Background Paper 2
Risk Assessment for Release and Marketing of GMOs in the European UnionPaper by the UK Joint Regulatory Authority and Secretariat to ACRE
Introduction
1. The purpose of this paper is to explain the current approach to assessing the risks to human health and the environment of releases of genetically modified organisms (GMOs). This paper will focus primarily on GM crops, since they account for the majority of GMOs being considered at the present time.
2. Risk assessment underpins the decision making process for granting consents for the release and marketing of GMOs, since the sole basis for making decisions on the granting of consents is safety. This paper should be considered in association with the companion paper describing the legal framework for regulating the deliberate release of GMOs.
Approach to Risk Assessment: Terminology
3. The approach to risk assessment in the UK has been to:
identify the intrinsic properties of an activity, substance or organism that may cause harmful effects to humans or the environment;
to estimate the likelihood of those effects occurring under the conditions of the proposed release or use;
to estimate the magnitude of the harm that may arise, assuming that the effects occur; and
on the basis of the above steps, evaluate the overall risk.
This general approach has been described in other recent documents8.
4. The intrinsic properties of an activity, substance or organism that may cause harm are termed ‘hazards’ and the likelihood of hazards being realised and the magnitude of the consequences are together termed ‘risks’. The process of risk assessment for GMOs is described in more detail in other specific DETR/ACRE/ guidance notes9.
5. The over-arching principle adopted in the risk assessment of releases of GMOs is the ‘case-by-case’ approach, where each GMO release is assessed according to the particular circumstances of the proposed release or use. This does not mean that every proposed GMO release is looked at in isolation of existing releases of GMOs; these are taken into account where appropriate.
Hazard Identification for GM Crops
6. Hazard identification is carried out using the information provided in applications for consent (see companion paper on the legal framework). The information requirements include information on the characteristics of the plant that has been modified (‘recipient’ plant), information on the genetic modification (including details of the sequences of nucleic acid inserted), and information on the genetically modified plant. This enables an understanding to be gained of the potential behaviour of the GM plant and its possible interactions in the environment and other organisms.
7. The potential hazards associated with a GM crop may be:
Expression of toxic or allergenic compounds; the genetic modification may result in the production of substances that are toxic to humans or other species, or allergenic.
Effects on biogeochemistry; the potential to cause changes in nitrogen and carbon recycling that depends on decomposition processes.
Increased persistence in the environment and invasiveness; a genetic modification may confer an ecological fitness advantage to the recipient plant, which potentially allows it to become persistent or invasive. Concerns about ‘superweeds’ have been raised regarding herbicide tolerance traits; however, traits such as disease, drought or insect resistance are more likely to confer an advantage to a recipient plant, since these pressures control natural plant populations.
Transfer of genetic material; cross-pollination with other crops of the same species or near-relatives can give rise to hybrids which express the traits introduced by the genetic modification. Such gene-transfer may not be a hazard in itself; this would depend on the trait being transferred.
Instability of the genetic modification; plants have the ability to inactivate inserted genetic material, particularly if there is a large number of copies inserted and if the constructs are large. In many cases, this may not pose a risk in itself, as the recipient plant is likely to revert to the wild-type. However, this would become an issue where a genetic modification was made to down-regulate a naturally occurring hazardous trait.
Unintended effects; while it is expected that inserted sequences of nucleic acid are well characterised, the exact positions of the insertion(s) cannot be predicted until more detailed analysis is completed. It is possible that the insertion can influence the expression of adjacent genes and their promoters, leading to unintended genetic modifications. These may not be hazardous, but the transformation event and its progeny would require careful monitoring.
8. If a hazard is realised in the course of a release, it may result in harm to human health if the GMO is toxic or allergenic, or to the environment, if population dynamics are affected, resulting in reductions of native species. The process of risk assessment examines whether or not it is likely that identified hazards are likely to be realised under the conditions of a proposed release.
Risk Assessment
9. The likelihood of a hazard being realised and the magnitude of the consequences will depend on the characteristics of the release or use of the GMO, including the features and position of the release site, or the scale of use. There is a significant difference between a small-scale research plot (covered under Part B of Directive 90/220 and from 17 October 2002 amended Directive 2001/18/EC) and unrestricted commercial use (covered under Part C).
10. Estimating the likelihood of an event in an ecological system is not an exact process, and extremely difficult to quantify in terms of probabilities. As with all risk assessment, there is an element of uncertainty. With respect to GMOs, this has been handled in a number of ways:
adopt a worse case scenario approach in the evaluation process, assuming the consequences will definitely occur, and assign a magnitude to those consequences;
request further information from the consent applicant to address specific areas of concern, and require an amended risk assessment;
impose a higher degree of risk management of the release until more information is available to address the issues; or
require monitoring during and after the release of the GMO.
11. As pointed out above, there is a significant difference between Part B release applications for small scale field plots and Part C notifications for unrestricted commercial cultivation of a GM crop. In the case of a small scale field plot, it is possible to have close control over the site and impose risk management measures to isolate the trial from important habitats and commercial crops to prevent or minimise cross-pollination and thus gene transfer by:
planting a border row around the GM plants to act as a pollen barrier;
isolation distances;
bagging plants during flowering; or
cutting the developing flower buds off; and
destroy all harvested material.
12. In accordance with the ‘step-by-step’ principle explained in the companion paper on the legal framework, small trials are necessary to gather more information on the crop and its safety, before a scaled up release can take place. More information reduces the uncertainty and enables ACRE to advise accordingly on subsequent releases. For Part C notifications, ACRE must consider the potentially wide use of a GM crop within the EU. ACRE is in the position of taking into account the realities of commercial agriculture, and the limited scope for onerous restrictions on use, in giving advice. If it is considered necessary to impose onerous risk management in commercial use order to address safety concerns, then it is likely that a consent would not be granted. In a Part C situation, the GM crop and its use must be intrinsically low risk.
13. In considering applications for both Part B and Part C consents, ACRE has often taken the approach of assuming that the hazard, if present, will be realised, and to focus on the consequences. For example, the release of GM oilseed rape raises the issue of gene transfer, since this crop readily cross-pollinates with nearby oilseed rape crops and there is some evidence of spontaneous hybridisation with some wild relatives. Experience in crop breeding and seed production has enabled separation distances that can be recommended to minimise cross-pollination and thus gene transfer. However, it has always been accepted that gene transfer between crops such as oilseed rape and maize can never be prevented, so ACRE has focussed attention on the consequences of the event occurring, even it occurs at a low frequency.
14. In the case of tolerance to glufosinate ammonium herbicides, ACRE concluded that any resulting feral populations or oilseed rape crops and wild relatives were unlikely to gain any selective advantage by inheriting this specific trait10. However, this would not necessarily be the case for other traits inserted by genetic modification. ACRE has considered applications for consent to release glyphosate tolerant oilseed rape on an experimental basis, and have concluded that it should be separated from other oilseed rape by 50m to minimise gene-transfer. The issues relating to multiple herbicide tolerance have been explored by ACRE11, but the committee has yet to advise on a marketing notification for GM glyphosate tolerant oilseed rape.
15. Similarly, the consequences of gene-transfer of traits such as reduced pod shatter, insect or fungal resistance could also be viewed very differently from herbicide tolerance, since it could result in decreased or increased ecological fitness of wild relatives, thus potentially affecting population dynamics. ACRE has taken the view that cross-pollination can only be minimised – not prevented – and if there are safety implications associated with the genetic modification, then the committee would advise that a consent should not be granted for the release or marketing of a GM crop12. This illustrates the case-by-case approach, and that existing GM crop consents are taken into account in the risk assessment process.
16. The estimation of overall risk takes into account the combination of the likelihood of hazards being realised and the magnitude of the consequences. The likelihood of harmful effects being realised may be high, medium, low or negligible. Similarly, the magnitude of the harmful effects may be severe, moderate, low or negligible. This approach illustrates the qualitative and descriptive nature of assessing the magnitude of harmful effects and the risk of releases of living organisms. Further guidance on what is meant by low, moderate or severe effects has been offered to applicants in Defra/ACRE Guidance. The behaviour of any organism in the environment is influenced by many interacting factors, and we are not yet in a position to take a quantitative approach with any confidence.
17. It should be stressed that a risk assessment is not a closed issue if a consent has been granted to allow a release or marketing to proceed in the EU. There is a statutory requirement that all consent holders keep themselves informed and alert the regulatory authorities of new information which has significance for the risk assessment. A risk assessment should only be carried out with the latest information available, and which is based on credible scientific studies. But the option for a re-appraisal is always available if new information becomes available.
Changes to the Risk Assessment Process
18. In the letter half of 1998, negotiations to amend Directive 90/220/EEC resulted in a new technical annex on risk assessment being agreed, which the UK played a key role in developing. This was an important step, since until then there had been no formally agreed approach to risk assessment for GMOs within the EU. The technical annex formalises some important principles, which reflect existing best practice among EU Member States13. These include:
That potential adverse effects may be direct, indirect, immediate or delayed;
The risk assessment should be carried out on a case by case basis; this implies that the required information may vary depending on the type of GMOs concerned, their intended use and the potential receiving environment, taking into account, among other things, GMOs already in the environment;
If new information on the GMO and its effects on human health and the environment becomes available, the risk assessment should be re-examined;
That the information required in notifications must include possible impacts of the specific techniques used for the management of GMO where these are different from those used for non-GMOs. This particular principle was a new addition to GMO risk assessment.
19. At the Environment Council in December 1998, Ministers agreed that until the revision was complete, competent authorities could take into account the underlying principles for risk assessment resulting from the work done, particularly when considering Part C notifications. Whilst the Directive as a whole was under the conciliation stage, this Annex was largely agreed and was formally adopted with these important principles intact on 14 February 2001 as Directive 2001/18/EC that will be replacing Directive 90/220/ EEC on 17 October 2002.
Baselines on Environmental Harm
20. The assessment of the risk of a GMO needs to be placed in the context of existing agricultural activities, whether non-GM or organic, and which also have the potential to cause adverse environmental effects. The baseline against which the risk of a GMO can be compared has been the subject of discussion in ACRE. All agricultural activities result in adverse effects on the environment, and there are serious concerns about farmland wildlife declines that have occurred during the last fifty years. With GMOs, it has been common to compare the risks of a GM crop with the unmodified crop. DETR and ACRE have commented that it is inappropriate to demand non-target effects from the use of GM crops, while tolerating them from the use of non-GM crops whether they are grown conventionally, organically, or in accordance with Integrated Crop Management (ICM) methods14.
21. Given the UK Biodiversity Action Plan commitments to stop and where possible reverse wildlife declines, it is important that the risk assessment process now identifies GM crops whose use may lead to increasingly intensive agricultural management practices. This is why the recent amendments to the risk assessment process to evaluate the effects of the management of GMOs as well as the GMO itself are so important. At present, the debate is focussed on whether GM crops tolerant to broad-spectrum herbicides will lead to increased weed control and less resources for farmland wildlife to feed on. This is why the farm-scale evaluation research programme was set up, and the results, whatever the final conclusions, will provide a greater understanding of arable ecology that can be used to underpin this process.
22. Work is currently being carried out by the ACRE Sub-Group on Wider Biodiversity Issues to develop guidance on how the effects of the management of GM crops should be assessed in the context of Biodiversity Action Plans and other farmland and habitat conservation targets. A draft was issued for consultation in September 2000.
Joint ACRE Secretariat: August 2000 (amended April 2001)
8 Guidance for Environmental Risk Assessment and Management; DETR 2000.
9 The Regulation & Control of the Deliberate Release of Genetically Modified Organisms; DoE/ACRE
Guidance Note 1, Chapter 4.10 Environmental Risks of Herbicide-Tolerant Oilseed Rape. A Review of the PGS Hybrid Oilseed Rape.
GMO Research Report No 15.11 Genetically Modified Herbicide Tolerant Crops; Chapter 3 of the Advisory Committee on Releases to the Environment Annual Report No 4: 1996/97
12 ACRE response to the MAFF Consultation on Separation Distances – July 2000.
13 Guidance on Principles of Risk Assessment and Monitoring for the Release of Genetically Modified Organisms. DETR/ACRE Guidance Note 12.
14 The Commercial Use of Genetically Modified Crops in the United Kingdom: the Potential Wider Impact on Farmland Wildlife; Chapter 3 of the Advisory Committee on Releases to the Environment Annual Report No 4: 1996/97.
Published 31 July 2001
Environmental Protection Index
Defra Home Page