QTL and genomic prediction accuracy for grain yield and secondary traits in a maize population under heat and heat-drought stresses

Abstract

Heat and drought stresses negatively affect maize (Zea mays L.) productivity. We aimed to identify the genetic basis of tolerance to heat stress (HS) and combined heat and drought stress (HS+DS) and compare how QTL and whole genome selection (GS) could be leveraged to improve tolerance to both stresses. A set of 97 testcross hybrids derived from a maize bi-parental doubled-haploid population was evaluated during the summer seasons of 2014, 2015, and 2016 in Ciudad Obregon, Sonora, Mexico, under HS and HS+DS. Grain yield (GY) reached 5.7 t ha−1 under HS and 3.0 t ha−1 under HS+DS. Twenty-six QTL were detected across six environments, with LOD scores ranging from 2.03 to 3.86; the QTL explained 8.6% to 18.6% of the observed phenotypic variation. Hyperspectral biomass and structural index (HBSI) had higher genetic correlation with GY for HS (r = 0.97) and HS+DS (r = 0.74), relative to the correlation with crop water mass or greenness indices. Genetic correlations between GY and canopy temperature for HS (r = −0.89) and HS+DS (r = −0.75) or vegetation indices, along with clusters of QTL in bins 1.02, 1.05, and 2.05, underline the importance of these genomic areas for secondary traits associated with general vigor and greenness. Prediction accuracy of the model used for GS had values below those found in previous studies. We found a high-yielding hybrid that was tolerant to HS and HS+DS.