摘要
本文利用灰色关联分析、逐步判别分析、单一性状分类、主成份分析4种方法探讨了亚洲栽培稻分类时程氏和冈氏所用分类性状的地位和作用。结果表明,将亚洲栽培稻分为籼、粳两亚种时,抽穗时稃色是最重要也是田间较易鉴别的性状,它的关联度、符合系数、共同度、负荷量的累计值均居第一位,其次为叶毛和苯酚反应;若将其分为籼、偏籼、中间、偏粳、粳五类,抽穗时稃色在分类上的重要性仍处于第一位,其次为苯酚反应,再次是稃尖茸毛。两种分类法均认为耐冷性和KCIO3抗性的作用较小。
The importance of 8 traits of Asian rice cultivars which were used as the classificatory criteria by Chen and Oka was evaluated through 4 statistical methods including grey relational analysis, stepwise discriminant method, single traits analysis and principle component analysis. The results showed that the husk colour at heading stage was the most important and easy to be discriminated trait for Asian rice cv. when they were devided into Hsien and Keng. Its incidence degree, coefficient of comformance, common degree and cummulated loading ranked the first,and those of the hair of grain tip and phenol reaction ranked the second. The importance of husk colour at heading stage still ranked the first, phenol reaction the second and finally the hair of grain tip in classification.When rice cv. were divided into Hsien, partial Hsien, medium, partial Keng and Keng types, cold tolerence and resistance to KClO3 played less role in both classifictory methods.
出处
《西南农业学报》
CSCD
1994年第1期1-5,共5页
Southwest China Journal of Agricultural Sciences
基金
四川省科委青年科技基金
关键词
亚洲栽培稻
分类
灰色关联分析
Asian cultivated rice
Grey relational analysis
Stepwise discriminant method
Principle component analysis