Drought stress is an important abiotic factor limiting common bean yield, with great impact on the production worldwide. Understanding the genetic basis regulating beans’ yield and seed weight (SW) is a fundamental prerequisite for the development of superior cultivars. The main objectives of this work were to conduct genome-wide marker discovery by genotyping a Mesoamerican panel of common bean germplasm, containing cultivated and landrace accessions of broad origin, followed by the identification of genomic regions associated with productivity under two water regimes using different genome-wide association study (GWAS) approaches. A total of 11,870 markers were genotyped for the 339 genotypes, of which 3,213 were SilicoDArT and 8,657 SNPs derived from DArT and CaptureSeq. The estimated linkage disequilibrium extension, corrected for structure and relatedness (r2sv), was 98.63 and 124.18 kb for landraces and breeding lines, respectively. Germplasm was structured into landraces and lines/cultivars. We carried out GWASs for 100-SW and yield in field environments with and without water stress for 3 consecutive years, using single-, segment-, and gene-based models. Higher number of associations at high stringency was identified for the SW trait under irrigation, totaling ∼185 QTLs for both single- and segment-based, whereas gene-based GWASs showed ∼220 genomic regions containing ∼650 genes. For SW under drought, 18 QTLs were identified for single- and segment-based and 35 genes by gene-based GWASs. For yield, under irrigation, 25 associations were identified, whereas under drought the total was 10 using both approaches. In addition to the consistent associations detected across experiments, these GWAS approaches provided important complementary QTL information (∼221 QTLs; 650 genes; r2 from 0.01% to 32%). Several QTLs were mined within or near candidate genes playing significant role in productivity, providing better understanding of the genetic mechanisms underlying these traits and making available molecular tools to be used in marker-assisted breeding. The findings also allowed the identification of genetic material (germplasm) with better yield performance under drought, promising to a common bean breeding program. Finally, the availability of this highly diverse Mesoamerican panel is of great scientific value for the analysis of any relevant traits in common bean.

In recurrent selection programs, progeny testing is done in multienvironment trials, which generates genotype × environment interaction (G × E). Therefore, modeling G × E is essential for genomic prediction in the context of recurrent genomic selection (RGS). Developing single‐step, best linear unbiased prediction‐based reaction norm models (termed RN‐HBLUP) using data from nongenotyped and genotyped progenies, can enhance predictive accuracy. Our objectives were to evaluate: (i) a class of RN‐HBLUP models accommodating combined relationship of pedigree and genomic data, environmental covariates, and their interactions for prediction of phenotypic responses; (ii) the predictive accuracy of these models and the relative importance of main effects and interaction components; and (iii) the influence of different grouping strategies of genetic–environmental data (within selection cycles or across cycles) on prediction accuracy of the merit for untested progenies. The genetic material comprised 667 S1:3 progenies of irrigated rice (Oryza sativa L.) and six check cultivars. These materials were evaluated in yield trials conducted in 10 environments during three selection cycles. Genomic information was derived from single‐nucleotide polymorphism markers genotyped on 174 progenies in the third cycle. We evaluated six predictive models. Environmental covariates and G × E interaction explained a significant portion of the phenotypic variance, increasing accuracy and decreasing the bias of phenotypic prediction. Within‐cycle data were sufficient for accurate prediction of untested progenies, even in untested environments. We concluded that the RN‐HBLUP model, with the comprehensive structure, could be useful in improving the prediction accuracy of quantitative traits in RGS programs.

No abstract

Plant breeding for quantitative traits is a complicated task; thus, the recurrent selection method has been used in the rice (Oryza sativa L.) breeding program at the Brazilian Agricultural Research Corporation (Embrapa). Our general objective was to assess the effectiveness of this method in achieving genetic progress, maintaining genetic variability, and increasing the potential for selection of superior lines. A genetically broad‐based population of irrigated rice, CNA12S, submitted to three selection cycles was used in this study. The dataset comprised 10 yield trials, in which 667 S1:3 progenies and six check cultivars were assessed for grain yield, plant height, and days to flowering. We measured effective population size in each cycle, using standard and linkage disequilibrium methods, and Nei's genetic diversity in the third cycle. Such analyses were performed using data of single‐nucleotide polymorphism markers from progenies of the third cycle. For estimating the genetic gain, we adapted a generalized linear regression method to the Bayesian approach. This approach was also used to estimate variance and covariance components, according to the multivariate linear mixed model. Magnitudes of genetic and relative variation coefficients, as well as Nei's genetic diversity, indicated maintenance of genetic variability over cycles. Mean genetic gain per year was 1.98% for grain yield and −1.29% for days to flowering. Genetic potential of the population for extraction of superior lines was increased, considering single‐, two‐, or three‐trait selection. Our results show the effectiveness of the recurrent selection method when applied in rice breeding, although some refinements in the selection strategy could further improve its efficiency.

Genomic selection (GS) is a promising approach to improve rice (Oryza sativa L.) populations by using genome‐wide markers for selection prior to phenotyping to estimate breeding values. In this study, our objectives were to compare certain prediction models with different structures of genetic relationship and statistical approaches for relevant traits in rice and to discuss some implications for integrating GS into a recurrent selection program of irrigated rice. We assessed nine models in terms of predictive potential, using empirical data from S1:3 progenies phenotyped for eight traits with different heritabilities and genotyped with 6174 high‐quality single nucleotide polymorphism markers. For all traits, marker‐based models outperformed prediction based on pedigree records alone. A similar level of accuracy was observed for many models, although the level of prediction stability and prediction bias varied widely. Random forest was slightly superior for less complex traits, although with high prediction bias, whereas the semiparametric RKHS method (reproducing kernel Hilbert spaces) was superior for many traits, showing high stability and low bias. Bayesian variable selection method Bayes Cπ showed acceptable accuracy and stability for several traits and thus could be useful for genomic prediction aiming at persisting accuracy for a long‐term recurrent selection.