College of Forestry, Northeast Forestry University/Key Laboratory of Sustainable Forest Ecosystem Management, Ministry of Education, Harbin 150040, China.
Ying Yong Sheng Tai Xue Bao. 2024 Jul 18;35(7):1735-1743. doi: 10.13287/j.1001-9332.202407.006.
In order to analyze the growth pattern of tree height of planted and screen the provenances with fastest growth, we grouped the provenances using the differences in tree height, diameter at breast height (DBH) and volume of timber of 234 individuals of planted from 26 provenances in Maoershan Experimental Forest Farm. We constructed the growth equation for tree height by combining the base models of Gompertz, Korf, Richards, Logistic, and Schumacher, and then selected the optimal one. We introduced the prove-nance grouping as a dummy variable into the base model, and evaluated the optimal tree height growth equation by a comprehensive evaluation of the model according to the coefficient of determination (), the root-mean-square error (RMSE), the Akaikei Information Criterion (AIC), and the model's predictive precision (). The results showed that the growth traits of the 26 provenances had significant difference among the groups, and that tree height and DBH showed significant differences among the provenances. According to the comprehensive consideration of different growth traits, the four groups of provenance growth were divided into group A (Wuying, Hebei, Linjiang, Dongfanghong, Huanan, Lushuihe, Fangzheng) >group B (Aihuisanzhan, Liangshui, Tieli, Qinghe) > group C (Wuyiling, Zhanhe, Liangzihe, Baihe, Chaihe, Caohekou, Bajiazi) >group D (Tongzigou, Dashitou, Wangqing, Helong, Yanshou, Dahailin, Xiaobeihu, Muling). The optimal base tree height growth model of the four groups was the Gompertz model, and the fitting accuracy of the model after the introduction of dummy variables (=0.9353) was higher than that of the base model (=0.9303), and the model prediction accuracy was also improved. The tree height growth curves of each provenance group conformed to the "S"-shaped rule of change. There were obvious differences among the groups, with the best performance of the provenances in group A. The growth of from different provenances was different, and the tree height growth model with dummy variables of provenance groups could effectively improve the prediction accuracy of the model, reflect the differences in height growth of of different provenances, which could provide the scientific basis for the selection and cultivation of plantations.
为了分析人工林树木树高生长模式，并筛选生长最快的种源，我们利用毛儿山实验林场 26 个种源的 234 株人工林树木的树高、胸径（DBH）和材积的差异，对种源进行了分组。我们通过结合 Gompertz、Korf、Richards、Logistic 和 Schumacher 基本模型，构建了树高生长方程，并根据决定系数 ()、均方根误差 (RMSE)、赤池信息量准则 (AIC) 和模型预测精度 ()，对基本模型引入种源分组作为虚拟变量进行综合评价，选择最优的树高生长方程。结果表明，26 个种源的生长特性在组间存在显著差异，树高和 DBH 在种源间存在显著差异。根据不同生长特性的综合考虑，将 4 组种源的生长分为组 A（五营、河北、临江、东方红、华南、露水河、方正）>组 B（爱辉山站、凉水、铁力、清河）>组 C（五营岭、沾河、亮子河、白河、柴河、曹大口、八家子）>组 D（佟子沟、大石头、汪清、和龙、延寿、大海林、小北湖、穆棱）。4 组的最优基本树高生长模型均为 Gompertz 模型，引入虚拟变量后模型的拟合精度（=0.9353）高于基本模型（=0.9303），且模型预测精度也有所提高。各种源组的树高生长曲线符合“S”型变化规律。组间差异明显，种源 A 组表现最好。不同种源的生长情况不同，引入种源组虚拟变量的树高生长模型可以有效提高模型的预测精度，反映不同种源树木的高度生长差异，为人工林树木的选择和培育提供科学依据。
