Multi-trait and multi-environment genomic prediction enhances yield components improvement in durum wheat

Area of Science:

• Agricultural Science
• Plant Breeding
• Genomics

Background:

• Durum wheat is vital for pasta and semolina, facing climate variability challenges in Mediterranean regions.
• Yield stability is crucial for durum wheat production in semi-arid and variable climates.

Purpose of the Study:

• Evaluate genomic prediction models (SE, MT, ME, MTME) for seven key durum wheat traits.
• Assess the impact of genomic (G) and gene-based (G2) relationship matrices on prediction accuracy.
• Identify optimal modeling strategies for durum wheat improvement under climate variability.

Main Methods:

• Compared single-environment (SE), multi-trait (MT), multi-environment (ME), and multi-trait-multi-environment (MTME) genomic prediction models.
• Utilized genomic (G) and target gene-based (G2) relationship matrices.
• Employed two cross-validation scenarios (CV1, CV2) simulating Mediterranean environments.

Main Results:

• MTME models demonstrated the highest prediction accuracies, especially under CV2 and sowing-by-season grouping.
• Incorporating G2 information improved predictions for morpho-phenological traits like heading date and plant height.
• MTME models effectively utilized inter-trait and inter-environment covariances for realistic genotype performance predictions.

Conclusions:

• MTME genomic prediction offers a robust framework for climate-resilient durum wheat improvement.
• This approach supports data-driven breeding pipelines to enhance genetic gain and stability.
• Leveraging G2 data and MTME models is key for durum wheat adaptation to environmental challenges.