A mixed model QTL analysis for sugarcane multiple-harvest-location trial data

Abstract

Sugarcane-breeding programs take at least 12 years to develop new commercial cultivars. Molecular markers offer a possibility to study the genetic architecture of quantitative traits in sugarcane, and they may be used in marker-assisted selection to speed up artificial selection. Although the performance of sugarcane progenies in breeding programs are commonly evaluated across a range of locations and harvest years, many of the QTL detection methods ignore two- and three-way interactions between QTL, harvest, and location. In this work, a strategy for QTL detection in multi-harvest-location trial data, based on interval mapping and mixed models, is proposed and applied to map QTL effects on a segregating progeny from a biparental cross of pre-commercial Brazilian cultivars, evaluated at two locations and three consecutive harvest years for cane yield (tonnes per hectare), sugar yield (tonnes per hectare), fiber percent, and sucrose content. In the mixed model, we have included appropriate (co)variance structures for modeling heterogeneity and correlation of genetic effects and non-genetic residual effects. Forty-six QTLs were found: 13 QTLs for cane yield, 14 for sugar yield, 11 for fiber percent, and 8 for sucrose content. In addition, QTL by harvest, QTL by location, and QTL by harvest by location interaction effects were significant for all evaluated traits (30 QTLs showed some interaction, and 16 none). Our results contribute to a better understanding of the genetic architecture of complex traits related to biomass production and sucrose content in sugarcane.Electronic supplementary materialThe online version of this article (doi:10.1007/s00122-011-1748-8) contains supplementary material, which is available to authorized users.

Mixed Modeling of Yield Components and Brown Rust Resistance in Sugarcane Families

“…Specifically, LMMs have the ability to consider variables as random rather than fixed and to use different variance-covariance (VCOV) structures for random effects to investigate the presence of heteroscedasticity and correlations. This approach allows the analysis of unbalanced data (Pastina et al, 2012;Smith et al, 2005) in addition to using more realistic models for residual variation (incomplete blocks and spatial correlation) and assuming sets of effects (e.g., genotypes) as random (Piepho et al, 2008;Smith et al, 2005). The estimation of variance component parameters is obtained preferably by restricted maximum likelihood (REML), and genotype effects may be obtained either by best linear unbiased estimation or best linear unbiased prediction (BLUP), depending on whether genotypes are considered fixed or random factors, respectively (Piepho et al, 2008;Smith et al, 2005).…”

“…This approach also means a great change in the analysis of breeding experiments because genotype observations may be grouped by levels of grouping factors generated from the experimental design, such as the harvest year and location (Pastina et al, 2012). The application of a mixed model approach is becoming increasingly popular in plant breeding, particularly in research involving the prediction of breeding values combined with genomic data (Beaulieu et al, 2014;Bevan and Uauy, 2013;Burgueño et al, 2012;Crossa et al, 2013;Muir, 2007;Wolc et al, 2011;Zhang et al, 2010).…”

“…The application of a mixed model approach is becoming increasingly popular in plant breeding, particularly in research involving the prediction of breeding values combined with genomic data (Beaulieu et al, 2014;Bevan and Uauy, 2013;Burgueño et al, 2012;Crossa et al, 2013;Muir, 2007;Wolc et al, 2011;Zhang et al, 2010). For sugarcane, the use of linear mixed models to map quantitative trait loci (Pastina et al, 2012) and genomic selection (Gouy et al, 2013) represents progress in crop improvement. However, the basis of genetic inheritance of all traits of economic interest and the genetic expression of related genes in cultivars that are used as parents by breeding programs should be better understood so that we can efficiently achieve higher gains with the selection process and apply genomic selection in sugarcane in the future.…”

“…N(0,G), where P is the number of locationharvest combinations and Ä represents the Kronecker direct product of both the genetic G P and identity I Ig matrices with the respective dimensions of P ´ P and I g ´ I g . Several structures for the G P matrix (Table 1) were examined and compared via Akaike (AIC; Akaike, 1974) and Bayesian (BIC; Schwarz, 1978) information criteria (Pastina et al, 2012). Abbreviations of each structure presented in Table 1 will be used hereafter in the text.…”

“…The mixed models approach is suitable for evaluating the heterogeneity of genetic variances and correlations across environments (Malosetti et al, 2013). Therefore, the use of a more sophisticated statistical model would permit data processing that is more appropriate for the experiment and that signifies a major change in the analysis (i.e., the genotype grouping factors, such as the harvest year and location, would be considered) (Pastina et al, 2012). Although the adjustments for VCOV structures were comprehensively studied for the data in the present study, it is important to note that the use of more locations and harvest years would probably permit the adjustment of other variance and covariance structures.…”