TL;DR
This study uses the SOILWAT2 process-based model to simulate evapotranspiration (ET) partitioning in a rainfed maize farmland in Northeast China over 30 years. The model accurately predicts ET and its components, revealing that drought years lead to higher transpiration (T) and lower soil evaporation (Es), suggesting drought limits maize yield by increasing water loss through T.
What Problem Does This Paper Address?
The Northeast China maize belt is a critical global food production region, but it's increasingly vulnerable to climate change, particularly drought. Understanding how water is lost from these farmlands through evapotranspiration (ET) and its components (transpiration from plants, soil evaporation, and canopy interception evaporation) is crucial for predicting crop yields and developing effective adaptation strategies. While ET is a major driver of water transfer, its partitioning is complex and not well-understood, especially under changing climate conditions. This paper aims to fill this knowledge gap by quantifying ET partitioning over a long period and exploring its relationship with drought and crop growth stages.
Methods
The researchers employed the SOILWAT2 model, a process-based ecosystem water budget model. This model simulates daily water movement within the soil profile, including precipitation, infiltration, deep drainage, and ET components. To validate the model's performance, they compared its simulated daily ET (ETsw) with ET data measured by eddy covariance (ETEC) from 2016-2018. The model was then used to simulate ET partitioning (transpiration (T), soil evaporation (Es), and canopy interception evaporation (Int)) from 1989 to 2018. Drought and humid years were identified using the standardized rainfall index (SPI). They also analyzed the seasonal and growth-stage variations of ET partitioning and investigated the relationships between ET partitioning and leaf area index (LAI) and above-ground biomass using regression analysis.
Datasets
- Meteorological Data: Daily precipitation, daily maximum and minimum air temperatures, mean monthly relative humidity, mean monthly wind speed, and mean monthly cloud cover from the Jinzhou meteorological station (1989-2018).
- Vegetation Data: Monthly above-ground plant biomass (total and live), and above-ground litter biomass (May-September, 2005-2018).
- Soil Data: Soil properties for eight layers, including texture (sand, silt, clay percentages), bulk density, and gravel content.
- Eddy Covariance Data: Measured ET (ETEC) for validation purposes (May-September, 2016-2018).
- Leaf Area Index (LAI) and Above-Ground Biomass (AGB): Measured during maize growth stages (2005-2018) for relationship analysis.
Key Insights and Ideas
- Drought Shifts Water Loss: The most striking finding is that during drought years, maize farmlands lose more water through transpiration (T) and less through soil evaporation (Es). This suggests that under drought stress, plants may increase their water uptake and loss, potentially limiting yield.
- Transpiration Dominates in Drought: While soil evaporation is generally higher in humid years, transpiration becomes the dominant water loss pathway in drought years, highlighting the plant's struggle to cope with water scarcity.
- Growth Stage Matters: ET partitioning isn't static; it changes significantly throughout the maize growing season. Transpiration is highest during the early growth stages (greening-jointing) and decreases as the plant matures, while soil evaporation follows a more complex pattern, being smallest during the critical greening-jointing stage.
- Biomass is a Stronger Predictor: Both LAI and above-ground biomass influence ET partitioning, but biomass appears to have a stronger effect. As biomass increases, transpiration tends to increase, and soil evaporation decreases.
- Model Reliability: The SOILWAT2 model proved to be a reliable tool for simulating ET and its components in this rainfed maize system, showing good agreement with eddy covariance measurements.
Key Results
- Mean Annual ET: 432.3 mm over 30 years.
- Mean Annual Partitioning: Transpiration (T) = 197.6 mm (45.9% of ET), Soil Evaporation (Es) = 204.7 mm (47.1% of ET), Canopy Interception Evaporation (Int) = 19.2 mm (4.4% of ET).
- Drought vs. Humid Years: Es/ET was 6.3% lower in drought years than humid years, while T/ET was 6.2% higher in drought years. Int/ET showed no significant difference.
- Growing Season Partitioning: T/ET ranged from 40.0% to 75.0%, Es/ET from 22.8% to 55.7%, and Int/ET from 0.7% to 7.0%.
- Maize Growth Stages: T/ET was greatest at the greening-jointing stage (53.1%) and smallest at the milking-harvest stage (49.7%). Es/ET was smallest at the greening-jointing stage (42.3%) and largest at the milking-harvest stage (46.8%).
- LAI/Biomass Relationships: Power function relationships were found between T/ET and LAI/biomass (R² = 0.65-0.78) and Es/ET and LAI/biomass (R² = 0.59-0.67).
Limitations and Future Work
- Model vs. Measurement: While the SOILWAT2 model showed good agreement with eddy covariance data (R²=0.64), it did underestimate ET compared to measurements. The authors note that eddy covariance itself has inherent uncertainties.
- LAI and Biomass Data: The LAI and biomass data were collected for specific growth stages and averaged, which might not capture the full dynamic range of these variables throughout the entire growing season.
- Future Work: The study suggests that securing food yield in the face of drought will require increased water-use efficiency and adaptive strategies like drought-resistant varieties and irrigation. Further research could explore the effectiveness of these strategies in detail. Investigating the impact of different management practices on ET partitioning would also be valuable.
Summary updated