云南桔小实蝇发生量预测模型研究

2003-2008年从田间系统地监测了云南元江和华宁桔小实蝇的发生规律.将诱虫资料和温度、湿度、雨量、光照等气象资料作为预测因子,采用简单逐步回归和多因子互作回归分别进行拟合,建立了不同发生量预测预报模型,并对其进行检验.结果表明发生量与多因子交互之间存在相关关系,可采用预报模型较为准确地预测出桔小实蝇的发生量,且多因子互作组建的模型比简单逐步回归的结果可靠.

Study on Predication Mathematical Model of Occurrence Quantity for Oriental Fruit Fly, Bactrocera dorsalis in Yunnan Province

Field occurrence regularity of oriental fruit fly (Bactrocera dorsalis) was monitored from 1990 to 1995 in Yuanjiang and Huaning county, Yunnan province. The datum can be regarded as prediction factors on model of occurrence quantity, included trapped pests and weather information such as temperature, hu-midity, precipitation, illumination, etc. The model was established using the method of simple regression and multifactorial stepwise regression respectively. The results showed that the occurrence quantity of ori-ental fruit fly was closely related to multi-factor interaction. The model was tested using the last year da-tum, and the findings indicated that the model by simple regression was credible than by multifactorial stepwise regression. The model could accurately predict the occurrence quantity of oriental fruit fly.