Extreme rainfall reduces one-twelfth of China’s rice yield over the last two decades

Jin Fu, Yiwei Jian, Xuhui Wang, Laurent Li, Philippe Ciais, Jakob Zscheischler, Yin Wang, Yanhong Tang, Christoph Müller, Heidi Webber, Bo Yang, Yali Wu, Qihui Wang, Xiaoqing Cui, Weichen Huang, Yongqiang Liu, Pengjun Zhao, Shilong Piao, Feng ZhouNature Food 2023 · DOI

Executive Summary

While climate change discussions often prioritize heatwaves and droughts, extreme rainfall has emerged as a significant, yet overlooked, threat to global food security. A recent study reveals that heavy precipitation has reduced China’s rice yields by approximately 8% (one-twelfth) over the past twenty years. The damage is primarily driven by nitrogen leaching during early growth stages and physical damage to flowers during pollination. These findings suggest that current agricultural models may be overly optimistic by failing to account for the intensity and timing of rainfall.

Introduction

Rice is traditionally viewed as a water-tolerant crop, often grown in flooded paddies. This reputation has led many researchers to focus on "the big three" climate stressors: rising temperatures, CO2 levels, and drought. However, previous studies often relied on coarse administrative data that masked the localized impacts of intense storms, leading to an underestimation of rainfall's negative effects (previously cited around -0.6%). By utilizing high-resolution data and stage-specific analysis, this research demonstrates that extreme rainfall poses a risk comparable to extreme heat.

Methodology

To establish a link between rainfall events and biological yield loss, the researchers employed a multi-scale approach:

  1. Observational Analysis: Data from 114 sites across China (1999–2012) were analyzed using a "window searching strategy" to isolate rainfall impacts from other variables like cultivar changes.
  2. Experimental Validation: Controlled experiments at the Jingzhou Agrometeorological Station used artificial rainfall systems to simulate real-world storm kinetics. Researchers tested the effects of rain during vegetative, reproductive (flowering), and ripening stages.
  3. Process-Based Modeling: The ORCHIDEE-crop model was updated with modules for physical disturbance and nitrogen-loss pathways to hindcast historical losses and project future risks under RCP 4.5 and 8.5 scenarios.

Mechanisms of Yield Loss

The study identifies two primary "smoking guns" behind the yield reductions:

  • Physical Disturbance (The "Firehose" Effect): During the reproductive phase (anthesis), heavy rain physically dislodges pollen or prevents successful pollination. This results in "empty" or shrunken grains, accounting for 71–75% of the total yield loss.
  • Nutrient Washout: During the vegetative phase, extreme rainfall triggers significant nitrogen leaching. This limits "tillering"—the production of side shoots that eventually bear grain—capping the plant's yield potential early in the season. This mechanism accounts for 22–25% of the loss.

Impact Assessment

The historical impact of extreme rainfall is substantial. Nationwide observations indicate a yield reduction of 7.3 ± 0.7%, a figure closely mirrored by the improved crop model's estimate of 8.1 ± 1.1%.

Comparison of rice yield reduction across different extreme weather events.

As shown in the data, the impact of extreme rainfall is significantly higher than that of drought (~4.4%) or extreme cold (~3.7%) in the studied regions, rivaling the damage caused by extreme heat.

Spatial distribution of rice yield loss due to extreme rainfall across China.

The spatial analysis reveals that Northeast China and the southeast coastal regions are particularly vulnerable to these events. Under moderate warming scenarios (RCP 4.5), an additional 8% yield reduction is projected by the end of the century due to intensifying rainfall patterns.

Discussion and Future Directions

The reported 8% reduction may be a conservative estimate, as the study's screening criteria likely excluded the most extreme events, such as typhoons. Furthermore, the experiments did not fully account for secondary storm effects like fungal diseases or "lodging" (plants falling over).

While this study focuses on China, the world's largest rice producer, the implications are global. Regions like Southeast Asia, which often experience higher rainfall intensities and use less nitrogen fertilizer, may face even greater risks. Future research should apply this integrated observation-experiment-model approach to other staple crops like wheat and maize to ensure global food systems are resilient to the full spectrum of climate extremes.

Summary updated

BibTeX
@inproceedings{fu2023extreme,
  author = {Jin Fu and Yiwei Jian and Xuhui Wang and Laurent Li and Philippe Ciais and Jakob Zscheischler and Yin Wang and Yanhong Tang and Christoph Müller and Heidi Webber and Bo Yang and Yali Wu and Qihui Wang and Xiaoqing Cui and Weichen Huang and Yongqiang Liu and Pengjun Zhao and Shilong Piao and Feng Zhou},
  title = {{Extreme rainfall reduces one-twelfth of China’s rice yield over the last two decades}},
  booktitle = {Nature Food},
  year = {2023},
  doi = {10.1038/s43016-023-00753-6},
}
Figure