May 13, 2005

Research Projects

Send

Concept for post-market monitoring of genetically modified plants – based on Brandenburg

(2001 – 2004) Leibniz-Centre for Agricultural Landscape Research (ZALF), Müncheberg

Topic

The cultivation of genetically modified plants (GM plants) will in future be accompanied by a monitoring system. This is prescribed by the Deliberate Release Directive (2001/18/EC).

This project has drawn up a post-market monitoring plan based on selected cultivation regions in Brandenburg. It contains elements that have never previously been linked together in this way in conceptual approaches. Previous studies have usually been monographic with no regional basis.

According to the Deliberate Release Directive, monitoring is designed to identify environmental effects of GM plants, particularly those effects that can be identified only during large-scale cultivation. It must also answer questions concerning suitable parameters for inspection, detection methods, measuring times and suitable measuring locations. In addition, these questions need to be asked in the context of different GM plants and gene constructs, heterogeneous agricultural landscapes, different cultivation systems and different infrastructure conditions. A systematic approach is needed to handle these complex tasks. It needs to be research-based, flexible enough for modifications and capable of being transferred to other landscapes. In particular, the regional monitoring plan must be flexible, so that it can be adapted to future variety and cultivation technique developments.

This project followed a comprehensive approach to monitoring possible environmental effects of the cultivation of transgenic plants compared with non-transgenic plants. Characteristic landscape and agricultural features of farming regions were taken into account and priorities were set regarding the parameters to be selected for measuring environmental effects. Representative monitoring networks were developed which make best use of available data and resources.

Summary

ZALF has developed a transparent, transferable and flexible plan for the systematic derivation of regional monitoring networks for genetically modified plants which takes account of the heterogeneity of the agricultural environment.

The plan is based on expert knowledge and on semi-quantitative and empirical data. These were translated into scores and rules which made it possible to optimise available monitoring networks on the basis of existing background information and with the application of a geographical information system (GIS).

Scenario analyses carried out in Brandenburg for various transgenic crops, possible measurement parameters and existing monitoring networks confirm the viability of the monitoring plan.

Methodology

Environmental effects are the result of interactions between the crop plant, environmental conditions and farming practice. In order to be able to develop a systematic monitoring plan for the environmental effects of GM plants, both the direct environmental effects, caused by the genetically modified plant, and the landscape and agricultural features of the location need to be taken into account.

The following components were linked in this project:

Landscape ecological features in Brandenburg A survey of landscape ecological features is an important basis for a representative monitoring network plan tailored to landscapes and cultivation regions. The features are available as geographical information layers (see Table 1). An ecological classification of Brandenburg was also produced. The results were generalised in order to be able to transfer them to other agricultural landscapes.

Survey of existing monitoring networks related to the agricultural environment

The idea was to use available resources, infrastructure and background data as a monitoring reference basis. Suitable monitoring networks were selected from the monitoring network land register of the Brandenburg State Environment Agency according to very specific criteria that were evaluated with the help of a special index.

• Geographical representativeness for Brandenburg
• (Species) number and relevance of monitored media and organisms
• Permanence and frequency of the measurements
• Links to other Land and national monitoring network programmes. Five monitoring networks were selected (Diagram 1).

Assessment of geographical information and development of a rule-based GIS analysis for regional measuring point selection

The meaningfulness of the geographical information levels was assessed by evaluating their relevance for the parameters to be measured in a systematic schema (Diag. 2). The assessment is based on empirical data and expert knowledge. Parameters that were discussed during the workshop on “Parameters for the monitoring of genetically modified plants in the agricultural ecosystem” (2002, Braunschweig) are used as examples. A group of experts from ZALF, the Federal Biological Research Centre for Agriculture and Forestry and the Federal Environmental Agency carried out the assessment for various GM plants using the following criteria: a) protection aims, b) different potential environmental impacts, c) possible features to be monitored and d) various kinds of geographical information.

From the overall assessment it was possible to derive weighting factors for the different types of geographical information. The selection of the measuring locations was based on the geographical information levels according to the weighting factors.

The assessments of the geographical information levels and monitoring networks provided the basis for a GIS analysis model. (Diag.3). The aim was to find out which of the existing measuring locations on agricultural land (approx. 100) best represent the various agro- environment zones.

Systematisation of current and potential agricultural use

The monitoring of transgenic crops must involve a representative monitoring plan adapted to cultivation regions and cover the environmental impacts caused by modified agricultural cultivation systems brought about by the introduction of GM plants. For this, the ZALF developed a system for describing agricultural use, which can deal with production technology features (e.g. plant protection measures, soil cultivation, fertilization) and can also be displayed in terms of land area (Diag. 4). The system can be applied to other agricultural landscapes.

Results

Development of a database structure for the monitoring of GM plants

A database structure was developed to record the very comprehensive information required for post-market monitoring.

It integrates the parameters relevant for monitoring (GM plants, environmental impacts from the cultivation of GM plants, investigation methods, features to be investigated, cultivation measures in farming, landscape areas and existing monitoring networks).

The network structure of the database makes it possible to carry out highly complex scenario calculations (Diag.5).

Landscape features: development of an ecological classification system for Brandenburg

To identify the geographical distribution of the features to be investigated, locations in the Brandenburg area were classified according to climatic elements, like air temperature, precipitation and other relevant variables, e.g. groundwater and soil type. This produced 13 classes of ecological zone for Brandenburg, which were collated and reduced to six classes for the GIS analysis model (Diag. 6).

Existing monitoring networks related to the agricultural environment

In Brandenburg there are four suitable fixed monitoring networks that record data relevant for the monitoring of GM plants (Diag.1).

There are also flexible monitoring structures, like those of the plant protection bureaux. Existing base data and structures of the fixed networks should be regarded as prerequisites for long-term monitoring because time series analysis is not possible without them.

Rule-based GIS analysis for regional measuring point selection

Different monitoring scenarios were developed using GIS analysis:

• Various transgenic crops (Bt maize, HR oilseed rape, HR sugar beet and starch-modified potatoes)
• Various monitoring network combinations
• Different measuring site densities

The optimum number of measuring sites varies for different scenarios (Diag. 7). The long-term soil observation areas and the variety trial sites could be operated as a stand-alone monitoring network with three to four monitoring locations per landscape unit, although a few additional parameters for the monitoring of GM plants would probably have to be recorded at these locations.

Joint monitoring of GM plants at a few locations would lead to a reduction in the overall number of measuring areas.

In addition, the project located suitable monitoring sites outside existing monitoring networks (Diag. 8).

Systematisation of current and potential agricultural use

Using the example of the introduction of genetically modified herbicide resistant (HR) oilseed rape in place of conventional oilseed rape, changes to the agro-ecosystem were systematised. The changes were limited to the system factors “agricultural practice” and biotic factors. Within biotics, the focus was on arable weeds and insects.

Changes in the agro-ecosystem also have a temporal component. The introduction of the new variety can alter agricultural practice, which can in turn influence biotics. Within biotics, changes can occur that can again have an effect on agricultural practice (Diag.9).

The systematisation covered:

• Farming changes of the first order, especially those caused by adaptations to cultivation practice to adapt to the crop characteristics
• Direct effects on biotic factors caused by the above changes that could affect the abundance and diversity of living organisms.
• Indirect effects on biotic factors resulting from the direct effects.
• Changes to farming practice linked to biotic factors
• Farming changes of the second order caused by adaptations of farming practice to the altered biotic factors

The extent of potential changes in the agro-ecosystem caused by the introduction of a transgenic variety varies according to region, location and farm. This needs to be taken into account in different scale levels when designing a monitoring system for GM plants. Whether the changes eventually lead to adverse environmental effects must be assessed using preset criteria or damage thresholds.

Final assessment of the monitoring plan

The implementation of the various crop, monitoring network and measuring intensity scenarios in monitoring practice is difficult to evaluate at present. In order to be able to derive statistically supported findings on the environmental effects of the genetically modified plants, the monitoring network design must take into account the variability of the parameters. The amount of work involved can be significantly reduced if a) existing resources are used and b) more precise geographical base information is available for the selection of monitoring locations – as shown in this study.

The scenarios analysed here and the results presented serve primarily to check and demonstrate the evaluation and optimisation concept. When decisions are taken concerning e.g. threshold values, responsibilities and costs, this monitoring plan can be flexibly adapted.