Li et al. (2026) Scale-dependent responses of water use efficiency to precipitation variability in a rain-fed maize field of Northeast China

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-09-26
• Authors: Junjie Li, YanHong Dong, Deyan Liu, Xian Wu, Tingting Zhu, 王慧琴 Wang Huiqin, Yi Liu, Ding WeiXin, Junji Yuan
• DOI: 10.1016/j.agwat.2026.110814

Research Groups

• State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China
• Center for Plant Water-use and Nutrition Regulation, State Key Laboratory of Agricultural and Forestry Biosecurity, College of Resources and Environment, Fujian Agriculture and Forestry University, Fuzhou, China
• Co-Innovation Center for Sustainable Forestry in Southern China, Nanjing Forestry University, Nanjing, China
• University of Chinese Academy of Sciences, Nanjing, China

Short Summary

This study investigated the scale-dependent responses of ecosystem (EWUE), transpiration (TWUE), and intrinsic (IWUE) water use efficiency to precipitation variability in a rain-fed maize field in Northeast China over three contrasting hydrological years. It found that both increased and decreased precipitation reduced EWUE and TWUE, primarily due to disproportionate declines in gross primary productivity (GPP) relative to water loss, while IWUE exhibited contrasting responses, decreasing with increased precipitation but increasing with decreased precipitation.

Objective

• To characterize the diurnal, seasonal, and interannual variations in cropland water use efficiency (WUE) at ecosystem, canopy, and leaf scales.
• To understand how ecosystem WUE (EWUE), transpiration WUE (TWUE), and intrinsic WUE (IWUE) respond to different hydroclimatic conditions induced by precipitation anomalies.
• To assess the extent to which the sensitivity of EWUE, TWUE, and IWUE to other environmental factors varies among hydrological years.

Study Configuration

• Spatial Scale: A single rain-fed maize (Zea mays L.) field located in Qixinghe Town, Baoqing County, Heilongjiang Province, Northeast China (46.64°N, 131.95°E). The field is situated on a flat plain with Mollisols (Aquic Hapludolls).
• Temporal Scale: A 3-year study period from June 2020 to June 2023, encompassing a wet year (2020–2021), a dry year (2021–2022), and a normal year (2022–2023). The study analyzed diurnal, seasonal (growing season), and interannual variations.

Methodology and Data

• Models used:
  • Eddy covariance technique for measuring ecosystem-scale CO2 (Net Ecosystem CO2 Exchange, NEE) and H2O (latent heat, LE; sensible heat, H) fluxes.
  • Marginal distribution sampling algorithm for gap-filling missing flux data.
  • Nighttime partitioning approach for separating NEE into GPP and ecosystem respiration (ER).
  • Three approaches for evapotranspiration (ET) partitioning into transpiration (T) and evaporation (E):
    • Underlying water-use efficiency (uWUE) method.
    • Transpiration Estimation Algorithm (TEA) (random forest algorithm).
    • Conductance-based method (inverted Penman-Monteith equation adapted with Medlyn model for partitioning ecosystem conductance into vegetation and soil components).
  • Random Forest (RF) algorithm for determining key environmental factors influencing WUE at different scales.
  • Ordinary linear regression analysis for examining relationships among WUEs and environmental variables.
• Data sources:
  • Eddy covariance system measurements of CO2 and H2O fluxes.
  • Meteorological sensors for auxiliary measurements:
    • Air temperature (TA) and relative humidity (VPD derived from these).
    • Precipitation (tipping-bucket rain gauge).
    • Wind speed (propeller anemometer).
    • Downward and upward shortwave and longwave radiation (net radiometer, for Net Radiation, Rn).
    • Incoming photosynthetically active radiation (PAR) (quantum sensor).
    • Soil temperature (TS) and volumetric water content (SWC) at 10 cm depth (multiparameter sensors).
    • Soil heat fluxes (G) at 10 cm depth (thermal sensors).

Main Results

• Impact of Increased Precipitation (Wet Year):
  • Reduced EWUE by 10.3% (GPP declined more than ET).
  • Marginally decreased TWUE by 5.4% (proportional reductions in GPP and T).
  • Lowered IWUE by 21.4% (larger reduction in GPP than canopy conductance, gc).
  • Dampened diel variations of EWUE, TWUE, and IWUE.
  • Steepened the positive EWUE-GPP relationship.
  • Strengthened the temperature dependence of EWUE.
  • Attenuated TWUE responses to GPP and vapor pressure deficit (VPD).
  • Intensified IWUE sensitivity to VPD and radiation.
• Impact of Decreased Precipitation (Dry Year):
  • Suppressed EWUE by 25.3% (GPP decline far exceeded ET decline).
  • Suppressed TWUE by 16.0% (GPP decline far exceeded T decline).
  • Increased IWUE by 8.2% (reduction in gc far exceeded GPP reduction).
  • Amplified diel variations of EWUE, TWUE, and IWUE.
  • Steepened the positive EWUE-GPP relationship.
  • Weakened the temperature dependence of EWUE.
  • Attenuated TWUE responses to GPP and VPD.
  • Intensified IWUE sensitivity to VPD and radiation.
• General Findings:
  • EWUE was primarily driven by GPP, soil temperature (TS), and air temperature (TA).
  • TWUE was mainly influenced by GPP, VPD, and T.
  • IWUE was primarily regulated by VPD, photosynthetically active radiation (PAR), and net radiation (Rn).
  • Diel patterns: EWUE peaked around 06:00, TWUE showed U-shaped patterns, and IWUE showed hump-shaped patterns.
  • Growing season cumulative precipitation: 642 mm (wet), 320 mm (dry), 432 mm (normal).
  • Growing season GPP: 1358 g C m⁻² (wet), 1065 g C m⁻² (dry), 1662 g C m⁻² (normal).
  • Growing season ET: 320 mm (wet), 301 mm (dry), 351 mm (normal).
  • Growing season EWUE: 4.24 g C mm⁻¹ H2O (wet), 3.53 g C mm⁻¹ H2O (dry), 4.73 g C mm⁻¹ H2O (normal).
  • Growing season TWUE: 7.65 g C mm⁻¹ H2O (wet), 6.79 g C mm⁻¹ H2O (dry), 8.08 g C mm⁻¹ H2O (normal).
  • Growing season IWUE: 47.73 µmol C mol⁻¹ H2O (wet), 65.67 µmol C mol⁻¹ H2O (dry), 60.72 µmol C mol⁻¹ H2O (normal).

Contributions

• Established a unified, multi-dimensional (ecosystem-canopy-leaf) and multi-temporal (diurnal-seasonal-interannual) framework to quantify cropland WUE using eddy covariance observations across contrasting hydrological years.
• Addressed a critical knowledge gap by explicitly linking cross-scale WUE dynamics to precipitation variability, revealing scale-dependent and asymmetric responses to both increased and decreased precipitation.
• Reconciled previously divergent evidence across spatial and temporal scales regarding WUE dynamics.
• Advanced a mechanistic understanding of carbon-water coupling under hydroclimatic variability.
• Provided broadly transferable insights into the functioning and resilience of rain-fed agroecosystems in a changing climate, suggesting that intensified precipitation anomalies may increase water consumption for crop production.

Funding

• National Natural Science Foundation of China (42322709, U24A20628, 42577358)
• Jiangsu Province (BK20230050)
• Chinese Academy of Sciences Project for Young Scientists in Basic Research (YSBR-089)

Citation

@article{Li2026Scaledependent,
  author = {Li, Junjie and Dong, YanHong and Liu, Deyan and Wu, Xian and Zhu, Tingting and Huiqin, 王慧琴 Wang and Liu, Yi and WeiXin, Ding and Yuan, Junji},
  title = {Scale-dependent responses of water use efficiency to precipitation variability in a rain-fed maize field of Northeast China},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110814},
  url = {https://doi.org/10.1016/j.agwat.2026.110814}
}