Search

Dementi vedrørende oplysninger i Per Clausens indlæg 11/04

Denne Supplerende rapport fra Udredningsgruppen omfatter en opdatering af "Udredningen fra 2003". Ny viden, som er fremkommet siden da, er inddraget i Udredningsgruppens faglige vurderinger.

Denne bog er for en stor dels vedkommende en revideret udgave TEMArapport fra DMU nr. 23, 1998: Gensplejsede planter

The increasing use of genetically modified plants (GMP) has opened up for a discussion about the problems related to the co-existence of GM and non-GM crops and especially the consequences for organic farming. One of the objectives of DARCOF project TOPRO (Tool for protection against contamination by GMO) has been to develop specific modelling tools, which can be used for the prediction of dispersal of GM pollen under different conditions and for investigating measures to limit the GM dispersal to organic fields.

*

Preface This publication is a first version of a manual identifying the data needs for ecological risk assessment of genetically modified higher plants (GMHP). It is the intention of the authors to stimulate further discussion of what data are needed in order to conduct a proper ecological risk assessment of GM plants when application for placing on the market is made. It is our hope that both the scientific community, the biotechnological industry and the regulatory bodies will participate in the process of improving the present draft, so that it can develop into a useful tool for both the industry as well as the national regulatory bodies. Furthermore, we hope that these efforts will improve the transparency of risk assessment and harmonisation of the requirements for data. The report suggests a structured way to identify the data need for risk assessment of GMHPs. It does not discuss the actual risk assessment procedures and the risk evaluation, which must proceed the data collection. The report use the terminology ecological risk assessment rather than environmental risk assessment because at present this work does not include bio-geochemical effects and environmental impacts from altered management of the fields, e.g. possible changes in for example leaching of pesticides or nitrogen, etc. Furthermore, we have abstained from suggesting number of species to test for specific issues because different risk assessment procedures have been developed which add a safety factor accounting for uncertainties in the extrapolation from limited laboratory studies to the species rich field environment. The relationship between the size of the safety factor and the number of species is therefore an issue of the risk assessment. Some of the issues raised in this report overlap with data needs to the assessment of agricultural risks or health risks. The work was overseen by a steering committee, which consisted of Gitte Silberg Poulsen, Jan Grundtvig Højland, and Hans Erik Svart from the Ministry of Environment and Energy, National Forest and Nature Agency and was performed within the framework of the project Biotechnology: elements in environmental risk assessment of genetically modified plants. December 1999 Christian Kjær Introduction In ecological risk assessment of transgenic plants, information on a wide range of subjects is needed for an effective and reliable assessment procedure. The information obtained from literature, field trials, laboratory and greenhouse tests have to be interpreted in a structured and well-defined manner. This guidance paper intends to assist the risk assessment procedure, as a working tool that stipulates the specific type of information needed in relation to the biotechnologically inserted or modified trait. Furthermore, reference to relevant test methods must be made for each type of information required. The present guidelines cover the need for information relevant to ecological risk assessment, raised in EU directive 90/220/EEC, including new issues raised in the amendment to the directive. This report suggests a structured way to identify the type of data needed to perform a sound ecological risk assessment for genetically modified higher plants (GMHP). The identified data types are intended to support the evaluation of the following risks: risk of invasion and establishment of the modified plant in natural habitats; risk of introgression of the inserted or modified traits to other plant species; and the risk of adverse effects on non-target organisms. The guidance paper considers only aspects of ecological risk assessment. Possible risks to human health, livestock and weediness of the transgenic crop plants as well as biogeochemical and environmental impacts ascribed to altered management should be considered separately. At present, no accepted test guideline exists for the issues mentioned above. However, it is agreed that a tiered approach is desirable for a standardised and effective treatment (Illueca, 1996; Rissler and Melon, 1993; Strandberg et al., 1998) and hierarchical systems are generally used in risk assessment (Suter, 1993; US-EPA, 1998; van Leeuwen and Hermens, 1995). The present guideline suggests a five-tiered approach with complexity of problems and required tests increasing from Tier to Tier. The guidance paper is structured so that first a summary of the Tier structure is presented (Chapter 2), hereafter chapter 3 gives a more detailed description of each Tier, and then test details are presented (chapter 4). Finally, chapter 5 describes the framework for establishing monitoring programmes. Summary of the Tier structure Tier I. General information on the receiver plant and the inserted trait. Tier II. Evaluation of general information and identification of additional data needs of the receiver plant A. Identification of capacity for reproduction of the GMHP B. Identification of possible effects on non-target organisms C. Detection of potential hybridisation, and assessment of the selective force acting on the inserted gene in a natural plant population Tier III. Laboratory measurements and small scale trials on the GMHP A. Test of changes in critical fitness components for the establishment of GMHP's in natural habitats B. Assessment of effects on non-target organisms C. Measurement of the rate of hybridisation and estimation of the probability of introgression Tier IV. Field assessments A. Identification of habitats/ecosystems that may be invaded by the GMHP and field tests for invasiveness B. Measures of effects on non-target organisms in the field Tier V. Investigation/calculations of regional and landscape effects. Tier I and II are both obligatory, and the design of Tier II experiments should be based on Tier I data. The progression from one Tier to the next depends on the results in the former Tiers. If a preceeding Tier recommends further investigation progression to the next Tier is obligatory. If the plant is able to complete a full life cycle under normal environmental conditions in the region of potential use within the EU it is necessary to proceed to Tier IIIA. Contrary, if the GMHP is unable to reproduce and survive for example winter conditions the Tier IIIA tests are not necessary, and so forth for the other issues raised. In any new case, it is obligatory to establish a monitoring program when the test programme has been carried through according to the tiered structure. The criterion for procedence to the next Tier depends on differences between the transgenic plant and the receiver plant. Such a comparison require that data are of a certain quality standard, because other-wise such a procedure would penalise good studies with a low variability. Therefore, the type of statistical tests needed and the number of replicates required for an adequate significance level (normally P < 0.05) must be carefully analysed when the trials are planned. This will require that a statistical power analysis is carried out. Statistical power is defined as 1-b, where b is the probability of not rejecting the null hypothesis when it is false. It must be tested that the power of the conducted experiments is above a specified value depending on the demands of the type of trial. A value of at least 80% is suggested. The power can be estimated before the experiment is performed if the design and variance is known. The power of a test will increase with an increased number of replicates. In some cases additional pretrials are needed to estimate the level of statistical variation. For methods, see standard statistics textbooks, e.g.Sokal and Rohlf (1995) and Cohen (1988). The effect size is defined as the size of change in the test parameter in relation to the expected value, incorporating variation, i.e. the difference between the null and the alternative hypothesis. An effect size of at least 0.25 will normally be required for detection of effects. Definitions of effect sizes for different statistical tests can be found in Cohen (1988). Figure 1. Schematic presentation of the Tier structure. The full line arrows indicate the flow through the Tier structure. The broken arrow from Tier IIIC to Tier I indicates a new assessment procedure for formed hybrids. Tier I and II are obligatory, which mean that they must always be performed, whereas the higher Tiers (III-VI) are conditional. This means that they are only necessary if an evaluation at a preceeding level indicates this positively.

*

OCDE/GD(94)41. Er supplerende rapport til rapport: The OECD Workshop on Methods for Monitoring Organisms in the Environment, Ottawa 14-17 September 1992. Environment Monograph 90.

*