Regulation should be based on data, not just models

As scientists, we prefer to test ideas with experiments before making recommendations or policy decisions. However, there are many questions of interest where the potential interactions are complex, where critical events are rare, and where long time frames are required for meaningful results, particularly for ecological systems that naturally change with time. Such questions are difficult and costly to resolve with convincing experiments. However, even in the absence of experiments that can be generally accepted as informative, decisions on key questions must be often be made, especially by regulators, because indecision and continuation of the status quo also has its own risks. Because of this reality, mathematical models gained a critical role in regulatory processes, long before transgenic crops and other controversial issues reached the public consciousness. Still, regulatory decisions should be based on sound data, whenever possible. The aim of this chapter is to suggest a perspective by which data collection, experiments, and modeling might be most usefully linked for examining transgenic crops in general, and ferality, in particular.