按Griffing双列杂交方法Ⅳ，对9个热带和温带玉米群体进行组配，获得36个杂交组合。2002年和2003年，采用混合线性（AD）模型和最小范数二阶无偏估算法（1）法（minimum norm quadratic unbiased estimation, MINQUE），分别在河南安阳和湖北十堰进行田间鉴定，对产量和产量构成因素的遗传方差分量、杂交组合的基因型值进行了分析。结果显示，产量和产量构成因素的遗传方差分量及其与环境的互作效应大多达到极显著水平，但各项遗传方差分量对总表型变异的贡献有差异。对小区产量而言，各效应的贡献是显性>显性与环境互作>加性与环境互作>加性；而加性方差分量对总表型变异的贡献为穗行数>穗粒数>百粒重>小区产量。对这36个组合的群体平均优势和群体超亲优势进行了估计，并在此基础上引入群体间遗传差异的度量参数ω/(2μ)，ω/(2μ) = 群体平均优势－群体超亲优势。对9个群体的比较表明，遗传差异与杂种优势之间不存在线性关系，亲本的遗传差异过大和过小都不利于产生强优势组合，具有中等遗传差异的亲本群体表现出了强的杂种优势。如组合3´6 (BSSS C9´Stay Green C4)、2´3 (BS16´BSSS C9)和1´3 (Suwan1´BSSS C9)都具有中等的遗传差别，有较好的杂种优势表现；BSSS C9与Stay Green C4和Suwan1、BS16和Suwan1在温带育种中值得关注，有可能形成新的杂种优势模式。
Nowadays the most common methods for evaluating heterosis population and its genetic pattern in maize are based on analyses of parentage, SCA and molecular markers, but each method has its limitation. In the present study, a mixed linear (AD) model and the Minimum Norm Quadratic Unbiased Estimation (MINQUE) (1) method were introduced to estimate the variance components of yield and the values of F1 heterosis, using 2 tropical and 7 temperate maize (Zea mays L.) populations. The 9 populations were crossed by Griffing Ⅳ design, and 36 combinations were made. All hybrids were planted by random block with 4 replications in Anyang of Henan Province and Shiyan of Hubei Province in 2002 and 2003, respectively. The number of rows per ear (RPE), number of kernels per row (KPR), hundred-kernel weight (HKW) and plot-kernel weight (PKW) were measured with 24 plants in the middle of each plot. A parameter for genetic differences measurement between parent populations, i.e. ω/(2μ) = heterosis over mid-parent value—heterosis over better-parent value, was introduced. The analysis of AD model showed that additive (VA) and dominative (VD) effects of yield and its components as well as their interactions with environment (VAE and VDE) reached highly significant level (P<0.01) (Table 2). For PKW, the ratios of variance components to total variances were sequenced as dominance > dominance ´ environment > additive ´ environment > additive, while the ratios of additive effect to total variances were ranked as RPE > KPR > HKW > PKW (Table 2). The estimation of genetic differences (GD) among the 9 populations was ranged from 0.02% to 21.61% (Table 4), and there was no linear relationship between the GD and heterosis. Three crosses, 3´6 (BSSS C9´Stay Green C4), 2´3 (BS16´BSSS C9) and 1´3 (Suwan1´ BSSS C9), with parents of medium GD in this study showed strong heterosis and great potential in maize breeding in temperate zones. This study provided valuable information for developing new heterosis pattern between tropical and temperate maize germplasm.
全 文 ：