Survey of Misdescription of Tuna species in Tuna products (Number 01/00)

1. It is difficult to identify the different species of tuna and/or bonito used in the preparation of processed products, since many of the distinguishing morphological features are no longer identifiable after processing. Price differences between the various species of tuna, and between tuna and bonito, coupled with EU import levies, serve to generally make bonito much cheaper in price than tuna. This raises concerns about the possibility of the substitution of cheaper species for more expensive ones in 'tuna' products, without the necessary label declaration. Such undeclared substitution would not only be in contravention of food legislation but would mislead consumers as to the content of products they were buying. The Working Party on Food Authenticity, as part of its continuing surveillance programme, therefore carried out a survey to investigate the accuracy of labelling descriptions of tuna species on labels of retail tuna products.

2. The survey aimed to identify whether tuna products had been correctly described. Tuna products (607 samples) were collected in January 1999 from retail outlets and the laboratory analysis of samples was completed in March 2000.

3. The survey found that virtually all of the tuna products analysed were correctly described and only one sample was identified as having a label which did not declare a species of tuna detected within it. This sample contained Katsuwonus pelamis (Skipjack tuna) but only had reference to Albacore (long-finned tuna) on the product label. The misdescribed product was a canned 'tuna salad' type product obtained from a supermarket identified as sample code TB147 in Table 1.

4. In the UK, it is an offence under Sections 14 and 15 of the Food Safety Act 1990 ¹ to sell food that is not of the nature, substance or quality demanded by the consumer or to falsely or misleadingly describe or present food. The misdescription of fish species in food may constitute such an offence. Similarly, failure to adequately describe certain fish species in a food may constitute an offence under the Food Labelling Regulations 1996² which require that the name used for the food shall be sufficiently precise to inform the purchaser of the true nature of the food and to enable the food to be distinguished from products with which it could be confused. A list of commercial names that can be used with fish species is given in Schedule 1 of the Food Labelling Regulations 1996.

5. Additionally, marketing standards for preserved tuna and bonito products are laid down in EC Council Regulation, 1536/92 (June 1992)³ stipulating that canned tuna products must be prepared from only one tuna fish species, which may be named with the species name or simply called 'tuna'.

6. The collective name 'tuna' can be given to the main commercially exploited species of the genus Thunnus and include the yellowfin tuna (Thunnus albacares), albacore or longfinned tuna (T. alalunga), bluefin (T. thynnus), big-eye tuna (T. obesus) and other species of the genus Thunnus. Skipjack or stripe-bellied tuna (Katsuwonus pelamis) is also included in the collective term 'tuna'. These are all large species and are found mostly in oceanic habitats and exploited mainly by larger commercial fisheries.

7. The collective term 'bonito' comprises the smaller species of the genera Sarda, Euthynnus and Auxis and include Atlantic bonito (Sarda sarda), Pacific bonito (S. chiliensis), Oriental bonito (S. orientalis), other species of the genus Sarda, Atlantic little tuna (Euthynnus alleteratus), Eastern little tuna (E. affinis), Black skipjack (E. lineatus), other species of the genus Euthynnus, frigate mackerel (Auxis thazard) and bullet tuna (A. rochei). These species are generally found in shallower waters than the larger-sized tunas, bringing them within the range of inshore fishing vessels.

8. Many of the distinguishing features between the tuna and bonito species are in the external morphological characteristics of the fish. It is therefore unlikely that whole fish could be substituted to any degree as it would still be possible to identify the particular features.

9. However, it is much more difficult to differentiate between species after processing when the external distinguishing features are not visible. The survey therefore concentrated on canned and processed products where there is potential for substitution, either partly or wholly, of one species of tuna and/or bonito with another undeclared species, especially in products which contain coloured and flavoured sauces, other ingredients such as vegetables and sandwich fillings (which are usually prepared from canned tuna products).

10. The total number of samples included in the survey was 607 and all were purchased in January 1999. These comprised branded and own label products from national, regional and local retailers. Collection of the samples was by local authorities from the following areas of the UK: Bedfordshire; Buckinghamshire; Cumbria; Denbighshire; Essex; London; Newcastle Upon Tyne; Northern Ireland; Shropshire; Somerset; Staffordshire; Surrey; West Lothian; West Sussex; Yorkshire.

11. In order to develop a representative sampling plan, JFSSG commissioned a survey of the UK tuna market in October 1998, from the Mintel Information Consultancy, through IIS Information Network. This market survey identified all the tuna products currently available for retail sale from various different types of retailers and provided information about market shares of both particular types and particular brands of products. The report from this study was used to ensure that in JFSSG¿s survey the samples collected in each of the geographical regions were as representative of the market as possible.

12. Tuna products were collected in 5 broad categories: canned products (including solid packs/flakes/chunks in brine/oil/sauces/spring water with mayonnaise/vegetables, etc.), sandwiches/sandwich fillings, tuna pates (excluding spreads and mousses), recipe dishes (e.g. 'tuna and pasta bake', 'tuna supreme'), and other (including tuna fish fingers, lunch kits/salad packs (plain or in sauce/mayonnaise or with sweetcorn/other vegetables). All samples were pre-packaged and included a mixture of fresh/chilled, frozen, bottled and canned products. Products were obtained from retail outlets including a range of supermarkets, delicatessens, take-away sandwich bars and other small independent retailers.

13. As canned tuna comprises by far the vast majority of the tuna market, this was the largest category in the sampling plan. Canned products were also further sub-divided to collect specific numbers of particular branded products in order to reflect market share values i.e. Princes, John West, retailer/own label, Gerber Pride, Osprey, Statesman, Glenryck, Heinz. The proportion of each category of product collected and the percentage share of particular brands of canned tuna collected are shown below:

14. Catergories of product collected in survey with percentage share of each category

Category	Percentage Share
Canned	75
Sandwiches	10
Pates	5
Recipe Dishes	5
Other	5

15. Percentage share of particular brands of tuna collected in the canned category:

Canned Tuna Brands	Percentage Share
Retailer/own label	44
John West	23
Princes	17
Other: (Gerber Pride, Osprey, Statesman, Glenryck, Heinz)	16

16. All the products were pre-packaged and product information was obtained from the product labels and recorded on a specially designed sample collection form by the collecting officers. Participating local authorities were supplied with a written sampling protocol and sampling plan outlining the information required and the details to be recorded for each sample. Details of the results of the analyses for each sample can be found in the attached Table 1.

17. Although fish species have been determined successfully by electrophoretic separation of their solubilised proteins, the determination of species in canned products has been made possible only by the relatively recent developments in DNA techniques. The methodology used in this survey is based on the findings of an EU project, "The identification of canned tuna species by characterisation of the nucleic acids, (1994-1997)". ⁴

Samples were tested using DNA analysis techniques using the polymerase chain reaction (PCR) and single-strand conformation polymorphism (SSCP). The PCR-SSCP technique has been shown to be a rapid and sensitive method in the detection of even single base changes in DNA sequences, and has been successfully applied to fish species identification⁵ and differentiation of fish populations. ⁶

Additionally, a recent study has shown that sequencing of PCR products and analysis of genetic distances, a measure of similarity between sequences of different species, allows the unambiguous identification of canned tuna samples.⁷

This technique was used for several samples in the survey where the initial analysis proved inconclusive.

18. The PCR-SSCP methodology involves extracting the DNA from samples and authentic species and purifying it, as far as possible, to remove any protein and potential PCR inhibitors. The extracted DNA is amplified using specific mitochondrial primers to amplify a portion of the cytochrome b gene. The PCR products, including positive and negative controls, are visualised on an agarose gel to check PCR product concentration and purified to remove residual components of the PCR reaction. An aliquot of the PCR product is then heated, chemically denatured and subjected to single strand polymorphism conformation analysis to determine the identity of the samples in relation to authenticated standards. The characteristic banding patterns can be used to distinguish between all the tuna and bonito species, except between the closely related bluefin and yellowfin species. To differentiate between these two species it is necessary to adopt another procedure, restricted fragment length polymorphism (RFLP) using different PCR primers. These SSCP and RFLP techniques are qualitative rather than quantitative methods.

19. SSCP can differentiate all species of tuna from bonito used in canned and processed products. The method can also discriminate between all species of tuna with the exception of between bluefin tuna (Thunnus thynnus) and yellowfin tuna (Thunnus albacares), because these two species are identical over the region of the DNA sequence amplified. Bluefin and yellowfin tuna can be differentiated using different PCR primers and RFLP profiles. However, these are premium tunas and therefore more likely to be substituted with cheaper species than be used as adulterants themselves. RFLP was used to confirm the presence of yellowfin tuna in appropriate cases in the survey.

Control standards
20. Authenticated tuna and bonito standards were essential for the analysis as the band profiles obtained from the samples were identified by comparison with these standards. All gels were run with twelve authentic tuna and bonito species standards, including those species listed at paragraphs 6 and 7. The standards were obtained from a comprehensive data bank of reference standards compiled by one of the analysing laboratories.

21. The sensitivity of the DNA method is dependent on the type of food matrix being examined. The method in the survey can detect levels of samples prepared from pure raw product and diluted to 0.01 per cent where only one tuna species present. However, in mixtures of tunas it is not possible to positively identify less than 10 per cent of one species blended with another. The JFSSG survey validated this using mixtures and pure samples accessed at different numbers of PCR cycles.

22. Although large fragments of DNA can survive quite harsh food manufacturing processes the DNA can, in some circumstances, be severely degraded by chemical, physical and enzymic factors. For example, prolonged heat treatment such as retorting used in some canning processes may result in DNA hydrolysis which fragments the DNA, or modifies the chemistry of the DNA in such a way that the PCR process may not perform efficiently or may not even function at all. This may occur also at low pH (e.g. product in vinegar).

23. Thus, for some tuna products that contain very degraded DNA and/or PCR inhibitors, there may be a substantial decrease in the assay efficiency, and even to the extent that there may be no reaction at all. This could result in the reporting of false negative results, where a declared species is not detected. Nevertheless, the failure to detect the presence of a declared species in this survey is not regarded as evidence that that species was not present in the product. (see also 'Interpretation of Results')

24. The method is species-specific, giving characteristic banding patterns within the diagnostic region of the gel for each species. Since the method will also amplify cytochrome b DNA from other fish and meat species, JFSSG conducted tests to check the response of meat and vegetable products to the PCR-SSCP analysis. The result of these trials was that the method does not amplify DNA from cucumber, sweetcorn, hydrolysed vegetable protein (soya products) or TVP (textured vegetable protein). It was, therefore, concluded that vegetables and vegetable products would not interfere with the analysis. The analytical protocol also dictated that tuna products, especially those in mayonnaise with sweetcorn etc., were thoroughly washed and any sweetcorn, vegetables and/or pasta removed to further ensure that only the fish component was analysed.

25. The full results are in Table 1. In the majority of samples tested, the analytical method was able to positively identify the species present in the product, and these species were as declared on the product label.

26. However, in a few instances (44 samples), it was not possible to fully identify the species present in the product. In some of these cases (12 samples, 2 per cent of the total), inadequate PCR product was obtained from the samples resulting in profiles/patterns which were not clear enough to be able to positively identify the individual species present. This can occur when the composition and/or the degree of processing of a product affects the extractability of the DNA (see also paras 22-23) from a sample and hence the availability of the specific DNA fragments that are used for the species identification. In practice, this may result in the occasional false negative result from certain products, but a positive result is always indicative of the presence of a particular species.

27. In the other cases (32 samples, 5.3 per cent of the total) good banding patterns were observed from the PCR products, but it was not possible to fully match these with any of the standard species controls used. This is most likely to be due to genetic polymorphism. Polymorphism can be defined as ¿the presence in a population of two or more very rare alleles of a particular gene. It may be the result of selective advantage of the heterozygotes or of the rarer forms. (Chambers "Science & Technology Dictionary", 1988 ed.). In polymorphism, the taxonomic classification of species does not always match up with a tight genetic definition. It is possible for variations to occur within specific gene or DNA sequences, and these variations are often associated with species from different geographical locations. While the method used in the survey is sensitive enough to detect these small variations, it was not possible, without excessive additional research and sequencing of species from all geographic locations, to make a positive identification of the species in some products.

28. The individual Home Authority for the manufacturer of the misdescribed sample has been informed of the results of the survey so that they may consider appropriate follow-up action. The manufacturer has also been informed of their results.

29. The results of the survey have been circulated to the Food Advisory Committee for information.

1. Food Safety Act 1990 (1990, Ch. 16). London: HMSO.
2. The Food Labelling Regulations 1996 (SI 1996 No. 1499). London: HMSO.
3. Council Regulation EEC No.1536/92 of 9 June 1992 laying down common marketing standards for preserved tuna and bonito. OJ No.L 163/1-3 (17.6.92).
4. The identification of canned tuna species by characterisation of the nucleic acids, EU Project, AIR 2-CT94-1126 (1994-1997).
5. Hara, M., Noguchi, M., Naito, E., Dewa, K., Yamanouchi, H. Bull. Jap. Sea Natl. Fish. Res. Inst. 44, 131-138 (1994).
6. Oohara, I. Detection of single strand conformation polymorphism (SSCPs) on mitochondrial DNA fragments between two domesticated strains of rainbow trout Oncorhynchus mykiss. Fisheries Science 63, 151-152 (1997).
7. Quinterio, J. et al. The identification of canned tuna using PCR-based methods. Western European Fish Technologists' Association Meeting. (October 1997, Madrid, Spain).

Dr Kevin Hargin
Food Standards Agency
Room 115b Aviation House
125 Kingsway
London WC2B 6NH

Tel: +44 (0) 20 7276 8953
Fax: +44 (0) 20 7276 8193

E-mail: kevin.hargin@foodstandards.gsi.gov.uk