Yang et al. (2025) Seasonal drought during the soybean growth period and agricultural water stress in southern China eased in the context of global warming

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-11-24
• Authors: Xiaolong Yang, Dongli She, Yong Jing, Daming Yang, Xiaoyu Jiang, Yong Zhong, Lei Gao
• DOI: 10.1016/j.agwat.2025.110012

Research Groups

• College of Agricultural Science and Engineering, Hohai University, Nanjing, China
• Jiangsu Province Engineering Research Center for Agricultural Soil-Water Efficient Utilization, Carbon Sequestration and Emission Reduction, Nanjing, China
• State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China

Short Summary

This study investigated soybean water requirements and seasonal drought characteristics in the middle reaches of the Yangtze River Basin from 1961 to 2021. It revealed a decreasing trend in both drought severity and soybean water requirements amidst global warming, with net radiation and vapor pressure deficit identified as dominant drivers.

Objective

• To investigate drought evolution during the soybean growth period by employing the daily Standardised Precipitation Evapotranspiration Index (SPEI).
• To assess drought impacts on soybean water requirements (SWR) at each crop growth stage.
• To investigate the key driving mechanisms and spatial heterogeneity in SWR under different drought and wet scenarios using Random Forest, Shapley Additive exPlanations (SHAP), and GeoDetector models.

Study Configuration

• Spatial Scale: Middle reaches of the Yangtze River Basin (MYRB), Central China (24°29′N–34°12′N, 106°5′E–118°36′E), including the Poyang Lake Basin, Dongting Lake Basin, Han River Basin, and Yangtze River Mainstream area.
• Temporal Scale: 1961 to 2021 (61 years).

Methodology and Data

• Models used:
  • Penman–Monteith equation (for reference crop water requirement, ET0)
  • Single crop coefficient method (for crop water requirement, CWR)
  • Sen–Mann–Kendall (MK) method (for trend analysis)
  • Trend-Free Pre-Whitening (TFPW) method (for mitigating serial correlation)
  • Daily Standardised Precipitation Evapotranspiration Index (SPEI) (for drought assessment, 30-day timescale)
  • Triple-threshold-based Run theory (for drought event identification)
  • Random Forest (RF) model (for regression prediction of SWR)
  • Shapley Additive exPlanations (SHAP) (for model interpretability and factor importance)
  • Optimal-parameter GeoDetector model (for spatial differentiation and driving factors)
• Data sources:
  • Daily meteorological data from 213 stations (1961-2021) from the National Meteorological Information Center of the China Meteorological Administration (CMA): maximum, minimum, and average temperatures (°C), relative humidity (%), sunshine duration (h), wind speed (m/s), and precipitation (mm).
  • Crop phenology data from provincial agriculture and rural departments and seed industry data websites.
  • Crop coefficients from FAO Irrigation and Drainage Paper No. 56 (Allen et al., 1998).
  • Historical drought information from the Yearbook of China Water Resources and the China Flood and Drought Disaster Prevention Bulletin.

Main Results

• The multiyear average Crop Water Requirement (CWR) for soybeans was 433.82 mm, and Irrigation Water Requirement (IWR) was 356.81 mm.
• Both CWR and IWR showed decreasing trends from 1961 to 2021, with annual rates of 0.197 mm/year and 0.235 mm/year, respectively.
• The regional drought characteristics exhibited a decreasing trend, with the average daily SPEI showing a significant increasing trend (p < 0.01), indicating a gradual transition towards wetter conditions.
• Light (46.40%) and moderate (33.68%) droughts dominated the soybean growth period, while severe (15.83%) and extreme (4.09%) droughts were less frequent.
• Drought count (DC), duration (DD), severity (DS), and peak (DP) all decreased over the past 61 years at rates of -0.002 events/year, -0.247 days/year, -0.381/year, and -0.006/year, respectively.
• The rapid growth and mid-season stages of soybean exhibited the highest vulnerability to drought (correlation coefficients between SPEI and SWR < -0.4), with the mid-season stage showing the strongest negative correlation (-0.4 to -0.8).
• Net radiation (Rn) was consistently the strongest driver of spatial heterogeneity in SWR (q > 0.7) across all drought and wet scenarios.
• The wet scenario was mainly driven by net radiation-temperature interactions, whereas other scenarios (drought, light-drought, normal) were predominantly driven by net radiation-vapour pressure deficit interactions.
• The primary meteorological factors contributing to CWR were Rn (46.20%), VPD (20.52%), Tem (13.34%), U2 (12.57%), RHU (6.55%), and TP (0.80%). A similar pattern was observed for IWR.

Contributions

• Provides a comprehensive analysis of soybean water requirements and seasonal drought evolution in the humid, monsoon-affected Yangtze River Basin, addressing a critical knowledge gap in such regions.
• Integrates daily SPEI monitoring with process-based water requirement analysis and advanced machine learning (Random Forest, SHAP, GeoDetector) to systematically examine spatiotemporal dynamics and driving mechanisms.
• Quantifies the decreasing trends in seasonal drought and soybean water requirements in the context of global warming in the study area, contrasting with trends observed in arid/semi-arid regions.
• Identifies the mid-season growth stage as the most critical period for drought impact on SWR and highlights net radiation and vapor pressure deficit as dominant, scenario-dependent drivers.
• Offers crucial insights and theoretical support for formulating drought mitigation strategies and optimizing regional agricultural water management under climate change.

Funding

• Strategic Priority Research Program of the Chinese Academy of Sciences (XDA0440202)
• Natural Science Foundation of Jiangsu Province (BK20241510)
• Fundamental Research Funds for the Central Universities (B240201052)

Citation

@article{Yang2025Seasonal,
  author = {Yang, Xiaolong and She, Dongli and Jing, Yong and Yang, Daming and Jiang, Xiaoyu and Zhong, Yong and Gao, Lei},
  title = {Seasonal drought during the soybean growth period and agricultural water stress in southern China eased in the context of global warming},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.110012},
  url = {https://doi.org/10.1016/j.agwat.2025.110012}
}