In crop species, adaptation to different agroclimatic regions creates useful variation but also leads to unwanted genetic correlations. Bouchet.......

Core Ideas A reference single nucleotide polymorphism dataset covering over 10,000 sorghum genotypes was developed for the crop research community. The dataset was validated through gene‐resolution mapping of known genes for pigmentation, plant height, and flowering time. The value of the dataset was demonstrated with novel candidate genes for mesocarp thickness (Z), plant height (qPHT7.1), and other traits. Mining crop genomic variation can facilitate the genetic research of complex traits and molecular breeding. In sorghum [Sorghum bicolor L. (Moench)], several large‐scale single nucleotide polymorphism (SNP) datasets have been generated using genotyping‐by‐sequencing of ApeKI reduced representation libraries. However, data reuse has been impeded by differences in reference genome coordinates among datasets. To facilitate reuse of these data, we constructed and characterized an integrated 459,304‐SNP dataset for 10,323 sorghum genotypes on the version 3.1 reference genome. The SNP distribution showed high enrichment in subtelomeric chromosome arms and in genic regions (48% of SNPs) and was highly correlated (r = 0.82) to the distribution of ApeKI restriction sites. The genetic structure reflected population differences by botanical race, as well as familial structure among recombinant inbred lines (RILs). Faster linkage disequilibrium decay was observed in the diversity panel than in the RILs, as expected, given the greater opportunity for recombination in diverse populations. To validate the quality and utility of the integrated SNP dataset, we used genome‐wide association studies (GWAS) of genebank phenotype data, precisely mapping several known genes (e.g Y1 and Z) and identifying novel associations for other traits. We further validated the dataset with GWAS of new and published plant height and flowering time data in a nested association mapping population, precisely mapping known genes and identifying epistatic interactions underlying both traits. These findings validate this integrated SNP dataset as a useful genomics resource for sorghum genetics and breeding.

Dissecting the genetic architecture of stress tolerance in crops is critical to understand and improve adaptation. In temperate climates, early planting of chilling-tolerant varieties could provide longer growing seasons and drought escape, but chilling tolerance (<15°) is generally lacking in tropical-origin crops. Here we developed a nested association mapping (NAM) population to dissect the genetic architecture of early-season chilling tolerance in the tropical-origin cereal sorghum (Sorghum bicolor [L.] Moench). The NAM resource, developed from reference line BTx623 and three chilling-tolerant Chinese lines, is comprised of 771 recombinant inbred lines genotyped by sequencing at 43,320 single nucleotide polymorphisms. We phenotyped the NAM population for emergence, seedling vigor, and agronomic traits (>75,000 data points from ∼16,000 plots) in multi-environment field trials in Kansas under natural chilling stress (sown 30–45 days early) and normal growing conditions. Joint linkage mapping with early-planted field phenotypes revealed an oligogenic architecture, with 5–10 chilling tolerance loci explaining 20–41% of variation. Surprisingly, several of the major chilling tolerance loci co-localize precisely with the classical grain tannin (Tan1 and Tan2) and dwarfing genes (Dw1 and Dw3) that were under strong directional selection in the US during the 20th century. These findings suggest that chilling sensitivity was inadvertently selected due to coinheritance with desired nontannin and dwarfing alleles. The characterization of genetic architecture with NAM reveals why past chilling tolerance breeding was stymied and provides a path for genomics-enabled breeding of chilling tolerance.

Aims Western Africa (WA) sorghums are predominantly cultivated under low plant available phosphorus (P) soil conditions with a diverse population of arbuscular mycorrhizal fungi (AMF) present. This study aims to determine whether sorghum breeding programs should target higher colonization by AMF through understanding the genotypic variation of sorghum for AMF-root colonization (AMF-RC) under different P-fertility conditions at different growth stages and assessing the genetics underlying AMF-RC using genome-wide association study (GWAS). Method A sorghum diversity panel of 187 WA genotypes was grown in low-P soil in a pot trial for 38 days and a subset of 13 genotypes was grown in a low-and high-P field until maturity at ICRISAT-Samanko in Mali, WA. Root samples were taken at 38 days from the pot trial plants and at flowering time in the field trials. Shoot biomass was analyzed for P concentration and dry matter yield. GWAS was conducted for shoot-P-content and AMF-RC. Results Significant genotypic variation was observed for AMF-RC, but the repeatability estimates were only low (w 2 =0.15 at 38 days) to moderate (w 2 =0.54-0.56 at flowering time). AMF-RC was significantly higher in low-P versus high-P field conditions. Large residual variation was observed for AMF-RC in both pot and field trials. None of the genotypic groups, contrasting for selection history, race and grain yield performance across multiple field trials, differed significantly for AMF-RC. AMF-RC showed no or negative relationships to shoot-P-content and grain yield, irrespective of soil-P level or plant developmental stage. AMF-RC at 38 days was significantly correlated (r=67**) to AMF-RC at flowering. However, GWAS did not detect significant genomic regions for AMF-RC but did for shoot-P content. Conclusion Although genetic differences for AMF-RC were detected, the trait appears to be highly polygenic. Genotypic selection for higher AMF-RC in WA sorghums is not promising due to the low heritability and the lack of positive relationships with P acquisition.

Evolution of plants under climatic gradients may lead to clinal adaptation. Understanding the genomic basis of clinal adaptation in crops species could facilitate breeding for climate resilience. We investigated signatures of clinal adaptation in the cereal crop sorghum (Sorghum bicolor L. [Moench]) to the precipitation gradient in West Africa using a panel (n = 607) of sorghum accessions from diverse agroclimatic zones of Nigeria. Significant correlations were observed between common-garden phenotypes of three putative climate-adaptive traits (flowering time, plant height, and panicle length) and climatic variables. The panel was characterized at >400,000 single nucleotide polymorphisms (SNPs) using genotyping-by-sequencing (GBS). Redundancy analysis indicated that a small proportion of SNP variation can be explained by climate (1%), space (1%), and climate collinear with space (3%). Discriminant analysis of principal components identified three genetic groups that are distributed differently along the precipitation gradient. Genome-wide association studies were conducted with phenotypes and three climatic variables (annual mean precipitation, precipitation in the driest quarter, and annual mean temperature). There was no overall enrichment of associations near a priori candidate genes implicated in flowering time, height, and inflorescence architecture in cereals, but several significant associations were found near a priori candidates including photoperiodic flowering regulators SbCN12 and Ma6. Together, the findings suggest that a small (3%) but significant proportion of nucleotide variation in Nigerian sorghum landraces reflects clinal adaptation along the West African precipitation gradient.