. 2001. Water re-use: hygienic and technical aspects related to the occurrence of enteric pathogens. Rapporti Istituto Superiore de Sanitá (ISTISAN) 01/34, Italy (in Italian), 38pp.

CFIA. 2003. Animal Feeds. Canadian Food Inspection Agency. Animal Products. Animal Health and Production. www.inspection.gc.ca/english/anima/feebet/feebete.shtml.

Croci, L., De Medici, D., Ciccozzi, M., Di Pasquale, S., Suffredini, E. & Toti, L. 2003. Contamination of mussels by Hepatitis A virus: a public-health problem in southern Italy. Food Control, 14(2003):559-563.

Den Hartog, J. 2003. Feed for Food: HACCP in the animal feed industry. Food Control, 14(2003):95-99.

Easton, M.D.L., Luszniak, D. & Von der Geest, E. 2002. Preliminary examination of contaminant loadings in farmed salmon, wild salmon and commercial salmon feed. Chemosphere, 46(2002):1053-1074.

EU. 1979. Council Directive 79/923/EEC - 30 October 1979 on the quality required for shellfish waters.

FAO. 2003. The state of world fisheries and aquaculture 2002. FAO, Rome, Italy.

FDA. 2001a. Draft Risk Assessment on the Public Health Impact of Vibrio Parahaemolyticus in Raw Molluscan Shellfish. US FDA, Center for Food Safety and Applied Nutrition, 118pp. www.cfsan.fda.gov/~dms/vprisk.html.

FDA. 2001b. Fish and Fisheries Products Hazards and Controls Guidance. Center for Food Safety & Applied nutrition; US FDA, June 2001, 3rd Edition. www.cfsan.fda.gov/~comm/haccp4.html.

Gomez-Couso, H., Freire-Santos, F., Martinez-Urtaza, J., García-Martín, O. & Ares-Mazás, M.E. 2003. Contamination of bivalve molluscs by Cryptosporidium oocysts: the need for new quality control methods. Int. J. of Food Microbiology, 87(2003):97-105.

Harrison, P.F. & Lederberg, J. 1998 (Editors). Antimicrobial Resistance: Issues and Options. Workshop Report; Institute of Medicine. National Academy Press; Washington, D.C. 115 pp.

INFOYU. 2002. FAO/INFOYU/SHFU (Shanghai Fisheries University) National Workshop on Introduction to Risk Analysis in the Fishery Industries, Shanghai, P.R. of China, 12- 23 August 2002. Final report (in Chinese).

Lagos, N., Onodera, H., Zagatto, P.A., Andrinolo, D., Azevedo, S. & Oshima, Y. 1999. The first evidence of paralytic shellfish toxins in the freshwater cyanobacterium Cylindrospermopsis raciborskii, isolated from Brazil. Toxicon, 37(1999):1359-1373.

Lupin, H.M., Subasinghe, R. & Alderman, D. 2003. Antibiotic residues in aquaculture products; in "The state of world fisheries and aquaculture 2002". FAO, Rome, Italy.

Millar, C. B., Finn, M., Xiao, L., Lowery, C.J., Dolley, J.S.G. & Moore, J.E. 2002. Cryptosporidium in foodstuffs - an emerging aetiological route of human foodborne illness. Trends in Food Science and Technology, 13(2002):168-187.

MIP. 2000. FAO/MIP (Ministerio de la Industria Pesquera) National Workshop on Introduction to Risk Analysis, Havana, Cuba, 20-22 March 2000. Final Report (in Spanish).

Miranda, R. & Zemelman, C.D. 2002. Antimicrobial multi-resistance in bacteria isolated from freshwater Chilean salmon farms. The Science of the Total Environment, 293(2002):207-218

Negri, A.P. & Jones, G.J. 1995. Bioaccumulation of paralytic shellfish poisoning (PSP) toxins from the cyanobacterium Anabena circinalis by the freshwater mussel Alathyria condola. Toxicon, 33(1995):667-678.

NRC (National Research Council). 1999. The Use of Drugs in Food Animals. Benefits and Risks. Committee on Drug Use in Food Animals. Board on Agriculture (NRC) and Food and Nutrition Board (Institute of Medicine); National Academy Press, Washington, D.C. 253pp.

Okhuysen, P.C., Chapell, C.L., Crabb, J.H., Sterling, C.R. & DuPont, H.L. 1999. Virulence of three distinct Cryptosporidium parvum isolates for healthy adults. Journal of Infectious Disease, 180:1275-1281.

Pereira, P., Onodera, H., Andrinolo, D., Franca, S., Araujo, F., Lagos, N. & Oshima, Y. 2000. Paralytic shellfish toxins in the freshwater cyanobacterium, Aphanizomen flosaquae, isolated from Montargil reservoir, Portugal. Toxicon, 38(2000):1689-1702.

Petersen, A. & Dalsgaard, A. 2003. Antimicrobial resistance of intestinal Aeromonas spp. and Enterococcus spp. in fish cultured in integrated broiler-fish farms in Thailand. Aquaculture, 219(2003):71-82.

Radu, S., Ahmad, N., Ling, F.H. & Reexal, A. 2003. Prevalence and resistance to antibiotics for Aeromonas species from retail fish in Malaysia; Int. J. of Food Microb., 81(2003):261-266.

Romalde, J.L., Area, E., Sanchez, G., Ribao, C., Torrado, I., Abad, X., Pinto, R.M., Barja, J.L. & Bosch, A. 2002. Prevalence of enterovirus and hepatitis A virus in bivalve molluscs from Galicia (NW Spain): inadequacy of the EU standards of microbiological quality. Int. J. of Food Microbiology, 74(2002):1 19-130.

Saker, M.L., Nogueira, I.C.G., Vasconcelos, V.M., Neilan, B.A., Eaglesham, G.K. & Pereira, P. 2003. First report and toxicological assessment of the cyanobacterium Cylindrospermopsis raciborskii from Portuguese freshwaters. Ecotoxicology and Environmental Safety, 55(2):243-250.

Subasinghe, R. 2003. Inland Water Resources and Aquaculture Service (FIRI), FAO. Personal communication.

Tamburrini, A. & Pozio, E. 1999. Long-term survival of Cryptosporidium parvum oocysts in seawater and in experimentally infected mussels (Mytilus galloprovincialis). International Journal of Parasitology, 29:711-715.

Tango, M.S. & Gagnon G.A. 2003. Impact of ozonation on water quality in marine recirculation systems. Aquacultural Engineering, 29(3-4):125-137.

WHO. 1995. Control of foodborne trematode infections. Report of a WHO Study Group. WHO Technical Report Series 849, Geneva. 157pp.

WHO. 1999. Food safety issues associated with products from aquaculture. Report of a Joint FAO/NACA/WHO Study Group. WHO Technical Report Series 883; Geneva; 55 p.

Zhao, S., Datta, A.R., Ayers, S., Friedman, S., Walker, R.D. & White, D.G. 2003. Antimicrobial-resistant Salmonella serovars isolated from imported foods. Int. J. of Food Microbiology, 84(2003):87-92.

Good Aquaculture Practices for farmers - an update

Brett Koonse
United States Food and Drug Administration, College Park, United States of America

Consumers are more concerned than ever about the safety of the food they eat and demand that their food not be contaminated. Whether it is corn, eggs, apples, various meat products or farm-raised shrimp, food safety and quality truly begins on the farm.

GOOD AQUACULTURE PRACTICES

Aquacultured seafood currently enjoys a reputation worldwide as a safe and wholesome food. To maintain this status, researchers and investigators from the US Food and Drug Administration's Office of Seafood are leaving their offices to visit aquaculture farms worldwide. Their goal is to determine what preventive controls (Good Aquaculture Practices - GAPs) are necessary on farms to minimize the potential for farm raised fishery products to be contaminated with pathogens, filth and unapproved or mis-used animal drugs.

PROACTIVE APPROACH

To emphasize the need for GAPs, it is fair to say that if they were established and implemented worldwide, the aquaculture industry could have avoided the current situation with aquaculture drugs. There would be a worldwide standard for approved drugs and the proper documentation requirements.

GAPs are intended to not only be proactive in reducing the current potential problems of aquacultured products being contaminated with Salmonella and filth, but also to prevent the next surprise issue that will inevitably come around.

COLLABORATIVE RESEARCH PROJECT

As background on the GAP project, the FDA is conducting a collaborative research project with the aquaculture industry and the government agencies of seven other countries to determine what the microbiological, chemical and drug use hazards are on aquaculture farms, and how to best control these hazards.

The next step will be to use this information to develop and present training for farm workers as a proactive measure to ensure aquacultured seafood continues to be recognized as a safe wholesome food.

GAPs will only work if they give aquaculture farmers practical and reasonable controls to help them minimize the occurrence of contamination in their products. That is why the FDA is not sitting in their offices or research laboratories, but is proactively developing GAPs.

STUDY SUPPORT AND APPROVAL

There is currently universal support and approval for the collaborative study from the international aquaculture industry, government agencies, processors and importers. In fact now that word has spread, FDA has been overwhelmed with requests from aquaculture associations and governments to do studies in their countries. Several countries have even offered to pay FDA's expenses if they would come to do a study in their country.

The reason for this strong support by the aquaculture industry is simple: this industry clearly understands that it is receiving critical advice and cutting edge recommendations on how to directly and substantially improve the likelihood that its products are safe and wholesome. Each farmer that has participated in this study has benefited directly through the interaction with FDA and other researchers and we all have gained new friends and a better understanding of each other.

Thus far, farms in three countries have been studied. Approximately 100 farms have been sampled and their respective farm managers interviewed. The overwhelming observational and analytical evidence to date shows that there are reasonable and practical solutions farmers can implement to reduce faecal coliforms and the potential of Salmonella on their farms. The two most important solutions are a holding pond to reduce the number of faecal coliforms (and, consequently, the chance for Salmonella) in the source water and protecting ponds from animal and human activity.

UPCOMING REPORT

A factual report on the possible sources of microbiological and chemical contamination at aquaculture sites is expected in the fall of 2003. "Train the trainer" education for farmers will follow and will be a collaborative effort by the US FDA, other country government agencies, academia, GAA and, of course, the aquaculture industry that is heavily involved in this effort worldwide.

CONCLUSION

Perceived safety, wholesomeness and the nutritional value of a food product greatly influence consumer-buying decisions. The entire aquaculture industry would benefit from widespread producer commitment to implementing GAPs. Implementing GAPs on every farm will help the industry protect - and even expand - its markets by reducing the risk of incidents that could erode consumer confidence in the safety, quality and wholesomeness of aquacultured products.