Abstract:

To study the high temperature and humidity tolerance of different hot pepper varieties at germination stage, screen index for tolerance to high temperature and humidity, and establish a reliable mathematical evaluation model for pepper tolerance to high temperature and humidity, 13 physiological indexes under high temperature and humidity stress were tested in this paper with 10 pepper materials as materials. Taking high temperature and humidity tolerance coefficient of each index, and seed dry weight and electrolyte leakage rate, etc. indexes as measuring gists for high temperature and humidity, using principal components analysis, clustering analysis and stepwise regression, etc. methods, we carried out comprehensive evaluation on hot peppers cold tolerance. The results showed that the 15 single indexes could be converted to 5 independent comprehensive components. The membership function method and cluster analysis were used to divide 10 hot pepper materials into 4 kinds according to their tolerance degrees to high temperature and humidity. The mathematical evaluation model for hot pepper high temperature and humidity tolerance was established by stepwise regression, D=(-141.548-0.071X2+5.580X7+4.195X13-1.363X14-3.432X15)×10-3（R2=0.999 36). And the evaluation accuracy of the equation was estimated over 97.39%. Five identification indexes related to high temperature and humidity tolerance at germination stage were screened. They were electrolyte leakage rate, dry weight of seedling, mean germination days, germination value and average daily germination rate. These 5 indexes could be used under the same stress condition for rapid identification and prediction of hot pepper varieties at germination stage. These results will lay a solid foundation for large scale evaluation on pepper germplasm tolerance to high temperature and humidity.

Key words: hot pepper, germination stage, high temperature and humidity tolerance, principal component analysis, comprehensive evaluation, stepwise regression

摘要：

为研究辣椒不同品种芽期耐高温多湿特性，筛选耐高温多湿指标，建立可靠的辣椒耐高温多湿数学评价模型，以10个辣椒材料为试验材料，对其在高温多湿胁迫等13个生理指标进行测定，以各单项指标的耐高温多湿系数和种子干重、种子浸出液电导率等指标作为衡量耐高温多湿性的依据，运用主成分分析、聚类分析和逐步回归等方法对其耐寒性进行综合评价。通过主成分分析将15个单项指标转换为5个相互独立的综合指标；通过隶属函数法和聚类分析将10个辣椒材料按耐高温多湿性强弱划分为4类；通过逐步回归建立辣椒芽期耐高温多湿性评价数学模型，D=(-141.548-0.071X2+5.580X7+4.195X13-1.363X14-3.432X15)×10-3 （R2=0.999 36），估计精度大于97.39%，并筛选到5个耐高温多湿性芽期鉴定指标，分别是电解质外渗率、苗干重、平均发芽天数、发芽值和日平均发芽率，在相同逆境下可用这5个指标进行芽期辣椒材料快速鉴定和预测。该结果为辣椒种质资源大规模耐高温多湿性评价奠定了基础。

关键词: 辣椒, 芽期, 耐高温多湿性, 主成分分析, 逐步回归;综合评价