Authors' abstract:

Gene flow is a major determinant of impacts of forest tree plantations on surrounding populations and ecosystems. Realistic predictions of gene flow are therefore essential for scientifically credible assessments of the impacts of transgenic plantations.

The choice of methods for measuring gene flow will be dictated by the organism and environment under study, the availability of tools and resources, and the desired scope of inference. Methods include direct tracking of propagule movement, parentage analysis, and analysis of genetic diversity in seeds, seedlings, and pollen. Gene flow estimates can be integrated with ecological and demographic data in spatially explicit simulation models to allow projections of transgene dispersal over large areas and long time frames. Such models allow exploration of a large number of scenarios of plantation cultivation, and help to identify the key parameters controlling gene flow. We describe a simulation study of gene flow from hybrid poplar plantations that illustrates some of the key issues in estimating and modeling transgene flow from plantations. For example, the results suggested that accurate estimates of long-distance gene flow are considerably more important to prediction of transgene dispersal than are estimates of local (<1 km) dispersal patterns.