GIS-FA: an approach to integrating thematic maps, factor-analytic, and envirotyping for cultivar targeting

KEY MESSAGE

We propose an "enviromics" prediction model for recommending cultivars based on thematic maps aimed at decision-makers. Parsimonious methods that capture genotype-by-environment interaction (GEI) in multi-environment trials (MET) are important in breeding programs. Understanding the causes and factors of GEI allows the utilization of genotype adaptations in the target population of environments through environmental features and factor-analytic (FA) models. Here, we present a novel predictive breeding approach called GIS-FA, which integrates geographic information systems (GIS) techniques, FA models, partial least squares (PLS) regression, and enviromics to predict phenotypic performance in untested environments. The GIS-FA approach enables: (i) the prediction of the phenotypic performance of tested genotypes in untested environments, (ii) the selection of the best-ranking genotypes based on their overall performance and stability using the FA selection tools, and (iii) the creation of thematic maps showing overall or pairwise performance and stability for decision-making. We exemplify the usage of the GIS-FA approach using two datasets of rice [Oryza sativa (L.)] and soybean [Glycine max (L.) Merr.] in MET spread over tropical areas. In summary, our novel predictive method allows the identification of new breeding scenarios by pinpointing groups of environments where genotypes demonstrate superior predicted performance. It also facilitates and optimizes cultivar recommendations by utilizing thematic maps.

Association mapping and genomic selection for sorghum adaptation to tropical soils of Brazil in a sorghum multiparental random mating population

A multiparental random mating population used in sorghum breeding is amenable for the detection of QTLs related to tropical soil adaptation, fine mapping of underlying genes and genomic selection approaches. Tropical soils where low phosphorus (P) and aluminum (Al) toxicity limit sorghum [Sorghum bicolor (L.) Moench] production are widespread in the developing world. We report on BRP13R, a multiparental random mating population (MP-RMP), which is commonly used in sorghum recurrent selection targeting tropical soil adaptation. Recombination dissipated much of BRP13R's likely original population structure and average linkage disequilibrium (LD) persisted up to 2.5 Mb, establishing BRP13R as a middle ground between biparental populations and sorghum association panels. Genome-wide association mapping (GWAS) identified conserved QTL from previous studies, such as for root morphology and grain yield under low-P, and indicated the importance of dominance in the genetic architecture of grain yield. By overlapping consensus QTL regions, we mapped two candidate P efficiency genes to a ~ 5 Mb region on chromosomes 6 (ALMT) and 9 (PHO2). Remarkably, we find that only 200 progeny genotyped with ~ 45,000 markers in BRP13R can lead to GWAS-based positional cloning of naturally rare, subpopulation-specific alleles, such as for SbMATE-conditioned Al tolerance. Genomic selection was found to be useful in such MP-RMP, particularly if markers in LD with major genes are fitted as fixed effects into GBLUP models accommodating dominance. Shifts in allele frequencies in progeny contrasting for grain yield indicated that intermediate to minor-effect genes on P efficiency, such as SbPSTOL1 genes, can be employed in pre-breeding via allele mining in the base population. Therefore, MP-RMPs such as BRP13R emerge as multipurpose resources for efficient gene discovery and deployment for breeding sorghum cultivars adapted to tropical soils.

The genetic architecture of phosphorus efficiency in sorghum involves pleiotropic QTL for root morphology and grain yield under low phosphorus availability in the soil

BACKGROUND

Phosphorus (P) fixation on aluminum (Al) and iron (Fe) oxides in soil clays restricts P availability for crops cultivated on highly weathered tropical soils, which are common in developing countries. Hence, P deficiency becomes a major obstacle for global food security. We used multi-trait quantitative trait loci (QTL) mapping to study the genetic architecture of P efficiency and to explore the importance of root traits on sorghum grain yield on a tropical low-P soil.

RESULTS

P acquisition efficiency was the most important component of P efficiency, and both traits were highly correlated with grain yield under low P availability. Root surface area was positively associated with grain yield. The guinea parent, SC283, contributed 58% of all favorable alleles detected by single-trait mapping. Multi-trait mapping detected 14 grain yield and/or root morphology QTLs. Tightly linked or pleiotropic QTL underlying the surface area of fine roots (1-2 mm in diameter) and grain yield were detected at positions 1-7 megabase pairs (Mb) and 71 Mb on chromosome 3, respectively, and a root diameter/grain yield QTL was detected at 7 Mb on chromosome 7. All these QTLs were near sorghum homologs of the rice serine/threonine kinase, OsPSTOL1. The SbPSTOL1 genes on chromosome 3, Sb03g006765 at 7 Mb and Sb03g031690 at 60 Mb were more highly expressed in SC283, which donated the favorable alleles at all QTLs found nearby SbPSTOL1 genes. The Al tolerance gene, SbMATE, may also influence a grain yield QTL on chromosome 3. Another PSTOL1-like gene, Sb07g02840, appears to enhance grain yield via small increases in root diameter. Co-localization analyses suggested a role for other genes, such as a sorghum homolog of the Arabidopsis ubiquitin-conjugating E2 enzyme, phosphate 2 (PHO2), on grain yield advantage conferred by the elite parent, BR007 allele.

CONCLUSIONS

Genetic determinants conferring higher root surface area and slight increases in fine root diameter may favor P uptake, thereby enhancing grain yield under low-P availability in the soil. Molecular markers for SbPSTOL1 genes and for QTL increasing grain yield by non-root morphology-based mechanisms hold promise in breeding strategies aimed at developing sorghum cultivars adapted to low-P soils.

Artificial neural networks as auxiliary tools for the improvement of bean plant architecture

Classification using a scale of visual notes is a strategy used to select erect bean plants in order to improve bean plant architectures. Use of morphological traits associated with the phenotypic expression of bean architecture in classification procedures may enhance selection. The objective of this study was to evaluate the potential of artificial neural networks (ANNs) as auxiliary tools in the improvement of bean plant architecture. Data from 19 lines were evaluated for 22 traits, in 2007 and 2009 winter crops. Hypocotyl diameter and plant height were selected for analysis through ANNs. For classification purposes, these lines were separated into two groups, determined by the plant architecture notes. The predictive ability of ANNs was evaluated according to two scenarios to predict the plant architecture - training with 2007 data and validating in 2009 data (scenario 1), and vice versa (scenario 2). For this, ANNs were trained and validated using data from replicates of the evaluated lines for hypocotyl diameter individually, or together with the mean height of plants in the plot. In each scenario, the use of data from replicates or line means was evaluated for prediction through previously trained and validated ANNs. In both scenarios, ANNs based on hypocotyl diameter and mean height of plants were superior, since the error rates obtained were lower than those obtained using hypocotyl diameter only. Lower apparent error rates were verified in both scenarios for prediction when data on the means of the evaluated traits were submitted to better trained and validated ANNs.

Diallel analysis to choose parents for black bean (Phaseolus vulgaris L.) breeding

In this study, conducted in two different seasons, we aimed to choose parents to obtain promising segregating populations for the extraction of black bean (Phaseolus vulgaris L.) lines that are superior in terms of disease resistance, plant architecture, and grain yield. Twelve parents were arranged in two groups to compose a partial diallel in a 5 x 7 scheme. Group 1 was composed of parents with black grains and erect plant architecture, while group 2 was composed of parents that had carioca grains and were resistant to the main fungal diseases that occur in the common bean. The following traits were evaluated: severity of angular leaf spot (ALS), plant architecture (PAG), and grain yield (YIELD). The data were analyzed according to a partial diallel model using parents and F₁ hybrids. In the genetic control of ALS and PAG, additive effects were predominant, while for YIELD, additive effects were predominant in one season and dominance effects were in another season, because it is a more complex trait than ALS and PAG. For YIELD, we observed an interaction between general combining ability and specific combining ability between seasons. The genes that control ALS, PAG, and YIELD were in eight of the 12 parents evaluated in the diallel. The cultivar 'BRS Estilo' is suitable to use as a parent in common bean breeding in terms of ALS, PAG and YIELD. Recurrent selection is the most recommended option for simultaneously breeding for PAG, YIELD, and resistance to angular leaf spot in bean culture.

Morphoagronomic characterization and genetic diversity of a common bean RIL mapping population derived from the cross Rudá x AND 277

Recombinant inbred lines (RILs) are a valuable resource for building genetic linkage maps. The presence of genetic variability in the RILs is essential for detecting associations between molecular markers and loci controlling agronomic traits of interest. The main goal of this study was to quantify the genetic diversity of a common bean RIL population derived from a cross between Rudá (Mesoamerican gene pool) and AND 277 (Andean gene pool). This population was developed by the single seed descent method from 500 F2 plants until the F10 generation. Seven quantitative traits were evaluated in the field in 393 RILs, the parental lines, and five control cultivars. The plants were grown using a randomized block design with additional controls and three replicates. Significant differences were observed among the RILs for all evaluated traits (P < 0.01). A comparison of the RILs and parental lines showed significant differences (P < 0.01) for the number of days to flowering (DFL) and to harvest (DH), productivity (PROD) and mass of 100 beans (M100); however, there were no significant differences for plant architecture, degree of seed flatness, or seed shape. These results indicate the occurrence of additive x additive epistatic interactions for DFL, DH, PROD, and M100. The 393 RILs were shown to fall into 10 clusters using Tocher's method. This RIL population clearly contained genetic variability for the evaluated traits, and this variability will be crucial for future studies involving genetic mapping and quantitative trait locus identification and analysis.

Genetic potential of segregating populations of red beans conducted by the bulk method with selection

This study evaluated the performance and estimated the genetic potential of segregating populations of red bean. Twenty populations of the second cycle of recurrent selection for red bean breeding at Universidade Federal de Viçosa were advanced to the F5 generation in bulk with selection for grain appearance. Populations plus five controls were evaluated for grain yield in three dry seasons (2004, 2005, and 2007) in a 5 x 5 lattice design with three replications, in four 4-m long rows. In the mean of the three crops, populations formed three distinct groups in which some populations had a clearly superior grain yield. Populations 288RVCI, 291RVCI, 295RVCI, 297RVCI, 300RVCI, and 303RVCI were the most promising. Over generations, segregating populations by the bulk method with selection for grain appearance may have reduced variability in grain yield. The 300RVCI population deserves further attention as it combines variability, high mean, and a high probability of generating above-standard lines.

Potential of hypocotyl diameter in family selection aiming at plant architecture improvement of common bean

Cultivars of common bean with more erect plant architecture and greater tolerance to degree of lodging are required by producers. Thus, to evaluate the potential of hypocotyl diameter (HD) in family selection for plant architecture improvement of common bean, the HDs of 32 F2 plants were measured in 3 distinct populations, and the characteristics related to plant architecture were analyzed in their progenies. Ninety-six F2:3 families and 4 controls were evaluated in a randomized block design, with 3 replications, analyzing plant architecture grade, HD, and grain yield during the winter 2010 and drought 2011 seasons. We found that the correlation between the HD of F2 plants and traits related to plant architecture of F2:3 progenies were of low magnitude compared to the estimates for correlations considering the parents, indicating a high environmental influence on HD in bean plants. There was a predominance of additive genetic effects on the determination of hypocotyl diameter, which showed higher precision and accuracy compared to plant architecture grade. Thus, this characteristic can be used to select progenies in plant architecture improvement of common beans; however, selection must be based on the means of at least 39 plants in the plot, according to the results of repeatability analysis.

Genetic improvement of plant architecture in the common bean

Knowledge of genetic control of plant architecture in the common bean can help breeders define the most adequate breeding strategy to optimize gains. We examined genetic control of plant architecture in the common bean by means of partial diallel crosses. Fourteen bean lines were crossed under a partial diallel scheme, in which group 1 was composed of 8 erect plant lines and group 2 of 6 carioca-type grain lines. The F1 plants from the crosses and the 14 parents were evaluated during spring (March sowing) for the characteristics plant architecture score, hypocotyl diameter, mean plant height, and grain yield. The additive effects were found to predominate in the genetic control of plant architecture score and hypocotyl diameter. Thus, selection of more erect plants can be done in early generations. Heritability estimate in a wide sense was greater for hypocotyl diameter (0.81) than for plant architecture score (0.60), showing that the latter is a more complex characteristic than the former. Hypocotyl diameter has the potential to be used in the selection of more erect plants, either by considering selection between and within families, or by combined or phenotypic recurrent selection. The reason for this is that the diameter measurements are taken from individual plants inside the plots.

Distinction between plant samples according to allele dosage by semiquantitative polymerase chain reaction

The lack of informativity of samples from heterozygotic individuals is one of the hindrances in the mapping of quantitative trait loci of outbred populations, since it is not normally possible to identify the origin of each allele. One way to include these individuals in analyses would be to genotype their endosperm, considering that a heterozygote (Aa) has AAa or Aaa endosperm, when the female genitor donates the A or a allele, respectively. We used semiquantitative polymerase chain reaction to determine allele dosages in DNA mixtures, by simulating the observed conditions for endospermic tissue. Semiquantitative polymerase chain reaction on agarose gels, along with regression analysis, allowed differentiation of the samples according to the amount of DNA. This type of information will help decrease the number of non-informative individuals in quantitative trait locus mapping of outbred populations, thereby increasing mapping accuracy.