MULTI ENVIRONMENT HAPLOTYPE-BASED GWAS TO INVESTIGATE HEAT STRESS REGULATION ON DURUM WHEAT PANEL

Heat stress is one of the major constraints for the Triticum turgidum spp. Durum wheat production. It accounts for the reduction of growth rate, photosynthetic capacity, and spike fertility. Additionally, climate change in Mediterranean areas is expected to become a major challenge for durum wheat production. Herein, a durum wheat panel (UNIBO-DURUM ) of 182 cultivars and breeding lines has been assessed for response to long-term heat stress (HS), nonstress (NS) under field conditions. In total 23 traits were measured related to phenology, grain yield, and its components, at CENEB CIMMYT’s experimental station (Mexico) across HS and NS combined (2018/19) environments. To avoid the confounding effect of Heading Date (HD), we focused on Loci not affected by HD. Haplotype Genome Wide Association Studies (GWAS) for both HS and NS identified 11 Quantitative Trait Loci (QTL) clusters under HS and 10 under NS. In the HS, we found Q-ubo-Cluster1 (chr1A) at 2.21 Mb associated with four traits linked to the anthesis stage LOD =7.95; R2%= 7.17 (NDTH30A). Q-ubo-Cluster-7 (chr5A) at 43.34 Mb is also associated with four traits linked to reproductive stage, highest LOD score = 3.29 for Kernel Number per Spike (KNS) and the highest R2% = 5.63 for Test weight (TWT), although statistical significance was very low. Moreover, the Q-ubo-Cluster-8 (chr5B) at 642.87 Mb proved to be associated with five physiological yield traits, with the highest LOD score = 5.93 for Kernel Width Index (KWI) and highest R2% = 8.63 for Grain Filling period (GFP). On the other hand, in the NS condition, the Q-ubo-Cluster-3 (chr2A) at 36.59 Mb was found to be associated with seven traits linked to physiological, phenological, and reproductive stages. Both Plant Height (PH) and Spikelets per Spike (SKT) were statistically significant while the highest LOD score =12.45 and R2% = 10.39 were only observed for SKT. The Q-ubo-cluster-9 (chr5A) on 534.09 Mb is associated with yield component trait, highest LOD score = 8.44) and R2% = 7.05 at Thousand Kernel Weight (TKW), while Q-ubo-cluster-10 chr(5B) at 699.90 Mb was found associated with physiological, environmental stress and yeild component. Highest LOD score = 8.97 and R2% = 6.13 was observed for SKT. The overlapping of the traits at these clusters suggests the presence of tightly linked genes that concurrently regulate multiple physiological and yeild- releated traits under both stress and non-stress environments, or some locus with pleiotropic effect on several traits. Moreover, for the functional characterization of the genes underlying these QTLs, KASP marker development and gene editing will be done on the candidate genes of the investigated traits. And, the latest results will be presented in the poster at the conference.