Abstract
Forest management relies on forest spatial structure, but studies on the spatial structure of Vietnamese forests are limited. Therefore, two one–hectare plots were established in tropical rainforest in Tuyen Quang province to evaluate the characteristics and spatial structures of plant communities. Descriptive statistics and species importance value indices were calculated for each plot. Then, to assess differences between plots, detrended correspondence analysis was applied. Species mingling, uniform angle, diameter dominance and carbon stock dominance indices were also computed to analyze forest spatial structures. The results indicated that 65–70% of trees occurred in areas with high diversity, and 55–58% of trees were randomly distributed. The number of dominant and non–dominant individuals, in terms of diameter and carbon stock, was similar at the stand level. In all estimated patterns, units with high species mixtures predominated. As biodiversity increased, the distribution of forest trees became more random. The quadrivariate distribution illustrated considerable differences between plots, the study area and other regions. Findings from multivariate distributions provide accurate information in each group of spatial indices and these can be used to select appropriate silvicultural measures for stand management.