Butterflies and moths barely affected, even with large-scale cultivation of Bt maize
Scientists from five European countries have developed a mathematical model for calculating the risk to butterflies and moths from genetically modified Bt maize.
Genetically modified Bt maize expresses the Bt protein Cry1Ab in all parts of the plant, including the pollen. The protein targets a pest, the European corn borer, which is a moth. It is therefore possible that other species of butterfly or moth could also be at risk from Bt maize.
However, with the exception of a few pests like the European corn borer , maize is not an important source of food for butterflies and moths in Europe. This means that they come into contact with Bt protein only if they ingest it along with maize pollen on their host plants, e.g. stinging nettles.
|
Scientists from Germany, the UK, Spain, Italy and Hungary have used a mathematical model to calculate to what extent various species of butterfly and moth come into contact with maize pollen containing Bt protein and whether the amount of pollen they ingest is enough to harm them. The only type of Bt maize currently authorised for cultivation in the EU is MON810. It is cultivated primarily in Spain. Some European countries, including Germany and France, have suspended the 1998 EU authorisation and banned the cultivation of MON810 maize for the time being. In summer 2009, EFSA’s GMO Panel reassessed the safety of MON810 maize as part of the legally prescribed renewal process, taking into account the results of the model calculation of potential effects on butterflies and moths, which have since been published. The model: worst case scenario Three butterfly and moth species were selected for the model calculation: two protected butterfly species (admiral and peacock) and the diamond-back moth. The diamond-back moth is a serious pest for cultivated cruciferous plants, including oilseed rape. In laboratory experiments, the diamond-back moth was found to be one of the species most sensitive to the Bt protein Cry1Ab . The common stinging nettle was set as the host plant for the peacock and admiral butterflies, while a range of cruciferous plants were chosen for the diamond-back moth, including shepherd’s purse and wild radish. For simplicity, only one larval stage was studied for each of the butterfly and moth species: the first and most sensitive larval stage (L1) for the butterflies, and the L4 stage for the diamond-back moth, which reacted most sensitively to Bt protein in laboratory experiments. In order to make the model as realistic as possible, eleven representative European maize-growing regions were selected, of which six were in Germany (Aachen, Berkatal, Bonn, Grebbin, Oderbruch and the upper Rhine), two in the Po Valley in Italy, two in Spain (Madrid and the Ebro Valley) and one in Hungary (the Tolna region). |
The model calculates what quantity of pollen the larvae of the three butterfly and moth species ingest under different conditions. To find out whether these quantities are sufficient to harm butterflies and moths, the researchers needed to know what quantities of Bt protein trigger certain effects, e.g. increased mortality or delayed development. To find this out, the researchers evaluated published data from various laboratory feeding studies and from pollen measurements in the field. They took into account the worst case data in each case, for which the potential harm to the butterflies and moths is likely to be greatest.
A total of eleven different parameters were used in order to model pollen exposure under natural conditions.
These included:
-
various ‘physical effects’, e.g. the fact that pollen is washed off by rain, that the caterpillars eat the underside of the leaves and not the mid-vein where most of the pollen is found;
-
the timing of the maize-flowering period and the presence of sensitive larval stages of the butterflies and moths, which do not always coincide;
-
the density of the host plants in the field and in the field margins;
-
the average size of a maize field in hectares and the average width of the field margin in metres;
-
regional geographical factors, such as the proportion of maize-growing fields in a region, the proportion of host plants growing in the fields and in the field margins;
-
the proportion of maize fields growing MON810 Bt maize. Since Spain is the only country where Bt maize is grown on any appreciable scale, and no predictions are possible for the other EU countries, the model was also run using the highest percentage possible (80%) if Bt maize were authorised for unrestricted cultivation.
Results: Mortality almost unchanged
The impacts on butterflies and moths calculated using the model were very small. The maximum mortality rate calculated for peacock and admiral butterflies in all regions was less than one in 1572 individuals. For the diamond-back moth the maximum mortality rate was one in 392. The average mortality rate for all regions was one in 5000 for the two butterfly species, and one in 4367 for the diamond-back moth.
If the simulation is also set to take account of the fact that pollen can be deposited very unevenly, resulting in unusually high concentrations in some places, the mortality rate rises by no more than a third.
The authors concede that only limited data was available for some of the factors, but claim that the predictions are still sufficiently robust, since the estimates were, if anything, over-cautious and represent worst case scenarios. The simulation also used the most sensitive species and larval stages and set the Bt maize proportion at an unrealistically high level. Actual exposure could therefore be much lower.
![[jump to top]](/images/icons/icon_gotop_9x9.gif){kind=icon}