我国甜菜夜蛾大尺度暴发频度与广域温度和广域降雨量关系的预测模型

甜菜夜蛾Spodoptera exigua (Hübner)是我国多种农作物上的重要害虫, 在我国许多地区频繁暴发成灾。为探索甜菜夜蛾种群动态规律并建立种群数量发生趋势预测模型, 作者应用时间序列分析和逐步回归分析方法研究了我国广域（较大范围）温度和广域降雨量变化趋势对我国广域甜菜夜蛾年暴发频度的影响规律。结果表明: 甜菜夜蛾发生的长期趋势和年间波动状况均与广域温度和广域降雨量具有复杂的影响关系。在1979-2008年间, 我国甜菜夜蛾暴发频度呈现出波浪式上升趋势, 其暴发指数平均年递增率为0.076, 而我国广域温度（以27个省市级气象台数据统计为例）在1990-2008年间的平均年递升率为0.039℃, 即我国甜菜夜蛾暴发频度上升趋势与我国广域温度升高趋势同向而行。作者从52个因素（当年和上年1-12月各月及全年日均温和月均降雨量）中筛选出了具有显著回归影响（P<0.05或0.01）的10个因素进入回归模型, 初步找出了能够预测广域甜菜夜蛾暴发趋势指数的温度与降雨量或其组合因素, 并使其模型达到99%以上的历史符合率和预测准确度。作者认为, 广域温、 雨因素与广域甜菜夜蛾暴发趋势指数的这种密切相关性, 不是偶然的巧合, 而是必然的环境（温度和降雨量）作用于生物（甜菜夜蛾）的因果关系。

Modelling of the relationship between the frequency of large-scale outbreak of the beet armyworm,Spodoptera exigua(Lepidoptera:Noctuidae)and the wide-area temperature and rainfall trends in China

The beet armyworm, Spodoptera exigua (Hübner), is a major agricultural pest in China, which frequently breaks out in many parts of China in the last 20 years. To explore the rule of the population dynamics and to establish the forecasting mode of population trends of S. exigua in quantity, the time series analysis and stepwise regression analysis methods were used for studying the law of wide-area (greater range) atmospheric temperature and rainfall influence on outbreak frequency of S. exigua in China. The results indicated that the wide-area temperature and rainfall had a complex effect on the long-term trends and fluctuations of S. exigua outbreak frequency. During 1979-2008, the frequency of the beet armyworm outbreaks in China showed an upward trend wave, and its average annual growth rate of the outbreak index was 0.076, while the wide area temperature of China (taking the 27 city-level observation points as a statistics example) in the years 1990-2008 increased at the average annual incremental rate of 0.039℃, showing a rising frequency of the beet armyworm outbreak with wide-area temperature rising trend. Ten factors, which had a significant effect (P<0.05 or 0.01) on the regression forecasting model and could be used to forecast quantitatively a wide-area outbreak trend index of S. exigua (the forecast accuracy of the simulation model was more than 99%), were screened out from 52 factors (the temperature and rainfall of the monthly and annual average of January to December of the current year and last year). The author believes that this close relationship between the wide-area temperature, rainfall factors and the beet armyworm outbreak trend index is not a mere coincidence, but a necessary consequence of environments (temperature and rainfall) acting on creatures (beet armyworm).