思茅松天然林树冠结构模型

以云南省普洱市思茅区思茅松天然林为研究对象,采用枝解析调查了34株思茅松样木的树冠数据,分析了一级枝枝长、枝径、着枝角度、弦长和树冠半径5个树冠形状变量的变化规律,分别构建其预估模型;分析了树冠结构变化,分别构建了一级枝轮枝高度预估模型、一级枝枝条数量预估模型和一级枝枝条数量累积预估模型,并采用独立样本进行模型统计精度检验。结果表明:8个预估模型的预测效果良好,精度达到91%以上,尤其是一级枝着枝角度模型和一级枝轮枝高度模型预测精度达到97%以上。研究结果合理准确描述思茅松树冠结构的变化,为思茅松天然林的经营管理提供科学依据。

Modeling tree crown structure of Simao pine (Pinus kesiya var. langbianensis) natural forest

Tree crown structure plays an important role in tree growth and forest management. Simao pine (Pinus kesiya var. langbianensis) is an important fast-growing coniferous tree and an important source of timber. Studying the crown structure of Simao pine natural forest has important theoretical and practical significance for forest management. Branch analysis of the crown structure of 34 sample trees was investigated in Simao District, Pu'er, Yunnan Province, China. The sample trees ranged in age (A) from 16 to 76 a, diameter at breast height (DBH) from 6.0 to 51.3 cm, tree height (H) from 6.3 to 27.4 m, crown width (CW) from 2.0 to 15.7 m, and crown length (CL) from 2.3 to 16.4 m. Models of the tree crown shape and structure variables were built by stepwise regression analysis using SAS statistical software. Three predictive models were established based on five independent variables of tree crown shape using a logarithmic linearization power function. Those variables were: the length of primary branches (BL), the diameter of primary branches (BD), the angle of primary branches (AB), the chord length of primary branches (BCL) and the crown radius (CR). Meanwhile, predictive models were established based on the three independent variables related to tree crown structure using three multivariate linear models: the growth height of primary branches (HGB), number of whorl branches (NWB) and the cumulative number of whorl branches (CNWB) models. Tests used to check the statistical accuracy of the models were carried out using independent samples. The total relative error (RS), mean relative error (EE), mean absolute relative error (RMA) and predictive accuracy (P) were selected to evaluate the models. The result showed that the eight predictive models performed well and the predictive accuracies of all models exceeded 91%; in particular, the accuracies of both the AB and HGB models were above 97%. Moreover, in the modeling of tree age (A), H and DBH resulted in different values in the different models. A did not have a significant effect on the crown shape variables in the models, but the crown structure variables did have a significant effect on A, indicating that the effect of age on the crown shape variables was not significant, but A did have a significant influence on crown structure variables. Furthermore, H did not have a significant effect on the models except for the modeled value of AB which did have a significant effect on the crown shape models; also, A had a significant effect on the modeling of crown structure. Meanwhile all modeled values for variables except for AB were significantly correlated with DBH, and all modeled values were positively correlated with DBH except for when the CNWB predictive model was used. This showed that when DBH was larger, BL, BD, BCL, CR were also larger, and more whorled branches were present. The negative correlation between DBH and stand density also showed that when stand density was higher, BL, BD, BCL, CR and NWB were all smaller. Overall, the models were suitable for describing the trends and inherent variability of the crown shape and the structure, and provided a valuable reference for the management of Simao pine natural forest.