论文部分内容阅读
根据我院和广东垦区代表性农场1959—1983年橡胶白粉病流行观察材料和历年流行学的研究成果,选出了橡胶越冬落叶量、越冬期的温度、降雨、5％抽芽期、抽叶早期的抽叶速度及该时的温度等20个预报因子,并用电子计算机逐步回归程序,选出了广东植胶区各代表性地区的预测模式。1979—1985年在广东不同植胶区进行了测报验证试验。结果表明:海南南部、中部的中期预测模式平均准确率为92.5～96.3％,东部、西部和湛江中部为85.5～91.7％,准确率都较高。短期流行强度预测,南部、东部、中部的模式较准确,平均准确率88.6～92.2％;西部和湛江中部的模式平均准确率为80～84.4％。施药期预测,南部、东部的准确率为84.3～85.2％,西部为77.7％。上述中、短期预测模式,经过7年试用,方法简单易行,调查次数少,省工省钱工效高,预见性强,可以在生产上应用。
Based on the epidemiological observations of the epidemic of powdery mildew from 1959 to 1983 and the representative farms in Guangdong Reclamation Farm, we selected the wintering deciduous rubber, the temperature during winter, the rainfall, the 5% budding period and the early leaf drawing stage Of the leaf pumping speed and the temperature at the time of 20 predictors, and the use of electronic computer step by step regression program to select the Guangdong rubber plantations in various regions of the prediction model. 1979-1985 in different plasticized areas in Guangdong reported test verification test. The results show that the average accuracy of the medium-term prediction models in southern and central Hainan is 92.5-96.3%, while that in the eastern, western and central Zhanjiang ranges from 85.5% to 91.7%. The prediction of short-term prevalence intensity shows that the patterns in the southern, eastern and central regions are more accurate with an average accuracy of 88.6-92.2%. The average accuracy of the models in the western and central Zhanjiang ranges from 80-84.4%. The application period predicted that the accuracy rate of southern and eastern regions was 84.3-85.2% and that of western regions was 77.7%. The above short-term and medium-term forecasting models, after seven years of probation, have a simple and easy method, a small number of investigations, and save labor and money on work efficiency and foresight. They can be applied in production.