Molecular Marker Information in the Analysis of Multi‐Environment Trials Helps Differentiate Superior Genotypes from Promising Parents

Abstract

P lant breeding programs aim to develop new crop varieties with improved characteristics over existing ones available to farmers. Appropriate methods of selection are crucial to successfully identify the best possible parents for crossing, and to choose the best performing genotypes to progress in the program. Both objectives generally involve evaluating the genotypes under consideration in series of designed experiments across several locations and years.Multi-environment trials (MET) are series of field experiments where genotypes grow under different conditions (environments) determined by combinations of year, location and sometimes management factors (Smith et al., 2001). The statistical analysis of MET data aims to provide reliable and accurate predictions of genotype performance across the target environments and information on specific performance from the interaction of genotypes with the environments (G ´ E) (Smith et al., 2001). Molecular Marker Information in the Analysis of Multi-Environment Trials Helps DifferentiateSuperior Genotypes from Promising Parents M. Gabriela Borgognone,* David G. Butler, Francis C. Ogbonnaya, and M. Fernanda Dreccer ABSTRACTThe statistical analysis of multi-environment trial data aims to provide reliable and accurate predictions of genotype performance across the target environments and information on specific performance from the interaction of genotypes with the environments. Genetic gain can be achieved faster when selections are based on predictions from a model that accounts for the relationships among genotypes rather than from a model that assumes unrelated genotypes. Yield and plant height data from 37 international wheat trials were analyzed using a linear mixed model that accounted for relationships among the genotypes via a genomic relationship matrix G derived from 2487 polymorphic DArT molecular markers for 197 genotypes. The elements of this matrix reflect the actual proportion of the parts of the genome surveyed that is identical by state between pairs of individuals, and including it into the model resulted in generally lower average prediction error variances of individual trials in the analyses. Partitioning the total genetic effects into additive and residual non-additive genetic effects has familiar interpretations for plant breeders and facilitates exploring genotype by environment interactions for additive and total effects. This interpretation is still possible with the form of G used in thispaper. This method of analysis could be readily implemented to accelerate genetic gain by plant breeding programs that have molecular markers for the genotypes under study.

Genetic parameter estimation and selection in advanced breeding population of white Guinea yam

“…The data collected were first subjected to a linear mixed model by residual maximum likelihood (REML) procedure (Patterson and Thompson 1971) to estimate the variance parameters and the empirical Best Linear Unbiased Predictions (E-BLUPs) for random effects using ASReml-R 4 (Butler et al 2018). The univariate Genomic-BLUP model, including the random interaction term between the genomic effect of the i th clone and the j th site, is represented by the model described in Borgognone et al (2016) and Smith and Cullis (2018), which is as follows:…”

“…The vectors of random effects u p , u g and e are assumed pairwise independent with Gaussian distribution, with a mean of zero. The variance matrices for u p , u g and e were as described in Borgognone et al (2016) and Smith and Cullis (2018). The random genotype effect (u g Þ in the model comprised the clone effects nested within sites and hence referred to as clone by site effect.…”

“…The random genotype effect (u g Þ in the model comprised the clone effects nested within sites and hence referred to as clone by site effect. The random genetic effect was then partitioned into additive genetic component associated with the covariance structure proportional to genetic relationships derived from SNP markers and non-additive (or residual) genetic effect explained by individual identity following the approach described in Borgognone et al (2016) and Ovenden et al (2018), which is as follows:…”

Improved methods of predicting genetic merit in plant breeding programs using linear mixed models

“…Burgueno et al (2012) state that MET models can boost predictive power in acrossenvironment prediction and also showed that modeling G×E using information on molecular markers and/or pedigree gives better prediction accuracy than not using molecular markers and pedigree information. Borgognone et al (2016) and Tolhurst et al (2019) recently discussed fitting MET models that incorporated genomic additive relationship matrices. They discuss the superiority of these models over their pedigree counterparts.…”

“…Smith and Cullis (2018) also present a factor analytic selection tool (FAST) which examines measures of overall performance and stability across environments. The FAST method is applicable to MET analyses where the first order FA loadings are positive and represent the majority of the explained variation.Authors such asOakey et al (2016),Borgognone et al (2016) andTolhurst et al (2019) discuss using marker based additive relationship matrices in a mixed model MET analysis incorporating spatial effects and factor analytic variance structures for both the additive effects and residual genetic effects.…”