A trait-based model ensemble approach to design rice plant types for future climate

Area of Science:

• Agricultural Science
• Climate Change Adaptation
• Computational Biology

Background:

• Crop models aid breeding by simulating genotype × environment × management interactions for climate change adaptation.
• Current crop models face limitations in plant breeding due to complex, non-direct trait-parameter relationships.
• Integrating crop modeling with breeding is crucial for developing climate-resilient crops.

Purpose of the Study:

• To develop a novel trait-based multi-model ensemble approach for designing improved rice plant types under future climate projections.
• To overcome limitations of single crop models by aggregating results into breeder-defined phenotypic traits.
• To integrate breeders' knowledge for clear mapping between designed plant types and breeding traits.

Main Methods:

• Utilized a trait-based multi-model ensemble approach with nine crop models from the AgMIP-Rice Project.
• Conducted multi-model simulations targeting enhanced productivity under various climate and management scenarios.
• Employed sensitivity analysis techniques and integrated breeders' perspectives for trait-parameter mapping.

Main Results:

• Demonstrated potential yield improvements ranging from 15.8% to 41.5% compared to current rice cultivars under mid-century climate projections.
• Identified phenological traits and those related to canopy development/structure as primary for climate change adaptation.
• Highlighted the importance of context-specific breeding programs due to site-specific climate condition effects on key traits.

Conclusions:

• The trait-based ensemble approach significantly advances the integration of crop modeling and breeding for climate change adaptation.
• Phenological and canopy-related traits are key for improving rice productivity and adaptation.
• Further model development is needed, but this approach offers a major step towards developing effective adaptation options for agriculture.