Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Flowering Time Under Drought and Heat Stress Conditions in Maize

Drought stress is recognized as a major constraint to maize yield production, the heat stress alone and in combination with drought stress are likely to become the increasing constraints. The association mapping and genomic prediction analyses were conducted in a collection of 300 tropical and subtropical maize inbred lines to reveal the genetic architecture of grain yield and flowering time under well-watered, drought stress, heat stress, and combined drought and heat stress conditions. Out of the 381,165 genotyping-by-sequencing SNPs, 1661 were significantly associated with all the 12 trait-environment combinations, the average PVE (phenotypic variation explained) value of theses associations was 4.33%, and 586 of them had a PVE value greater than 5%. These associations were clustered into 446 genomic regions with a window size of 20 Mb per region, and 671 candidate genes containing the significantly associated SNPs were identified. In addition, 33 hotspots were identified for 12 trait-condition combinations and most were located on chromosomes 1 and 8. Compared with single SNP-based association mapping, the haplotype-based associated mapping detected less number of significant associations and candidate genes with higher PVE values. All the candidate genes were enriched into 15 gene ontology terms, and 46 candidate genes showed significant differential expression under the well-watered and drought stress conditions. Association mapping results identified few overlapped significant associations and candidate genes for the same traits evaluated under different conditions, indicating the genetic divergence between the individual stress tolerance and the combined drought and heat stress tolerance. The genomic prediction accuracies obtained from the marker-trait associated SNPs were relatively higher than those obtained from the genome-wide SNPs for most of the target traits. The genetic architecture information of the grain yield and flowering time revealed in this study, and the genomic regions identified for the different trait-environment combinations are helpful accelerating the efforts on rapid development of the stress-tolerant maize germplasm through marker-assisted selection or genomic selection.