Abstract:
Forest management is essential for maintaining environmental stability and ecological biodiversity. The high species diversity of tropical rainforest forests obstructs the development of forest dynamic models. A lot of tree species exist in the forest for which each type of species will have insufficient data for reliable parameter estimation. The best way to avoid bias prediction is to group the trees based on their characteristics similarity. In a tropical rain forest in Koh Kong province, Cambodia, four species groups have been classified using statistical analyses of principal component analysis (PCA) and cluster analysis. Some indices related to diameter structure, growth, mortality and recruitment of species were formed from measurement data rather than the parameter estimates of some predetermined growth regression functions.
