Positive effects of public breeding on U.S. rice yields under future climate scenarios.

Authors: WANG, DIANE - Purdue University; JAMSHIDI, SAJAD - Purdue University; HAN, RONGKUI - Basf Corporation; Edwards, Jeremy; McClung, Anna; MCCOUCH, SUSAN - Cornell University

Submitted to: Proceedings of the National Academy of Sciences (PNAS)
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/2/2024
Publication Date: 3/18/2024
Citation: Wang, D.R., Jamshidi, S., Han, R., Edwards, J., McClung, A.M., Mccouch, S.R. 2024. Positive effects of public breeding on U.S. rice yields under future climate scenarios.. Proceedings of the National Academy of Sciences (PNAS). https://doi.org/10.1073/pnas.2309969121.
DOI: https://doi.org/10.1073/pnas.2309969121

Interpretive Summary: To support the production of food for a growing global population, agricultural systems must be able to adapt to a changing climate. One way to do this is to evaluate how weather has affected yields in the past and then predict how crops will respond in the future under forecasted climate scenarios. This study focuses on the Southern U.S. rice-growing area and uses information on genetic variation, crop productivity, and past weather data to model and predict rice yields. Based on county-level data from 1970 to 2015 that includes yields, acreage by variety, and weather data, along with molecular marker information for the varieties grown in each year for each county, machine learning models were developed to predict historic yields. Results indicated that the models developed from this approach were highly correlated with the actual observed historic yields (r=0.74). These models were then used to predict yields under forecasted future climates. The results show that recent varieties developed from public breeding programs that have included genetic resources from outside of the southern US gene pool are more resistant to predicted future climate conditions. This modeling approach can be applied to other crops besides rice and will serve as a guide for adjusting breeding strategies so that varieties developed from them are more resilient to climate change.

Technical Abstract: In this study, we model and predict rice yields by integrating molecular marker variation, varietal productivity, and climate, focusing on the Southern U.S. rice growing region. This region spans the states of Arkansas, Louisiana, Texas, Mississippi, and Missouri and accounts for 85% of total U.S. rice production. By digitizing and combining four decades of county-level variety acreage data (1970 - 2015) with varietal information from genotyping-by-sequencing data, we estimate annual county-level allele frequencies. These allele frequencies are used together with county-level weather and yield data to develop ten machine learning models for yield prediction. A two-layer meta-learner ensemble model that combines all ten methods is externally evaluated against observations from historical Uniform Regional Rice Nursery trials (1980 - 2018) conducted in the same states. Finally, the ensemble model is used with forecasted weather from the Coupled Model Intercomparison Project across the 110 rice-growing counties to predict production in the coming decades for Composite Variety Groups assembled based on year-of-release, breeding program, and several breeding trends. Results indicate positive effects over time of public breeding on rice resilience to future climates and potential reasons are discussed.