黑龙江省红松和长白落叶松人工林树冠外部轮廓模拟

【目的】红松和长白落叶松是黑龙江省主要的造林树种,构建两个针叶树树冠外部轮廓预估模型,为进一步科学合理地经营人工林提供参考。【方法】采用黑龙江省红松和长白落叶松人工林中各50株解析木的枝条解析数据,以分段抛物线方程、修正Kozak和Weibull方程为基础模型,构建以样木为单一水平树冠外部轮廓的非线性混合效应模型,并对两树种的树冠轮廓进行比较。【结果】两树种树冠外部轮廓的分段抛物线预估模型包含的变量均为D_BH(胸径)、C_L(冠长)、C_R(冠长率)、H_D(高径比),修正Kozak和Weibull预估模型包含的变量为D_BH、C_R、H_D。加入混合效应后,各模型的拟合效果较基础模型均有所提高。修正Weibull方程对红松和长白落叶松的拟合效果最好,但与分段抛物线和Kozak方程之间的差别不是很大。对于红松,分段抛物线方程预估的树冠半径在靠近树冠基部要低于修正的Kozak方程和修正的Weibull方程; 对于长白落叶松,各模型之间的拟合效果差别较小。【结论】分段抛物线方程、修正Kozak和Weibull方程均能够对黑龙江省红松和长白落叶松人工林树冠外部轮廓进行比较准确的模拟。修正Kozak方程和Weibull方程包含的树木变量较少,参数估计相对简单,容易通过积分计算树冠体积及表面积。

Modelling outer crown profile for planted Pinus koraiensis and Larix olgensis trees in Heilongjiang Province,China

Abstract: 【Objective】The Pinus koraiensis Sieb. et Zucc(Korean pine)and Larix olgensis Henry(Olgan larch)are the main afforesting tree species in Heilongjiang Province. Developing outer crown profile models for the two tree species will provide references for forest management of plantations. 【Methods】The branches from 50 trees of planted Korean pine and Olgan larches were used. The segmented polynomial equation, modified Kozak equation, and modified Weibull equation were used as the basic model, and a nonlinear mixed-effects model at the tree level was developed. The crown profiles of the two tree species were compared. 【Results】For both tree species, the segmented polynomial predicted model, included diameter at breast height(DBH), crown length(CL), crown ratio(CR), and height to diameter ratio(HD), the modified Kozak predicted model included DBH, CR and HD, and the modified Weibull predicted model included DBH, CR and HD. By including mixed-effects into the models, the goodness of fit for all the models was improved compared with the basic models. The modified Weibull equation performed best for both Korean pine and Olgan larch. However, there was not a large difference between the modified Kozak equation, Weibull equation, and the segmented polynomial equation. As for the Korean pine, the crown radius of the lower crown estimated by the segmented polynomial predicted model was lower than that of the modified Kozak and Weibull equation, whereas the differences between the three models for the Olgan larch were relatively less. 【Conclusion】 The segmented polynomial equation, modified Kozak, and modified Weibull equation performed well in crown profile modeling with high accuracy. Fewer variables were included in the modified Kozak equation and modified Weibull equation than the segmented polynomial equation. The parameter estimation for the modified Kozak and Weibull equation were also stable, and the crown volume and surface area were easy to calculate by integration.