基于黄花菜产量与品质特征的优异种质筛选及评价

Screening and Evaluation of Superior Germplasm Based on Yield and Quality Characteristics of Hemerocallis citrina

摘要:

为筛选黄花菜(Hemerocallis citrina)优异种质资源，采用相关分析、逐步回归、通径分析和基于逼近理想解排序(TOPSIS)分析等方法，对20份菜用黄花菜种质的产量形成与品质进行了综合评价。结果表明，不同黄花菜种质的农艺、产量和品质性状变异丰富，变异系数为1.9%~75.6%，其中以花蕾色度a^*值的变异系数最大，总黄酮含量、花蕾质量次之，花蕾含水量最低。花蕾质量、花蕾粗和花薹数与产量均呈显著或极显著相关，相关系数分别为0.737、0.583和0.517。营养品质间的相关性较弱，而外观颜色指标间及其与部分营养品质存在极显著相关性。影响黄花菜产量的主要决定因子为花蕾质量和花薹数,其决策系数分别为0.534和0.239。营养品质和外观颜色的权重分别为0.523和0.477，结合TOPSIS法可以对不同种质的品质进行有效区分。根据综合得分与产量绘制二维散点分布图可以快速筛选到产量与品质均佳的优异资源。

Abstract:

The aim was to screen the superior germplasm resources of Hemerocallis citrina. The yield formation and quality of 20 H. citrina germplasms were evaluated by using correlation analysis, stepwise regression, path analysis and TOPSIS analysis. The results showed that variations in agronomic, yield and quality traits among different H. citrina germplasms were rich with coefficient of variation (CV) from 1.9% to 75.6%. The CV of bud chroma a^* (redness) was the highest (75.6%), followed by total flavone content (67.9%) and bud weight (41.0%), and water content of bud was the lowest (1.9%). Bud weight, bud diameter and flower moss number were significantly correlated with yield with CV of 0.737, 0.583 and 0.517, respectively. The correlation among nutrient quality traits was weak, while the correlation between appearance color index and some nutritional quality was extremely significant. The results of stepwise regression and path analysis suggested that floral bud weight and scape number were the main determinant of yield formation across all investigated germplasms, with decision coefficients of 0.534 and 0.239, respectively. The combination weights of nutritional quality and color parameters, calculated by using the method of coefficient of variation-entropy weight, were 0.523 and 0.477, respectively, and the results could effectively distinguish the best quality rank of different germplasm resources with application of combined the TOPSIS analysis method. Two-dimensional scatter distribution diagrams, a visualized overall assessment result simultaneously considering yield and quality, could be derived for use in the high yield and bud quality quickly screening of H. citrina.