摘要
为探寻黑龙江省第二积温带水稻育种多目标性状优化方案,利用7个主推水稻品种的22个农艺性状值和NSGA-II遗传算法、熵权综合评价法,对适宜的育种目标性状参数值进行分析。结果表明:黑龙江省第二积温带水稻育种各优化参数值为:食味品质分值90.9~94.5分,较供试品种均值增加8.5~13.0分;产量9 916.3~10 959.8kg/hm^(2),较供试品种均值增加799.0~1 842.5kg/hm^(2)。本研究提出的育种多目标优化设计方法可统筹设计不容易兼顾的育种多目标性状,以便获得更加合理的育种方案。综上,遗传算法和熵权评价法结合是一种可用于不同生态区水稻新品种育种多目标科学量化并高效设计的优化方法。
To investigate the optimum breeding scheme of multi-object characters in the second accumulated temperature zone of Heilongjiang Province,the parameters pf suitable breeding objective traits were analyzed by using22 agronomic character values of seven main rice varieties.NSGA-II genetic algorithm and entropy weight comprehensive evaluation method were adopted in this study.The results showed that Optimization parameter values of multi-object traits in rice breeding were as follows:The taste quality score was 90.9-94.5,which was 8.5-13.0 higher than the average of tested varieties;The yield was 9 916.3-10 959.8,which was 799.0-1 842.5 kg/hm^(2) higher than the average of tested varieties.A more reasonable breeding scheme can be obtained by multi-objective optimization design method proposed in this study which can coordinate the design of breeding multi-objective traits that are not easily balanced.To sum up,the combination of genetic algorithm and entropy weight comprehensive evaluation method was an optimum method that can be used for multi-object scientific quantification and efficient design of new crop varieties such as rice in different ecological areas.
作者
刘宝海
聂守军
高世伟
刘晴
刘宇强
马成
常汇琳
张佳柠
薛英会
白瑞
LIU Baohai;NIE Shoujun;GAO Shiwei;LIU Qing;LIU Yuqiang;MA Cheng;CHANG Huilin;ZHANG Jianing;XUE Yinghui;BAI Rui(Suihua Branch of the Heilongjiang Academy of Agricultural Sciences,Suihua 152052,China;College of Agronomy,Heilongjiang Bayi Agricultural University,Daqing 163319,China)
出处
《中国农业大学学报》
CAS
CSCD
北大核心
2022年第1期38-49,共12页
Journal of China Agricultural University
基金
黑龙江省“百千万”工程生物育种重大科技专项(2020ZX16B01)
黑龙江省农业科学院“农业科技创新跨越工程”专项(HNK2019CX02)
黑龙江省农业科学院科研项目(2019CGJL003)。
关键词
寒地
水稻育种
多目标遗传算法
优化设计
cold region
rice breeding
multi-objective genetic algorithm
optimization design
