The Innovation Team of Improvement and Amelioration of Soil Fertility at the Institute of Agricultural Resources and Regional Planning (IARRP) of the Chinese Academy of Agricultural Sciences (CAAS) has made significant progress in the quantitative assessment of the impacts of future extreme climates on wheat yields, based on the national soil quality observation and experimental station network. The study revealed the spatiotemporal response characteristics of wheat yields and the threshold effects of key driving factors under different climate scenarios. The findings were recently published in the "Field Crops Research", under the title "Risks of wheat yields reduction under future climate extremes".

As the world's third most important staple crop, wheat plays a vital role in ensuring food security. However, climate variability—manifested through increasingly frequent extreme heat, drought, and heavy rainfall events—poses serious challenges to agricultural production. Previous studies have paid limited attention to the nonlinear relationships and critical thresholds between multiple climate factors, particularly extreme events, and wheat yields.

The results reveal regional differences in wheat yield responses under future climate scenarios. The Northwest region is the most vulnerable to extreme climate events, with yields projected to decline by as much as 30–40 percent by the end of the century. In the Yangtze River Basin, the combined effects of high temperatures and intense rainfall could reduce yields by 19–36 percent, although yield increases remain possible under low-emission scenarios. The North China Plain shows relatively stronger resilience, but under high-emission scenarios, it may still face yield reductions of 17–25 percent.

Importantly, the study identified key inflection points where wheat yield shifts from growth to decline, highlighting the "threshold effects" of climate factors. For example, in the Northwest region, the extreme precipitation index during wheat growing season (R95p) has a critical threshold of 75.5 mm. When heavy rainfall remains below this level, additional precipitation can benefit crop growth and increase yields. However, once the threshold is exceeded, the positive effect quickly reverses and becomes a risk factor for yield loss. Similarly, cumulative precipitation during wheat growing season (PRE) shows a threshold of 285.8 mm.

By linking field-level observations with regional-scale projections, the study provides deeper insights into the nonlinear mechanisms through which extreme climate events influence crop productivity. The findings underscore the importance of identifying climate thresholds in intensive agricultural systems and offer scientific guidance for developing region-specific climate adaptation strategies and optimizing field management practices to enhance the resilience of China's grain production systems.

By linking field-level observations with regional-scale projections, the study provides deeper insights into the nonlinear mechanisms through which extreme climate events influence crop productivity. The findings underscore the importance of identifying climate thresholds in intensive agricultural systems and offer scientific guidance for developing region-specific climate adaptation strategies and optimizing field management practices to enhance the resilience of China's grain production systems.

Li Guanmo, a jointly trained doctoral student of IARRP and the University of Liege, is the first author of the study, while Professor Zhang Wenju serves as the corresponding author. The research was supported by the National Key Laboratory for Efficient Utilization of Arid and Semi-Arid Cropland in Northern China, the National Science & Technology Fundamental Resources Investigation Project of China, the National Key Research and Development Program of China, and the China Scholarship Council.

Figure 1: Historical and projected wheat yield under different climate scenarios in different periods.

Figure 2: SHAP summary plots showing the contributions of the top four climate indicators to wheat yield predictions based on the Random Forest model.