二代棉铃虫种群动态的地统计学分析

Geostatistic analysis on population dynamics of second generation cotton bollworm

运用地统计学方法，分析了山东省滨州地区二代棉铃虫在1965-2004年间种群动态的时间序列结构特征和自相关性，根据决定系数尺。和Akaike’s Information Criteria（AIC）判别参数，筛选出最优拟合模型。结果表明：棉铃虫越冬蛹量、二代棉铃虫百株累计卵量、发生程度的最优拟合模型均是球形模型，呈聚集分布；三者的变程分别为10．41、12和12．06年，高度一致；三者在时间序列上的自相关性分别是53．22％、52．62％和69．06％；三者的结构特征引起的种群变异介于25％～75％，为中等程度的空间自相关性，其相关强度顺序是百株累计卵量〈越冬蛹量〈发生程度，随机部分引起的空间变异则相反。

By the methods of geostatistics, this paper analyzed the temporal structural characters and autocorrelations of the population dynamics of second generation cotton bollworm （ Helicoverpa armigera） in Binzou City of Shandong Province in 1965-2004. Optimal simulation models were selected based on the decision coefficient R2 and Akaike＇ s Information Criteria （AIC）. The resuits showed that the optimal simulation models of the amount of overwintering pupae （ Ⅰ ）, accumulated amount of eggs on 100 cotton plants （ Ⅱ ）, and occurrence degree （ Ⅲ ） of second generation H. armigera were all of spherical. These three indices had an aggregative distribution, and their change ranges were 10. 41, 12 and 12. 06 years, respectively, with a high accordance. The temporal autocorrelations of the three indices were 53.22%, 52. 62% and 69.06%, respectively, and the population variance resulted from the structural characters of the indices ranged from 25 % to 75 %, being a medium degree of spatial autocorrelation with the sequence of Ⅱ 〈 Ⅰ 〈 Ⅲ, while the spatial variance resulted from the random parts ranked in adverse.