Genomic Selection For Yield-relevant Traits And Stalk Strength In Maize

Maize(Zea mays L.),one of the most important crops for food and forage,has a vital role in the development of national economy and ensurance of food security.It has great significance to perform genetic dissection and genomic selection to accelerate breeding process and improve maize production.Genomic selection(GS),a promising strategy for molecular breeding,is widely used by plant breeders to enhance breeding efficiency by shortening breeding cycle.In present study,phenotypic and genotypic data from four experimental populations were used to evaluate the effect of trait-relevant markers(TRMs),nonadditive effects,and multivariate model on prediction accuracy.Consequently,these results can provide pertinent advices for improving GS schemes in practical breeding programs.1.Compared to use stochastically selected markers,the prediction accuracy can reach plateau when 500-1,000 TRMs are used in GS program.The prediction accuracy can be significantly improved by including nonadditive effects and TRMs in GS model when genotypic data with high proportions of heterozygous alleles and complex agronomic traits with lower heritability are used to perform GS.In addition,taking information on population structure into consideration can slightly improve prediction performance when genetic relationship between training and testing sets is affected by population stratification that is attributed to different allele frequencies.2.Two populations consisting of recombinant inbred lines(RILs)were used to dissect the genetic basis of rind penetrometer resistance(RPR)that was evaluated in seven stages.The optimal stages to measure stalk strength are the silking stage and stages after silking.A total of sixty-six and forty-five quantitative trait loci(QTL)were detected in each RIL population,respectively.Several candidate genes were predicted according to the maize gene annotation database and were closely related to the biosynthesis of cell wall components.Moreover,analysis of gene ontology(GO)enrichment and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway further implied that candidate genes relevant to cell wall formation were involved in the RPR determination.In addition,the multivariate model can efficiently improve the prediction accuracy relative to the univariate model and the model considering RPR-related loci as fixed effects.