基于部分NCII设计的陆地棉F1表现预测

【目的】探寻预测陆地棉F₁表现的方法,降低育种成本,提高育种效率。【方法】采用60个陆地棉亲本,通过部分NCII交配设计,形成一个由亲本和180个F₁杂交组合组成的部分NCII群体,用亲本加性效应（additive effect,A）、一般配合力（general combining ability,GCA）和表型中亲值（mid-parent value,MP）等3种预测方法对F₁的产量和纤维品质性状进行预测。【结果】陆地棉皮棉产量杂种优势明显。皮棉产量正向中亲优势的组合占97.78%,中亲优势平均达19.63%;正向超亲优势的组合占79.44%,超亲优势平均达8.47%。3种预测方法对F₁表现具有不同的预测效果,其中,以亲本加性效应对F₁的表现预测效果最优,其对皮棉产量、铃数、铃重、衣分、纤维长度、纤维强度和马克隆值等7个目标性状预测精度（pearson相关系数）达到0.738—0.928。性状的加性方差分量和亲本杂交次数对预测效果有影响。加性方差分量越大,所有方法的预测精度都越高;随着每个亲本杂交次数的增加,加性效应预测和GCA预测的精度提高,但表型中亲值预测的效果基本无变化。【结论】陆地棉F₁的表现可以通过利用包含亲本的部分NCII设计群体和亲本的加性效应进行有效预测,采用“大群体、少杂交”的策略可以在保持预测效果的同时降低预测的工作量。

F1 Performance Prediction of Upland Cotton Based on Partial NCII Design

【Objective】Develop the method for predicting F₁ performance of upland cotton, and reduce breeding cost and improve breeding efficiency. 【Method】A partial NCII population composed of 60 parents and 180 F₁ crosses, and three prediction methods with parent additive effect (A), general combining ability (GCA) and mid-parent value (MP) were employed to predict F₁ performance on yield and fiber quality traits. 【Result】The Heterosis of lint yield of upland cotton was obvious. All crosses showed 19.63% average mid-parent heterosis and 8.47% average over-parent heterosis. 97.78% and 79.44% crosses showed positive mid-parent heterosis and positive over-parent heterosis respectively. Three prediction methods showed different prediction effects for F₁ performance. Prediction with additive effect of parents showed the highest prediction accuracy (pearson correlation coefficient, 0.738-0.928) for seven target traits,lint yield, boll numbers, boll weight, lint percentage, fiber length, fiber strength and micronaire value. The additive variance component of target traits and the crossing times for every parent influenced the prediction effect. The higher additive variance component, the higher the prediction accuracy of three methods; with the increase of crossing times of each parent, the prediction accuracy of A and GCA increased, but did not change for MP prediction. 【Conclusion】 The performance of upland cotton F₁ can be effectively predicted by using additive effect of parents based on partial NCII design, and ‘large parent population and less crossing times’ is the preferable strategy to maintain reasonable prediction effect and reduce the workload.