Understanding the genetic and environmental basis of genotype × environment interaction (G×E) is of fundamental importance in plant breeding. If we consider G×E in the context of genotype × year interactions (G×Y), predicting which lines will have stable and superior performance across years is an important challenge for breeders. A better understanding of the factors that contribute to the overall grain yield and quality of rice ( Oryza sativa L.) will lay the foundation for developing new breeding and selection strategies for combining high quality, with high yield. In this study, we used molecular marker data and environmental covariates (EC) simultaneously to predict rice yield, milling quality traits and plant height in untested environments (years), using both reaction norm models and partial least squares (PLS), in two rice breeding populations ( indica and tropical japonica ). We also sought to explain G×E by differential quantitative trait loci (QTL) expression in relation to EC. Our results showed that PLS models trained with both molecular markers and EC gave better prediction accuracies than reaction norm models when predicting future years. We also detected milling quality QTL that showed a differential expression conditional on humidity and solar radiation, providing insight for the main environmental factors affecting milling quality in subtropical and temperate rice growing areas.

Genotype × environment interaction (G × E) is the differential response of genotypes in different environments and represents a major challenge for breeders. Genotype × year‐interaction (G × Y) is a relevant component of G × E, and accounting for it is an important strategy for identifying lines with stable and superior performance across years. In this study, we compared the prediction accuracy of modeling G × Y using covariance structures that differ in their ability to accommodate correlation among environments. We present the use of these approaches in two different rice (Oryza sativa L.) breeding populations (indica and tropical japonica) for predicting grain yield, plant height, and three milling quality traits—milling yield, head rice percentage, and grain chalkiness—under different cross‐validation (CV) scenarios. We also compared model performance in the context of global predictions (i.e., predictions across years). Most of the benefits of multienvironment models come from modeling genetic correlations between environments when predicting performance of lines that have been tested in some environments but not others (CV2). For predicting the performance of newly developed lines (CV1), modeling between environment correlations has no effect compared with considering environments independently. Response to selection of multienvironment models when modeling covariance structures that accommodate covariances between environments was always beneficial when predicting the performance of lines across years. We also show that, for some traits, high prediction accuracies can be obtained in untested years, which is important for resource allocation in small breeding programs.

Core Ideas Genome‐wide association study (GWAS) for rice quality was performed in two breeding populations. Twenty‐two putative quantitative trait loci (QTL) were associated to rice quality. A genomic region on chromosome 6 was associated with all quality traits in the tropical japonica population. Markers for favorable haplotypes are ready for immediate use for selection. Rice (Oryza sativa L.) is one of the most important staple food crops in the world; however, there has recently been a shift in consumer demand for higher grain quality. Therefore, understanding the genetic architecture of grain quality has become a key objective of rice breeding programs. Genome‐wide association studies (GWAS) using large diversity panels have successfully identified genomic regions associated with complex traits in diverse crop species. Our main objective was to identify genomic regions associated with grain quality and to identify and characterize favorable haplotypes for selection. We used two locally adapted rice breeding populations and historical phenotypic data for three rice quality traits: yield after milling, percentage of head rice recovery, and percentage of chalky grain. We detected 22 putative quantitative trait loci (QTL) in the same genomic regions as starch synthesis, starch metabolism, and cell wall synthesis‐related genes are found. Additionally, we found a genomic region on chromosome 6 in the tropical japonica population that was associated with all quality traits and we identified favorable haplotypes. Furthermore, this region is linked to the OsBEI gene that codes for a starch branching enzyme I, which is implicated in starch granule formation. In tropical japonica, we also found two putative QTL linked to OsBEII, OsDEP1, and OsDEP2. Our study provides an insight into the genetic basis of rice grain chalkiness, yield after milling, and head rice, identifying favorable haplotypes and molecular markers for selection in breeding programs.

Stem rot and aggregated sheath spot are the two major stem and sheath diseases affecting rice (Oryza sativa L.) in temperate areas. A third fungal disease, sheath blight, is a major disease in tropical areas. Resistance to these diseases is a key objective in rice breeding programs but phenotyping is challenged by the confounding effects of phenological and morphological traits such as flowering time (FT) and plant height (PH). This study sought to identify quantitative trait loci (QTL) for resistance to these three diseases after removing the confounding effects of FT and PH. Two populations of advanced breeding germplasm, one with 316 tropical japonica and the other with 325 indica genotypes, were evaluated in field and greenhouse trials for resistance to the diseases. Phenotypic means for field and greenhouse disease resistance, adjusted by FT and PH, were analyzed for associations with 29,000 single nucleotide polymorphisms (SNPs) in tropical japonica and 50,000 SNPs in indica. A total of 29 QTL were found for resistance that were not associated with FT or PH. Multilocus models with selected resistanceassociated SNPs were fitted for each disease to estimate their effects on the other diseases. A QTL on chromosome 9 accounted for more than 15% of the phenotypic variance for the three diseases. When resistance-associated SNPs at this locus from both the tropical japonica and indica populations were incorporated into the model, resistance was improved for all three diseases with little impact on FT and PH.

The possibility of introducing metabolic/biochemical phenotyping to complement genomics-based predictions in breeding pipelines has been considered for years. Here we examine to what extent and under what environmental conditions metabolic/biochemical traits can effectively contribute to understanding and predicting plant performance. In this study, multivariable statistical models based on flag leaf central metabolism and oxidative stress status were used to predict grain yield performance for 271 indica rice (Oryza sativa) accessions grown in the field under well-watered and reproductive stage drought conditions. The resulting models displayed significantly higher predictability than multivariable models based on genomic data for the prediction of grain yield under drought (Q2 = 0.54-0.56 versus 0.35) and for stress-induced grain yield loss (Q2 = 0.59-0.64 versus 0.03-0.06). Models based on the combined datasets showed predictabilities similar to metabolic/biochemical-based models alone. In contrast to genetic markers, models with enzyme activities and metabolite values also quantitatively integrated the effect of physiological differences such as plant height on grain yield. The models highlighted antioxidant enzymes of the ascorbate-glutathione cycle and a lipid oxidation stress marker as important predictors of rice grain yield stability under drought at the reproductive stage, and these stress-related variables were more predictive than leaf central metabolites. These findings provide evidence that metabolic/biochemical traits can integrate dynamic cellular and physiological responses to the environment and can help bridge the gap between the genome and the phenome of crops as predictors of grain yield performance under drought.