相对病情指数划分棉花品种抗病性的统计学基础

对多年棉花品种的黄萎病抗病性鉴定资料分析表明,同一个品种的黄萎病病情指数年度间差异变化较大。8年的鉴定资料统计结果表明,中棉所12的病情指数年度间变异系数为37.4％,冀棉11为33.3％,方差分析表明它们的病情指数年度间变异极显著。对1992—1997年中棉所12的病情指数进行概率统计,其总体平均数的估计值为62.0,95％置信区间为58.4,65.6];99％置信区间为57.2,66.8]。利用本文方法计算的相对病情指数来表示棉花品种的黄萎病感病性,与利用病情指数表示方法相比,可以将年度间变异降低60％,发生概率p=0.997。本方法划分的抗病性不同于目前通用的以一个年度田间病情指数划分的品种抗病性,而是概率保证下的抗病性,具有深层的数理统计的意义。

Statistical Basis for Determining Verticillium Wilt Resistance of Cotton Cultivar/Line according to Relative Disease Index

Results of statistical analysis on multi- year data of disease index (DI) for Verticillium wilt of cotton caused by Verticillium dahliae showed that it was highly significant (p = 0. 01) different for the same cotton cultivar/line among different years. Over eight years, coefficients of variation (CV) of disease indices for CCRI 12 and Jimian 11 were 37. 4% and 33. 3%, respectively, and the result of F - tests showed that the disease indices were different among years for each cultivar at p = 0. 01. The unbiased estimate of disease index for CCRI 12 for 8-year period was 62. 0, and the 95% and 99% confidence limits were 58. 4,65.6] and 57. 2, 66. 8] , respectively. Relative disease index (RDI) for each line or cultivar was an adjusted figure of DI for a single year, based on the coefficient calculated from the unbiased estimate of DI for CCRI 12. Comparison of the differences of RDI and DI for nine cotton lines in two years concluded that the method of determining resistant cotton culti-vars/lines, according to RDI, increased stability and accuracy by 60. 0% more than according to DI at p = 0. 997. Thus, using RDI based on multi-year data to compared cotton cultivars/ lines resistance to Verticillium wilt is statistically sound.