摘要
为了能够对异地鉴定的大豆品系进行综合性评价提供依据,本文采用方差分析、灰色关联度、DTOPSIS法和同异分析4种方法对12个自育品种(系)进行综合分析、优劣排序,并对4种分析方法结果进行了比较研究。结果表明:方差分析法与灰色关联度法秩显著相关,相关系数为0.769,达极显著水平,与同异分析(0.650~*)显著相关,与DTOPSIS法分析(0.539)相关不显著;说明4种分析方法都能反映大豆品种的优劣表现且统计趋于一致,但DTOPSIS法分析结果受品种产量的影响较小。在离散程度越大说明分析方法越准确的区分度检验中,DTOPSIS法的区分度最大(0.726 9),方差分析法区分度最小(0.073 2),进一步验证了本试验条件下DTOPSIS法综合评价效果更适合用来综合评价大豆新品种异地鉴定。
In order to provide a basis for the comprehensive evaluation of soybean lines in different places, variance analysis, grey correlation degree, DTOPSIS and identity and similarity degree method were used to make a comprehensive analysis and rank of the advantages and disadvantages of 12 varieties(lines). The results of four analysis methods were compared. The results showed that the rank correlation coefficient of ANOVA and grey correlation method was 0.769, the correlation was extremely significant. It was significantly correlated with the same and different analysis(0.650);it was not significantly correlated with DTOPSIS analysis(0.539), which indicated that the four analysis methods could reflect the performance of soybean varieties and the statistics were consistent. The results of DTOPSIS analysis were less affected by the yield of varieties. The greater the degree of dispersion, the more accurate the method of analysis, the maximum differentiation of DTOPSIS(0.726 9), the minimum differentiation of variance analysis(0.073 2). The comprehensive evaluation effect of DTOPSIS method under the experimental conditions was more suitable for comprehensive evaluation of new soybean varieties in different places.
作者
周长军
陈井生
田中艳
李建英
吴耀坤
于吉东
马兰
李泽宇
ZHOU Chang-jun;CHEN Jing-sheng;TIAN Zhong-yan(Daqing Branch,Heilongjiang Academy of Agricultural Sciences,Daqing,Heilongjiang 163316)
出处
《安徽农业科学》
CAS
2020年第16期39-41,81,共4页
Journal of Anhui Agricultural Sciences
基金
现代农业大豆产业技术体系大庆综合试验站项目(CARS-004-CES07)
黑龙江省农业科学院跨越工程项目“大豆优异品种创制及高产高效栽培配套技术集成”(HNK2019CX01)
黑龙江省百千万工程科技重大专项(2019ZX16B01)。
关键词
灰色关联度
DTOPSIS法
综合性状评价
Grey correlation degree
DTOPSIS method
Comprehensive character evaluation
