1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References

Trade in genetically modified products is a longstanding and contentious issue in agricultural trade. One issue has not, as yet, received much attention. This is the mingling of unapproved GM products with conventional products. This issue is likely to gain more prominence in the future as new GM product development accelerates. Until recently, problems with mingling were largely one-off events. Recently, however, an ongoing case of mingling has arisen – the case of Canadian GM flax in the EU. The case led to an embargo of imports from Canada and subsequently the bilateral negotiation of a protocol to allow exports to resume. The case raises a number of important issues pertaining to the objective of zero tolerance policies for GM products, the operationalisation of zero tolerance, the role of the testing industry, the design of testing regimes and the risks associated with the absence of transparency and/or international standardisation. It is concluded that mingling is a topic that is deserving of multilateral attention.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References

International trade in genetically modified organisms (GMOs) and agricultural products has been a contentious issue since the technology was first commercialised in the latter half of the 1990s (Isaac and Kerr, 2003; Wugar and Cottier, 2008). While imports of these products are embargoed or heavily restricted by a number of countries, the major protagonists in the international debate over trade in genetically modified (GM) products have been the European Union (EU), which virtually excludes market access for such goods, and the United States, which is a major producer and exporter (Isaac, 2007). Trade in agricultural products is increasingly polarised between those countries that have adopted and can trade GM products and those that have chosen to eschew the use of the technology and exclude products derived from biotechnology from their markets (Smyth et al., 2009). From the perspective of international trade law, the heart of the issue is whether the EU's regulatory regime for GM products is compliant with obligations under the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS) of the World Trade Organization (WTO), which only allows import bans when based on scientific evidence that imports can harm human, animal or plant health and the environment (Isaac and Kerr, 2007). To date, this issue in trade law remains unresolved (Viju et al., 2011a) and agricultural markets remain effectively segregated between those that accept the technology and trade its products, and those that do not.

The effective segregation of those countries that grant market access to GM products and those that do not, however, has created externalities that can significantly impact trade in non-GM agricultural products. The mingling of GM and non-GM crops can lead to import bans, shipment refusals and costly delays – barriers to trade. Unintended mingling1 can take place at any point along the supply chain from the seeds farmers in exporting countries plant, to farmer's fields, to the grain handling and transport system, to port facilities, to ships, to food processing and distribution firms in importing countries. Hence, assuring the non-GM status of any agricultural products requires monitoring and testing all along the international supply chain, which can add considerably to the costs of engaging in international trade. The issue has, as yet, received little attention in multinational negotiations, and there is no international standardisation of contamination tolerances, testing regimes or how to assess risks.2 As a result, issues related to mingling are dealt with bilaterally. The net effect is an increasing level of risk for the trade in agricultural products.

The problem of mingling is likely to increase over the next few years as the adoption of GM products continues to expand in those parts of the world where GM technology has been accepted. The number of hectares planted to GM crops, the range of GM crops and the number of countries growing GM crops can all be expected to expand. In 2010, there were approximately 30 commercial GM crops cultivated worldwide, but it is forecasted that by 2015, there will be more than 120 varieties in commercial production (Stein and Rodriguez-Cerezo, 2010). Since commercial cultivation began in 1996, there has been a continuous increase in the area of cropland where GM varieties are grown, reaching 125 million hectares in 2008. GM crops are grown in more than 25 countries and 12 million farmers plant them (James, 2008). As a result of this expansion of GM agriculture, unintended mingling can be expected to increase apace, particularly because cross-species mingling of crops (e.g. GM corn in non-GM wheat – where no commercial varieties of GM wheat exist) are just as unacceptable as within-species mingling of crops (e.g. GM corn with non-GM corn).

A search of the NGO-managed GM Contamination Register indicates 223 cases of non-GM contamination by unauthorised GM material worldwide from 1997 to 2010, with 141 cases occurring in Europe. Cases of mingling GM material with non-GM material include Canadian rapeseed/canola, 1997; US corn, 1998; StarLink corn (processed corn approved for animal feed but not for food in the US), 2000; Bt10 corn (in corn gluten for feed), 2005; Liberty Link rice 601 and 604, 2006; Herculex maize, 2006/2007; Roundup Ready II and Liberty Link soy, 2008; BT 63 rice, 2008; and MON88017, MON89034 and MIR 604 (corn in soya), 2009. Import bans and refused shipment were generally the result.3 In many of these cases, however, the mingling was with experimental and/or unapproved varieties and a one-off event. While trade was disrupted over the short run, removal of the unwanted material from future shipments was relatively simple. In 2009, there was a major case of mingling where the exporter had a clear desire to continue to have access to the EU market for its non-GM crop and importing firms in the EU wished to continue to have access to the product. Further, the potential for mingling was endemic throughout the supply chain. This case provides an excellent opportunity to explore the implications of the failure to deal with standards and procedures pertaining to the mingling of GM and non-GM crops given the likely increase trade disruptions as GM technology and adoption continues to progress. The case was the discovery of Canadian GM flax in the EU. Before proceeding to the discussion of the Canadian GM flax case, it is important to understand the EU regulatory regime on the mingling of GM with non-GM crops in the case of imports.

2. The European Union Policy of Zero Tolerance

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References

The use of genetic modification became a major political issue in the EU in the latter half of the 1990s and in June, 1999 the EU's Council of Ministers voted to cease approvals of GM products until a new regulatory regime was developed. This de facto moratorium was also applied to imports. Devising a new regulatory regime proved extremely difficult.4 In 2003, a new EU regulatory regime for GM products was finally agreed. The processes for approving new GM varieties are time-consuming and require both scientific and political acceptance. As a result, the first new product to be approved, BASF's amylopectin potato, only worked its way through the regulatory procedures in March 2010, approval that was first sought in 2005.5

The result is that GM product approvals have been on hold for a decade. As approval of new GM products in many parts of the world has continued, agricultural production is increasingly characterised by asynchronous authorisations whereby a GM crop may be fully approved for commercial use in one country, but not in others. As with most countries, the EU does not allow the import of non-authorised GM products.

The problem of mingling pertains to both approved and unapproved GM products. The EU is tolerant when mingling takes place between authorised GM products and non-GM products but officially has a zero tolerance for the mingling of unauthorised GM products with non-GM products. In the case of the mingling of authorised GM products and non-GM products, a 0.9 per cent threshold has been set for food and feed. Any conventional product found to contain in excess of 0.9 per cent GM product must be labelled as GM.

In the EU, there are two classifications for the mingling of unauthorised GM products and conventional products – low level presence (LLP) and adventitious presence (AP). Low level presence occurs when a GM product approved by an exporter but not by the EU is found to be mingled with a conventional product. On the other hand, an AP GM event occurs when the GM product is not approved in any market (i.e. is an experimental product or is cultivated under confined field trials). When either an LLP or AP event occurs, the product is considered an illegal substance by EU officials and can trigger a significant reaction including emergency measures (EC, n.d.). European Commission food safety Regulation 178/2002 allows the enactment of emergency measures for imported food and feed if an unauthorised GM product is found (EC, n.d.). Emergency measures can include the assignment of special conditions, temporary or long-term import suspensions or the closing of markets.

The official EU policy towards unauthorised GM products found in the EU, or are found in shipments wishing to enter the EU market, is zero tolerance. While zero appears to be an absolute, or fixed, limit, in reality, it is defined by a number of factors that are changeable or arbitrary. This is because zero must be operationalised for those engaged in transactions for conventional products that could have GM products mingled in with them. Given the expanding localities and number of GM products being produced around the world, and the potential for mingling across species so that zero must apply to imports where no GM products exist, the scope of the policy extends to almost all potential imports of agri-food products. Hence, the only way to ensure zero mingling in imports would be to test every single grain of wheat, every single soy bean and every bit of foreign plant material in a shipment of apples. As this is neither technically nor economically feasible, unless imports of agricultural products from outside the EU are to be totally eschewed, some form of testing regime is required. A total ban on agricultural imports is not a viable political option and would run counter to the EU's international trade commitments (Gaisford et al., 2003).

The design and implementation of a testing regime, however, can impose costs and risks on international transactions that, in effect, act as barriers to trade. The primary reason for negotiating trade agreements – rules of trade – is to provide firms conducting international transactions with transparency regarding the ways in which governments can intervene in those transactions. Engaging in international trading activities is a risky activity. One of the major risks is that, after investing in developing a market for a product and enjoying some commercial success, a government could intervene to ban, tax or impose costly regulations on the transboundary movement of the product. If governments can intervene in international markets without constraint, then the risk of investing in international trading activities rises, these investments are inhibited and the benefits of international trade foregone. The EU's zero tolerance policy provides little transparency either for firms that wish to export agricultural products to the EU or those that wish to engage in importing.

A testing regime is, first and foremost, defined by the technology of testing. Zero is effectively delineated by the efficacy of the test that is currently available. As in most cases, the use of biotechnology in production does not alter agricultural products in ways that are physically discernible – they are credence attributes6 – the only way to know whether they are present is to have a test that can identify their presence. If no test exists, then zero has no effective limit and imports should be unrestricted.7 As the GM product cannot be found, it can enter a market where it is considered an illegal substance.

There are, however, considerable risks for firms that wish to enter into international transactions for agricultural products. This is because testing technology is not static. If a new test for a GM product is developed, then a product that had enjoyed open access would, in effect, be embargoed. Given that new GM varieties are expected to proliferate, tests for them are unlikely to exist, at least initially. At the moment, devising a test is a private commercial decision which depends on the expected return on making an investment in developing a test. As developing a test cannot always be expected to yield a positive expected return, tests may or may not be developed. There also may be a considerable lag between when a GM product is commercialised and when a test is developed. If a test does not exist, and imports are allowed, the question is raised regarding the point of closing markets once a test is developed. Presumably, the reason that a product is excluded from a market is that it imposes a risk, or the potential of a risk,8 to human health or the environment. If GM products have entered a market in mingled shipments prior to a test existing, then the population has likely consumed the product and/or it will have become part of the environment. While in some cases, it may be possible to remove GM products from supply chains and/or the environment, in other cases, it will not and banning imports is a bit like closing the barn door after the horse has bolted. Hence, a blanket policy of banning imports ex post to a test being developed is open to question.

If a method of testing exists, then zero tolerance implies that any positive test result would lead to the refusal of a shipment when it attempts to enter the customs territory of the importer. Sampling, however, implies that there is only a probability of exclusion, not exclusion – thus tolerance is non-zero. In effect, the importing authority must agree upon an implicit, if not explicit, probability that the objective of zero tolerance has been achieved. The efficacy of testing, however, is a function of sampling effort. Sampling is a costly activity, both in collecting samples and analysing them. There are complex trade-offs that must be made (e.g. does one take one sample of 10 kg from a shipment or five samples of 2 kg from different points in a shipment?). Testing may also have costs of time associated with them (e.g. a ship containing wheat being sampled once it arrives in port at Antwerp may have to await unloading until the samples have been analysed). If the laboratories take two weeks to return, the results significant demurrage charges will be incurred. Lags associated with waiting for tests to be undertaken will impose costs anywhere along a supply chain where a sample is taken, but may differ in the costs imposed (e.g. the opportunity cost of grain remaining in a bin in a primary grain handling facility in an exporting country may be considerable less than the demurrage charge for a ship awaiting unloading).

The point of sampling along the supply chain can also influence the costs and risks associated with a testing regime. If, for example, testing is mandated to take place at the point of entry into the customs territory (e.g. when a ship docks at a port of entry) and a positive test result for mingling is returned, the exporter is faced with finding a new customer for its product, paying the demurrage charges while the new customer is found and paying for transporting the shipment to the alternative point of delivery.9 The importing firm will also have to find an alternative source of supply and await the arrival of the shipment, which could lead to delays in fulfilling customers’ orders or even temporary shutdowns of processing plants or other facilities. On the other hand, if a cargo could be cleared for entry into the importing market based on samples taken prior to the ship being loaded, then these costs could be largely avoided (Viju et al., 2011b).

Testing itself is not infallible. In testing for genetic material, both false positives and false negatives are possible. Thus, one test cannot be relied upon to give a definitive answer to the presence or absence of an illegal substance. This implies more tests and the establishment of tolerances for false positives.

Beyond the official position of zero tolerance, there is sometimes a tacit recognition by regulators that the achievement of zero is likely technically infeasible. Thus, only near zero tolerances may actually be set. Such tolerances are common for a range of contaminants in foodstuffs such as faecal matter, insect fragments, etc.10 In the case of the mingling of authorised or approved GM products and conventional products, there has been a considerable debate over the levels set by the EU, Japan and other importers. Members of international supply chains for grains and oilseeds have argued that what governments consider technically feasible may not be commercially viable. The threshold that the industry generally considers commercially acceptable is 5 per cent.11 Regulators may wish to establish some non-zero tolerance levels for unauthorised GM products, but as yet, there has been little discussion of what these tolerances will be. Of course, any movement away from total refusals for imports when any mingling is found will be welcomed by those involved in international supply chains.12

None of the regulatory options outlined above are pre-established in a transparent way in the EU or multilaterally. As a result, those wishing to engage in the international trade of agricultural products face considerable risks directly as a result of the regulatory inertia. The ad hoc nature of the implementation of the zero tolerance policy for dealing with LLP and AP of GM products would seem to run counter to commitments under the SPS. In the SPS, control, inspection and approval procedures are dealt with in Annex C. Annex C 1. (e) states:

Members shall ensure, with respect to any procedure to check and ensure the fulfilment of sanitary or phytosanitary measures, that

(e) any requirements for control, inspection and approval of individual specimens of a product are limited to what is reasonable and necessary.

Given the large number of unknowns in the regulatory regime meant to deal with mingling of GM products with conventional products, it is hard to determine whether what may be required is limited to what is reasonable and necessary. If the issues outlined above were simply speculative, then they might be of only academic interest. There is, however, a recent mingling event that had a significant impact on the international trade of an agricultural product which incorporated elements of most of the issues; this was the discovery in the EU of GM flax from Canada mingled with conventional flax.

3. Genetically Modified Flaxseed in the EU

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References

On 8 September 2009, Germany issued an EU-wide Rapid Alert notification confirming the presence of GM flax in some samples of flax imports from Canada. At a meeting of the Standing Committee on the Food Chain and Animal Health of the European Commission, held in Brussels on 16 November 2009, Member States agreed that illegal flaxseed should not be allowed to enter the EU market. Imports were embargoed until Canadian exporters could satisfy the EU regulators that shipments conformed to EU standards. The process entailed the development of a detailed sampling and testing regime. The closing of the market imposed significant costs on both the Canadian flax industry and EU users of flax. Given the preponderance of Canadian flax in the global market, EU users of flax had no alternative sources of supply and suffered considerable disruptions to their production. A bilateral testing protocol was eventually agreed which allowed imports of Canadian flax to resume. The protocol, however, imposes significant costs on the export supply chain.

Oil seed flax (also known as linseed) is largely grown for industrial use, with the oil used in the manufacture of linoleum and paint.13 The Canadian GM flax is an oil seed variety. The flaxseed is crushed to extract the oil, with the residual meal used as an animal feed. Small quantities of oil seed flax are also consumed by humans.14 There is no segregation of seed to be used for industrial use from seed destined for human consumption in export shipments. In the case of oil seed flax shipments from Canada to the EU, flax for human consumption is sourced from common cargo.

Canada is the world largest producer and exporter of flax, with an average of 746,000 metric tonnes of production per year between 2000 and 2007, accounting for 40 per cent of global output.15 Only 20 per cent of Canadian production is used domestically. Prior to the mingling event, the largest market for Canadian flax was the EU, which normally took over 70 per cent of Canadian flax exports.

The GM flax developed in Canada is formally known as FP967, but is commonly known as CDC16 Triffid. The genetic modification made the Triffid variety resistant to soil residues of sulfonylurea-based herbicide – an herbicide commonly used on cereal crops which persists in the soil until the next growing season. Field trials of Triffid flax began at the end of the 1980s, and the regulatory approval process commenced in 1994. Triffid received official Canadian approval for use as animal feed in 1996 and for human consumption in 1998. Approval was also obtained in the United States.

By the late 1990s, however, GM products were becoming an important political issue in the EU. As future developments in the EU were not clear, the Canadian flax industry became concerned about continued EU market access. Triffid's developer decided to voluntarily deregister the variety to pre-empt potential problems (EC, 2010). Triffid had not yet been commercially grown, but seed companies were in the process of growing the Triffid variety to produce seed destined for commercial sale to farmers. A recall and crush of these initial seed stocks was undertaken in 2001.17 The germ plasm and other materials held by the developer were incinerated. The Triffid variety was deregistered in 2001. Thus, Triffid flax was never grown commercially and was thought to have been removed from the ecosystem. No tests existed that could detect the presence of Triffid and as far as anyone knew GM flax no longer existed. Exports of Canadian flax to the EU continued as normal until 2009.

In July 2009, GM material was found in a shipment of flax in the EU. Initially, it was thought that the material was the result of cross-contamination from Canadian GM canola (rapeseed). This event, however, led to an increased detection effort in the EU. In September 2009, Triffid was identified in bakery goods in Germany, the first of in excess of one hundred positive tests reported through the EU's Rapid Alert System for Food and Feed (RASFF). Contamination was widely dispersed geographically in cereal and bakery products made by EU firms (Viju et al., 2011b). The EU market was closed to Canadian flaxseed. The Canadian Grain Commission (CGC), which is the agency in Canada responsible for ensuring quality, initiated its own testing which confirmed the presence of trace amounts of Triffid material in some Canadian flaxseed shipments. The Canadian industry and government were surprised and, as yet, have not been able to determine the source for the Triffid flax that entered the supply chain. Triffid was found all along the supply chain starting from seed suppliers.18

As a result, GM flax may have been in the EU market for nearly a decade. Its presence, while only in trace amounts, is endemic and geographically widely dispersed. From an international trade perspective, this calls into question the rationale for first embargoing imports of Canadian flax and then imposing a costly testing regime. In international trade law, it is generally accepted that trade measures imposed for SPS reasons should only be put in place to achieve some objective relating to human health, animal health or plant health (Gaisford et al., 2001; Holtby et al., 2007). Zero presence is the EU's objective for unauthorised GM products. In the case of Triffid flax, it is already widely diffused in EU food supply chains. There is no indication that excluding Canadian flax or a testing regime will return the domestic EU market to a pristine state. The EU has not taken extra measures to purge food supply chains of Triffid flax. Given that Triffid has not ever formally entered the supply chain, it appears to have a persistent presence. As in the EU, Triffid is only present in Canada in trace amounts. As a result, future imports can be expected to contain trace amounts – those missed in sampling. Hence, it would seem that the objective of EU policy can be no better achieved with barriers to trade than without them.

The costs imposed by the testing regime have no foreseeable end date. The only way that Canadian flax exports can escape the requirements of the EU policy of zero tolerance would be for Triffid to become an authorised GM product in the EU. This would require someone to pay for Triffid to undergo the very expensive and time-consuming EU registration process.19 Given that Triffid is no longer registered in Canada and not grown, no one will shoulder this burden. Further, Triffid is now agronomically obsolete; thus, it has no potential economic value. While this problem may on the surface appear to be unique to Triffid; in fact, it has considerable implications for the future of agricultural trade. As GM products proliferate in the future, there may be many instances where GM products are not agronomically suitable for agricultural production in the EU and, hence, authorisation will not be sought from the EU. If they are, however, grown in other countries, they can become mingled with conventional varieties exported to the EU or mingled with different conventional crops (e.g. unapproved GM rapeseed mingled with conventional sunflower seeds). Similar costly testing procedures to those in place for Triffid flax could have to be imposed on a wide variety of imports – and as tests are specific to each GM product, there may need to be multiple testing regimes for imports that can contain more than one GM product.20

Prior to the Rapid Alert notification, in August 2009, German authorities transmitted to the EU's Community Reference Laboratories Network,21 a construct-specific method for detecting Triffid flax using a real-time polymerase chain reaction. The detection method was developed by Genetic ID NA, Inc., of Augsburg, Germany. The method was specific to the Triffid genetic modification as it targeted a construction found only in Triffid flax. The original method was replaced in October 2009 with a newer version developed by the same company. These tests are used both in the EU and Canada. As reported above, once the test existed, trace amounts were found in over 100 food samples in the EU. The presence of this illegal material led to imports being embargoed.

From the Canadian perspective, a mechanism was required that would allow renewed exports to the EU. A protocol was developed by the CGC in consultation with the Flax Council of Canada22 and DG SANCO, the European Commission Directorate for Health and Consumer Affairs. The protocol is not a government-to-government agreement; thus, it may be technically outside the ambit of the SPS. The protocol puts in place a mechanism for documenting, sampling and testing for the presence of Triffid flax in the supply chain of Canadian flaxseed destined for the EU. The protocol satisfies the zero tolerance policy for LLP GMOs as currently interpreted in the EU, that is, maximum acceptable level of risk of Triffid is at the 0.01 per cent level – thus, the EU has chosen not to reject all shipments that show non-zero evidence of mingling. The 0.01 per cent detection level established by the protocol is linked to a level that can be accurately and reliably detected by the Triffid-specific test. However, the threshold represents a very high commercial risk as it has proven nearly impossible for the grain storing and handling companies to achieve it and, thus, is of considerable economic importance.

Testing all along the Canadian supply chain began in the fall of 2009 and by April 2010, over 5,000 tests had been conducted. A widespread but extremely low level presence of Triffid was found in the Canadian flax production and distribution system. One hundred and seventy-four (3.4 per cent) of the samples tested positive at 0.01 per cent while another 300 (6 per cent) tested positive at less than 0.01 per cent. The loads of flax that showed the latter results could be imported into the EU. Of the 213 tests conducted on pedigreed seed, 6.5 per cent tested positive at 0.01 per cent (Stephens, 2010). However, the level of detection has dropped in 2011. Only 7 per cent of 4,003 samples in 2010–11 were found positive compared to 10 per cent positive samples of 6,013 farm samples tested in 2009–10 (Vakulabharanam, 2011). Thus, it would appear that the presence of Triffid can be reduced through monitoring and subsequent preventive activities, but not reduced to zero.

The EU has shown some flexibility in implementation of the protocol, which may be of considerable importance for future cases of mingling of unapproved GM products in export shipments. The original 2009 protocol was found to impose a large level of risk for both Canadian exporters and EU importers, as the final test results were made available only after ships with flax cargos departed from Canada. This lag time resulted in the quarantining of some flaxseed shipments in port in Belgium after samples tested positive for FP967 – a very costly point in the supply chain for refusals to take place. In response, the protocol was revised in March 2010, and a preload test at the port of export (in-store samples) was added, allowing for the final results to be known before transport vessels were loaded (Hall, 2011). Although there could be, there is not necessarily any further testing once ships have arrived in EU ports. While encouraging for supply chain participants faced with future mingling events, the potential for testing upon arrival leaves a degree of opaqueness in the importing regime that increases risk to a considerable degree.

The protocol specifies that three samples must be collected at different points along the supply chain: by the grain handling company from each producer delivery from a farm; before loading the railcar, with each railcar being sampled and the composition of samples comprised of not more than five railcars; and at the terminal elevators prior to loading on ships. The latter tests are undertaken by CGC personnel.23

If the composite sample tests positive for FP967 at the grain handling level, all railcars that tested positive are removed from the aggregate flaxseed shipment destined for export to the EU and the list of railcars that tested negative is transmitted to the CGC. The CGC samples all railcars carrying flax destined for the EU, monitors the unloading of each railcar in a designated silo, then seals and records the silo and seal number. A 2.5 kg composite sample is prepared from each silo and sent to an ISO 17025 accredited laboratory for testing. The laboratory must be from the list of laboratories approved for testing flaxseed shipments to the EU.24 These approved laboratories operate in accordance with the ISO 17025 standard on general requirements for competence and testing and calibration laboratories (CGC, 2010a). Flax from any grain handling or storage facility that tested positive is diverted from the EU flaxseed supply. The CGC prepares an official Letter of Analysis which is presented to the Canadian flaxseed exporter, who, in turn, provides it to the appropriate EU authorities. The number of samples taken, the size of the lots from which samples are taken and the size of the samples to be taken are the result of one-on-one negotiations between officials from the CGC and EU officials from DG SANCO. While it was clearly possible for a mutually acceptable testing regime to allow Canadian exports of flax to resume to be reached, there is little that can be learned that will be useful for those who are faced with a mingling event in the future – how the outcome is reached is not transparent and there is no assurance that it would be replicated in the future. The Canadian flax supply chain is also faced with considerable costs, in perpetuity.

Canadian exports to the EU began to recover in May 2010 but have only reached approximately 60 per cent of pre-embargo volumes (Viju et al., 2011b).25 The EU flax market was the price-setting market prior to the mingling event. The average farm price received for flax dropped from Can$490.80/tonne in December 2008 to Can$357.40/tonne in December 2009 (Statistics Canada, 2011). As the 2009 flax harvest was near completion when the mingling event took place, a market had to be found for the crop. China entered the market in a major way for the first time and, taking advantage of the lower prices, replaced the EU as the major export destination (Viju et al., 2011b).

Industrial firms in the EU that use linseed oil as a major input to linoleum and paint have suffered from disruptions in input supplies as, at least in the short run, alternative sources of supply were simply not available. According to a study conducted by COCERAL26 and FEDIOL,27 the total additional costs incurred by the EU flaxseed industry due to the Triffid event amounted to €23,530,000 (Dayananda, 2011). A sharp increase in flaxseed prices in the EU followed. As a large number of products were recalled, extra costs were incurred by traders due to freight, storage, sampling and monitoring as well as the destruction of some products.

Each sample tested costs Can$240.00. Testing takes place all along the supply chain with complex sampling requirements at each stage.28 Dayananda (2011) estimates the total cost of testing for the 2009–10 crop year at Can$1,280,000. Segregating those shipments that fail the tests from shipments that are cleared for export to the EU is estimated to have cost Can$4,240,000 (Dayananda, 2011). Flax is a relatively minor crop. If mingling were to be found occurring in a major export crop such as wheat or maize, these testing costs would have to be scaled up accordingly. Further, they will be ongoing because the only way they can be reduced is if the GM product is approved in the EU. In some cases, approval will never be sought.

4. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References

The prospects for increased future mingling between GM products and conventional agri-food products appear to be high. Asynchronous authorisations seem likely to accelerate as new GM products are developed at an increasing pace. In many instances, approval may not be sought where the process is demanding and expensive. Hence, how mingling is dealt with may have considerable importance for trade in agricultural products.

As yet, the question of mingling has not been dealt with formally at the multilateral level. As a result, cases of mingling will be dealt with bilaterally on a case-by-case basis. This means that the regulatory regime that will apply will not be known ex ante to the commercialisation of new GM products in exporting countries. Hence, transparency in the rules of trade is less than what might be considered reasonable for international agri-food supply chains. The examination of the case of the mingling of Triffid flax with conventional flax destined for the EU provides considerable insights into how mingling may be dealt with, thus contributing to transparency. Of course, given that there are no agreed international procedures for mingling cases, there is no assurance that future cases will be dealt with in the same way.

It seems clear that the degree of mingling that will be acceptable in the EU is non-zero. Mingled shipments are not rejected outright as loads with samples that are non-zero can be approved for import. It is also clear that commercial considerations can inform, to some degree, the design of the regulatory regime, the decision to accept tests undertaken prior to loading ships at port of export and not testing shiploads at port of entry. There is some additional evidence that commercial factors can influence the EU's import policy for GM products. The spread of GM technology has made it more costly for livestock producers in the EU to import sufficient animal feed. Due to the technical and commercial infeasibility of a zero tolerance policy, the European Commission Regulation (EU) No. 619/2011 that came into force on 15 July 2011 sets out a minimum required performance limit (MRPL) of unauthorised GM material for crops for feed, with the threshold being established at 0.1 per cent (EUR-Lex, 2011).29 However, this new regulation suffers from a number of limitations. First, it covers only a limited number of GMOs, and second, it is restricted to feed crops. Also, Article 6 of the regulation specifies the measures Member States are allowed to implement once the presence of unauthorised GM material is identified as present: if GM material is at or above the MRPL, Member States should notify this information through the EU RASFF (Article 6.1); if GM material is below the MRPL, the Commission and the other Member States should be notified (Article 6.2); and the Commission or Member States can implement emergency measures when they consider it appropriate (with no specification of the level of GM material) (Article 6.3). Thus, Article 6.3 raises a number of questions regarding whether this regulation increases the legal certainty of the market operators as it allows the European Commission or the Member States to adopt emergency measures even if the level of GM material is below the MRPL. Testing will play a crucial role in whether trade barriers will arise. As the Triffid case illustrates, zero is effectively defined by the availability of a test; and presumably, the efficacy of any tests that exist. The development of tests for genetic material is undertaken by the private sector. In the case of Triffid, the developer of the test benefitted from the use of its original, and subsequently improved, test through widespread use of its test in the EU and Canada and eventually in a number of other countries. Thus, private firms may have an incentive to invest in developing tests. While the development of tests increases transparency, it also presents considerable risks for those that are involved in international agri-food supply chains.

The Canadian flax industry and regulatory institutions initially had to develop a monitoring and testing regime to propose to DG SANCO as the basis of the protocol. The EU was not, however, the only country with which Canada had to deal in the wake of the closure of the EU market to Canadian flax. Brazil,30 Japan31 and New Zealand (Smyth and Ryan, 2010) also imposed monitoring and testing requirements for imports. Each of these protocols differed from the protocol implemented for shipments to the EU. While there were no major issues of conflicting regulations among the protocols, it does raise the spectre of exporters facing conflicting requirements for future GM crops, which could force exporters to choose among export markets. Of course, in the Canadian case, flax shipments had to be monitored to determine which protocol would apply.

The experience with Triffid flax suggests that mingling can lead to the imposition of trade barriers through mechanisms that are not transparent and that can impose considerable disruptions to trade and ongoing costs. As GM crops proliferate, the incidents of mingling are likely to increase. Hence, the issue of mingling is one that countries that are accepting of genetic modification technology may wish to put on the agenda for consideration at the WTO's SPS Committee.

  1. 1

    For example, unintended mingling (or involuntary genetic contamination) at the farmer's level can happen when traditional and GM crops are planted in the same region and can be caused by wind, birds or insects (Choi, 2013).

  2. 2

    In July 2008, the Codex Alimentarius Commission (Codex) adopted an amendment in Annex 3 – Food Safety Assessment in Situation of Low-Level Presence of Recombinant-DNA Plant Material in Food – of its guidelines for GM food assessments (Codex, 2008) that introduced a practical set of simplified risk assessment procedures when mingling of a GM product approved by the exporter but not the importer is found in a non-GM shipment (Gruere, 2009). Individual countries are, however, unconstrained in how they chose to implement the Codex procedures.

  3. 3

    See Viju et al. (2011b) for a discussion of these cases.

  4. 4

    The delays associated with the development of the new regulatory regime led trading partners faced with the moratorium to complain that it represented a barrier to trade motivated by economic protection and to eventually launch a formal dispute at the WTO. The dispute was initiated in 2003 by the United States and Canada. In 2006, a WTO Disputes Panel found in favour of the complainants but by that time, the new EU regulatory framework was in place. The EU responded by indicating that its regulatory regime would be WTO compliant, but that time would be required to accomplish the task (Viju et al., 2011a).

  5. 5

    While EU's regulatory approval has been granted to BASF's amylopectin (‘Amflora’) potato, some individual Member States wish to deny approval for cultivation in their territory, setting the stage for a major legal confrontation between the European Commission and Member States. Thus, the EU regulatory regime for GM approvals seems set for a further period of flux (Viju et al., 2011a).

  6. 6

    Hobbs (1996) defines three types of goods or their attributes. Search attributes are those that can be discerned by visual inspection – the colouring of a distinctive type of apple. Experience attributes are those that can only be discerned through the consumption experience – the tenderness of a steak. Credence attributes are those that cannot be determined either by visual inspection or a consumption experience – the country of origin of olive oil or the presence of modified genetic material.

  7. 7

    The only alternative is to ban imports but with the potential for mingling across agricultural products this would, in effect, mean any imports of a product where the potential for mingling exists would have to be banned (i.e. when a commercially approved GM crop is grown).

  8. 8

    This is the rationale for imposing trade barriers under the precautionary principle. The precautionary principle is, of course, itself a contentious issue in international trade. See Van den Belt (2003) and Phillips et al. (2006) for a discussion of the precautionary principle.

  9. 9

    Alternatively, the shipment could be unloaded and destroyed, an alternative with its own set of costs.

  10. 10

    See Olsen et al. (2001) for a discussion of tolerances for contaminants.

  11. 11

    The EU level is currently set at a contentious 0.9 per cent. See Baker and Smyth (2012) and Phillips and Smyth (2004) for discussions of mingling levels for authorised GM products.

  12. 12

    It is difficult to see what the basis of this discussion would be without the broader questions of the EU's scientific justifications for import barriers and the conduct of risk assessments surrounding imports of GM products being dealt with. These questions remain in legal limbo in international law. See Viju et al. (2011b) and Dayananda (2011) for a discussion of the EU's compliance with SPS obligations for unauthorised GM products where no scientific justifications are given or risk assessments undertaken prior to the impositions of import bans.

  13. 13

    A separate group of flax varieties – fibre flax – is grown primarily for its long fibres that are used in the production of linen cloth.

  14. 14

    The human market for flax seed has increased in recent years as it has been discovered that consuming flax can impart considerable health benefits.

  15. 15

    The other three large producers of flax, China, the US and India, collectively account for 40 per cent of world production. India is largely self-sufficient while the US and China are net importers.

  16. 16

    CDC stands for the Crop Development Centre at the University of Saskatchewan in Canada where the Triffid variety was developed using public funding.

  17. 17

    In less than a year, 5,000 metric tonnes of seed were destroyed at a cost of Can$3.2 million.

  18. 18

    Tests also revealed traces of Triffid in flax grown around the world (Dayananda, 2011).

  19. 19

    See Viju et al. (2011a) for a discussion of the EU approval process for GM products.

  20. 20

    If tests do not exist when commercialisation takes place, a presence may be established in the EU just as in the Triffid case.

  21. 21

    Now, the European Union Reference Laboratories (EURL).

  22. 22

    Agriculture and Agri-Food Canada (AAFC), Foreign Affairs and International Trade Canada and the Canadian Food Inspection Agency (CFIA) were also involved (CGC, 2010a).

  23. 23

    The CGC has published a guide on sampling methods: Sampling Systems Handbook and Approval Guide (CGC, 2010b).

  24. 24

    The CGC confirms the proficiency of laboratories to test flaxseed samples using the qualitative polymerase chain reaction assay as per the construct-specific method approved within the protocol for the EU. CGC maintains a list of laboratories (laboratories approved for testing flaxseed shipments to the EU) that are ISO 17025-accredited laboratories that are able to test bulk shipments of Canadian flaxseed destined for the EU.

  25. 25

    Given the major recession over the same period, it is not likely correct to attribute all of the failure to recover export volumes to the remaining risks associated with exporting flax to the EU and the additional costs imposed by the protocol.

  26. 26

    The committee of cereals, oilseeds, animal feed, oils and fats, olive oil and agrosupply trade of the EU.

  27. 27

    The European oil and protein-meal industry federation.

  28. 28

    See Dayananda, 2011 for a full description of the testing procedures.

  29. 29

    The regulation refers to GM material that is authorised in a third country, and the EU authorisation procedure is pending for at least three months. In addition, the European Food Safety Authority (EFSA) must not have identified potential adverse effects on human or animal health or the environment. The regulation refers as well to GM material for which authorisation has expired and there is no intention for a renewal.

  30. 30

    Brazil has zero tolerance for unapproved GM products. According to the Flax Council of Canada (2010), as of 4 November 2009, the Government of Brazil announced the mandatory holding and testing of all flax shipments entering Brazil from Canada. If the tests prove negative, the flaxseed can enter Brazil. Several Canadian flaxseed containers have tested positive for Triffid and these shipments have been halted at the border by Brazilian Ministry of Agriculture officials.

  31. 31

    Japan imports Canadian flaxseed mainly for industrial purposes. After the GM flax event in the EU, Japan's Ministry of Agriculture, Forestry and Fisheries requested that Canada's flaxseed industry implements appropriate preventive measures. The Canadian Government has developed a protocol to describe the system of sampling, testing and documentation pertaining to the presence of Triffid in bulk vessel shipments of Canadian flaxseed to Japan. This protocol is specific to bulk shipments of flaxseed destined for industrial or feed use and incorporates an allowable tolerance of one per cent for Triffid (CGC, 2010c).


  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. The European Union Policy of Zero Tolerance
  5. 3. Genetically Modified Flaxseed in the EU
  6. 4. Conclusion
  7. References
  • Baker, A. and S. J. Smyth (2012), ‘Managing Opportunism in Value-Added Supply Chains: Lessons from Organics’, Journal of International Food and Agribusiness Marketing, 24, 1, 125.
  • CGC (2010a), ‘Sampling and Testing Protocol for Canadian Flaxseed Exported to the European Union’ (Winnipeg: Canadian Grain Commission), Available at: (accessed 14 October 2011).
  • CGC (2010b), ‘Background Information on Genetically Modified Material Found in Canadian Flaxseed: Low Levels of Genetically Modified Material Found in Canadian Flaxseed’ (Winnipeg: Canadian Grain Commission), Available at: (accessed 14 October 2011).
  • CGC (2010c), ‘Sampling and Testing Protocol for Bulk Shipments of Canadian Flaxseed Exported to Japan for Feed or Industrial Use’ (Winnipeg: Canadian Grain Commission), Available at: (accessed 19 October 2011).
  • Choi, E. K. (2013), ‘Genetic Contamination of Traditional Products’, International Review of Economics and Finance, 27, 29197.
  • Codex Alimentarius Commission (2008), Guideline for the Food Safety Assessment of Foods Derived from Recombinant-DNA Plants, CAC-GL 45-2003, Amended in 2008, (Rome: United Nations World Health Organization and United Nations Food and Agriculture Organization).
  • Dayananda, B. (2011), ‘The European Union Policy of Zero Tolerance: Insights from the Discovery of CDC Triffid’, Unpublished MSc Thesis, University of Saskatchewan, Available at: (accessed 3 December 2012).
  • EC (2010) Final Report of a Mission Carried out in Canada From Oct 04 to 13 October 2010 in Order to Evaluate the Control Systems for Genetically Modified Organisms (GMOs) in Respect of Seed, Food and Feed Intended for Export to the EU, Health and Consumers Directorate General, 2010-8788 (Brussels: European Commission).
  • EC (n.d.), ‘GMOs in a Nutshell’, Available at: Http:// (accessed 12 November 2011).
  • EUR-Lex (2011), ‘Commission Regulation (EU) No. 619/2011 of 24 June 2011’, Available at: Http:// (accessed 22 May 2013).
  • Flax Council of Canada (2010), ‘GMO Flax Update 20 January 2010: Canadian Flax in the Brazilian Marketplace’, Available at: Http:// Brazil.pdf. (accessed 18 December 2011).
  • Gaisford, J. D., J. E. Hobbs, W. A. Kerr, N. Perdikis and M. D. Plunkett (2001), The Economics of Biotechnology (Cheltenham: Edward Elgar Press).
  • Gaisford, J. D., W. A. Kerr and N. Perdikis (2003), Economic Analysis for EU Accession Negotiations – Agri-food Issues in the EU's Eastward Expansion (Cheltenham: Edward Elgar Press).
  • Gruere, G. P. (2009), ‘Asynchronous Approvals of GM Products, Price Inflation, and the Codex Annex: What Low Level Presence Policy for APEC Countries?’ Paper presented at International Agricultural Trade Research Consortium (IATRC), Seattle, Available at: Http:// (accessed 23 November 2011).
  • Hall, B. (2011), ‘Triffid: A Year Later’, presentation at the Flax Council of Canada's Flax Day 2011, Saskatoon, 10 January 2011, Available at: (accessed 16 October 2011).
  • Hobbs, J. E. (1996), ‘A Transaction Cost Approach to Supply Chain Management’, Supply Chain Management, 1, 2, 1527.
  • Holtby, K. L., W. A. Kerr and J. E. Hobbs (2007), International Environmental Liability and Barriers to Trade (Cheltenham: Edward Elgar).
  • Isaac, G. E. (2007), ‘Sanitary and Phytosanitary Issues’, In W. A. Kerr and J. D. Gaisford (eds.), Handbook on International Trade Policy (Cheltenham: Edward Elgar), 38393.
  • Isaac, G. E. and W. A. Kerr (2003), ‘Genetically Modified Organisms and Trade Rules: Identifying Important Challenges for the WTO’, The World Economy, 26, 1, 2942.
  • Isaac, G. E. and W. A. Kerr (2007), ‘Whose Vision of the Future? The Entrenched International Conflict Over Genetic Modification’, The Geneva Post Quarterly, 2, 1, 87107.
  • James, C. (2008), ‘Global Status of Commercialized Biotech/GM crops: 2008’, ISAAA Brief 39, (Ithaca: International Service for the Acquisition of Agri-biotech Applications), Available at: Http:// (accessed 11 December 2011).
  • Olsen, A. R., J. S. Gecan, G. C. Ziobro and J. R. Bryce (2001), ‘Regulatory Action Criteria for Filth and Other Extraneous Materials V. Strategy for Evaluating Hazardous and Nonhazardous Filth’, Regulatory Toxicology and Pharmacology, 33, 3, 36392.
  • Phillips, P. W. B. and S. Smyth (2004), ‘Managing the Value of New-trait Varieties in the Canola Supply Chain in Canada’, Supply Chain Management, 9, 4, 31322.
  • Phillips, P. W. B., S. J. Smyth and W. A. Kerr (eds.) (2006), Governing Risk in the 21st Century (New York: Nova Science Publishers), 99102.
  • Smyth, S. and C. Ryan (2010), ‘Liability in the Canadian Marketplace’, presentation at the Total Utilization of Flax Genomics (TUFGEN) workshop, 23 February 2010. Available at: (accessed 1 November 2011).
  • Smyth, S., P. W. B. Phillips and W. A. Kerr (2009), ‘Global Governance Quandaries Regarding Transformative Technologies for Bioproducts, Crops and Foods’, Journal of World Trade, 43, 6, 1299323.
  • Statistics Canada (2011), Farm Product Prices, Crops and Livestock, Monthly, CANSIM @ CHASS, Tables: Manitoba (v31212177), Saskatchewan (v31212213), Alberta (v31212249), (Ottawa: Statistics Canada).
  • Stein, A. J. and E. Rodriguez-Cerezo (2010), ‘Low-level Presence of New GM Crops: An Issue on the Rise for Countries Where They Lack Approval’, AgBioForum, 13, 2, Article 8.
  • Stephens, D. (2010), ‘Risk Management Studies: The European Union Perspective’, presentation. Canada Grains Council 41st Annual Meeting, Fairmont Hotel Winnipeg, Winnipeg, Canada, April 19.
  • Vakulabharanam, V. (2011), Flax Testing Results – A Positive Outcome Due to Producers’ Efforts (Regina: Government of Saskatchewan), Available at: Http:// (accessed 29 November 2012).
  • Van den Belt, H. (2003), ‘Debating the Precautionary Principle: “Guilty Until Proven Innocent” or “Innocent Until Proven Guilty”?’, Plant Physiology, 132, 3, 11226.
  • Viju, C., M. T. Yeung and W. A. Kerr (2011a) ‘Post-Moratorium EU Regulation of Genetically Modified Products: Trade Concerns’, CATPRN Commissioned Paper, No 2011-02, Canadian Agricultural Trade and Competitiveness Research Network, 2011, Available at: Http:// (accessed 22 November 2011).
  • Viju, C., M. T. Yeung and W. A. Kerr (2011b), ‘Post Moratorium EU Regulation of Genetically Modified Products – Triffid Flax’, CATPRN Commissioned Paper, No. 2011-03, Canadian Agricultural Trade Policy and Competitiveness Research Network, Available at: Http:// (accessed 7 December 2012).
  • Wugar, D. and T. Cottier (eds.) (2008), Genetic Engineering and the World Trade System (Cambridge: Cambridge University Press).