蒙古栎天然林单木生长模型研究:Ⅲ.单木枯死模型
Growth models for natural Mongolian oak forests:Ⅲ.Individual-tree mortality model

【目的】研究蒙古栎天然林的单木枯死模型，为蒙古栎天然林的合理经营提供依据。【方法】以蒙古栎天然林为研究对象，基于吉林省汪清林业局195块固定样地的2期复测数据，采用二分类的Logistic回归方法建立蒙古栎天然林的单木枯死模型。【结果】期初胸径、竞争指数、林分密度都是显著影响树木枯死的因子，林木直径是影响林木枯死的重要因子，直径越小，枯死概率越大。χ2检验结果以及ROC(Receiver operating characteristic)曲线显示，使用Logistic回归可以有效预测树木的枯死情况，所建模型较合理。【结论】建立的单木枯死模型统计可靠，可为吉林省汪清地区蒙古栎天然林的生长预测提供依据。
【Objective】Aiming to provide foundation for reasonable management of Mongolian oak natural forests,individual-tree mortality model was researched.【Method】The data used to develop individual-tree mortality model for predicting Mongolian oak natural forests were collected from 195 re measured permanent sample plots in Wangqing Forest Bureau,Jilin province.Then,the individual tree mortality model for Mongolian oak natural forests was developed using binary logistic regression method.【Result】The initial diameter at breast height,competition index,and stand density (number of trees per hectare) had significant effects on tree mortality.The mortality probability increased as the decrease of diameter.Chisquare-Test (χ2) and receiver operating characteristic (ROC) curve showed that the binary logistic could effectively predict death of trees.【Conclusion】The established individual-tree mortality model was reliable to predict the growth of Mongolian oak natural forests in Wangqing region,Jilin province