论文部分内容阅读
落叶松毛虫是大兴安岭林区的主要食叶害虫之一,利用模型模拟其景观尺度的空间分布、揭示其发生机理对落叶松毛虫灾害的预防和治理具有重要意义。本研究基于黑龙江省大兴安岭林区下辖3个林业局2008—2012年落叶松毛虫发生数据(小班水平),应用广义线性模型和最大熵模型预测了落叶松毛虫在立地条件、林分结构等环境变量共同作用下的空间分布,对比分析了环境变量的相对重要性,量化了落叶松毛虫发生概率对环境变量的响应规律。结果表明:广义线性模型与最大熵模型均取得了较好的预测结果,两种模型预测落叶松毛虫发生概率的空间分布大体一致,但通过设定不同的虫害发生概率阈值,适宜发生面积存在明显差别。海拔、龄组、土壤厚度、落叶松百分比是影响落叶松毛虫分布的重要环境变量。其中,落叶松毛虫发生概率在海拔300 m附近最高;中龄林、幼龄林的发生概率显著高于其他龄组;土壤厚度薄的小班更易爆发成灾;虫害发生概率与落叶松比例存在非线性正相关关系。
Dendrolimus cochinchinensis is one of the major leaf-eating pests in the forest of Daxinganling. Using the model to simulate the spatial distribution of its landscape scale, it is of great significance to reveal the mechanism of its occurrence on the prevention and control of the pine caterpillar disaster. In this study, based on the data of the occurrence of small-scale caterpillars of the three forestry bureaus in Daxing’an Mountains, Heilongjiang Province from 2008 to 2012, the generalized linear model and the maximum entropy model were used to predict the environmental conditions The spatial distribution under the joint action of variables, the relative importance of environmental variables was analyzed and compared, and the response rule of the probability of occurrence to environmental variables was quantified. The results show that both the generalized linear model and the maximum entropy model have achieved good prediction results. The spatial distribution of probability of occurrence of the two models are roughly consistent, but the appropriate area of occurrence is obvious by setting different probability of pest occurrence threshold difference. Elevation, age group, soil thickness and percentage of larch are the important environmental variables that affect the distribution of the pine caterpillar. Among them, the occurrence probability of Larix caterpillar was the highest around 300 m above sea level; the incidence of middle-aged forest and young forest was significantly higher than that of other age groups; small classes with thin soil thickness were easier to break out; the probability of occurrence of pests and the proportion of larch Linear positive correlation.