棉花枯萎病预报方法研究与应用

摘要

Since 1987, the occurred probability of fusarium wilt of cotton was more than 90 percent in Pengze, Jiangxi, and it oeeured at every stage from the seedling stage to the boll period, annual variety of the grades had a 4a main period, but the oscillation was weaker, the ＂subsequent ecological effect＂ was not remarkable, the occurrence grades had a close relation- ship with meteorological factors such as illumination, temperature, water, and so on. A long- term prediction model which was established by ＂teleeonneetion patterns＂ to forecast the annual occurrence level of fusarium wilt of cotton and mid - short term prediction model which was established by stepwise regression method to predict the grades of meteorological condition about fusarium wilt of cotton both displayed better accuracy and prospect during both test and application.%1987年以来江西省彭泽县棉花枯萎病中等以上发生的概率在90%以上,发生时段包括育苗期至结铃期的各个阶段,发生等级年际变化存在着振荡强度较弱的4a主周期,＂年际生态后续效应＂不够显著,发生等级轻重与光、温、水等气象要素密切相关。采用＂遥相关＂方法建立的棉花枯萎病年发生等级长期预报模式以及采用逐步回归方法建立的棉花枯萎病发生的气象条件等级中短期预报模式在试验与应用中表现出较好的前瞻性与准确性。