Abstract: Based on 234 Larix kaempferi tree samples from 78 sampling plots across three regions, including Liao-ning, Hubei, and Gansu provinces, we selected the optimal one among the six commonly used taper equations in forestry research to construct a basic model, a climate response model, and a nonlinear mixed-effects model suitable for different climatic regions with the climatic factors being added in an exponential form. We further investigated the stem form indices of L. kaempferi in different regions, analyzed the differences in stem form responses to climatic variables, and selected the optimal universal equation based on the fitting accuracy of the three models. The results showed that Kozak model had the best universality and was the optimal basic model. After simultaneously introducing annual average temperature and annual average precipitation into the optimal basic model to the climate response model, the fitting accuracy of the model was improved by considering the impact of various climatic factors on stem form. Additionally, when different regions were introduced into the nonlinear mixed-effects model as random factors, the fitting accuracy reached the highest (R²=0.9874) among the three models with the lowest AIC (6426.04) and BIC (6512.88) values according to the evaluation indicators and residual plots. The root-mean-square error was reduced by 4.9% and 4.0% compared with the basic model and climate response model, respectively. Therefore, the nonlinear mixed-effects model could be the optimal universal taper equation for describing stem form of L. kaempferi in the three regions.

Key words: Larix kaempferi, taper equation, climate response model, nonlinear mixed-effects model