Objective: This study aimed to construct a growth model of diameter at breast height(DBH) for Chinese fir plantations based on competition and climate factors, and analyze effects of competition, climate factors and their interactions on DBH growth. The results were expected to provide a reference for the growth of reserved trees after thinning and selective cutting under climatic changes. Method: Based on the permanent sample plots in the south of Jiangxi Province, the potential increment equation with multiplicative modifiers was used to construct DBH growth model. The quantile regression technique was used to model the potential DBH increment, and seven environmental factors(mean temperature, maximum temperature, minimum temperature, precipitation, accumulated temperature greater than 5℃, elevation, slope) were used to describe the effects of site quality on potential increment. Parameter significance and variance inflation factor were chosen to determine environmental factors, which could be the independent variables. Exponential function form was used to construct modifier functions. Competition variables included in the modifier functions were three stand-level and three tree-level competition indices(2 distance-independent indices and 1 distance-dependent index). The interactions between the selected best competition index and five climate factors on good-of-fit of growth models were also considered. According to model evaluation, the model with the best good-of-fit was selected to add plot-level random effect parameters, and then used to analyze the effects of competition, climate and their interactions on DBH growth of Chinese fir plantations. Model evaluation criteria included mean absolute error(MAE), relative mean absolute error (RMAE), and mean predicted error(MPE). Result: Elevation, minimum temperature and precipitation during investigation intervals showed significant effects on the potential DBH increment for Chinese fir. The potential increment would reach the maximum when DBH was 20 cm. Minimum temperature and precipitation during investigation intervals showed positive effects on the maximum of potential DBH increment, but elevation showed a negative effect. Compared with DBH growth models based on the other competition indices, the models with distance-independent indices showed the best good-of-fit, and followed by the models with distance-dependent indices. After containing the interactions between climate and competition, the good-of-fit of models was improved. Compared with the models which only included competition indices in modifier functions, MAE of growth models with interactions decreased 0.60%-18.69%, MPE decreased 0.12%-9.72%. The model included the interactions between basal area in larger trees as the competition index and mean temperature, maximum temperature and minimum temperature as climate factors showed the best good-of-fit, thus the model was selected to add plot-level random effect parameters to construct the final model. The final model had a good good-of-fit, and MAE was 0.071 1 cm, RMAE was 20.37%, MPE was 4.90%. The final model showed that temperature could increase the modifying effects of competition on potential DBH increment, except for the trees with a little competitive pressure or with a large competitive pressure. Conclusion: The quantile regression could present a good effect when used to model the potential increment. The interactions between climate and competition could improve the good-of-fit of DBH growth model. Therefore, if climate changed dramatically, we recommended that the interactions between competition and climate should be considered into individual tree growth model. © 2021, Editorial Department of Scientia Silvae Sinicae. All right reserved.
