麦田天敌消长演替规律及超长期预测的研究

The successive regulation and super long-time prediction of natural enemies in wheat fields

通过聚类分析 ,在明确麦田天敌消长演变规律的基础上 ,采用时间序列分析法 ,建立麦田总体天敌和优势天敌七星瓢虫的超长期预测模型 ,经过 2 0 0 1年和 2 0 0 2年实际应用 ,季节水平模型、ARIMA模型的预测效果较好 ,并用这两种模型对未来三年麦田天敌的消长进行了超长期预测。

The successive regulation of natural enemies in wheat fields was demonstrated through clustering. The super long-time models for predicting the natural enemies and the major enemy Coccinella septempunctata were built up on the base of those researches by use of the time-sequence. Those models were put into use in 2001 and 2002. The testing results showed the models of season-level and ARIMA could match the observing values well. The occurrence of natural enemies in next three years was predicted by the equation of season-level and ARIMA.