Genetic Basics of Seed Traits in Soybean with Bayes Hierarchical Generalized Linear Model Method
闫宁,谢尚潜,耿青春,徐宇,李广军,刘兵,汪霞,李其刚,章元明
DOI: https://doi.org/10.3724/sp.j.1006.2013.00258
2013-01-01
ACTA AGRONOMICA SINICA
Abstract:Seed size and shape traits in soybean play a crucial role in yield and appearance quality. In this study, an experiment was performed to detect main-effect quantitative trait loci (MQTL), QTL-by-environment (QE), QTL-by-cytoplasm (QC), and QTL-by-QTL (QQ) interactions for the soybean seed traits (length, width, thickness, length-to-width, length-to-thickness, width-to-thickness, and 100-seed weight) using Bayes hiearchical generalized linear model approach. Evaluation of these traits for the 504 F2:3–F2:7 families from the direct and reciprocal crosses of Lishuizhongzihuangdou × Nannong 493-1 was carried out in 2007–2011, respectively, and the 504 F2 plants were scanned by 152 SSR markers. As a result, a total of 89 MQTL, 35 QQ inter-actions, 33 QE interactions and 20 QC interactions were detected. As for the above seven traits, there were 7, 10, 10, 19, 19, 17, and 7 MQTL;1, 10, 6, 0, 6, 9, and 3 QQ interactions;5, 7, 6, 3, 6, 2, and 4 QE interactions;and 2, 1, 3, 8, 4, 2, and 0 QC interac-tions respectively. The total proportion of phenotypic variance explained by the above four types of QTL for each trait is 12.42–61.79%, 0–23.21%, 0.35–1.51%, and 0–14.16%, respectively, indicating that the most important genetic component is MQTL, the second one is epistasis, and the last one is QE interaction. Pleiotropic effects were observed in all kinds of QTL, while various types of QTL shared with one same locus were found to be response for a seed trait as well. These results revealed genetic basis of seed size and shape traits in soybean, and provide reference information for marker assisted breeding.