摘要： 为明确云南不同稻作生态区水稻高产育种性状的选择目标.以粳稻丽江新团黑谷孕穗期耐冷近等基因系(十和田⁴//丽江新团黑谷/十和田)BC₄F₈群体的105个家系为材料,采用方差分析、主成分分析、逐步回归和通径分析,深入探讨云南寒冷粳稻区、冷凉粳稻区和籼粳交错区水稻育种中各农艺性状对单株籽粒产量的调控规律.结果表明,单株穗数、穗颈长、穗下节长、穗长、一次枝梗数、二次枝梗数和穗粒数以基因型控制为主,株高、剑叶长、秕粒数、单株稻草干质量和单株生物产量变异主要是环境差异所致,剑叶宽、倒2叶宽、实粒数、结实率、千粒质量和单株籽粒产量以基因型×环境互作为主.主成分分析表明,影响单株籽粒产量的主成分因子是单株穗数、穗粒数、千粒质量、结实率、剑叶长宽、花药发育、穗抽出度和干物质积累.逐步回归和通径分析表明,寒冷粳稻区决定单株籽粒产量的主导因子依次为千粒质量 >结实率 >一次枝梗数,其共同决定单株籽粒产量83.1%的变异,且均有较大的直接正效应调控作用;冷凉粳稻区为结实率 >1-2节长 >剑叶长 >穗颈长,共同决定单株籽粒产量87.2%的变异,其中结实率和剑叶长对单株籽粒产量有较大的正效应;籼粳交错区为穗粒数 >千粒质量 >单株稻草干质量 >剑叶宽>单株穗数 >花药体积,共同决定单株籽粒产量85.5%的变异,6个性状均对单株籽粒产量有直接的正效应.根据产量结构因子的多重统计分析,在云南寒冷和冷凉粳稻区实现高产的途径是选择重穗与穗粒协调的品种;而在籼粳交错区是选择大穗、大粒及多穗的品种.

关键词: 水稻, 近等基因系, 生态区, 农艺性状, 主成分分析, 通径分析

Abstract: In order to explicit objective breeding traits in various alpine-cold ecological regions of japonica rice in Yunnan Province.A set of 105 near-isogenic lines,BC₄F₈ was developed by backcrossing between Lijiangxintuanheigu(the stongly cold-tolerant japonica landrace,grant No.2)as a donor parent and Towada(cold-sensitive japonica cultivar)as a recurrent parent,and was used as materials,and were grown in Baiyi(alpine-cold Japonica rice region,AJRR),Xundian(cool-warm Japonica rice region,CWJRR)and Yuxi(Indica and Japonica rice interlace region,IJRIR)in Yunnan,respectively.The regulation pattern for 22 growth and grain yield-related traits in three locations was discussed by two-way ANOVA,principle component analysis,stepwise regression analysis and path analysis.The results indicated that No.of panicles per plant(NPP),peduncle length(PDL),node length under spike(NLUS),panicle length(PL),No.of primary rachis branches(NPRB),number of secondary rachis branches(NSRB)and number of spikelets per panicle(NSPP)were mainly attributed by genotypes.The difference of environmental conditions was the main factor affecting plant height(PH),length of flag leaf(LFL),blighted grains per panicle(BGPP),dry straw weight per plant(DSWP)and biomass yield per plant(BYP).Width of flag leaf(WFF),width of the 2nd leaf from the top(WWLF),filled grains per panicle(FGPP),seed setting rate(SSR),thousand-grain weight(TGW)and grains yield per plant(GPP)were mainly influenced by genotypes×environments interaction.The results of principle components analysis showed that the main factors influencing GPP were NPP,NSPP,TGW,SSR,LFL,WFF,anther development,panicle exsertion and dry matter accumulation.The conclusions according to multiple stepwise regression analysis and path analysis were consistent,which showed that there were different dominant factors in various locations.The comprehensive effects of 3 main agronomic traits on GPP ranked as follows:TGW > SSR > NPRB,determination coefficient of three factors was 0.831,and the direct effects of all traits was positive in AJRR.The coefficient of determination reached 0.872,and the magnitude of each factor affecting GPP was SSR,Length between first to second node,LFL,PDL,whereas the biggest direct positive effects on grain weight per plant was SSR and LFL in CWJRR.The coefficient of determination reached 0.855,and the magnitude of each factor affecting GPP was NSPP,TGW,DSWP,WFF,NPP and anther volume,which direct effects of grain weight per plant was positive in IJRIR.The high yield structure was obtained according to multiple statistics in three regions,respectively.The coordinating relationship between grain weight and grain number per panicle should regard as an important criteria in AJRR and CWJRR.Moreover,grain number per panicle,massive-spike and panicle number per plant were selected in IJRIR.

Key words: Rice, Near-isogenic lines, Ecological region, Agronomic trait, Principle component analysis, Path analysis

中图分类号: 

• S511.01