国有用材林害虫和害鼠的定量测报模型

收集了黑龙江省４个国有林区（哈尔滨、牡丹江、佳木斯和伊春林区）３９个林业局１１种主要害虫和１类害鼠的发生资料和气象资料。这１１种害虫和１类害鼠基本代表了黑龙江林区的虫害和鼠达发生的优势种类。通过相关分析和多元线性回归。建立了２１个预报模型。模型的Ｆ检验相关系数为０．８５以上。回报准确率达８５％以上。这些模型可用于黑龙江省国有林区主要害虫和害鼠发生量趋势预报。

Quantitative Prediction Models for Occurrence of Pests and Mice in State Commercial Forests

The occurrence data and weather information of 11 species of pests and one sort of mice, which basically represent the dominant species of insect and mice damages occurring in the forest areas of Heilongjiang province were collected from 39 forestry bureaus belonging to four state forest areas(Harbin, Mudanjiang. Jiamusi and Yichun forest areas). Through correlation analysis and multivariate linear regression, 21 prediction models were set up. The correlation coefficients for the models are above 0. 85 by Fs-test. and the accuracy of return prediction are above 85%. Those models can be used in prediction of occurring tendency of the major pest and mice in state forest areas of Heilongjiang provmce.