Individual tree growth models are flexible and commonly used to represent growth dynamics for heterogeneous and structurally complex uneven-aged stands. Besides traditional statistical models, the rapid development of nonparametric and nonlinear machine learning methods, such as random forest (RF), boosted regression tree (BRT), cubist (Cubist) and multivariate adaptive regression splines (MARS), provides a new way for predicting individual tree growth. However, the application of these approaches to individual tree growth modelling is still limited and short of a comparison of their performance. The objectives of this study were to compare and evaluate the performance of the RF, BRT, Cubist and MARS models for modelling the individual tree diameter growth based on tree size, competition, site condition and climate factors for larch–spruce–fir mixed forests in northeast China. Totally, 16,619 observations from long-term sample plots were used. Based on tenfold cross-validation, we found that the RF, BRT and Cubist models had a distinct advantage over the MARS model in predicting individual tree diameter growth. The Cubist model ranked the highest in terms of model performance (RMSE_cv [0.1351 cm], MAE_cv [0.0972 cm] and R²_cv [0.5734]), followed by BRT and RF models, whereas the MARS ranked the lowest (RMSE_cv [0.1462 cm], MAE_cv [0.1086 cm] and R²_cv [0.4993]). Relative importance of predictors determined from the RF and BRT models demonstrated that the competition and tree size were the main drivers to diameter growth, and climate had limited capacity in explaining the variation in tree diameter growth at local scale. In general, the RF, BRT and Cubist models are effective and powerful modelling methods for predicting the individual tree diameter growth.

Alternative methods have been used for developing tree basic and generalized height–diameter models, including non-linear least-squares regression, non-linear mixed-effects method, quantile regression (QR) and a generalized additive model (GAM). However, these approaches have not been thoroughly compared. The purpose of this study was to compare these methods applied to tree height–diameter models for larch plantations in northern and northeastern China. The data were obtained from an extensive database curated by the China National Forest Inventory (NFI) that covers a wide distribution and range of stand conditions, including 5684 paired heights and diameters in 342 plots. Random effects at the Province, species and plot levels, and a combination of Province and species into the basic and generalized H–D models were incorporated. This study demonstrated that the mixed-effects basic model provided the best performance for model fitting and validation; QR could represent variable H–D relationships at different quantiles and the GAM was a suitable alternative for flexible modelling of the H–D relationship, especially for generalized H–D model. This finding indicates that the mixed-effects basic model does not require additional variables and is an adequate tool for generating missing height measurements for larch plantations in northern and northeastern China.

Forests play a key role in regulating the global carbon cycle, a substantial portion of which is stored in aboveground biomass (AGB). It is well understood that biodiversity can increase the biomass through complementarity and mass‐ratio effects, and the contribution of environmental factors and stand structure attributes to AGB was also observed. However, the relative influence of these factors in determining the AGB of Quercus forests remains poorly understood. Using a large dataset retrieved from 523 permanent forest inventory plots across Northeast China, we examined the effects of integrated multiple tree species diversity components (i.e., species richness, functional, and phylogenetic diversity), functional traits composition, environmental factors (climate and soil), stand age, and structure attributes (stand density, tree size diversity) on AGB based on structural equation models. We found that species richness and phylogenetic diversity both were not correlated with AGB. However, functional diversity positively affected AGB via an indirect effect in line with the complementarity effect. Moreover, the community‐weighted mean of specific leaf area and height increased AGB directly and indirectly, respectively; demonstrating the mass‐ratio effect. Furthermore, stand age, density, and tree size diversity were more important modulators of AGB than biodiversity. Our study highlights that biodiversity–AGB interaction is dependent on the regulation of stand structure that can be even more important for maintaining high biomass than biodiversity in temperate Quercus forests.