Response to Wilkinson & Tepfer's "Fitness and beyond: preparing for the arrival of GM crops with ecologically important novel characters". Fuzzy reasoning and unacceptable change: defining and assessing an ambiguous endpoint.

In the context of ecological risk assessment, we require an effective definition of the undesirable endpoint. These may be defined according to scientific, economic and amenity value sets. Unfortunately, since the processes we are dealing with are continuous over space and time and also dynamic (in that ecological population sizes fluctuate), unambiguous definitions are elusive. In the case of enhanced fitness, an organism that is fitter than a direct competitor will continuously increase its numbers at the expense of its competitor. Unless a change in fitness or environment occurs, the competitor will often eventually be driven to local extinction. It could therefore be argued that any enhanced fitness is undesirable. We can be concerned about the decline of a particular species, assemblage or visual landscape, but at what point does it become unacceptable? At one extreme, one could argue that any decline is unacceptable. At the other, a definite endpoint would be the “rapid” extinction of whatever we are trying to protect. In between, we have a shift in the ecological balance, leading to a continuous decline of the protectee. Here we have to define a timescale, and some form of measurable degree of change. Whilst we can ignore the problem and use true but unhelpful statements such as “every situation will have its own unique criteria”, this will not lead to a satisfactory definition. Instead, we require a concrete framework to deal with what is essentially a qualitative judgement. It seems to me that the crux of the problem lies in its inherent ambiguity. Acceptability is continuous in itself. An ecological situation will become less acceptable and more unacceptable as it moves in the undesirable