“We can clearly show varietal differences.”
The living conditions for insects and other small organisms are not the same throughout the maize field: conditions in the middle are different from the field margin, a hollow is damper than a raised patch of ground. Micro differences like this can have a significant impact on a population. When scientists are investigating the possible effects of Bt maize on the maize ecosystem, how can they distinguish between natural fluctuations and a possible effect caused by the genetic modification? GMO Safety discussed this with Stefan Rauschen from RWTH Aachen University.
|
GMO Safety: In your project you are examining the effects on the maize ecosystem of Bt maize which is resistant to the Western corn rootworm . The basic approach is to compare the occurrence of selected species in different maize variants, so as to identify possible Bt effects. However, the abundance of certain species on a specific site depends on a very large number of factors, such as the weather or the soil type, for instance; even the collecting method and the time of sampling can influence the results. This being the case, how can you possibly make a prediction about the potential effects of Bt maize? Stefan Rauschen: Since the micro conditions at each site result in quite natural variations, it is best when conducting field studies for the trial site to be as uniform as possible, with similar conditions in all plots, so that differences between the varieties being studied can be identified. But even if we have a perfectly homogenous site, the first question is always: which plot shall I put where? If the plots are distributed on a random basis, two or three identical plots could potentially end up next to one another. So we have attached two conditions to the distribution of our 32 plots. Plots with the same variety should not be placed next to one another other and there should be the same number of plots of each variety on the field margin and in the middle, to ensure that the edge effects are evenly distributed. In our plot trial we have eight repetitions, in other words eight plots of each variety. This is really the minimum number; ten or twelve would be better as it would enable us to evaluate the data a bit better. |
GMO Safety: Were you able to find clear differences between the individual maize variants, despite the micro factors?
Stefan Rauschen: In addition to Bt maize and the isogenic variety, we also have two conventional varieties in this trial and these two varieties differ considerably with regard to one mirid which is central to our study. At certain times of the year we have found seven to ten times more insects in one of the conventional varieties than in the other. In all probability it can be said that this is definitely a varietal effect. The beauty of it is that we have found this effect, this significant difference relating to this mirid species, in all three years.If we had found this difference between Bt maize and the isogenic variety, it would have given us cause to take a closer look. But we have found no differences between Bt maize and the isogenic variety for this mirid, for example. Basically this shows very clearly that we can identify varietal differences. Although we don't know what causes them, we can clearly show a varietal effect on this particular species.
GMO Safety: If there was a similarly large Bt effect, would it also stand out?
Stefan Rauschen: Yes, if there was such an effect, it would stand out.
GMO Safety: Might it be also possible that Bt does have an effect that cannot be detected because it is camouflaged by natural fluctuations caused by a wide range of factors?
Stefan Rauschen: That is possible in principle. If there were a small affect, it would certainly be lost among the natural fluctuations. But the question is, does it then have any ecological relevance? If the thrips population, for example, fluctuates by a factor of 100 or 200 because it rained very heavily one year, then I have to ask myself if a ten percent reduction found in the Bt maize is relevant when the standard deviation is already so great. And then there is the question of how to assess the fact that one conventional variety is so much worse than another one for a particular mirid. Would I not have to assess all other conventional varieties as well ?
GMO Safety: The maize group projects are being supported by statisticians to make sure that the evaluation of the trials and in particular the conclusions are statistically reliable, in other words to provide watertight proof of safety. What does this mean for your trial?
Stefan Rauschen: The statisticians in our group use the special "proof of safety" approach, which was developed originally for clinical trials. This involves defining safety zones. Deviations within the zones are unproblematic and not relevant, but deviations that exceed these zones are problematic. For us, however, it is difficult to define the safety zones. We don't have the necessary data. As a biologist, I cannot say with a clear conscience that a certain species of mirid occurring here or there with a density of x insects per plot is problematic, or that it is okay if their numbers increase by 250 percent. I can’t define these zones. What’s more, I don’t know enough about the biology of these insects. All I can do is use existing data to try to estimate zones which contain the actual average values, in order to compare these and see how great the variations are. One important point made by the statisticians is that wherever possible we select insect species which occur with a relatively high average density, i.e. 20 to 25 insects per sample on average, to make it possible for statistical evaluation to take place. Insects which are found five or six times a year or only occur sporadically are non-starters in terms of statistics.
GMO Safety: How do you deal with other factors such as soil values etc. from a statistical point of view?
Stefan Rauschen: If you know the factors which could affect insect numbers, then statistical methods can be used to determine what proportion of the variability in numbers can be attributed to a specific factor. And then you can say that x proportion of the variations observed in the field can be explained by the fluctuation of individual parameters that I have measured. In this way I can make a more accurate statement about differences between individual varieties.
GMO Safety: Is there such a thing as a baseline with regard to certain groups of organisms? Could you say for example that x number of mirids generally occur in a maize ecosystem and use that as a starting point, or are you always dealing with a relative comparison between different variants on a specific site at a specific time?
Stefan Rauschen: Yes, it’s always relative, there is no baseline. To start with, we don’t know for certain which organisms can generally be found in which regions. Cicada xy may be more common in Bavaria or in Baden-Württemberg, but if you go further north you might not find any in the maize field because they may already have reached their distribution limit. Then there are seasonal differences, and significant differences can occur from year to year. However, there are a couple of species which you can always find with certainty, or with a probability bordering on certainty. I can name five species of cicada which we have found on our first site, which we have found on the current trial field and which I am likely to find in Oderbruch as well.
GMO Safety: Could you then say, we recommend such and such organisms as indicator organisms, they are always present in the maize field and are suitable for monitoring?
Stefan Rauschen: There are a couple of species which I could safely say would be relevant for the whole of Germany or for Europe. Take Zyginidia scutellaris, which we are also working with. I know that it is found in Spain, it is found in Hungary, it is found here in Germany, so it is probably found in France and Italy as well. But the next question is how relevant is the organism? And how representative is it for its group? I can suggest indicator species, but I would have to add that I can do so with certainty only for their range of distribution and that further sensitivity tests would have to be carried out in the laboratory. That’s exactly what we have decided to do for the next project. In my view, the correct sequence would be to go to a specific region where maize is to be grown, identify the key species which occur in sufficient numbers to allow a statistically reliable prediction to be made. See what their function is within the species community. Then we study these organisms in the laboratory, effectively subject them to a worst-case exposure over several generations if possible, using standardised methods. If we find nothing in the laboratory, then the matter is settled. If we do find something in the laboratory, we investigate further with parallel field studies.
GMO Safety: For the Bt maize you are investigating, would it be useful to start by studying certain species of beetle, since the Western corn rootworm is itself a beetle?
Stefan Rauschen: In principle that sounds like the obvious thing to do, but in reality the beetles don’t have such a prominent position in the maize field here, at least not in terms of abundance and diversity. We have three or four species here, but they are very scarce. It is very important for us to look carefully at which species come into contact with Bt. If I discover, as with this mirid which we have been discussing, that it actually ingests a large amount of Bt protein - it sucks the green leaves and on average one fifth of what was measured in the leaf can still be detected in the insect – and I still can’t find anything, then it really does look as though Bt has no effect here.
GMO Safety: Thank you for talking to us.
More from GMO Safety
![[jump to top]](/images/icons/icon_gotop_9x9.gif){kind=icon}