[Objective] The aim was to enhance specificity and efficiency of quality breeding of millet. [Method] Difference analysis, correlation analysis, principal compo- nent analysis, and path analysis were conducted on nutr...[Objective] The aim was to enhance specificity and efficiency of quality breeding of millet. [Method] Difference analysis, correlation analysis, principal compo- nent analysis, and path analysis were conducted on nutritional quality characters of 29 millet varieties. [Result] Variance coefficients of nutritional quality characters among different varieties were in descending order of crude protein 〉 crude fiber 〉 fat 〉 water 〉 total carbohydrate. The results of correlation analysis suggested that crude protein exhibited extremely significant negative correlation with total carbohydrate （correlation coefficient of -0.822 4） and significantly negative correlation with fat （correlation coefficient of -0.380 7）; the rest characters showed no significant correla- tion. In addition, according to PCA and path analysis, four characters, including crude protein, fat, total carbohydrate and crude fiber, determined 97.35% of variance of nutritional quality and could be taken as references for selection of breeding ma- terials, with a determination coefficient of 1.000 0. Fat and carbohydrate were typical nutritional quality characters of breeding materials, which should be selected in low generation. However, crude protein and crude fiber should be selected for millet in high generation. [Conclusion] The research provides references for breeding of new millet varieties with high nutritional quality.展开更多
基金Supported by Special Funds of Modern Agricultural Industry Technology System(nycytx-13)~~
文摘[Objective] The aim was to enhance specificity and efficiency of quality breeding of millet. [Method] Difference analysis, correlation analysis, principal compo- nent analysis, and path analysis were conducted on nutritional quality characters of 29 millet varieties. [Result] Variance coefficients of nutritional quality characters among different varieties were in descending order of crude protein 〉 crude fiber 〉 fat 〉 water 〉 total carbohydrate. The results of correlation analysis suggested that crude protein exhibited extremely significant negative correlation with total carbohydrate （correlation coefficient of -0.822 4） and significantly negative correlation with fat （correlation coefficient of -0.380 7）; the rest characters showed no significant correla- tion. In addition, according to PCA and path analysis, four characters, including crude protein, fat, total carbohydrate and crude fiber, determined 97.35% of variance of nutritional quality and could be taken as references for selection of breeding ma- terials, with a determination coefficient of 1.000 0. Fat and carbohydrate were typical nutritional quality characters of breeding materials, which should be selected in low generation. However, crude protein and crude fiber should be selected for millet in high generation. [Conclusion] The research provides references for breeding of new millet varieties with high nutritional quality.
