Summary document of the FAO e-mail conference: "Regulating GMOs in developing and transition countries"

Executive Summary

It is important for developing countries to regulate genetically modified organisms (GMOs). Two main motivations are that GMO regulation allows developing countries to protect consumer health and the environment and/or to harness the benefits of these new technologies. While there is agreement about the need for a GMO regulatory framework, there are differences in opinion about how strict it should be, as this is influenced by issues such as costs, perceived risks and benefits of GMO release, enforceability and credibility of the regulatory framework. Regulation of some stages or components of the process can be stricter than others. Developing countries currently lack the resources and capacity to adequately regulate GMOs, although there are notable differences between individual countries in this respect, and there is an important need for capacity building activities in this area. Methodology for risk assessment is well described, but there is disagreement whether it can be appropriately applied to GMOs, given their novelty. The social, ethical and economic aspects of GMOs are important but it is not certain that they should be included in the regulatory framework. The risks of GMOs should be weighed against their benefits as well as the risks of alternative options. There is a strong division over whether GMOs should be regulated differently to non-GM varieties, with participants disagreeing whether it is the process (i.e. genetic modification or not) or the product (the kind of traits expressed) that should be the "regulatory trigger". Particular attention is needed for regulation of GMOs in countries that are also the centres of origin or diversity of agricultural species. There is general consensus that harmonization of regulatory systems across countries is important (and that existing international agreements/guidelines can assist in this context), but that it should also be possible to retain some country-specific elements in the systems. Co-ordination and harmonization of GMO regulation between the different relevant government ministries within a country is also important. Developing countries wishing to establish a GMO regulatory framework can learn a lot from, but do not need to model it on, the existing regulatory frameworks in developed countries. There is general support for involving the public in GMO regulatory processes; informing the public about GMOs (including labelling of GM products); and ensuring transparency of the regulatory processes. Monitoring implementation of a GMO regulatory framework may be especially difficult in developing countries due to lack of resources, although some issues are difficult to monitor even for resource-strong developed countries. The cost of regulation, including post-release monitoring of GMOs, is an issue of concern for developing countries, although strategies to reduce it can be considered. The question of liability is important and should be covered in the GMO regulatory framework.

These were some of the main themes and outcomes of a moderated e-mail conference entitled "Regulating GMOs in developing and transition countries", hosted by the FAO Biotechnology Forum from 28 April to 1 June 2003. About 400 people subscribed to the conference and 93 messages were posted over the 5-week period by 44 participants from 20 countries, with half of the messages coming from people living in developing countries. Most messages came from people working in research centres, universities, NGOs and as private consultants, with the remainder from people working in government bodies, UN agencies or farmers' organizations.

1. Introduction

The theme of the 9th conference of the FAO Biotechnology Forum, which took place from 28 April to 1 June 2003, was "Regulating GMOs in developing and transition countries". As stated in the Background Document to the conference, regulation has always been a central part of the general debate about genetically modified organisms (GMOs) and it is especially important because of the potential implications of GMOs for the environment and human health and because of the impacts that regulation can have on the trade of GM products and on the research and development climate for GMOs, in what is still a relatively new field. The topic has been given considerable attention by policy-makers, both at the national and international level. For example, over 100 countries have ratified the Cartagena Protocol on Biosafety, a supplementary agreement to the Convention of Biological Diversity that entered into force in September 2003, which seeks to protect biological diversity from the potential risks posed by living modified organisms (LMOs, i.e. living GMOs), specifically focusing on transboundary movements.

Regulation of GMOs in developing countries presents serious challenges for policy-makers. Vasanthi (in the 8th message posted in the conference) described them well, writing "Much debate is going on as to what, how and when to regulate GMOs, in international circles. Under these circumstances, developing countries are faced with the challenge to put in place a regulatory system that can meet the changing context/developments in scientific and international fields as well as to build the capacity required to establish such a system. Regulations for GMOs have been established in many countries adopting these foods, but differences in the basic philosophy of regulation between these countries has created much difficulties in assessing the extent and type of regulation that is required for GM foods. Added to this, are the complexities of regulatory guidelines established by international bodies like WHO, FAO, WTO, OECD and CBD which are affecting international trade and the release of GMOs in individual countries. These developments are likely to have much influence on the acceptance and release of GMOs in developing countries, most of which are as yet to establish regulatory frameworks for GM foods".

During the 5-week conference, a total of 93 messages were posted, each one numbered in order of posting. The aim of this Summary Document is to provide a summary of the main arguments and issues discussed during the conference, based on the participants' messages. Specific references to messages posted, giving the participant's surname and message number, are provided. All messages can be viewed at the Archives of Conference 9. Note, in the Forum, participants are always assumed to be speaking on their own personal behalf and not on behalf of their employers, unless they state otherwise.

A total of 401 people subscribed, which was the highest number for any of the conferences held up to that date, reflecting the large interest in this subject, and roughly half of the messages posted came from participants living in developing countries. Although the conference aimed to cover the agro-industry, crop, fisheries, forestry and livestock sectors, most of the discussions, when referring to specific agricultural situations, considered the crop sector. The large areas cultivated with GM crops (overview given in the Background Document), and the resultant public and media debate about them, are no doubt responsible for this focus. In GM livestock for example, Wollny (87) said they are lagging far behind the scientific and public discussion in crops.

In Section 2 of this document, the main elements of the discussions are summarized under 11 topics (2.1-2.11). Section 3 provides some information about participation in the conference and Section 4 gives the name and country of the people that sent referenced messages. Section 5 provides an explanation of abbreviations used in the document. Note, the term 'developing countries' used in this document includes countries with transition economies.

2. Main topics discussed

2.1 Is it important for developing countries to regulate GMOs?

Participants agreed that development of a regulatory framework for GMOs was important for developing countries. Two main motivations were provided. The first was that it would allow developing countries to harness the benefits of these new technologies i.e. to avoid being bypassed by the "gene revolution", countries needed "the stable and predictable regulatory regimes necessary to create an enabling environment for the application of agricultural biotechnology"(MacKenzie, 5). Similarly, Morris (73) felt that if developing countries saw potential benefit from introducing GM technology, then the effort of developing a regulatory framework would be worth it. She also suggested that subsequent familiarization with GM technology could benefit science in the country as a whole.

The second, more frequently cited, motivation was protection of consumer health and the environment in developing countries (e.g. Bhat, 34; Rajaratnam, 80). According to Villaverde (25), if the regulatory framework of a developing country was weak or inexistent "its consumers could be exposed to potential risks, and this is unacceptable". Particular mention was made of the importance of protecting consumers in developing countries from non-approved GMOs in food aid (Vasanthi, 8; Bhat 14, 34). Kambikambi (31) also highlighted the situation in her country, Zambia, when GM maize provided as food aid was refused, writing "I suppose that underscores the importance of having a regulatory framework because what really happened is that we did not know how to handle that product in the absence of appropriate legislation". McCowen (93) noted that developing countries might be affected by foreign assistance policies of donor countries linking acceptance of medical aid to acceptance of food aid that might be GM.

2.2 How strict should the regulatory framework be in developing countries?

While participants agreed about the need for a regulatory framework for GMOs, they disagreed about how strict it should be. Several participants warned that strict regulation in developing countries, requiring substantial financial and human resources inputs, would penalize public sector GMO research initiatives directed toward developing country needs and would allow multi-national corporations (MNCs) to continue dominating the area with their focus on major crops and traits (Strauss, 1; Quemada, 4; MacKenzie, 6; Morris, 13). Quemada (4) pointed out that in the United States the cost of regulatory data collection and compliance had already driven most small players out of GM crop work, so that it was now being carried out by "major companies who can afford to spend the money and who have the appropriate staff to deal with the regulatory requirements". While not suggesting that regulation be abandoned he, supporting Strauss (1), proposed streamlining the process for traits that could be considered low risk and for crops with traits that could solve major problems. Morris (13) argued that there was enormous potential in Africa for improvement of crops such as sorghum and millet using GM technology, but asked "who will carry the cost of undertaking the biosafety research to ensure their safe introduction?". Morris (73) concluded that the regulatory framework should not be "large and cumbersome" because it only introduces extra costs and because the number of applications for use/release of GMOs in developing countries was likely to be relatively low, at least in the beginning.

Other participants, while recognizing that GMOs have a lot of potential, felt that rigorous regulation was nevertheless needed in developing countries (Ombori, 75; Muralidharan, 88; Pena-Neira, 89). In addition, Muralidharan (57) pointed out that as an alternative to relaxing regulations and allowing ready access to GMOs in developing countries, other "simple, cheap and safe" technologies were available. Richardson (79) was adamant about the need for strict regulation, arguing that, because of the harm they could cause if released into the ecosystem and found to be detrimental, introduction of new GMOs "requires more conservative regulation than either novel pesticides or novel medical treatments". The differences in perspective on this issue were also reflected in the exchange between Ashton (56) and Blanchfield (58) about current GMO regulations in different countries, with Ashton (56) arguing that most merely facilitate introduction of GM crops, whereas "regulations must regulate, not facilitate" and Blanchfield (58) responding that they should regulate but also facilitate.

A couple of participants also raised the issue of the relationship between strictness of the regulations and the ability to enforce them. Prakoso (50) argued the strict regulations could be adopted but, given financial and technical limitations of developing countries, they might not be enforceable. Similarly, Jackson (33) maintained it was important to have legal regulations that stipulate conditions for production or marketing of GMOs, but felt that "if these conditions are unenforceable then the regulatory framework has only limited relevance to what is occurring on the ground". Prakoso (50) argued therefore for regulation that should be practical, low cost and implementable. Also related to this issue of enforceability is the question of whether legislation governing the subject is binding or non-binding (Vapnek, 22; Kambikambi, 31; Jackson, 33).

For Willemse (17), the question of how strict regulatory frameworks should be in developing countries was not a simple one to answer and "would mostly be determined by each country's specific needs and circumstances". He felt that one of the most important aspects of a national regulatory framework was the credibility that the implementation of such a framework would enjoy at national level and that, to achieve this, regulatory frameworks in developing countries would probably need to be, at least initially, stricter than in developed countries to ensure acceptance not only of the framework, but also of subsequent approvals and/or refusals. He suggested that over time the framework might then evolve to become less strict.

As mentioned in Section 3 of the Background Document, regulations governing GMOs can cover activities at a number of key stages, including research and development (R&D) of GMOs, commercialization of GM products or the import of GM material or food. It can be a lengthy process. According to Rao (37), "it takes about 11-13 years for a specific transgenic variety to get into commercial cultivation. Five years to develop the transgenic event, such as pest resistance or herbicide resistance, 2 or 3 years of controlled greenhouse trials on approval by a regulatory agency, and 3 or more years of controlled field trials". Regulation of some stages or components of the process could be stricter than others. For example, MacKenzie (5) argued for a clear distinction to be made between experimental field trials, allowing GM crops to be assessed prior to commercial release, and the subsequent commercial release of GMOs into the environment, maintaining that the focus for the former should be on implementation of risk mitigation strategies (the terms and conditions necessary to safely permit confined trials) instead of on rigorous risk assessment, which should be the case for the latter. He argued that a "permissive environment for the conduct of experimental trials" was important for local R&D investment and for providing biosafety committees and regulatory officials with experience and expertise. Rao (37, 46) also argued that if a given GM event had been approved for commercial release in a given crop, then GMOs with the same GM event in different varieties of the same crop should be allowed to follow a shorter trial stage. Muralidharan (57) agreed with this, although McCowen (41) argued that such a system might deter companies from being the first to introduce new traits onto the market. Strauss (1) also proposed that GMO regulations should distinguish between GM products resulting from modification of native genes and GMOs expressing novel proteins or anti-pest toxins, arguing that the former should not be regulated as if they were 'potential environmental menaces'. While not opposing this proposal, MacKenzie (6) noted that incorporating this "categories of risk" argument into regulatory systems required a fundamental rethink of the existing approach to risk assessment of GM plants.

2.3 Developing countries lack the resources and capacity to adequately regulate GMOs

This was a common refrain from participants in the conference (e.g. Mog, 16; Kuhn, 29), with Bhat (48) suggesting that individual developing countries might not have the "resources, infrastructure and technical manpower" to even review the regulatory data.

The poor financial resources of developing countries and the potentially high costs of regulation were issues of concern. Ashton (35, 56) described the cost of rigorous regulation as "onerous", involving payments for inspectors, transport, laboratory procedures and consultations. Badr (38) suggested that developed countries could assist by providing funding while Morris (73), in a similar vein, proposed that developing countries should make maximum use of available resources internationally and in the developed world to assist them with risk assessment. Ashton (35, 56) also felt it was important to consider who should pay. He argued that it was wrong for the taxpayer in developing countries to foot the bill and proposed, instead, that parties wishing to introduce GM products onto the market should pay for the costs of regulation.

The lack of knowledge, experience and capacity required for GMO regulation was raised on many occasions throughout the conference. In addition, capacity levels can differ greatly between developing countries. As Morris (13) wrote, many developing counties, "do not have the basic tools of molecular biology in place", while Bhat (34) suggested that for GM crop/food issues, developing countries could be placed in two categories, those with 1) infrastructure for biotechnology (e.g. Brazil, China, Egypt, India, Malaysia, Mexico, South Africa) or 2) practically no work or expertise in the area of biotechnology. In contrast to most developed countries that could build their regulatory expertise together with advances in biotechnology, Lekoape (12) pointed out that developing countries lack this advantage, although Willemse (21) suggested that it was becoming more frequent for technological capacity and regulatory framework development to go hand in hand in developing countries. Prakoso (50) described some of the technical difficulties (lack of suitable equipment, reagents) developing countries face for detecting GMOs, while Vasanthi (8) and Bhat (48) also emphasized the importance of developing countries having functional laboratory facilities for detection of GMOs. Kuta (45) said that Nigeria lacked "the required quantity and quality of human capacity for scientific assessment of possible environmental and health risks associated with GE-products". Bhat's (14) conclusion was therefore that "the expertise and infrastructure needed to undertake a critical, transparent, valid scientific assessment of the food and environmental safety [of GMOs] are either not existing or are in a rudimentary state of development in most of the [developing] countries".

Given this situation, some participants emphasized the need for capacity building activities (e.g. Lekoape, 12; Olutogun, 15; Kuta, 45; Nath, 77). There were, however, different ideas about how this should be done. Olutogun (15) stressed the contribution that the developed world could make in helping the developing world with capacity building. Willemse (21) also noted that most capacity building initiatives currently modelled development of regulatory frameworks on the existing frameworks in developed countries. Lekoape (23), however, argued that most developing countries did not have the resources, or perhaps the need, to establish similar regulatory frameworks to those in developed countries and concluded "it is therefore imperative that capacity building initiatives are demand-driven. Developing countries should not feel obliged to follow in the footprints of the developed nations". This viewpoint was supported by Mog (26), calling for "locally-controlled and demand-driven" capacity building initiatives, who noted that developing countries could learn from the experiences of other countries but did not have to follow their model. Rajaratnam (91), echoing Acikgoz (83), proposed that international organizations like FAO could help developing countries to draft their regulations by providing expertise, advice and training.

2.4 The approach to risk assessment and risk management

As described in the Background Document, one of the main elements to be addressed in a GMO regulatory framework is the approach to risk assessment and risk management, involving issues such as how to assess the risk from GMOs, whether to weigh potential risks against potential benefits and whether (in addition to environmental and human health aspects) to consider economic issues, social impacts or ethical concerns. These topics were amply discussed in the conference. Methodology for risk assessment was described, but there was disagreement whether it could be appropriately applied in this situation, given the novelty of GMOs. While some participants argued that social, ethical and, in particular, economic, aspects should be included in the regulatory framework, a small number argued this was not appropriate. Several participants also noted that the risks of GMOs should be weighed against possible benefits of GMOs and the risks of alternative options.

Muir (59) reminded participants that there was an entire field of science devoted to methodology for risk assessment, whereby a number of ways were available to estimate risks of potential hazards before eventually releasing any new product onto the market. His mention of cars and airplanes in this context, evoked a series of comments regarding the differences and similarities between their risk assessment and those of GMOs (Doebel, 62; Blanchfield 63; Wuerthele, 66; Willemse, 86). Muir (59) noted that there is no such thing as a zero risk of anything and that risks are expressed in terms of probabilities, e.g. there is an 80% probability of rain (if rain is identified as a hazard). Doebel (62) argued, however, that because genetic modification was new, there were no precedents that would allow the probabilities to be established and so proper risk assessment could not be carried out. Similarly, Muralidharan (88) felt it was too early to claim that a fairly good assessment of risks was possible. Muir (70) maintained, however, that probabilities of environmental hazards could be quantified, based on our knowledge of how natural selection works. Phillips (82) also argued that the environmental impact of releasing a GM crop "could vary widely depending on the indigenous flora and fauna and on the nature of the cultivated and uncultivated areas, and their interaction", thus requiring location-specific assessments. Regarding the probabilities associated with human health hazards, Phillips (82) suggested these should be assessed within cultural-economic-social populations, as diets and susceptibility to allergens can vary widely between populations.

Wuerthele (66), on the other hand, agreed with Doebel (62) that the lack of experience with genetic modification meant there was no basis for proper assessment of the risks associated with GMOs, arguing, in addition, that all of the hazards of GMOs were not yet known. Phillips (82) maintained, however, that characterization of the hazards related to GMOs was unlikely to vary greatly between populations or ecosystems, and that it would seem appropriate to carry out one good characterization effort. Regarding human health, he proposed that the key hazards are toxicity, allergenicity and compositional change. Regarding environmental impacts, he proposed that invasiveness, outcrossing and harm to non-target organisms would be examples of the hazards. His conclusion was that "regulation of the health and safety aspects should rely fundamentally on internationally characterized hazards, generally accepted methods and locationally relevant assessments of exposure".

[Note, application of a science-based risk assessment framework was one of the main topics of discussion in a previous conference (on GMOs and gene flow) hosted by the FAO Biotechnology Forum in 2002 (discussions summarized here].

Several participants mentioned the importance of including these aspects, in addition to human health and environmental impacts, in the regulatory framework. For Morris (73), while they might not be safety issues per se, they were important for the cost-benefit equation as well as ensuring the long term acceptance of the technology. Bucchini (74) argued that safety issues regarding GMOs could be resolved at the international level whereas social, economic and ethical issues should be debated and decisions made at the national level. Villaverde (72) argued that regulation of socio-economic risks from GM foods and organisms was the main regulatory gap in developing countries, emphasizing the need to consider seed monopolies held by MNCs and the economic cost/benefits of introducing GM crops. Newman (2) felt that as introduction of a GM variety to a country was supposedly based on economic reasons, it was essential to include a "comprehensive, unbiased economic assessment as part of the risk analysis process. This risk assessment must include the impact on the non-GM grower and associated industries that may be affected". She was appalled that developed countries had excluded this aspect in their GMO legislation. Vasanthi (9) agreed with her on the need for economic risk assessment of GMOs, arguing that it could be included as part of the post-market monitoring procedures where, ideally, "data on the economics of the entire process of cultivation, harvesting, marketing, traceability and consumption of GM crops would be needed".

Richardson (51) also felt that a regulatory framework should include the aspect of "who benefits", a point considered in detail by Mog (16, 84), who maintained that because GMOs developed for profit might "make poor farmers dependent upon GM products that have been designed to increase the profits of foreign corporations", the regulatory framework should be stricter on these GMOs (potentially restricting their access to domestic markets) than on GMOs not developed for profit. Willemse (86) disagreed. Arguing that regulation should be risk-based, he maintained that a distinction between for-profit and not-for-profit GMOs for regulatory purposes would incorrectly assume that the risk lies with the objective or purpose of the GMO rather than the technology or the product. Phillips (82) felt these kinds of concerns should not be added to health and safety regulations as this would raise the risk that safe and possibly beneficial GMOs would be rejected or that unsafe GMOs would be approved by the regulatory framework in developing countries. He concluded that "most other aspects that concern people about GM foods--e.g. industrial structure; distribution of winners and losers; social impacts; moral aspects--while raised in the context of GM foods are not unique to GM foods and probably are better handled in the context of a broader development policy".

Several participants noted that carrying out a risk assessment of GMOs was not the final result. On the one hand, risks should be weighed against possible benefits (Richardson, 11; Muir, 59; Efaw, 67), with Blanchfield (78) arguing that potential benefits should be measured primarily for those suffering hunger and malnutrition. On the other hand, they should be weighed against possible alternatives, such as doing nothing (Muir, 59; Blanchfield, 78) or using a different technology (Richardson, 51; Muir, 59; Blanchfield, 78; Hongladoram, 92). For example, Richardson (51) suggested that a regulatory framework needed to consider whether alternative solutions that cost little with less risk had been examined.

2.5 Regulation of GM versus non-GM products

In most current regulatory systems, GM crops are more strictly regulated than non-GM crops. This can be an incentive for employing other biotechnologies (such as genomics or tissue culture) instead of genetic modification (Newman, 7). In addition, if GM crops require lengthy trial periods this may give enough time to non-GM varieties to overtake them (Rao, 37). Participants were strongly divided on whether GMOs should be regulated differently to non-GM varieties.

Muir (65) referred to a 2002 publication from the United States National Academy of Sciences which concluded that specific traits produced by either conventionally bred or GM plants could pose unique risks and that conventionally bred plants should therefore be evaluated using the same regulatory process as GM plants. He noted that it was possible to use the same methodology to evaluate the risks from GM and conventional plants and that the regulatory framework could be the same. Willemse (86) highlighted the negative impacts that some non-GMO related activities could have, such as the import of invasive alien species or the introduction of agricultural pests carried by crops. For him, this showed that "the same regulatory criteria are not being applied for GMOs and for their non-GMO equivalents" and he argued "if we continue to apply different criteria in risk assessment and risk management based on our perceptions and individual likes and dislikes, we will continue to generate disasters, while stifling development that is needed above all by developing countries".

A number of participants pointed out that the range of potential techniques available for conventional breeding included some with potentially large effects on the genetic material, such as hybridization, mutagenesis (e.g. using irradiation) and polyploidization and that crop products developed using these techniques should not be regulated differently than GMOs (Rao 28, 43; Blanchfield, 63; Muir, 65). Doebel (69) disagreed arguing, inter alia, that mutagenesis and polyploidization were not part of traditional or conventional breeding. Regarding hybridization, Richardson (51) argued that the consequences (at the cell and organism level) of genes moved by hybridization were not equivalent to those introduced by molecular techniques (and so the regulatory implications/consequences were not the same).

Some participants maintained that the process of genetic modification was unique, requiring special regulation, and was not comparable to conventional breeding (e.g. Muralidharan, 88). Richardson (51) argued that insertion of a DNA segment "cannot be assumed to be neutral or equivalent to any normal cellular process until appropriately tested" and that GMOs should be strictly regulated because they could cause non-reversible damage (Richardson 51, 79). Doebel (62, 69) maintained that genetic modification was not similar to conventional breeding as non-targeted insertion of DNA from one species into the DNA molecule of another species was "problematic in an unprecedented way" because of uncertainty about interaction of the inserted DNA with the recipient DNA molecule and interaction between the resulting DNA molecule and the environment. Regarding regulation, he therefore urged caution. Blanchfield (63, 78), however, argued that conventional breeding resulted in random insertion of unspecified and unknown numbers of genes and therefore "whatever the problems of genetic modification, they are at least matched if not surpassed by those of 'breeding'". Efaw (67) maintained instead that conventional breeding randomly combined genes already present in the species and that the genes had already been through countless iterations of checks and balances.

As seen above, discussions basically come down to whether it is the process (i.e. genetic modification or not) or the product (the kind of traits expressed) that is the "regulatory trigger" i.e. the criteria making a product subject to regulatory assessment. Participants arguing for GMOs and non-GM products to be regulated in the same way, obviously felt that "product" should be the trigger (Muir 65, 70; Prakoso, 76), whereas those arguing for stricter regulation of GMOs than non-GM products (even if the traits produced (e.g. herbicide tolerance) are the same) felt "process" should be the trigger. Most existing regulatory systems use process rather than product as the trigger (MacKenzie, 24; Willemse, 30). This distinction has interesting implications for a specific case raised in the conference. If a "stacked" GMO variety (i.e. with two or more transgenes inserted) is developed by crossing two GM parental lines that are already approved, it was argued that to have a consistent regulatory policy, the stacked variety should not require regulatory approval (Willemse 17, 21, 30; MacKenzie, 24). As MacKenzie (24) concluded, "for countries with "process-based" regulatory systems to invoke the "product risk" argument only for the special case of stacked events, is confusing at the least".

2.6 Centres of origin or diversity

Some specific attention was given during the conference to the topic of regulating GMOs in countries that are also the centres of origin or diversity of crop species. Specific examples mentioned were work on development of GM potatoes in the Andean regions of Peru (Buijs, 49) and, especially, reports of GM maize in Southern Mexico (Bucchini, 74; Pena-Neira, 89). Willemse (21), echoed by Diaz (39), felt that the latter example could influence development of regulatory frameworks in countries with centres of origin of agricultural species (as well as the revision of existing frameworks of countries exporting to such countries). For Acikgoz (83), centres of genetic diversity should be considered as a key point in GMO legislation, although he felt it would be difficult to decide whether to ban or permit cultivation of economically useful GMOs in these areas. Willemse (86) argued, instead, that no compromise should be made on the principle of "not allowing a GMO into the species center of origin". Morris (13), supporting Quemada's (4) comments about the high cost of getting adequate data for regulatory purposes, argued this was becoming a critical issue in any developing country where crops have their centre of origin and where there is little documentation of the potential for cross-pollination with wild relatives.

2.7 Co-ordination and harmonization of GMO regulations between countries

As mentioned in the Background Document, the majority of developing countries, in contrast to developed countries, currently lack regulatory systems for GMOs. Should developing countries aim to establish similar and harmonized systems or can they be unique and country-specific? Should they be modeled on existing systems in developed countries and what role can international agreements play? In discussions on these issues, most participants seemed to feel that harmonization of regulatory systems across countries was important (and that existing international agreements/guidelines could assist in this context), but that it should also be possible to retain some country-specific elements in the systems.

For Villaverde (25), a strong reference regulatory framework was needed worldwide because "this discussion on GM foods is global, and [we] therefore need to have global solutions". A number of different arguments were provided in favour of cross-country harmonization. Because of the widely divergent views held regarding safety of GM foods, Vasanthi (8) argued that international harmonization of risk assessment procedures was needed urgently. Bucchini (74) felt that as prevention of gene flow and movement of GM material within and between countries was not feasible in most parts of the world (i.e. "low level flow of GMOs cannot be prevented"), appropriate safety levels should therefore be determined at the international or regional level. Regional co-operation was also promoted by Morris (73) as a way of pooling limited resources and reducing the necessity for creating individual regulatory mechanisms in each country. Phillips (82) argued that the hazards related to GMOs and GM food were unlikely to vary greatly between ecosystems or human populations and that international harmonization should therefore be the standard.

In order to achieve harmonization of GMO regulations across countries, some participants proposed that existing international agreements/guidelines could be used, such as the Principles and Guidelines adopted by the Codex Alimentarius Commission in July 2003 regarding GM foods (Olutogun, 90; Phillips 82; Villaverde 25, 55, 64; Vasanthi, 44) and the Cartagena Protocol on Biosafety (Villaverde, 25). Vasanthi (8, 44) and Rao (52) pointed out that uncertainty exists about the relationship between a number of multilateral agreements that are relevant for GMOs. Referring to the Codex Guidelines, the Cartagena Protocol and the WTO Agreement on the Application of Sanitary and Phytosanitary Measures, Vasanthi (44) wondered whether international harmonization of GM food safety standards might help in reducing such conflicts. Bhat (34), supported by Rajaratnam (80), called on FAO, IAEA, UNEP and WHO to work together to develop a model regulatory framework that individual developing countries could then adapt for their purposes. Similarly, Olutogun (90) called for development of a global regulatory mechanism, involving the Codex Alimentarius Commission and CGIAR centres, to assist developing countries.

Although promoting development of harmonized regulatory systems in developing countries, participants maintained that it should be possible to include some country-specific elements/needs in the framework (e.g. Vasanthi, 8; Mog, 26; Bhat, 34; Rajaratnam, 80). Willemse (17) argued that flexibility for individual country needs should be allowed, given the importance of credible implementation of the framework at the national level, a point also made by Mog (26). Bucchini (74) proposed that whereas human health and environmental safety of GMOs should be decided at the international level, the social, economic and ethical issues regarding GMOs should be debated and decisions made at the country level. Hongladarom (92) concluded that "no matter what kind of regulatory schemes be put in place in the so-called Third World countries, those schemes need to be in accordance with the need for those countries to find a way to flourish and prosper in their own terms".

Lekoape (23) emphasized the importance of each country's unique set of circumstances, i.e. their "level of development, capabilities, aspirations, cultures and traditions", and that these should influence adoption and regulation of any technology. Willemse (27) supported this, proposing that "adoption of technologies (or their products) and regulation should always be driven by domestic demand and influenced by domestic circumstances". Willemse (30) was also adamant that for countries receiving GMOs "one principle that should not be negotiable is the right of the recipient country to decide its own criteria and level of biosafety". McCowen (32), however, responding to Lekoape (23), questioned how independent and individual the regulatory systems of developing countries could be, given their reliance on trade and the importance of WTO agreements on trade. Similarly, Villaverde (25) maintained that for a developing country wishing to export GM food to the developed world, "the regulatory framework in a developing country does not have much degree of freedom in relation to the regulatory framework of the importing country".

As mentioned earlier, most current regulatory frameworks have been established in developed and not developing countries. For developing countries wishing to establish a regulatory framework, there was consensus that, although they could learn a lot from the experiences of developed countries, they did not have to follow in their footsteps (e.g. Lekoape, 23).

Bhat (48) provided some details about differences between the regulatory systems established in the United States and the European Union (systems that were increasingly diverging, according to Hongladarom (92)), the two primary regulatory models available in developed countries. He concluded that "developing countries need general comprehensive regulations and separate regulations evolved on a case-by-case basis depending on the need of each country". Participants emphasized that experience with implementation of GMO regulatory systems in developed countries could provide valuable information for developing countries wishing to establish their own systems (Mog, 26; Ashton, 56; Bucchini, 74). For example, Wuerthele (66) suggested a number of possible lessons that developing countries could learn from experiences in the United States, such as the need to consider legal frameworks which will be flexible enough to regulate the wide range of GMOs that might be developed in the future. In addition, Willemse (21, 27) pointed out that the experiences and developments in developing countries with existing regulatory frameworks could also provide some important lessons for other developing countries in establishing regulatory frameworks as well as benefiting subsequent evolution of frameworks in developed countries. Willemse (27) concluded therefore that "the global GMO regulatory scenario is evolving into a dynamic interdependent network that could only benefit from cross-fertilization of all experiences and lessons learnt".

2.8 Co-ordination and harmonization of GMO regulations within a country

In the same way that participants felt that harmonization of regulatory systems across countries was important, there was also an appreciation of the need for regulatory harmonization within countries. Regulation of GMOs touches on issues relevant to several different government ministries within a single country. According to Bhat (14), the issues of GM food/feed/crops "cannot be tackled by a single ministry and coordination between ministries of agriculture, health, environment, science and technology, commerce is essential". For example, Kuta (45) pointed out that there would be five major federal agencies, from four different ministries, involved in regulation of GM crops in Nigeria. Rajaratnam (91) suggested that the GMO regulatory framework would need to be drafted by a body involving ministries and experts working in areas such as agriculture, forestry, livestock, health, nutrition and environment. This may present some challenges, as different ministries or ministerial committees may differ in their approaches to GMOs (e.g. Rao, 28) leading to situations where "one ministry is pushing for more research on GMOs, while the other is trying to put a damper on it" (Hongladarom, 92).

Wuerthele (66) argued that the experiences from GMO regulation in the United States demonstrated the importance of determining in advance who will review the health, environmental, food safety and social effects of GMOs and how those reviews will be coordinated. Bhat (14) suggested that a single co-ordinating agency be established that would "act as a single window for interacting with the risk assessors, risk managers and risk communicators and all the stakeholders, including the industry, farmers and consumers". Similarly, Rajaratnam (91) argued for a single body to monitor GMO regulation rather than having different monitoring body systems for different aspects of GMOs. In the Background Document, it was suggested that development and enforcement of a GMO regulatory framework might need to be co-ordinated within a cross-sectorial national approach to the management of biological risks associated with food and agriculture, a concept referred to as Biosecurity by FAO. Willemse (17, 86) supported the suggestion, arguing that the 'one-stop-shop' envisaged in the concept could provide a solution to the issue of capacity constraints.

2.9 Public participation/awareness and GMO regulations

As described in the Background Document, one of the key elements to be considered when developing a regulatory framework relates to transparency and public involvement in the decision making processes. In the conference there was general support for involving the public in these processes; informing the public about GMOs (including labelling GM products); and ensuring transparency of the regulatory process.

For Lekoape (23) the most important lesson that developing countries could learn from developed country experiences with GMO regulation was to engage the public at all levels, from drawing up a research policy to making decisions about individual approvals, arguing that "this bottom-up approach means consumers are not only a part of the process, they identify with it and assume responsibility for the end result, thus endorsing the credibility of a regulatory framework". Mog (26), based on his research and experiences in southern Philippines, was convinced of the importance of involving local people in the process of researching and developing both technologies (like GMOs) and the policy measures necessary to regulate them because it tailors them to local circumstances; increases local credibility; and helps the local people to face unknown future challenges in this area. Morris (73) pointed out that in Africa, the traditional decision-making processes were community-based, and decisions made in this way were more likely to achieve buy-in than decisions imposed by scientists.

Although the importance of public involvement in regulating GMOs was supported during the conference, Kuta (60) also noted that, at least in his country Nigeria, there was still low public awareness about issues surrounding GM crops. He therefore called for more funding to be provided for public awareness projects in this area. Nath (77) also urged that farmers and the public be provided with information to enable them to make an informed choice about the use of GM crops and that the information should not come only from GMO firms. Regarding public awareness about GMOs, there was also specific discussion in the conference about the importance of labelling GM products, where there was general agreement that labelling was needed in developing countries (e.g. Bhat 34, 38; Ombori, 75). For example, Nath (77) argued that labelling was essential because, if unlabelled, GM crops could enter the markets without a conscious decision being made about them by the agro-intermediaries. Labelling practices can vary between countries (Prakoso, 50; Wuerthele, 66; Rao, 68) and Ashton (56) and Rao (68) supported international harmonization of labelling requirements. Prakoso (50) noted the technical difficulties that developing countries with limited resources face regarding GMO detection and proposed that reasonable enforceable labelling requirements should cover GMOs as raw materials but not as food products. Prakoso (78) also proposed that consumers should be informed about the safety of released GMOs so that labelling of GM materials would not negatively impact their marketing.

Many participants called for transparency in the regulatory processes (e.g. Bhat, 14; Ashton, 35; Muralidharan, 88). For Rao (54), regulatory data should be made available to scientists and the interested public and not considered in secrecy, arguing that there were anti-technology lobbies throughout the world who pressurized governments and regulatory committees to impede the import or release of GMOs. Similarly, Ashton (35, 56) complained about closed and secretive regulatory processes, arguing that powerful lobby groups were operating to facilitate introduction of GM crops. Diaz (39) and Ashton (56) emphasized that regulators and regulatory bodies should not have any conflicts of interest. Diaz (39) was also concerned about use of confidential information in the regulatory processes concluding that, for any GMO application, the public had the right to know as much as possible about the GMO involved.

2.10 Monitoring implementation of GMO regulations

Once a regulatory framework has been put in place (defining the procedures for GMO approval; the kinds of GMOs that might be approved; how and where they may be released etc.) and applications for release of individual GMOs have been approved, monitoring of implementation of the regulations is needed. Participants highlighted the fact that monitoring may be especially difficult in developing countries due to lack of resources, although some issues can be difficult to monitor even for resource-strong developed countries.

Among others, Vasanthi (8) pointed out the importance of monitoring of GMOs, particularly in the field during and after cultivation, and during marketing, and that attention needed to be given to the kind of methods required, particularly approaches for preventing/checking unapproved cultivation of GM crops and checking for compliance with cultivation procedures of GM crops. Morris (73) noted, however, that the costs of monitoring compliance with any legislation could be high, and that GMOs were no exception in this respect. Ashton (35) argued that even for an "advanced developing nation" such as South Africa, there was insufficient capacity to properly and independently monitor or regulate all the trials and general releases of GM crops. A number of solutions to this problem of resources was proposed. Kuhn (29) suggested "industry self-policing" for developing countries that lack the necessary resources and expertise, whereby biotechnology companies selling GM crops would provide some oversight of post-market use of GM crops by farmers. He argued that this could be used at least until the developing country's government was able to assume a greater oversight role. McCowen (40), however, was not convinced that such a proposal might be suitable for countries other than the United States. Another potential solution came from Morris (73) who proposed that, to avoid duplicating functions and incurring additional costs due to monitoring GM crops, people working in the existing agricultural inspection service, as well as extension officers, could be trained in issues regarding GM technology to take over these tasks. Another solution proposed was that the costs could be met by the GMO producer. For Richardson (79), the regulatory agencies should be responsible for enforcement, but they should charge the seller for the costs of testing and enforcement. Similarly, Rajaratnam (81) suggested that any party who imports GM seeds/seedlings or micro-organisms into a country should have the responsibility for setting up the monitoring system.

Although lack of resources was considered a problem for monitoring implementation of GMO regulations in developing countries, participants noted that some areas of GMO regulations were hard to monitor regardless of the resources available. One area is implementation of a refuge strategy, where Bt crops (i.e. producing insecticidal toxins using genes from the bacterium Bacillus thuringiensis (Bt)) are planted together with non-Bt 'refuge' areas of the same crop, to delay the development of resistance to Bt toxins among the pests. According to McCowen (10), implementing refuge strategies in North America had proven to be "basically impossible" as, although the seed dealers ensured that farmers buy the correct ratio of Bt and non-Bt seeds, planting was ultimately left up to the farmers and many did not know the correct refuge size or shape (McCowen, 93). Morris (13) and Muralidharan (88) pointed out that implementation might be even more problematic in developing countries, where illiterate or poor farmers on small farms might ignore the need for adequate non-Bt refuges. Monitoring the implementation of refuge strategies is, however, burdensome at the farm level as there is no quick mechanism for distinguishing GM from non-GM products (Jackson, 33). A second area concerns regulations covering tolerable limits for GM material in products identified as "non-GM". The regulations might be difficult to implement and monitor because of the diversity of products that could contain GM material (Jackson, 33) and because of technical limitations of procedures for detection of GM material (Prakoso, 50; Muralidharan, 57). A third area concerns regulations covering GMO gene flow, where Ramsaroop (42), supported by Buijs (49), noted that in developing countries such as Guyana there was prolific exchange of plant genetic materials between farmers making it difficult to control the movement of GM crops.

2.11 GMO regulations and liability

If GMO regulations have been infringed and some economic damage has been done, who is liable? Participants argued that the issue of liability is important and should be covered in the GMO regulatory framework.

Diaz (39) argued that the national solution to issues about legal responsibility for GMO introduction and payment for potential environmental or health damage should be given in the regulatory framework and that it was important to clearly assign such responsibilities before any undesirable events occurred. Wuerthele (66) bemoaned that in the United States, many legal, ethical and societal issues raised by GMOs were still unresolved, so for example, there were no regulations on liability for the consequences of GMO gene flow to non-GM crops. In South Africa, Ashton (35) maintained that the responsibility for negative impacts (financial, environmental) of GMOs fell on 'the user' (i.e. farmers, retailers and consumers) while Newman (2, 61) said non-GMO farmers in Australia were responsible for any negative consequences resulting from GMO gene flow to their crops. She (61), like Bhat (48), concluded that "the GM product provider must be legally responsible for containing and controlling their product and for any economic damage that would occur". While Efaw (78) suggested the GMO producers seemed to be very deliberate about avoiding responsibility for negative economic or environmental consequences, Blanchfield (78) noted that in a recent case (involving the finding of Starlink corn, a variety approved for animal feed but not for human consumption, in food products in 2000) the company involved accepted full liability. Blanchfield (78) also predicted that in the future the nature of the GMO producers would change as there would be more GMO R&D by government agencies, charitable foundations and international organizations and less by the private sector. Mog (84) disagreed, predicting that in the future private companies would continue to dominate, with governments struggling to respond adequately to the technologies they introduce.

3. Participation in the conference

The conference ran for five weeks, from 28 April to 1 June 2003, and a total of 401 people subscribed. Of the 401 people, 44 (i.e. 11 %) submitted at least one message. Messages came from all major regions of the world - 27 of the 93 messages posted (i.e. 29%) came from participants in North America, 23% from Asia, 19% from Africa, 13% from Europe, 9% from Oceania and 8% from Latin America and the Caribbean. Messages came from people in 20 different countries - the greatest proportion coming from India and the United States (18% each), South Africa (12%), Canada (11%) and Australia (9%). A total of 46 (i.e. 49%) messages were from participants in developing countries and 47 (51%) from developed countries. The greatest proportion of messages came from people working in research centres (27%) and universities (26%), with the remainder from private consultants (16%) and from people in NGOs (15%), government bodies (8%), farmers' organizations (4%) and UN organizations (4%).

4. Name and country of participants with referenced messages

5. Abbreviations

Bt = Bacillus thuringiensis; CBD = Convention on Biological Diversity; CGIAR = Consultative Group on International Agricultural Research; FAO = Food and Agriculture Organization of the United Nations; GE = Genetically engineered; GMO = Genetically modified organism; IAEA = International Atomic Energy Agency; MNC = Multi-national corporation; NGO = Non-governmental organization; OECD = Organisation for Economic Co-operation and Development; R&D = Research and development; UNEP = United Nations Environment Programme; WHO = World Health Organization; WTO = World Trade Organization.

6. Acknowledgements

To each and all of the people who submitted messages, a very special thanks.

*************************************
Recommended citation for this publication:
FAO. Regulating GMOs in developing and transition countries. Summary Document to Conference 9 of the FAO Biotechnology Forum (28 April to 1 June 2003):
http://www.fao.org/biotech/logs/C9/summary.htm