摘要
谷类作物种子胚乳性状的遗传表达,可能受三倍体的胚乳基因型和/或二倍体的母体基因型控制,还可能存在细胞质效应。Foolad等[6]和pooni等[11]提出的分析模型,由于遗传参数的系数间存在线性相关,不能有效地区分上述遗传控制系统。本文的交配设计和相应统计方法,可对一胚乳性状是否存在胚乳基因型、母体基因型或细胞质效应分别作出测验,从而可确定合适遗传模型和估计有关遗传参数。这些方法也可方便地推广于鉴别非谷类作物种子的胚(子叶)性状的遗传控制。
Genetic expression for an endosperm trait in seeds of cereal crops may be controlled by triploid endosperm genotypes or/and diploid maternal genotypes, and additionally possible cytoplasm difference. Analysed models proposed by Foolad et al. (1992) ̄[6] and Pooni et al. (1992) ̄[11] cannot distinguish the maternal effect from theendosperm effect due to the linear dependence among coefficients of genetic parameters in their models and the singular information matrix.A mating design and correspondingly statistical method are presented in this study, in which the genetic effects of endosperm and maternal genotypes as well as cytoplasm difference can beindependently tested. Consequently, the adequately genetic model for an endospermtrait may be found and the appropriately genetic parameters may be estimated. Apractical example about amylose content of rice grain is used for the illustration,of which results indicate that endosperm effect is. the only source of variation for thecrosses under consideration.The proposed method can be easily extended to embryo(cotyledon)traits, such as oil and protein content in soybean grain, cotton-seed andrape-seed. Received September 17, 1993. This project supported by National Natural Science Foundation ofChina
基金
国家自然科学基金
关键词
胚乳性状
遗传控制
禾谷类作物
Endosperm trait, Genetic control,Endosperm genotype, Maternal genotype, Cytoplasm