| 摘要: |
| 【目的】分析陇东黄土高原不同高粱新品种间的农艺性状及其差异，为高粱新品种选育评价、引种推广和产业可持续发展提供参考。【方法】以国家谷子高粱产业技术体系岗站科学家培育的30个高粱新品种为材料，于2021-2022年连续2年在甘肃省平凉市农业科学院高平试验基地进行品种生态适应性评价试验，在生长期间记载生育期，并在成熟期测定了植株农艺性状和产量及其构成因素，基于多样性分析、相关性分析、主成分分析、聚类分析和隶属函数分析方法进行了综合性评价。【结果】30个高粱品种的7个农艺性状指标中株高和茎粗的变异系数较大，分别为23.66%和21.36%。相关性分析发现，高粱各项农艺性状指标间表现出不同程度的相关性，其中株高与茎粗、穗长呈极显著正相关，与穗粒质量、生育期呈显著正相关；穗粒质量与穗长、千粒质量、产量均呈显著正相关。基于主成分分析确定生育期、株高、穗长和千粒质量可作为评价高粱产量性状的代表性指标，其4个主成分因子的方差贡献率分别为32.06%，26.59%，16.98%和11.56%，累计方差贡献率达87.19%；综合得分结果表明，辽粘3、晋杂34、晋中4554、晋杂22和瑞杂8综合适应性较好。通过聚类分析将供试高粱品种划分为3大类群，第Ⅰ类群包含5个品种，主要表现为高秆、产量低和生育期较长；第Ⅱ类群包含9个品种，主要表现为矮秆、生育期较短和产量较低；第Ⅲ类群包含16个品种，主要表现为株高中等、生育期适中和产量高。隶属函数分析结果表明，辽粘3、晋杂34、晋中4554、晋杂22和平杂14表现较好。主成分分析与隶属函数法评价结果基本一致。【结论】高粱品种辽粘3、晋杂34、晋中4554、晋杂22的综合农艺性状优于其他品种，适宜在陇东黄土高原旱作雨养区推广种植。 |
| 关键词: 高粱 品种综合评价 主成分分析 聚类分析 隶属函数 陇东黄土高原 |
| DOI：10.13207/j.cnki.jnwafu.2025.05.005 |
| 分类号: |
| 基金项目:财政部和农业农村部国家现代农业产业技术体系项目(CARS-06-14.5-B29)；甘肃省青年科技基金项目（23JRRL0005） |
|
| Comprehensive evaluation and screening of adaptability of different sorghum varieties |
|
FU Jiangpeng, LIU Facai, YAN Baoqin, ZHAO Zhihui, LI Lili, WANG Yongdong, WEI Wei, ZHOU Yingxia
|
|
Pingliang Academy of Agricultural Sciences,Pingliang,Gansu 744000,China
|
| Abstract: |
| 【Objective】The agronomic traits and differences of different sorghum varieties in the Loess Plateau of Eastern Gansu were analyzed to provide reference for breeding evaluation,introduction and promotion of new sorghum varieties and sustainable industrial development.【Method】With 30 new sorghum varieties cultivated by scientists of the National Millet Sorghum Industrial Technical System Station as materials,the ecological adaptability evaluation test of varieties was conducted for two consecutive years from 2021 to 2022 in Gaoping Test Base of Academy of Agricultural Sciences,Pingliang City,Gansu Province.The growth period was recorded during growth,and the agronomic traits,yield and component factors of plants were measured during the maturity period.A comprehensive evaluation was carried out based on diversity analysis,correlation analysis,principal component analysis,cluster analysis and membership function analysis.【Result】Among the 7 agronomic traits of 30 sorghum varieties,the variation coefficients of plant height and stem diameter were 23.66% and 21.36%,respectively,which were higher than the other traits.Correlation analysis indicated that various agronomic indexes of sorghum showed different degrees of correlation,among which plant height was highly positively correlated with stem diameter and ear length,and was positively correlated with ear grain quality and growth period.Panicle quality was positively correlated with panicle length,thousand grain quality and yield.Based on principal component analysis,it was determined that growth period, plant height,ear length and thousand grain weight could be used as representative indicators to evaluate the yield traits of sorghum.The variance contribution rates of the four principal component factors were 32.06%,26.59%,16.98% and 11.56%,respectively,with a cumulative contribution rate of 87.19%.According to the comprehensive score,Liaonian 3,Jinza 34,Jinzhong 4554,Jinza 22 and Ruiza 8 showed good comprehensive adaptability.The tested varieties were divided into 3 groups by cluster analysis.Group Ⅰ contained 5 varieties,which were characterized by high stem,low yield and long growth period.Group Ⅱ consisted of 9 varieties,which were mainly characterized by low stalk,short growth period and low yield.Group Ⅲ consists of 16 varieties,mainly characterized by medium plant height,moderate growth period and high yield.The results of subordination function analysis showed that Liaonian 3,Jinza 34,Jinzhong 4554,Jinza 22 and Pingza 14 performed better.The evaluation results of principal component analysis and membership function method were basically consistent.【Conclusion】Liaonian 3,Jinza 34,Jinzhong 4554,Jinza 22 have better comprehensive agronomic traits than other varieties,and are suitable for spreading planting in the rainfed area of the Loess Plateau of Eastern Gansu. |
| Key words: sorghum variety comprehensive evaluation principal component analysis clustering analysis membership function Loess Plateau of Eastern Gansu |
