Short-term And Medium-term Predicting For Occurrence Of Cnaphalocrocis Medinalis In Rice Producing Areas Of Southern China

As a thermophily, hygrophilous and migratory pest, rice leaf roller, Chaphalocrocis medinalis Guenee, is the second major pest of rice production in China and the meteorological conditions in the early stages play an important role on its immigration and endangering in the later stages. The medium and short-term forecasting of C. Medinalis has been a difficult problem in the predicting and management of rice pests in the world. In order to solve this problem and provide a decision-making support for its effective prevention and controlling, in this paper, for short term prediction, the pentad systematic investigation data of C. Medinalis at the four representative plant protection stations of four main rice -growing regions (including the rice-growing region of the south China, the rice-growing region of the southwestern China, the rice-growing region between the Nanling Mountains and the Yantze River Valley and the rice-growing region between the Yantze River Valley and the Huaihe River Valley) in China during the period from 1994 to 2014 was collected, the key meteorological factors influencing on C. Medinalis’occurrence amount and occurrence grades were screened out and Kalman Filter Algorithm, a non-linear statistical predicting method, and stepwise regression, a linear statistical method, were used to establish the short-term forecasting models of C. Medinalis’pentad occurrence amount and occurrence grades at the four plant protection stations such as Quazhou in the Guangxi Zhuang Autonomous Region, Xiushan in Chongqing City, Xiangyin in Hunan Province and Zhangjiagang in Jiangsu Province in the immigration and damage period of C. Medinalis respectively. Based on the back substitution fittings and forecasting tests of these models, the difference between the errors, stabilities and accuracy rates,of the two models were compared.For medium and long term forecast,the the correlation coefficents between the C. Medinalis’monthly occurrence amount and 74 atmospheric circulation indices were calculated, The indices associated with the highest correlation coeffcient were selected as the most influencing atmospheric circulation characteristic variables,then SCSA were applied to analysis the period of the coupling oscillation. And combined with the auto regression model,the medium and long term prediction model of C. Medinalis’ pentad occurrence amount were established.The results showed as follows:(1) For the four stations, the occurrence amounts and grades of C. medinalis in the present pentad had the significant positive correlations (P<0.01) with the C. medinalis’s moth amounts of the preceding pentad and the preceding two pentads in the field respectively. There were the significant positive correlations (P<0.01) between the occurrence amounts and grades of C. medinalis in the present pentad and the mean minimun air temperature, mean air temperature and mean maximun air temperature of in the preceding pentad, but the remarkable negative correlations existed between the pentad occurrence amounts and grades and the surface atmospheric pressure in the preceding pentad.(2) The occurrence amounts of Cmedinalis in the present month had the significant positive correlations(P<0.01) with atmospheric circulation characteristics. There were the significant positive correlations (P<0.01) between the occurrence amounts of C. medinalis in the present month and the northern boundary of North Africa subtropical high(20W-60E) (at quanzhou and xiangyin station), the northern Hemisphere polar vortex center strength(JQ)(at xiushan station),the North American Subtropical High Ridge(110W-60W)(at zhangjiagang station),the correlation coefficient of which are 0.7977(P<0.01),0.7720(P<0.01),0.7753(P<0.01) and 0.6389(P<0.01).(3) For the four representative plant protection stations of four main rice-growing regions, the key meteorological factors influencing on C. Medinalis’occurrence amount and occurrence grades were screened out and Kalman Filter Algorithm model and stepwise regression model were established.and the result showed that the prediction errors and of the Kalman model is less than the stepwise regression model,and the Kalman model is better than the stepwise regression model in stability and forecasting accuracy. The comprehensive mean accuracies (MA) of the Kalman model in the occurrence amounts and occurrence grades are 82.58 and 93.90%, therefore, the Kalman model were proved to be available for the short-term forecast of C. Medinalis’pentadant occurrence amount and occurrence grades.(4) According to the atmospheric circulation factors that had been researched and reported to have influence on Cmedinalis immigration and occurrence, then using singular cross spectrum analysis method(SCSA) to analysis the coupling cycle of Cnaphalocrocis Medinalis and the significant related atmospheric circulation factors, with the self regression method,the Cmedinalis’?, medium and long-term forecasting model were established. The prediction model of rice leaf roller in four representative stations can make good recursive prediction and the trend forecast was basically right. Therefore,the application of SCSA and AR model were available for the medium and long-term forecast of C. Medinalis’pentadant occurrence amount.