Abstract: To investigate the identification methods for rice low temperature tolerance among different varieties, taking 8 rice varieties and 3 self-bred rice strains as materials, and relative value of seed germination index and seedling growth index as assessment indicators for rice low temperature tolerance, this paper carried out comprehensive evaluation on low temperature tolerance of 11 rice materials by method of membership function in fuzzy mathematics on the basis of principal component analysis and classified by clustering methodology. The results showed that 5 principal components were screened among 16 identified indexes, among which germination potential, germination rate, germination index, plant height, dry weight of root, superficial area of root and root volume could be as the indexes for rice low temperature tolerance identification. The order of low temperature tolerance of 11 rice varieties was appraised by weighted membership function methods. Clustering analysis results showed that the low temperature tolerance of 11 rice varieties was classified into 3 categories: Class Ⅰ with strongest low temperature tolerance; Class Ⅱ with medium low temperature tolerance; Class Ⅲ with the weakest low temperature tolerance. These results could provide certain basis for identification methods of rice low temperature tolerance and production of 11 rice varieties.

Key words: Oryza sativa L., appraise of low temperature tolerance, principal component analysis, clustering analysis

摘要： 为探究水稻品种间耐低温性鉴定方法，以8个水稻品种和3个自育品系为材料，以种子萌发和幼苗生长指标的相对值作为耐低温性评价指标，在主成分分析的基础上，利用模糊数学中的隶属函数法对11个水稻品种进行耐低温性综合评价，16个鉴定指标中筛选出5个主成分，其中发芽势、发芽率、发芽指数、株高、根干重、根表面积和根体积可作为水稻品种耐低温性的鉴定指标。加权隶属函数法评价出11个水稻材料耐低温性强弱顺序，聚类分析结果将11个水稻材料耐低温性聚为3类，第Ⅰ类耐低温性较强，第Ⅱ类耐低温性中等，第Ⅲ类耐低温性最弱。结果为水稻材料耐低温性评价方法和这11个水稻品种的生产应用提供一定的依据。

关键词: 水稻, 耐低温性评价, 主成分分析, 聚类分析