Risk assessment and risk management of transgenic trees

My name is Beatrix Tappeser. I am the head of division "GMO-regulation and Biosafety" at the Federal Agency for Nature Conservation, Germany

Forests are of great significance for humans from a socio-economic, an environmental and a cultural point of view. An important economic factor is the production and use of wood as an industrial and energy resource, while from an ecological perspective, the conservation of biological diversity and of ecological functions of forests are highlighted and connected with their sustainable use.
A number of ecological risks are being discussed concerning single traits of transgenic trees and their overall use. This discussion is based upon experience with genetically modified agricultural crops, which are commercially grown since the 1990s (e.g. herbicide tolerance and pest resistance), and ex-tended to include the special characteristics of trees and the complexity of forest ecosystems. With other traits, such as modified wood composition, however, no experience from other plants exists. Transgene escape through out-crossing and vegetative propagation is the most widely acknowledged risk connected with genetically modified trees (i.a. Williams & Davis 2005, DiFazio et al. 2004). Different approaches are being followed to prevent transgene escape by means of bio and gene technology (e.g. inhibition of pollen production, GURTs). Suppressing of pollen production, on the other hand, may have impact on pollinators and pollen feeders.
The identified risks of transgenic trees were compiled by several authors and are being discussed controversially, mostly since commercial use of transgenic trees has become possible and is being propagated for different reasons. However, studies addressing these identified and discussed risks of genetically modified trees by exemplified experiments under laboratory and greenhouse conditions are still widely missing. One the other hand, attempts are being made to prevent risks from the outset, including inhibition of the reproductive abilities of trees. Results of this research, however, show that stability of the transformed genotype is of special concern (i.a. Frankenhuysen & Beardmore 2004)
In general, there is still considerable uncertainty of potential impacts and risks of transgenic trees on biodiversity. Therefore, also considering the still deficient understanding of forest ecology and of the causes of plant invasions, frequently raised claims to approve the commercial use of transgenic trees must be considered premature. Also, existing regulations for genetically modified plants must be amended to include the special characteristics of trees. Application of the step-wise approach must be ensured also in the environmental risk assessment of transgenic trees, i.e. a decision on whether release experiments can be approved is only possible after appropriate assessments under laboratory and greenhouse conditions. Modelling approaches may be of special help (i.a. Farnum et al. 2007).

DiFazio SP, Slavov GT, Burczyk J, Leonardi S & Strauss SH (2004): Gene flow from tree plantations and impli-cations for transgenic risk assessment. Plantation Forest Biotechnology for the 21st Century.
Farnum P, Lucier A & Meilan R (2007) Ecological and population genetics research imperatives for transgenic trees. Tree Genetics & Genomes, 3, 119-133.
van Frankenhuyzen K & Beardmore T (2004): Current status and environmental impact of transgenic forest trees. Canadian Journal of Forest Research, 34, 1163-1180.
Williams CG & Davis BH (2005): Rate of transgene spread via long-distance seed dispersal in Pinus taeda. Forest Ecology and Management, 217, 95–102.

I am Tom Nickson, a scientist working in the area of environmental risk assessment for Monsanto Company for about 15 years.  I would like to add to the idea of the "stepwise approach" used for environmental risk assessment.

Over the past 20 years, much has been learned about the process and conceptual basis that underpins environmental risk assessment.  In my experience, one of the most significance challenges facing a risk assessor is discerning the amount of information needed for the risk assessment.  Data requirements are dictated by the level of certainty required in the risk assessment, which is a matter of policy, not science.  No matter what the object of the risk assessment is (trees, crops, fish, etc), overall guidance for the risk assessor is environmental protection goals set within policy, which determines what attributes of the environment are valuable, in need of protection and at what cost they will be protected.  Typically, food and fiber production are an environmental attribute whose protection will be balanced against the desire for natural and urban areas. 

As a developer of products improved through biotechnology, I have seen problems arising from not following a stepwise approach that begins with problem formulation (Raybould, 2006).  However, I would also like to highlight another problem, which is the perception that science, particularly ecology, will define or describe "risk".  Recently, Raybould (2007) has commented that risk assessment must have a structure grounded in policy.  He described a commonly encountered problem as a conflict between an ecological versus and ecotoxicological approach.  After reading many postings on these on-line fora, I feel it important to point out this distinction.  Risk assessment should not be a surrogate for basic research.  Public policy must guide the degree and even nature of data needed for decision-making, not a scientist or particular group versed in ecological theory.  A risk assessment based in basic research does not serve the public since it creates excessive costs to the regulatory authority and developer, results in delays to introducing valuable products into the market and provides no greated certainty in decision-making (Raybould, 2007). 

Regardless of the LMO, I believe that environmental risk assessment should begin with proper problem formulation (Raybould, 2007).  This critical first step enables risk assessors to use a guided process that is as efficient as possible, and ensures that sufficient data are collected for a decision based on reasonable certainty and guided by policy.   

Raybould A. 2006. Problem formulation and hypothesis testing for environmental risk assessments of genetically modified crops. Envion.Biosafety Res. 5: 119-125

Raybould A (2007) Ecological versus ecotoxicological methods for assessing the environmental risks of transgenic crops. Plant Science: 173: 589-602