TY - JOUR
T1 - Past and future wheat yield losses in France's breadbasket
AU - Nóia-Júnior, Rogério de S.
AU - Martre, Pierre
AU - Deswarte, Jean Charles
AU - Cohan, Jean Pierre
AU - Van der Velde, Marijn
AU - Webber, Heidi
AU - Ewert, Frank
AU - Ruane, Alex C.
AU - Ben-Ari, Tamara
AU - Asseng, Senthold
N1 - Publisher Copyright:
© 2024 The Authors
PY - 2025/3/1
Y1 - 2025/3/1
N2 - Context or problem: In recent decades, compounding weather extremes and plant diseases have increased wheat yield variability in France, the largest wheat producer in the European Union. Objective or research question: How these extremes might affect future wheat production remains unclear. Methods: Based on department level wheat yields, disease, and climate indices from 1980 to 2019 in France, we combined an existing disease model with machine learning algorithms to estimate future grain yields. Results: This approach explains about 59% of historical yield variability. Projections from five CMIP6 climate models suggest that extreme low wheat yields, which used to occur once every 20 years, could occur every decade by the end of this century, but elevated CO2 levels might lessen these events. Conclusions: Heatwave-related yield losses are expected to double, while flooding-related yield losses will potentially decline by one third, depending on the representative concentration pathway. Ear blight disease is projected to contribute to 20% of the expected 400 kg ha-1 average yield losses by the end of the century, compared with 12% in the historical baseline period. These projections depend on the timing of anthesis, currently between late May and early June in most departments. Anthesis advancing to early May would shift losses primarily to heavy rainfall and low solar radiation. Implications or significance: French wheat production must adapt to these emerging threats, such as heat stress, which until recently had little impact but may become the primary cause of future yield losses.
AB - Context or problem: In recent decades, compounding weather extremes and plant diseases have increased wheat yield variability in France, the largest wheat producer in the European Union. Objective or research question: How these extremes might affect future wheat production remains unclear. Methods: Based on department level wheat yields, disease, and climate indices from 1980 to 2019 in France, we combined an existing disease model with machine learning algorithms to estimate future grain yields. Results: This approach explains about 59% of historical yield variability. Projections from five CMIP6 climate models suggest that extreme low wheat yields, which used to occur once every 20 years, could occur every decade by the end of this century, but elevated CO2 levels might lessen these events. Conclusions: Heatwave-related yield losses are expected to double, while flooding-related yield losses will potentially decline by one third, depending on the representative concentration pathway. Ear blight disease is projected to contribute to 20% of the expected 400 kg ha-1 average yield losses by the end of the century, compared with 12% in the historical baseline period. These projections depend on the timing of anthesis, currently between late May and early June in most departments. Anthesis advancing to early May would shift losses primarily to heavy rainfall and low solar radiation. Implications or significance: French wheat production must adapt to these emerging threats, such as heat stress, which until recently had little impact but may become the primary cause of future yield losses.
KW - Compounding factors
KW - Extreme weather
KW - Machine learning
KW - Plant diseases
KW - Wheat
KW - Yield failure
UR - http://www.scopus.com/inward/record.url?scp=85211486390&partnerID=8YFLogxK
U2 - 10.1016/j.fcr.2024.109703
DO - 10.1016/j.fcr.2024.109703
M3 - Article
AN - SCOPUS:85211486390
SN - 0378-4290
VL - 322
JO - Field Crops Research
JF - Field Crops Research
M1 - 109703
ER -