Predictive breeding and marker-assisted selection for grain quality and freezing tolerance in durum wheat

Yawar Habib¹Giuseppina Angione^2,3Paolo Vitale⁴Hassan Baneh^1,5Vincenzo Natoli⁶Concetta Lotti³Svetlana Dolaberidze⁷Ludmila Bespalova⁷Alexandra Mudrova⁷Aleksey Yanovsky⁷Salvatore Esposito^2*Pasquale De Vita^2*

• ¹Project Centre for Agro Technologies, Skolkovo Institute of Science and Technology, Moscow, Russia, Moscow, Russia
• ²Council for Agricultural Research and Agricultural Economy Analysis | CREA, Rome, Italy
• ³Universita degli Studi di Foggia Dipartimento di Scienze Agrarie Alimenti Risorse Naturali e Ingegneria, Foggia, Italy
• ⁴International Maize and Wheat Improvement Center (CIMMYT), Texcoco, Estado. de México, Mexico, Texcoco, Mexico
• ⁵Animal Science Research Department, Kurdistan Agricultural and Natural Resources Research and Education Center, Agricultural Research, Education and Extension Organization (AREEO), Sanandaj, Iran, Sanandaj, Iran
• ⁶Genetic Services SRL, Contrada Catenaccio, snc 71026, Deliceto (FG), Italy, Deliceto, Italy
• ⁷Federal State Budgetary Scientific Institution “NTsZ im. P.P. Lukyanenko", Krasnodar, Russia, Krasnodar, Russia

Durum wheat, a globally significant crop for high-quality pasta production, remains vulnerable to unseasonal freezing events, a risk that is intensified with climate variability. To address this challenge, we combined genome-wide association studies (GWAS), genomic prediction, and marker-assisted selection to improve both freezing tolerance and grain quality in durum wheat. A panel of 250 diverse accessions, comprising cold-adapted lines from Eastern Europe and high-quality genotypes from Southern Europe, was genotyped using a 25K SNP array. Clear genetic differentiation by geographical origin and growth habit highlighted contrasting allelic patterns for adaptation and quality traits. Phenotypic evaluations were carried out in experimental field trials over two consecutive growing seasons in Italy and Russia to assess the freezing tolerance and quality performance of the genetic materials. GWAS identified five significant marker-trait associations (MTAs) for freezing tolerance on chromosomes 2A, 2B, 3B, 4A, and 5A. Notably, a strong MTA on chromosome 5A (physical position 488.2 Mb) individually explained up to 27% of the phenotypic variance (PVE), co-localizing with the critical Fr-A2 cold-stress regulatory locus. Significant associations for grain-quality traits were localized on a 1B chromosome hotspot (541-652 Mb). A multi-trait genomic selection model integrating freezing tolerance, grain weight, and gluten traits enabled the identification of optimal parental lines, resulting in measurable gains across simulated generations. From the top-ranked crosses, BC₂F₂ populations were developed and genotyped with KASP markers targeting validated MTAs. Lines carrying favorable alleles for both freezing tolerance and gluten strength were successfully selected, confirming the predictive accuracy of the model. The integration of GWAS, diversity-preserving genomic prediction, and functional marker validation offers a robust and scalable pipeline for breeding cold-resilient, high-quality durum wheat, providing tangible tools to adapt Mediterranean and similar wheat systems to increasing climate variability.