“The processes are highly variable, so coexistence rules must be sufficiently robust.”

Dr Broder Breckling , adjunct professor at the University of Bremen specialising in general and theoretical ecology; spokesman for the specialist group on genetic engineering of the Gesellschaft für Ökologie (Ecological society of Germany, Austria and Switzerland). Coordinator of the GenEERA (generic analysis and extrapolation of oilseed rape dispersal) SiFo joint project: Modelling for the dispersal behaviour of oilseed rape on a landscape scale.

Is GM oilseed rape capable of coexisting with conventional rape? – Questions put to Broder Breckling.

GMO Safety: The cultivation of GM oilseed rape in Europe is a long time coming. So far, findings about the impacts of large-scale cultivation are available only from other countries. You have developed a model that makes it possible to estimate the rate of spread of oilseed rape for large regions. What are the main findings?

Broder Breckling: We used a model study for northern Germany and examined the details very carefully, placing them in a consistent context. The results show that if GM oilseed rape is produced, transgenes will be found, with a high degree of certainty, in small traces in oilseed rape crops and in the environment for years, even decades – even if GM rape is no longer being cultivated.

GMO Safety: What factors and what exceptions play a key role in this scenario?

Broder Breckling: Oilseed rape cultivation in northern Germany – and in particular Schleswig-Holstein – has very particular characteristics: rape cultivation in this area is highly concentrated and extremely fragmented. Other key factors are the fact that oilseed rape can persist in the seed bank for up to ten years, the fact that pollen is transferred over long distances and the existence of an independent, dispersal pattern that does not depend on oilseed rape cultivation. Wild rape populations are found in towns and cities, in industrial areas and on fallow land. These populations cannot be directly explained through seed losses along transport routes.

GMO Safety: To what extent can outcrossings be expected when GM and conventional oilseed rape are grown side by side? Long-term practical field studies with GM oilseed rape in Germany have shown that buffer plantings can reduce GMO pollen presence in the conventional fields to significantly below 0.9 per cent.

Broder Breckling: We reckon that the threshold of 0.9 per cent can be respected in many – probably most – cases. What is significant though is the degree of variability. Considerable uncertainties remain at this level. Even if, let’s say, the threshold value were to be respected on average throughout Schleswig-Holstein, this would not be inconsistent with the threshold value being exceeded in e.g. one-third of cases. So the fluctuation ranges should not be ignored.

GMO Safety: Studies commissioned by the European Commission see no fundamental difficulties for conventional farms in staying below the threshold if GM oilseed rape is cultivated.

Broder Breckling: The studies you mention are based on isolated local research projects and the opinions of experts based on them. It is not possible to derive generalised conclusions from them. Our study is the first and, so far, the only one that actually calculates regional interrelationships using empirical data. During our research we have shown for the first time that this kind of regional, data-based approach is actually possible. This is called ‘proof of concept’. We included data on crop rotation, farming practice, field geometries, climate effects etc. in northern Germany.

GMO Safety: Is it possible to derive recommended minimum distances for farming practice from your model calculations?

Broder Breckling: At the moment we cannot make many generalisations based on the model calculations. For that we need further evaluations. Because of the high variability of the processes involved there is a high potential for uncertainty.

GMO Safety: You also took long-distance pollen transport into account in your model. What were the results?

Broder Breckling: Even if farmers respect agreements on separation distances, for instance, GMO pollen still enters conventional fields through long-distance transport, which can cover several kilometres and which cannot be attributed to a single perpetrator or field. We have calculated that 50 per cent GM oilseed rape cultivation in Schleswig-Holstein for instance, would lead to pollen contamination across the area of 0.1 per cent. This would have to be added to other contamination routes such as seed impurities.

bioSicherheit: Oilseed rape seeds are capable of surviving for a long time in the soil and can keep germinating in subsequent crops. Practical studies show that such volunteer plants can be controlled very well in follower crops through adapted soil cultivation and suitable crop rotation.

Broder Breckling: Our model calculations have shown that in individual cases volunteer plants can lead to the threshold value being exceeded in follower crops for more than ten years. However good the crop management, it will not be able to control volunteers completely. There is also a problem in that we cannot use a single transgenic variety as a basis, but must take into account the combined effects of several authorised varieties. In the case of herbicide-resistant varieties, pollen transfer can lead to volunteers with single or multiple resistance in conventional crops.

bioSicherheit: But are volunteer plants still a coexistence problem when it is now possible to reduce volunteers by 99 per cent using suitable techniques?

Broder Breckling: Reductions of 99 per cent are not unusual results, but they do not completely solve the problem. You can see this from the following example. Volunteer rape – the quantity that is lost during harvest and remains on the field – can reach 15 or even 20 per cent under certain weather conditions. On average, an oilseed rape plant forms in the region of 1200 seeds. If the volunteer rape is reduced by 99 per cent, the remaining one per cent is still in the order of magnitude of the original sowing volume.

bioSicherheit: Cultivated oilseed rape that has gone wild is very often to be found on ruderal areas along paths and railway tracks. But it is not very constant there. Do such populations have any relevance for coexistence?

Broder Breckling: They have potential significance as a genetic reservoir that holds the possibility of the emergence of undesirable genotypes. Taken in relation to the threshold of 0.9 per cent, however, these are marginal populations and play no role over the short term. However, they constitute a certain potential for uncertainty because their characteristics are unpredictable and combinations with transgenic varieties that are developed in the future would therefore have to be checked. As a result of fundamental considerations, particularly with regard to possible combination effects, the persistence of transgenic individuals off the field is therefore undesirable and something to be avoided. There is a need for very conscious, responsible management at this level.

GMO Safety: Oilseed rape has wild relatives in Europe with which it could outcross. The incidence of these species, however, varies widely from one region to another and fertile hybrids have been found in only a few cases. How relevant are such findings for coexistence?

Broder Breckling: Since these are rare events, statements about them are associated with considerable uncertainty. The OECD presented the state of play on this topic as far back as the nineties in what I believe is a very good consensus paper. Since then, our knowledge about proven hybridisations has increased. The problem is that in principle a single incidence of an undesirable genotype is sufficient to set in process a further propagation process. We know this from invasion biology. But such results play not significant for coexistence in the narrower sense.

GMO Safety: Now to the conclusion of your research. What conclusions do you draw from the results of your model calculations? Can GM and conventional oilseed rape farming coexist?

Broder Breckling: We are dealing here with living systems. And we want to make statements about large cultivation regions. It is evident that the cultivation of GM oilseed rape should be approached with a high duty of care. If we look at the processes and events that play a role in cultivating oilseed rape on a landscape scale, they are highly variable. They fluctuate to such an extent that differences of more than 100 per cent are regarded as normal. This means that there are considerable uncertainties associated with the processes. So rules to ensure coexistence must be sufficiently robust and include safety margins. Deviations or errors can always occur in practice. For instance, when analysing satellite pictures from consecutive years we found that oilseed rape is occasionally grown on the same field in two consecutive years. This clearly flies in the face of all recommendations on good farming practice – but it still occurs.

GMO safety: Thank you for talking with us.