黑龙江省中东部地区主栽大豆品种产量指标筛选及评价

以黑龙江省中东部地区28个大豆主栽品种为试验材料,采用主成分分析、隶属函数分析和逐步回归分析方法,对该区域大豆产量进行综合评价及鉴定指标筛选。结果表明,通过主成分分析将10个产量相关单项指标转换为4个独立综合指标,性状累计贡献率为79.083%;通过隶属函数计算产量综合评价值（Y）,对参试品种进行产量综合评价,筛选出了产量较高的大豆品种合农71、合农85、绥农26、合农75、黑农61及合农76等;通过逐步回归分析法建立大豆产量评价数学模型,Y=0.122+0.736X₁+0.465X₂+0.168X₃+2.527X₄+0.326X₅+0.289X₆-0.313X₈,平均估计精度为92.16%;筛选了7个核心产量评价性状,分别为主茎节数、单株荚数、单株粒数、每荚粒数、单株粒重、百粒重及4粒荚数,该模型可作为大豆育种材料产量评价及品种产量鉴定的理论依据。

Evaluation and Determination of Yield Evaluation Indicators of Soybean Mainly Cultivated Varieties in the Central and Eastern of Heilongjiang Province

Twenty-eight soybean mainly cultivated varieties in the central and eastern of Heilongjiang province were used as materials, the analytical methods of the principal component analysis, membership function method, and stepwise regression analysis had been used to evaluate the yield and to select the determination of evaluation indicators. The results showed that ten single yield indexes were being transformed to four independent comprehensive components through principal component analysis representing 79.083% cumulative contribution rate. The membership function method was employed to calculate the comprehensive yield value (Y). After the yield comprehensive evaluation, high-yield soybean varieties Henong 71, Henong 85, Suinong 26, Henong 75, Heinong 61 and Henong 76 were screened out. A mathematical evaluation model for soybean yield was established by stepwise regression analysis, Y=0.122+0.736X₁+0.465X₂+0.168X₃+2.527X₄+0.326X₅+0.289X₆-0.313X₈, which mean accuracy was 92.16%. Seven indexes were closely related to the yield, including the nodes on the main stem, pod number per plant, seed number per plant, seed number per pod, seed weight per plant, 100-seed weight, and four-seed pod number. The model can be used as a theoretical basis for yield evaluation of soybean breeding materials and variety yield identification.