IMPROVING INDIVIDUAL CROWN BIOMASS ESTIMATION BY INCORPORATING COMPETITION FACTORS USING MIXED EFFECT MODELS FOR PINUS KESIYA.

Due to the high uncertainty of tree crown biomass modeling, it is crucial to estimate individual tree crown biomass by incorporating competition factors using mixed effect models. The crown biomass of 128 sampling trees was investigated at three typical sites of the natural Pinus kesiya forest in Pu'er city of Yunnan province, China. Considering the random effects of the site index and incorporating competition factors, the branch and needle biomass models were constructed using the nonlinear mixed effect model. The results showed that: (1) the mixed effects models, including the fixed effect of competition factors, had a better fitting performance than the ordinary mixed model for branch biomass, however, mixed effects models without the fixed effect of competition factors had the best-fit performance for the needle biomass; (2) mixed effect models incorporating competition factors had better prediction ability because of the highest precision. The increase in accuracy varied from 49.87 to 70.27% for branch biomass and from 66.19 to 66.57% for needle biomass. Mixed-effects models, considering site effect and competition factors, may provide a flexible and powerful tool for individual crown biomass estimation. [ABSTRACT FROM AUTHOR]