3. Border cases in the European Union, North America and Japan

3.1 Introduction

Each year the major importing countries of fish and fish products reject or detain imports for a variety of reasons. Indeed, the previous chapter outlined the regulations and procedures governing the import of fish into the three largest fish importers -European Union, North America and Japan.

In most cases, the statistics held for these detentions or rejections have not been easy to come by unless you had contacts with the various authorities responsible for food or fish inspection. Where the data are available, they are often similar, but not identical, in terms of the type of information collected or the classification of the causes for rejection/detention. Importantly, the ability to access these data easily as a member of the public (or industry) varies considerably between importing countries. This latter issue is discussed in some detail in Chapter 4, including recommendations as to how to improve the situation.

On a global scale, it would be nice to be able to present a table summarizing the total number of detentions and rejections of food at borders worldwide, but this is not possible as the data available are still limited.

The rest of this chapter considers fish and seafood imports into the European Union, United States of America, Japan and Canada markets. Each market is considered separately, laying out the facts on border cases, and also putting the border cases into perspective as regards import volumes. It analyses the data for trends or patterns in rejections/detentions across a range of parameters - problem type (microbial, chemical, other causes), species, geography and process type (fresh, frozen, cured, etc.). The case posed by aquaculture products is covered and the chapter also specifically examines the scientific basis for rejection/detentions.

3.1.1 Data sets used

There are two main sets of data used in this document - border case data and import data. Currently, border case data are not held centrally anywhere, and thus data have to be sourced from the importing countries themselves. Import data are held centrally by FAO and also by the countries themselves. However, a single data set for imports is not always examinable from both a product basis (i.e. imports broken down by products) and a country basis (i.e. imports broken down by exporting country).

The study covers the period 1999 to 2002 for which attempts were then made to get the border case and import data from the four countries/regions considered.

Border case data

As highlighted earlier, the types of data and the periods covered differed from one country/region to another. Thus, for the European Union and Canada, detailed line-by-line data for the border cases from 1999-2002 was compiled in spreadsheets with subsequent breakdown by risk category, products and exporting regions.

The data collected for the United States of America cover a two-year period from mid-2000 to mid-2002, but line-by-line data were not available.

The data collected for Japan were more varied. Publicly available data (Web based) were restricted to an annual summary for 2000 for foods in general. However, border case data for fish were obtained from the Japanese authorities directly but were restricted to two periods - April 2000/October 2001 and November 2001/0ctober 2002. For ease of comparisons, the 19-month period was averaged to provide a monthly figure and then aggregated to give a 12 month estimate. Again, the level of detail was not as good as for the European Union and Canada with only summary data available.

As has been noted in the previous chapter, the latest border case data is available on the internet for the European Union, United States of America and Canada. An annual summary for all foods (for 2000 at the time of writing) is available for Japan. Example extracts of the information available on the internet can be found in Annex A. 16 for the European Union, Annex A. 17 for the United States of America and Annex A. 18 for Canada. However, though data is available publicly, and this should be applauded, there are limitations with regard to the data fields recorded, the ease of extraction of the data for later analysis and access to archived information.

Putting border cases into perspective

In the present study, border case data were analysed considering two issues: impact on consumer safety and international economic and trade implications, with the understanding that protection of consumer safety cannot be compromised by any consideration whatsoever. Absolute number of border cases were used to assess the extent of fish safety and quality aspects, whereas relative comparisons between exporting and importing regions, or between risk categories and products types used border case data weighted in relation to the volume of trade.

For instance, if Region 1 exports to Region 2 and ends up with 100 border cases in Region 2, and Region 3 exports to Region 2, and ends up with only ten border cases, then it is inappropriate to say that Region 1 is performing poorly compared to Region 3 without knowing the quantities of imports involved. Thus we have used the volumes of trade in tonnes to provide a comparative figure. This is a crude figure, as the border cases do not indicate how many kilograms were involved in each border case - so a "case" could involve a shipment of 100 kg or 10 tonnes. When the latter were known, a rough calculation was made to estimate the total value of border cases to trade disruption. However, the number of border cases per unit volume does give an indication of the relative importance of various factors in border cases. This calculation is not possible in all circumstances due to data gaps.

3.2 European Union

In the European Union in 2003^[9] there were a total of 4 286 notifications of food related problems from internal production and imported foods. This rose from 3 024 in 2002. This is for all food and feed products and represents a 41.7 percent increase over 2002. Of these, 454 were alert notifications and 1 856 were information notifications (see Chapter 2.2 for an explanation of the differences).

Significant specific notifications during the year included aflatoxins in nuts (763), cadmium (103) and mercury (24) in fish (mostly swordfish), industrial dyes in chilli powder (119) and dioxins (26) mostly in animal feeds. Table 11 breaks down the notifications (which are mostly information notifications and these are mostly border cases of imported foods). As noted above, a large number of notifications were due to mycotoxins (95 percent were aflatoxins) in nuts. Of the remaining categories, chemical contaminants dominate (pesticides, veterinary drugs, heavy metals and others) with microbial contamination the next most important group.

TABLE 11
Information notifications in the European Union (EU) according to categories of source of contamination - 2003

Source: European Union RASFF Report 2003.

TABLE 12
Residues of veterinary medicinal products - 2003

Source: EU RASFF Report 2003.

TABLE 13
Information notifications in the European Union (EU) according to the food groups involved, 2001-2003

Source: EU RASFF Report 2003.

Veterinary drug residues were a special case. Table 12 gives a breakdown of the drug or metabolites and the food associated with the contamination, and it is apparent that fish and crustacea dominate as the carrier for these contaminants.

The main culprits as regards food groups were fish, crustacea and molluscs followed by nuts. The other main food groups responsible for notifications are shown in Table 13. It is very evident that the total numbers have also been increasing over the period quite significantly and across all the food groups.

Of course, care has to be exercised in taking too much from this data, as the import volumes can vary significantly, and thus as a percentage of trade, some of these groups may increase or decrease in importance. This being said, border cases impacting on consumer safety should be considered in absolute terms, especially in relation to the severity of the hazard.

3.2.1 Imports of fish and fish products

The European Union is a huge importer of fish and fish products. This is a necessity as total community production of fish, from both capture and farmed methods, falls way short of the demand in the member states. Over the last few years the European Union has imported around 3.5-3.8 million tonnes annually from all countries outside of the European Union. Intra European Union trade amounts to around 3 million tonnes annually.

The main exporter to the European Union is Norway with around 600 000 tonnes annually, nearly three times the next major exporter (in 2002), the United States of America. The trend for Norwegian exports, however, has been downward over the four- year period, 1999 to 2002. United States exports to the European Union on the other hand have increased over the same period, with the United States now ranked second in 2002 from 5th in 1999. China had a dramatic drop in 2002 after quite significant increases from 1999 to 2001. This was due to a ban on imports from China because of chloramphenicol contamination (European Union Decision 2002/69).

The total imports are also broken down by continent (Table 15), as this allows a later comparison with the border cases from these same regions. The most important exporting region was Europe (non EU), which accounted for some 36-40 percent of imports into the European Union. The next most important exporting regions were Africa, Asia and Central and South America.

Table 16 breaks down the imports into species groups and into main product types, again for comparative reasons when examining the border case data. The data also only cover the period from 1999 to 2001, as the FAO statistics provide this breakdown, and 2002 data were not available at the time of writing. The data also cover both extra and intra European Union trade. Annex A. 19 defines the product types (prepared, processed, etc.) according to EC Directive 91/493.

The dominant species group imported was fish, followed by cephalopods, shrimp and molluscs. The main fish species imported are canned and frozen tunas, fresh and frozen salmon and ground fish (cod, Alaska pollack, hake), in a mixture of forms (frozen, chilled, salted). Pelagic species are also important (mackerel, herring, sardines).

TABLE 14
Top ten exporters to the European Union 1999-2002 (2002 basis) (tonnes)

Source: European Commission.

TABLE 15
Total European Union imports by exporting continent 1999-2002 (tonnes)

Source: European Commission.

TABLE 16
Total European Union imports by product type and species group 1999-2001 (tonnes)

Source: FAO. Note that for some products several types, e.g. prepared and frozen, are used to categorize the product.

3.2.2 Border cases

General

From 1999-2002, the European Union recorded just under 900 border cases for fish and fish products reaching its borders. However, the distribution is skewed, with just under 50 percent of cases occurring in 2002 (Table 17). The reason for this is a dramatic increase in chemical risks that occurred in 2002. This will be discussed in more detail later. It is also possible that some non-reporting occurred during the first years as European Union member states were starting to implement the RASFF.

A detailed data set is available for the four years that allows a more detailed breakdown of the border cases. Table 17 shows the border cases by year (1999-2002), by risk and by exporting region.

The main exporting region that gave rise to border cases in the European Union was consistently Asia, varying from around 50 to 75 percent of cases each year with no trend evident. Africa and central and south America were the next main exporting regions to have problems with products imported into the European Union. There is a notable trend of decreasing relative importance for exports from Africa dropping from 34 percent of cases in 1999 to just over 11 percent in 2002. Overall, for the period from 1999-2002, Asia accounted for 66 percent, Africa for 18 percent and central and south America for 11 percent with these three regions accounting for 95 percent of border cases.

These figures do not take into account the volume of imports from these respective regions. A later section in this chapter puts the border cases into perspective, comparing the border cases with the volume of exports from each region.

For the European Union, it becomes apparent that, until 2002, the dominant cause of border cases was microbial in origin. Chemical risks though were becoming more important and by 2002, chemical risks dominated. It is worth noting that "other causes" only played a small role in the cause for border cases throughout the European Union.

Microbial risks

The microbial risks that caused most problems at European Union borders were Vibrio spp. and Salmonella accounting for around 66 percent of cases between them (Table 18). There were no significant rises or drops, except, perhaps, in the appearance of parasites as a cause in 2001 and 2002. However, it is interesting to note the complete absence of Listeria spp. as a cause for border cases. This is probably because the EC and several European Union countries, are not supportive of zero tolerance stance, do not have specific control programmes for Listeria monocytogenes.

TABLE 17
Border cases in the European Union from 1999 to 2002 by risk and exporting region

Source: European Rapid Alert System.

TABLE 18
Border cases in the European Union from 1999 to 2002 - microbial risks

Source: European Rapid Alert System.

It is important to note that the only harmonized microbial criteria in the European Union so far are for cooked crustaceans and molluscs and live bivalve molluscs, as noted in the previous chapter (see Annex A.4). For all other fish and fish products, the individual member states use their own criteria for the common indicator and specific bacteria, with France and Spain having the most detailed requirements for various products, but quite different in terms of the product groupings, and the criteria used (Eurofish, 1998). This causes a confusing picture for exporters who may be exporting to different countries within the European Union. This has been recognized by the EC which has initiated efforts to harmonize the microbial standards of fish products. These interim draft standards and guidelines are presented in Annex A.4. However, these efforts need to be expanded using scientifically based risk assessment for the products in question.

Vibrios provide an interesting case in this respect. Although the EC has no harmonized criteria for Vibrios yet, they are clearly being tested for on a regular basis. In examining the data, it is evident that mainly two members (Italy and Norway) of the Rapid Alert System are responsible for 75 percent of the Vibrio notifications from 1999-2002, with 75 percent of cases due to frozen shrimp. Indeed, the major shrimp import markets of the UK and Germany did not notify at all during this same period. This shows major differences within the European Union towards the testing of Vibrios and presents a confusing picture for exporters. The data does not allow determination of the criteria used for the notifications, so care must be taken in drawing conclusions as regards testing criteria used. However, products are being removed from trade based on Vibrio counts at some level. This contrasts with recommendations from the "EC Expert Opinion on V. parahaemolyticus and V. vulnificus in raw and undercooked seafood" (EC, European Commission 2001) which states the following:

• "The practice of judging seafood exclusively based on total Vibrio counts as indicative for the presence of pathogenic vibrios is not appropriate and should be discontinued.

• The practice of judging seafood exclusively based on total V. parahaemolyticus counts without consideration of the virulence factors TDH/TRH (or tdh/trh) is not appropriate and should be discontinued.

• Currently available scientific data do not support setting specific standards or microbiological criteria for pathogenic V. parahaemolyticus and V. vulnificus in seafood. Codes of practice should be established to ensure that GHP has been applied."

This question is elaborated upon further in this chapter (section 3.6.7).

Chemical risks

As regards the chemical cases, some notable trends appear (Table 19). Prior to 2001, the main risks were from heavy metals, with mercury and cadmium accounting for nearly 70 percent of border cases in 1999 and 2000. However, in 2001 and 2002, two new chemical agents, chloramphenicol and nitrofuran, appeared dramatically in border cases. In 2001 and 2002, these two chemicals accounted for over 65 percent of the border cases, with the heavy metals now accounting for only 14 percent of cases.

This increase in these two veterinary drugs is due to rigorous testing regimes imposed in 2001 and 2002 on shrimp (and other food) imports from various southeast Asian countries by the European Union. China was most affected as Commission Decision 2002/69 suspended the import into the European Union of Chinese products of animal origin intended for human consumption or for use in animal feed. The main products affected by the suspension in volume terms were honey, rabbit meat, poultry and crustaceans such as shrimps and prawns. During this period, Viet Nam, Thailand and Pakistan were also requested to submit each seafood shipment for analysis for chloramphenicol and nitrofurans.

Later in 2002, the European Union decided to lift the import restrictions due to guarantees by exporting authorities and the results of further tests. It is worth noting that similar stringent testing regimes were imposed by other importers, such as the United States of America, Canada and Japan. However, a significant difference was the lower limit of detection of the analytical method used, which was very low for the European Union and Canada (less than 1 part per billion (ppb)) as compared to 5 ppb level for the United States of America. Later, the United States detection limits were lowered to be more in line with the European Union and Canadian limits.

TABLE 19
Border cases in the European Union from 1999 to 2002 - chemical risks

* Biphenol A diglycidyl ether - a component in plastics manufacture.
Source: European Rapid Alert System.

Other causes

Other causes of border cases in the European Union only averaged at around 6 percent of incidents over the period studied, although the percentage varied between 4 and 12 percent year on year. Of the other causes for border cases, the main problem, and pretty consistently each year, was with the certificates that accompany imports, accounting for just over half of the cases (Table 20). From the data it is not possible to discover which certificates are involved in these incidents. It is likely that the problems centre on the health certificate. This is an absolute requirement and is prescribed by the European Union for all fish and fish product imports and the certificate is generally known to be examined in some detail. Given that these certificate problems only account for around 2-3 percent of all border cases, this is not a big concern for developing countries, and is one that is easily remedied should it become an issue for a particular country.

The other two types of problems were sensory tests revealing quality problems or with a temperature abuse of the imported product, for instance a thawed frozen product or a problem in the time-temperature history.

Border cases by product and species

Table 21 breaks down the border cases by product type and species group for the period 1999-2002. The predominant forms of product causing border cases are frozen and prepared products. It is interesting that for the European Union, canned and processed fish does not constitute a major problem for European Union importers, and the fresh and cured fish accounts for only around 5 percent of border cases during the four-year period and that caviar accounts for 2 percent.

The main problem species group is shrimp, not surprisingly given that they were the species responsible for the chloramphenicol and nitrofuran border cases that dominated in 2001 and 2002. Fish and cephalopod come in second and third. Other groups (crab, lobster and bivalves) account for only 7 percent of border cases. However, it will be interesting to note the relative frequency when trade volumes are considered in the next section.

A final analysis that is possible with the data collected for the European Union is to determine the causes for border cases according to the species or products imported (Table 22). Not surprisingly, histamine is restricted to fish species, and is mostly found in canned fish imported into the European Union. Shrimp border cases are split fairly equally between microbial and chemical causes, but, as discussed earlier, veterinary drugs account for a huge rise in border cases in 2001 and 2002, and if viewed by year, the dominant cause in 1999-2000 is microbial and chemical for 2001-02. Fish species also account for the bulk of the "other causes" detailed earlier in this section.

TABLE 20
Border cases in the European Union from 1999 to 2002 - other causes

Source: European Rapid Alert System.

TABLE 21
Border cases in the European Union from 1999 to 2002 - product types and species groups

Note that for some products several types e.g. prepared and frozen, are used to categorize the product and for some none are used.
Source: European Rapid Alert System.

TABLE 22
Combined border cases in the European Union from 1999 to 2002 - by cause and product type/ species

Note that for some products several types e.g. prepared and frozen, are used to categorize the product and for some none are used.
Source: European Rapid Alert System.

3.2.3 Border cases in the European Union in the context of import volume

Comparative studies of border cases between exporting and importing regions, or between risk categories and products types, etc. require the use of a figure that allows a relative comparison. Thus, we have used the volumes of trade in tonnes to provide a comparative figure and have expressed the data as the number of border cases per unit volume to indicate the relative importance of various factors in border cases.

We can compare the border cases arising from exporting regions for the four-year period, 1999-2002. For border cases arising from problems associated with products or species, comparisons were restricted to the years 1999-2001 because of data availability. Table 23 breaks down the border cases per 100 000 tonnes from various exporting regions.

The picture changes little from earlier indications of absolute numbers of border cases. Asia still tops the list with consistently over twice as many cases per unit volume as the next nearest regions (Africa and central and south America) from 1999 to 2001, and in 2002, following the dramatic increase in shrimp border cases, the figure jumps threefold to 52 cases per 100 000 tonnes of imports. Good performers are non-European Union Europe, North America and Oceania.

It is interesting to note that non-European Union Europe, the region that exports the most product to the European Union, has also the least border cases/trade volume, possibly because of the efficiency of their fish control systems. Indeed, these countries are renowned for their food control and surveillance systems. Another possibility can be linked to the fact that those exporting countries that trade the largest volumes are likely to have larger consignments, thus the number of border cases per unit volume would be lower, but absolute volumes of border cases (in kg of rejected product, for instance) per unit volume of import may be comparable.

A useful figure to note is the total number of border cases per unit volume (100 000 tonnes) arising in the European Union from all imports each year. This figure can be compared to other regions later in the analysis section. From 1999 to 2002, the figure ranges from 4-11 border cases/100 000 tonnes imports, with the higher figure appearing in 2002. This is due to the increase in the absolute number of border cases for 2002, as import volumes have remained essentially static.

As regards products and species, a different situation arises where the higher absolute numbers of border cases for a product category or species are no longer the main problem relative to the amount of trade in those species/products (Table 24), though the changes are not dramatic. Frozen products do now become a lesser problem, with processed products becoming more predominant in the relative importance of border cases in 2000 and 2001. Prepared products also generate a significant level of cases.

The high levels of caviar cases per unit volume of trade for 1999 is notable, but with the very low amounts traded, a single case dramatically inflates the figure. In 2000 and 2001, there were no border cases and hence a zero figure for these years. Again, cured and fresh products prove to be low risk products. For the European Union, canned products also pose very few problems.

Likewise for species trends. Fish are no longer the main problem, with shrimp dominating the tables each year, and becoming more pronounced in 2002. The reasons for this have already been discussed. Bivalves, crab and cephalopods also give rise to varying levels of border cases throughout the three-year period. The reasons for bivalve cases are mainly microbial (ten cases - elevated total counts, hepatitis virus, E. coli and Staphylococcus spp.) with two cases each of diarrheic shellfish poisoning (DSP) and cadmium.

Given that a significant amount of bivalves are eaten raw, controls are likely to be more stringent, given a higher risk factor in eating raw products. Indeed, Huss, Ababouch and Gram (2004) ranked the risk associated with seafoods by product type and identified raw or live molluscan shellfish, amongst others, as a high risk product. This is why the European Union (and other countries') legislation imposes a requirement for monitoring programmes of microbial and biotoxins contamination of the harvesting areas of bivalves and well defined conditions for processing and distribution, including traceability. Thus, the number of countries authorized to export are limited, and any exports from these approved areas should be carefully tested for biotoxins and bacterial counts.

TABLE 23
Border cases in European Union per unit volume of imports, 1999-2002 - by continent

Source: European Rapid Alert System.

TABLE 24
Border cases in the European Union per unit volume of imports, 1999-2001 - by product types and species groups

Note that for some products several types e.g. prepared and frozen, are used to categorize the product and for some none are used.
Source: European Rapid Alert System.

Salmonella spp. and Vibrio spp. dominate the cases for cephalopod species and various microbes for the limited number of crab cases.