Prediction model for the quantity and density of first-order branches of Larix kaempferi in eastern area of Liaoning Province, China

Ming-Qi Ni,Hui-Lin Gao,Jia-Teng Liu,Yi-Wen Tong,Yu Qiu,Hui Xing
DOI: https://doi.org/10.13287/j.1001-9332.202408.006
Abstract:As an important branch characteristic factor, the quantity of branches could influence crown structure, tree growth, and wood quality. Taking Larix kaempferi plantation in Dagujia Forest Farm, Qingyuan County, Liao-ning Province as the research object, we developed a mixed effect prediction model of the first-order branches quantity of L. kaempferi including sprouting branches based on the negative binomial distribution model, and a mixed effect prediction model of the first-order branches density of L. kaempferi including sprouting branches based on the negative exponential model. The results showed that the mixed effect model considering sample level as the random effect effectively decreased the heteroscedasticity and autocorrelation. The fitting goodness was better than the traditional model. The quantity of the first-order branches increased with increasing crown ratio. The mixed effect model with the basic model intercept of the first-order branches quantity as the random effect parameter was determined as the optimal model, with Ra2=0.552 and the RMSE=7.242. As for the density of the first-order branches, the heteroscedasticity and autocorrelation were also reduced when the random effect was added. The density of the first-order increased with increasing crown ratio. The mixed effect model with the basic model intercept of the first-order branches density model and branch depth as random effects was determined as the optimal model, with Ra2=0.792 and the RMSE=4.447. The model for branch quantity and density of L. kaempferi constructed would lay an important foundation for making scientific forest management plans and improving wood quality.
What problem does this paper attempt to address?