GenEERA III – Model calculations on dispersal of oilseed rape transgenes at landscape level – Schleswig-Holstein regional study

(2002 - 2004) University of Kiel, Ecology Centre

Topic

The project focused on simulations of dispersal and persistence of oilseed rape transgenes , the results of which were portrayed on a landscape scale. The calculation models underpinning the simulations worked on the basis of various assumptions. Amongst others, the following factors were taken into consideration:

Different degrees of intensity of agricultural cultivation of genetically modified oilseed rape varieties and
Geographical variations in factors such as climate, cultivation density and crop rotations, geographical structure and incidence of wild rape.

Results were presented in the form of survey maps representing the frequency of transgenes in harvested crops and seed stocks in the soil as well as possible outcrossing events between transgenic oilseed rape and its cross-pollination partners across the entire region.
The project is part of the research group GenEERA (Generic analysis and extrapolation of oilseed rape dispersal) and developed methodological approaches to describe the dispersal dynamics over wide geographical areas based on the results of individual projects within the group (“upscaling”).

Summary

The aim was to be able to make predictions regarding dispersal and persistence of transgenes in the landscape prior to large-scale cultivation of genetically modified (GM) oilseed rape. Extensive geographical data sets were generated for this purpose during the course of the project and used to carry out large-scale model calculations on the dispersal dynamics of oilseed rape. On the question of co-existence of GM oilseed rape and conventional rape cultivation, the results of the modelling indicate that:

Isolation distances are necessary between GM oilseed rape and conventional plots.
Conventional rape seed production, e.g. in Schleswig-Holstein, is systematically affected.
One-off cultivation of GM oilseed rape can preclude subsequent conventional cultivation of oilseed rape for approximately ten years.

For Schleswig-Holstein, this means that one should aim for regional separation of GM and GM-free production for the oilseed rape crop. Overall, methods were developed with which the relevance of local individual findings can be evaluated at regional level. Also, new methods are now available for other areas of research in the fields of nature conservation, GMO monitoring , impact analyses and cost-benefit analysis of GM cultivation.

Experiment description

The objective of the research project was to apply the methodological instrument developed by the joint project in order to predict the dispersal of oilseed rape over large areas. In the Schleswig-Holstein (SH) regional study, particular emphasis was placed on processing the available geographical data in such a way as to enable small-scale model results to be transferred on a 1:1 basis.

For example, using satellite data, a database was created depicting individual medium-sized plots across the entire area of Schleswig-Holstein. By reconciling a combination of expert knowledge, land use data (regional statistical office) and satellite pictures, a land use pattern covering several years was assigned to each individual plot, corresponding to a large degree with reality. Farm-specific aspects were also taken into account. With the aid of this database, it was possible to transfer modelled ranges of results to the area very specifically according to local farming structures. This means that a high level of local detail can be expected from the regional predictions regarding dispersal and persistence of transgenes in commercial oilseed rape cultivation.

Results

The methods employed on this project make it possible to evaluate the dispersal of oilseed rape transgenes over a large area before genetically modified oilseed rape is actually used. On the basis of specified geographical utilization data, the model can quantify the dispersal of oilseed rape transgenes from small-scale events up to events for an entire region. However, prediction accuracy can only ever be as good as the quality of the existing data sources allows.

Effects of GMO cultivation on adjacent conventional crops

The effect of GMO input into an adjacent conventional oilseed rape plot decreases as the size of the conventional plot increases. The results indicate a considerable fluctuation range due to naturally varying cross-fertilisation rates and geographical configurations. In the case of conventional oilseed rape plots considerably larger than 25 hectares, the compulsory labelling threshold of 0.9 percent is unlikely to be exceeded. The threshold value is exceeded, with diminishing frequency, for plots of five hectares and under, five to ten hectares, ten to 15 hectares and 15 to 20 hectares. Since approximately 97 per cent of plots in Schleswig-Holstein are smaller than ten hectares and the threshold may therefore be exceeded, cultivation negotiations have to be held, or isolation zones established, for almost all oilseed rape plots. These statements apply only to one-off adjacent GM cultivation. It is suspected that GMOs can accumulate over the course of several years and as a result of unintended foreign GM matter entering conventional seed stocks.

Input situation for GM oilseed rape pollen assuming 50% GM oilseed rape cultivation

GMO input as a result of long-distance pollen transport

Even if isolation distances between GM and conventional oilseed rape crops are respected, a small GMO input would occur across most of Schleswig-Holstein. In the case of ten or 50 percent GM oilseed rape cultivation, the model calculations indicate that GMO input of up to 0.1 percent can be expected in the harvested crop of conventional oilseed rape.

In particular, this has far-reaching consequences for the production of conventional rape seed in Schleswig-Holstein. Since hybrid seed currently accounts for 75 per cent of seed produced in this region, and has considerably higher incrossing rates in production compared with conventional rape seed, it can be assumed that seed produced in Schleswig-Holstein would be systematically mixed with GM seed.

Input situation for GM oilseed rape pollen assuWhen complying with a distance of e.g. one plot width from GM rape cultivation, GMO input into conventional cultivation is to be expected from seeds, pollen input from remote GM cultivation and roadside seed loss. Assuming that GM rape cultivation does not exceed 50 percent in the long term and that a GMO presence in conventional seed of up to 0.3 per cent is permissible, the model calculations show that the labelling threshold of 0.9 percent is exceeded in the harvested crop of only a few conventional oilseed rape plots (up to 0.5 percent), even in the intensive oilseed rape cultivation areas of Schleswig-Holstein. According to the model calculations, this proportion increases slowly but steadily over the years (accumulation rate of 0.035 per cent in ten years).

Proportion of conventional harvests (%) with more than 0.9% GM rape seed in the years after one-off GM rape cultivation. The model calculation is based on 354 2-course, 163 3-course und 148 4-course oilseed rape rotations in the Seegeberg area.
Vertical: Proportion of conventional harvests (%) with more than 0.9% GM rape seed
Horizontal: Simulation period (years after GM rape cultivation)

Effects of one-off cultivation of GM rape

The model calculations prove that, up to eight years after one-off GM rape cultivation, the GMO presence in the harvested crops exceeds the labelling threshold in more than half of the plots due to the persistence ability of rape seeds. High numbers of volunteer plants could continue to cause some outcrossing to adjacent oilseed rape plots in the first five years following GMO cultivation.

Conclusions for oilseed rape cultivation in Schleswig-Holstein

Cultivation of GM oilseed rape should take place in SH only if isolation distances are observed (e.g. one plot width). If regional separation between GM and conventional rape cultivation does not take place, conventional seed produced in SH would systematically contain foreign GMO matter even if such distances were observed.

In the case of 50% GM oilseed rape cultivation, for example, an average GMO presence in conventional rape harvests of between 0.3 and 0.5 per cent would be expected. Due to the ability of rape seeds to persist, farm planning and farm leasing would have to take into account the fact that one-off GMO cultivation can preclude conventional rape cultivation for more than ten years.

Outlook: Using the model calculations for cultivation practice

The ability of the model to portray diverse situations and variations in a very flexible and realistic way is important for the quantification of certain events. The method can be used for the following:

The planning of GMO monitoring or for investigating rare events in relation, for example, to high local intensities.
Holistic evaluation of expected effects. To this end, the methodological approaches would have to be extended to include GMO-specific effects and the portrayal of transport and marketing routes as well as other types of crop (potatoes, sugar beet, maize).
A cost-benefit analysis: A cost-benefit analysis of GMO cultivation is an essential step towards making co-existence feasible. The model approaches described can be used to calculate the efficiency of various solutions for individual farms or for regional associations. This may contribute towards establishing the direct costs of co-existence.