Diagnostic markers for vernalization and photoperiod loci improve genomic selection for grain yield and spectral reflectance in wheat

Authors: MASON, R - University Of Arkansas; ADDISON, CHRISTOPHER - University Of Arkansas; BABAR, ALI - University Of Florida; ACUNA, ANDREA - University Of Arkansas; LOZADA, DENNIS - University Of Arkansas; SUBRAMANIAN, NITHYA - University Of Arkansas; ARGUELLO, MARIA - University Of Arkansas; MILLER, RANDALL - University Of Arkansas; Brown-Guedira, Gina; GUEDIRA, MOHAMMED - North Carolina State University; JOHNSON, JERRY - University Of Georgia

Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/20/2017
Publication Date: 1/15/2018
Citation: Mason, R.E., Addison, C.K., Babar, A., Acuna, A., Lozada, D., Subramanian, N., Arguello, M.N., Miller, R.G., Brown Guedira, G.L., Guedira, M., Johnson, J. 2018. Diagnostic markers for vernalization and photoperiod loci improve genomic selection for grain yield and spectral reflectance in wheat. Crop Science. 58(1): 242-252.

Interpretive Summary: The objective of this study was to identify quantitative trait loci (QTL) associated with normalized difference vegetation index (NDVI) measured across different growth stages in a wheat recombinant inbred line population and to determine the predictability of NDVI and grain yield (GY) using a genomic selection approach. The population was grown over three seasons in 12 total site-years and NDVI was measured in seven site-years. Measurements of NDVI from tillering to physiological maturity showed low to moderate heritability (h2 = 0.06–0.68). Positive correlations were observed among NDVI, GY, and biomass, particularly in low-yielding environments. Quantitative trait loci analysis found 18 genomic regions associated with NDVI, with most affecting across multiple growth stages. The QTL were detected near markers for the major photoperios response (Ppd-B1, PpdD1) and vernalization response genes (vrn-A1, and vrn-B1), with Ppd-D1 having the largest effect. Multiple QTL models showed that interactions between Ppd and Vrn loci also significantly influenced NDVI. Genomic selection accuracy ranged from r = -0.10 to 0.54 for NDVI across growth stages. However, the inclusion of Vrn and Ppd loci as fixed effects increased GS accuracy for NDVI and GY in environments with the lowest heritability. The highest accuracy for GY (r = 0.58– 0.59) was observed in the environment group with the highest heritability (h2 = 0.85). Overall, these results will aid in future selection of optimal plant growth for target environments using both phenotypic and GS approaches.

Technical Abstract: The objective of this study was to identify quantitative trait loci (QTL) associated with normalized difference vegetation index (NDVI) measured across different growth stages in a wheat (Triticum aestivum L.) recombinant inbred line (RIL) population and to determine the predictability of NDVI and grain yield (GY) using a genomic selection (GS) approach. The RILs were grown over three seasons in 12 total site-years and NDVI was measured in seven site-years. Measurements of NDVI from tillering to physiological maturity showed low to moderate heritability (h2 = 0.06–0.68). Positive correlations were observed among NDVI, GY, and biomass, particularly in low-yielding site-years. Quantitative trait loci analysis found 18 genomic regions associated with NDVI, with most pleiotropic across multiple growth stages. The QTL were detected near markers for Ppd-B1, PpdD1, vrn-A1, and vrn-B1, with Ppd-D1 having the largest effect. Multiple QTL models showed that epistatic interactions between Ppd and Vrn loci also significantly influenced NDVI. Genomic selection accuracy ranged from r = -0.10 to 0.54 for NDVI across growth stages. However, the inclusion of Vrn and Ppd loci as fixed effect covariates increased GS accuracy for NDVI and GY in site-year groupings with the lowest heritability. The highest accuracy for GY (r = 0.58– 0.59) was observed in the site-year grouping with the highest heritability (h2 = 0.85). Overall, these results will aid in future selection of optimal plant growth for target environments using both phenotypic and GS approaches.