论文部分内容阅读
经过计算机对我国若干年越冬代、一代、三代粘虫发生程度的季节性迭加趋势时间序列分析,这三个代次有15—18年的周期现象,每个周期里中等至严重发生的11—14年,属于偏轻发生的6—9年。一年以上(1—5年)的预测结果经近5年的测试、验证、预测、测试和验证的准确率为77.3%,预测结果准确率86.4%。粘虫有自身的生命周期(卵、幼虫、蛹、成虫)和年发生周期(不同地区的不同代数)。近几十年来众多昆虫学家和植保工作者的大量深入研究,使我们对粘虫的这两类周期已知之较多,但若干年的粘虫发生有无内在规律,有无循而复始的周期现象,我们还知之甚少。由于清楚的了解了粘虫本身的生命周期,对于短期预测(预测防治适期)已较为易举了;又由于搞清了粘虫的年发生周期,揭开了不同地区不同世代的发生规律,加之近十年来发展综合分析、异地预测,利用计算机进行相关分析、回归分析、聚类分析、判别分析等,对粘虫的中长期预测也有了一套较为成熟的方法,使测报的时效性、准确率、预测水平等有了较大的提高,尽管八十年代以来,时有粘虫严重发生;还是得到有效的控制,并未造成大面积成灾和严重损失。然而,要发展粘虫的综合防治和更加有效的生态对策,则不仅需要及时准确的中长期(隔代次,跨地区的)预测预报,而且还需要更早一些的知道较长远未来的粘虫发生趋势,即有必要作出粘虫的超长期(跨年度跨代次的)预测。为此,我们从1986年开始偿试粘虫的超长期预测,并对若干年的粘虫发生规律及其周期现象进行分析研究,现将初步结果奉献给大家。
After several years of computer overwintering generations of China, the first and third generation of armyworm occurrence of seasonal superimposed trend of time series analysis, the three generations have a period of 15-18 years, every cycle of moderate to severe 11 - 14 years, belonging to the light of the 6-9 years. The accuracy of the prediction of more than one year (1-5 years) after testing, verification, prediction, testing and verification in the past 5 years was 77.3% and the prediction accuracy was 86.4%. Mythimna separata have their own life cycle (eggs, larvae, pupae, adults) and the annual cycle (different algebra in different regions). Numerous in-depth studies by many entomologists and plant protection workers in recent decades have led us to know more about these two types of cycles of armyworm, but for some years, The periodic phenomenon, we also know little. Owing to a clear understanding of the life cycle of the armyworm itself, it has been relatively easy to predict the short-term (control of the appropriate stage of control) insects. In addition, by clarifying the annual occurrence period of armyworm and revealing the law of occurrence of different generations in different regions, Coupled with the development of comprehensive analysis of the past 10 years, remote prediction, the use of computer-related analysis, regression analysis, cluster analysis, discriminant analysis, the long-term prediction of armyworms also have a more mature approach to measuring the timeliness, Accuracy rate and forecast level have been greatly improved. Although the armyworm pests occurred seriously since the 1980s, effective control has not resulted in large-scale disaster and serious losses. However, comprehensive development of armyworm and more effective ecological measures will not only require timely and accurate long-term (inter-generational and inter-regional) forecasts and forecasts, but also require earlier understanding of the long-term future armyworms The trend is that it is necessary to make very long-term (cross-annual intergenerational) predictions of armyworm. To this end, we started in 1986 to test the long-term prediction of armyworm, and a few years of armyworm occurrence and cycle phenomena were analyzed and studied, the initial results are now dedicated to everyone.