Risk Assessment of genetically modified (transgenic) trees, an example from the Netherlands
In the Netherlands we have up to this moment very limited experience with the environmental risk assessment of genetically modified trees. From this experience, however, we would like to present some thoughts that we hope are useful for the process of risk assessment for genetically modified trees.
It should be clear that our comments are presented as one would do this in an informal discussion between regulators and other interested parties, as points for discussion. They are not official standpoints of the Netherlands competent authority.
General considerations
As a point of departure we agree with the conclusion that was also drawn by the Canada-Norway Expert Workshop on Risk Assessment for Emerging Applications of Living Modified Organisms, that risk assessment of genetically modified trees can be done in a scientifically sound manner according to the methodology of Annex III of the Protocol (the report of the Canada-Norway workshop is included in the selected readings under the documents recommended by COP-MOP). The methodology of Annex III takes into account in the first step, inter alia, the specific characteristics of the receiving organism, the tree in this case. These characteristics are subsequently taken into account in the next steps.
The overall methodology of risk assessment of Annex III is a topic that will be discussed more extensively in the second session of this forum, when a further elaboration of the risk assessment process under Annex III of the protocol will be a topic for discussion.
Specific examples
At this moment there are two cases of deliberate release of transgenic trees in the Netherlands, one is an ongoing field trial (genetically modified apple trees expressing hordothionin), the other is an application for a field trial (genetically modified poplar trees with reduced lignin content)that is not yet permitted.
(1) Genetically modified apple trees expressing a hordothionin gene from barley (Hordeum vulgare).
Small scale confined field experiments with these trees have been permitted, under specific conditions (giberilin spraying to prevent flowering, and observation of effectiveness of this measure), to prevent outcrossing, so the potential effects of the field experiment will be restricted to the field plot where the apple trees have been planted.
Hordothionin expression is expected to lead to reduced sensitivity of the trees for fungal and bacterial infection. Hordothionin is known to have this function in barley, by reducing growth of fungi and bacteria. Thionins in general have this function in many plants.
From general experience with apple trees, it is not expected that reduced sensitivity to infectious disease will have a decisive impact on general fitness or invasiveness of apple, so no adverse effects on biodiversity are expected in this respect.
As another consideration, there may be effects on soil microflora, as those are the target organisms of hordothionin. The baseline experience is that no such effects have been observed for hordothionin in its natural environment, although we are familiar with barley in long term cultivation. It may be argued that in our agricultural practice barley will be cultivated in rotation with other crops, which may mask adverse effects to a certain extent. A typical effect of hordothionin expression in trees, e.g. in apples, would be that soil microflora would be exposed over longer continuous periods to hordothionin. In that respect, it might be interesting to see what effects are observed in no till cultivation of barley, as a baseline for a longer continuous period. Such data, if available, may be informative if at later stages of development of the transgenic apple trees larger areas will be planted in larger scale field experiments.
In the discussion on risk assessment, we focus on the potential adverse environmental effects, as they may arise from the deliberate release of the genetically modified trees into the environment. As we are dealing with a small scale field experiment in this case, we have not yet had extensive discussions on topics that may become important at later stages, such as changes in agronomic management, as these are not (yet) actual.
(2) A proposed field experiment with genetically modified poplar trees (Populus sp.), modified for reduced lignin content.
This is intended to be a small scale field experiment. The application is still under consideration.
In this case an OECD consensus document is available on the biology of poplar (
http://www.olis.oecd.org/olis/2000doc.nsf/LinkTo/NT00002EC2/$FILE/JT00103743.PDF), which provides baseline information on the behavior of poplar trees. A relevant important trait of poplar is that the trees are (normally) dioecious and obligatory outcrossing. Using female trees in a field experiment (as is done in this case) will reduce chances of outcrossing enormously, in the plants that are planted in the experiment.
What will be the most important area for discussion is the potential adverse effects that reduction of lignin content, the transgenic trait, may have on the interaction of the poplar trees with their environment: interactions with the (micro-) flora and fauna.
For this discussion we need to have knowledge of the baseline: an overview of the natural variation in lignin content within poplar species, and between species, and in general, between woody crops (and trees in general), and the typical environmental effects that various abundancy of lignin in crops will have, also compared to herbaceous crops; and on environmental effects that have been correlated with these variations.
The applicant of the field experiment has already provided information on these issues (e.g. the file attached to this contribution). At later stages it may be helpful if results of broader discussions on the trait of lignin content and its potential environmental effects would be available.
The so-called trait documents of OECD are a good source of this type of information, e.g. the document on safety information on transgenic plants expressing Bt toxins, for the Bt trait (
http://www.olis.oecd.org/olis/2007doc.nsf/LinkTo/NT00002DF6/$FILE/JT03230592.PDF).
No OECD consensus document is available for lignin, however, and from experience we know that the drafting of such a document is not easy and takes a considerable time.
Still, it would be very useful to have documents available at an early stage of the risk assessment discussion, that provide scientifically sound overviews on the ecological impact of trees with reduced lignin content. Halpin et al. (Tree Genetics & Genomes (2007) 3:101–110) have recently published a paper on this subject, that may be a good start in this direction. This paper is added to this contribution.
