亲本选配是水稻杂种优势利用的一个极其重要的环节。人们在长期的育种实践中,在这方面积累了不少宝贵经验,如双亲性状互补、亲缘关系及地理相距较远等,有些已被公认为育种上选配亲本的重要原则。但是,由于构成产量因素的复杂性,仅根据表型进行亲本选配,往往不易做到完全符合这些原则,因此,在强优组合的选配上,过去不得不采用每年配制数以百计乃至数以千计的组合进行鉴定筛选,这种做法不但工作量大,而且往往由于组合多,每组合的料材相对地较少,影响鉴定筛选的准确性。为了寻求科学的预测杂 Parental matching is an extremely important part of rice heterosis utilization. In the long-term practice of breeding, people have accumulated a lot of valuable experiences in this field. For example, their parents are complementary in their traits, their kinship and geography are far apart, and some of them have been widely recognized as the important principle of matching their parents in breeding. However, due to the complexity of the factors that make up the yield, it is not always easy to match these principles based solely on phenotypic parental selection. Therefore, in the selection of high-quality combinations, in the past, hundreds, Thousands of combinations for identification and screening, this approach is not only heavy workload, and often due to the combination of more, each combination of materials is relatively small, affecting the accuracy of identification screening. In order to seek scientific prediction mixed