GM Crop Farm-Scale Evaluations:
Background Papers
The Science of the Farm-Scale Evaluations
Paper by the DETR Biotechnology Safety Group and the Scottish Executive
Introduction
1. The Farm-Scale Evaluations (FSE) of genetically modified herbicide tolerant (GMHT) crops have become one of the most controversial ecological research programmes ever undertaken. The research has received great public attention partly because its outcome will influence Government policy on commercial use of these crops, and partly because of direct action by pressure groups concerned about genetically modification technology. This paper describes the rationale and science behind the FSE and should be read in association with the paper on the History of the FSE.
2. A small number of crop varieties genetically modified to be tolerant to broad spectrum herbicides has almost completed all the necessary approvals to permit them to be grown commercially in the UK. The furthest forward are oilseed rape, maize, sugar beet and fodder beet. The approval process includes a comprehensive assessment of the genetically modified (GM) plant and the possible risks it may pose to human health and the environment. However, a number of bodies, including the statutory nature conservation agencies, (such as English Nature and Scottish Natural Heritage) are concerned that the commercial planting of GMHT crops may lead to changes in agricultural practice that will adversely affect farmland wildlife. The increasing intensity of agricultural production systems has already had severe effects on farmland wildlife, causing decreases in many once common species such as brown hares and skylarks. The concern is that management (especially herbicide use) associated with GMHT crops could exacerbate these effects, further reducing the quality of the agricultural ecosystem for wildlife. Increases in agricultural intensity would make it impossible for the UK to meet Biodiversity Action Plan targets, agri-environment commitments and pesticide regulations. This has been discussed in detail in a discussion document published in February 1999.
3. In response to these concerns, the UK successfully introduced the need for the management of GMOs to be considered in risk assessments in the recent negotiations to amend Directive 90/220/EEC. This is discussed in more detail in the paper on risk assessment. However, there was an urgent need for the UK Government to address the potential effects of management of GMHT crops on farmland wildlife, because a number of products were close to being placed on the EU market.
4. Even though these GMHT crops are not considered directly harmful to humans and the environment, the FSE research will allow an evaluation of the potential indirect effects of the management associated with GMHT crops, especially herbicide use, on the farmland environment. Herbicides can indirectly affect wildlife by reducing or eliminating the food resources such as weeds and insects available to them in an ecosystem. This has knock-on effects through food chains to higher feeding (trophic) levels. Investigation is needed of the potential large-scale effects of herbicide management associated with these crops, were they to become a part of commercial agriculture. Small-scale evaluation in experimental plots or laboratory conditions cannot provide information on the complex ecology of the agricultural ecosystem and no existing research is available to help answer the question. The current pesticide legislation now includes provisions to address the issue of indirect effects of herbicides on farmland wildlife.
5. If GMHT herbicide management does exacerbate the adverse effects of conventional agriculture, the FSE results will provide the evidence needed to modify or delay EU directives covering the commercialisation of GMHT crops until risk associated with the management of these crops has been fully assessed. The FSE are probably one of, if not, the largest ecological investigation of their type in the world and will provide invaluable information about the indirect effects of non-GM conventional farming as well as those of new GM technology.
GMHT crops
6. The GMHT crops have been modified to be resistant to broad-spectrum herbicides such as glyphosate and glufosinate ammonium. This means that broad-spectrum herbicides that would normally kill the crop as well as the weeds can be applied to remove all weeds without damaging the crop. In practice it is difficult to control effectively broad-leaved weeds in broad-leaved crops because the crop is often susceptible to a herbicide. This means that existing herbicide programmes are complex and there is a need to control weeds either at a very early stage of growth, or use persistent soil acting herbicides to prevent weed seed germination in anticipation of them being a problem at a later stage when the crop is susceptible to competition. It is argued by the proponents of the GM crops that these crops have the potential to allow much greater flexibility in the timing of herbicide application, help control herbicide resistant weeds and reduce the quantity of persistent and more hazardous chemicals (such as Atrazine on maize) that are currently in wide use in the UK. To date, the Government has seen no evidence that this is the case; nor is there evidence that the use of herbicides on these GM crops will exacerbate farmland wildlife declines, hence the need for specific field studies.
The Farm Scale Evaluations
7. The farm scale evaluations of GMHT crops are designed to investigate the effect on the agricultural ecosystem of the management associated with their production. The crops themselves: forage maize, oilseed rape (winter and spring varieties) and beet (fodder and sugar varieties) have already cleared most of the regulatory procedures, designed to investigate their safety to human health. However, potential effects on the agricultural ecosystem resulting from the novel herbicide management that can be used with these crops have not been investigated. Industry has entered a voluntary agreement with the Government that commercial development of GMHT crops will not continue until biodiversity effects relating to their management are understood (see 'History of the Farm Scale Evaluations' paper).
8. The farm scale evaluations are comparing the impact on farmland biodiversity of two different herbicide regimes. They are not evaluating the safety of the GM plants themselves. There has been a widespread misunderstanding the FSE have been designed to investigate the potential effects on the environment of the GMHT crops themselves. This has already been done in the laboratory and in small-scale field trials. Consent to grow these crops at the scale of the farm scale evaluations would not have been granted if these plants were not already considered safe.
The research consortium
9. The Government invited 15 organisations to submit proposals for the evaluations, including the experimental design. As a result the ecological research is being carried out by a consortium of independent scientists from the Centre for Ecology and Hydrology (CEH), the Institute of Arable Crops Research (IACR) and The Scottish Crop Research Institute (SCRI). The scientific validity of the research is being overseen by a Scientific Steering Committee (SSC) of independent experts in agriculture and ecology drawn from Universities and organisations such as The RSPB and The Game Conservancy Trust (GCT).
Farm and site selection
10. The industry body SCIMAC (Supply Chain Initiative on Modified Agricultural Crops) finds farmers willing to offer fields for the evaluations. From these, a suitable number are selected by the researchers to be representative of regional geographical differences and the range of current farming methods, biodiversity and production intensities throughout the British Isles. The SSC approves the suitability of these sites for the FSE research. The experimental fields need to be large enough to allow commercial style crop management and to prevent edge effects from compromising the results. Field sizes are in the range 3-12 hectares.
Experimental design
11. The FSE experiment has been designed by the research consortium to test the statistical 'null hypothesis' that there is no significant difference in the effects on biodiversity from herbicide management in GM crops when compared with herbicide management in equivalent non-GM varieties of these crops. The use of a 'null hypothesis' is a standard part of designing scientific experiments and not unique to the FSE. It allows statistical analysis of data to seek for both potential positive and potential negative effects.
12. At each site, the field is split and one side sown at random with the GM crop while the other side is sown with an equivalent non-GM variety of the crop. This experimental design means that initial differences in biodiversity between the two halves are minimised and any differences in biodiversity detected during the course of the experiment will be the result of the herbicide treatment. As this is a manipulative experiment looking at differences between the crops, no baseline data on biodiversity are required, provided that the pairing is efficient. The grower is advised on herbicide use in the GM part of the field by the company providing the GM seed, but continues to manage the non-GM part in their usual way. The management in each FSE field is audited to ensure that herbicide use provides cost effective weed control, and that herbicide is not applied in a way that is unrepresentative of commercial practice.
Number of sites
13. The research consortium carried out a statistical power analysis to determine just how many fields would be needed in the experiment to allow the null hypothesis to be tested reliably and take into account natural variability. Overall, the analysis indicated that 60 sites would be necessary over the whole programme to give sufficient results. Ideally this meant 20 starts each year. However with the difficulty in obtaining sites, for the first year of research, the analysis indicated that 12-15 fields of oilseed rape, 12-15 fields of maize and 20-25 fields of beet would be a sufficient minimum to allow analysts to be confident that the hypothesis could be tested reliably. Ecological experiments are always subject to complications and the number of fields available in a given year of the FSE can fall below the optimum minimum, provided additional fields are added to the study in subsequent years.
Duration of the research programme
14. The evaluations are a four-year programme. In the first year, 1999, a small number of fields of each crop were sown to test the evaluation protocols. The main project started in spring 2000 and is due to end in autumn 2002 for the spring crops and summer 2003 for the autumn sown rape.
Species being monitored
15. It is not possible to monitor all biodiversity indicators in the FSE ecosystem, so a number of sensitive species that can be studied reliably have been selected. These are shown in the box below.
|
In practice, the variation in biodiversity can not be recorded for all species. The approach is to compare key indicators of biodiversity between the GM HT and non-GM HT cropping systems. In reporting these effects, they will be placed into the context of national recording schemes that can help to show the relationships between the biodiversity associated with the study sites and arable areas in general. The indicators being measured are:
|
16. At this stage the emphasis is on studying species at the lower end of the food chain, although monitoring of the birds and mammals using the FSE fields for foraging or breeding is also underway. This work is being done by the British Trust for Ornithology. The assessment methods are based on existing protocols, which have been modified appropriately for this project. In addition to studying animal indicator species, protocols are being used to investigate seed and plant biodiversity in and around the study fields. The study is limiting the number of confounding factors by using a split-field design and controlling for all the variables except the herbicide treatment. The monitoring takes place throughout the growing season in relation to herbicide applications. After the first year the studies will continue in the subsequent follow-on conventional crops.
Publication of the results
17. Publication of the results will be supervised by the SSC. They have advised that the results and analysis should published in peer-reviewed scientific journals at the end of the three years of research. All the data collected will also be made available at that time. The research contractors produce interim reports on progress with the evaluations every six-months. After review by the SSC these are published in the FSE web site.
Other studies
18. The opportunity is being taken to use the FSE fields for other work where this does not interfere with the evaluations themselves. The decision to allow such additional studies lies with the SSC.
19. The main additional project is monitoring gene flow from the GM half of the fields of rape and maize to the non-GM half. Both the flow of pollen and the extent of cross-pollination is being monitored. The work is being done by the Central Science laboratory and where appropriate they are also monitoring cross-pollination in nearby fields.
20. The Centre for Ecology and Hydrology is monitoring any cross-pollination with wild relatives of rape within the fields themselves and within a margin of 10metres. Maize has no wild relatives present in the UK agricultural ecosystem.
DETR Biotechnology Safety Unit: August 2000
Page published 20 February
2001;
Page last modified
10 August, 2002