Demir et al. (2026) Flexible water uptake across soil depths buffers transpiration in winter cereals during dry spells

Identification

• Journal: Agricultural and Forest Meteorology
• Year: 2026
• Date: 2026-10-10
• Authors: Gökben Demir, Christian Markwitz, David Dubbert, Anas Emad, Alexander Knohl, Maren Dubbert
• DOI: 10.1016/j.agrformet.2026.111515

Research Groups

• Ecophysiology of Water and Matter Cycling (ECO), ZALF, Müncheberg, Germany
• Bioclimatology, University of Göttingen, Göttingen, Germany

Short Summary

This study investigated root water uptake patterns of winter cereals (wheat, barley) across two contrasting growing seasons in a temperate cropland, revealing that water uptake depth is a conditional, reversible, and dynamic response to topsoil drying, rather than a systematic deepening, which helps buffer transpiration and maintain water-use efficiency during dry spells.

Objective

• How do timing and duration of dry spells, relative to crop phenology, modify water uptake patterns, and the median water uptake depth of winter wheat and winter barley?
• How changes in water uptake depth related to evapotranspiration demand, soil water availability and water-use efficiency during dry spells, and to what extent atmospheric demand and soil-plant interactions can explain seasonal variation in the median uptake depth?

Study Configuration

• Spatial Scale: A temperate cropland site (Reinshof experimental farm, 51.494°N, 9.932°E) in Central Germany, within the footprint of an eddy covariance tower. Soil moisture was monitored at three profiles, at ten depths between 5 cm and 100 cm. Soil and plant samples were collected from five random locations.
• Temporal Scale: Two contrasting growing seasons: winter wheat (15 March 2024 – 21 July 2024) and winter barley (15 March 2025 – 13 June 2025). Continuous eddy covariance and soil moisture measurements (30-minute and 5-minute resolution, respectively) were combined with weekly stable water isotope sampling.

Methodology and Data

• Models used:
  • Craig–Gordon model (for estimating xylem water isotope composition)
  • Multistep, multilayer regression (MSML) (for evapotranspiration and root water uptake from soil moisture)
  • Bayesian isotope mixing model (MixSIAR package in R) (for identifying transpiration water sources)
  • Priestley-Taylor equation (for potential evapotranspiration, PET)
  • Linear mixed-effects model (lme4 package in R) (for identifying drivers of water uptake strategies)
• Data sources:
  • Eddy covariance (EC) flux tower (DE-Rns, FLUXNET) for CO₂, momentum, latent and sensible heat fluxes, evapotranspiration (ETflux-tower), gross primary productivity (GPP), and intrinsic water-use efficiency (iWUE).
  • Meteorological variables: radiation components, air temperature, precipitation, relative humidity, air pressure, wind speed and direction (30-minute intervals).
  • Volumetric soil water content and soil temperature sensors (SMT100) at ten depths (5–100 cm) (5-minute resolution).
  • Stable water isotope analysis (δ²H and δ¹⁸O): weekly samples of plant leaves, precipitation, soil water (surface to 55 cm), and groundwater (110–150 cm), measured using cavity ring-down spectroscopy.
  • Crop phenology (BBCH stages) and Leaf Area Index (LAI).

Main Results

• In 2024 (winter wheat), more frequent rainfall maintained wetter soil layers and higher evapotranspiration (ET). Winter wheat systematically relied on deeper soil water sources as the topsoil dried, coinciding with higher LAI, with the median uptake depth (Z50) shifting from <20 cm to ~40–60 cm.
• In 2025 (winter barley), longer dry spells led to consistently lower ET. Winter barley predominantly met transpiration demand from 20–40 cm soil depth, and dry spells did not necessarily deepen the median uptake depth (Z50).
• Both crops dynamically shifted their water uptake depth in response to the interaction between dry-spell timing and duration, soil water availability, and phenology; shallow uptake increased rapidly after rainfall.
• Water uptake depth in rainfed cereals is a conditional, reversible, and dynamic response to topsoil drying rather than a systematic deepening.
• Seasonal average intrinsic water-use efficiency (iWUE) remained similar across both years, but with greater variability in 2025.
• A linear mixed-effects model (R² = 0.42) indicated that soil moisture and soil temperature at corresponding depths were the strongest predictors of proportional water uptake depth. Interactions between crop type, growth stage, LAI, and soil moisture were also significant, while direct atmospheric forcing (e.g., wetness index, vapor pressure deficit) was not a primary individual driver.

Contributions

• First study to combine continuous eddy covariance and soil moisture measurements with high-resolution stable water isotope sampling to link temporal changes in water uptake depth with evapotranspiration dynamics in an open cropland in a humid-temperate climate.
• Demonstrated that water uptake depth in rainfed cereals is a conditional, reversible, and dynamic response to topsoil drying and rewetting, rather than a systematic monotonic deepening.
• Provided novel insights into how the timing and duration of dry spells, relative to crop phenology, interact with soil water availability and plant traits to govern water uptake strategies and maintain ecosystem water and carbon fluxes.
• Highlighted the value of a multi-method monitoring framework for resolving event- to seasonal-scale variability in transpiration signals and assessing how water uptake depth shifts among different sources in response to changing rainfall patterns.

Funding

• Leibniz SAW collaborative Excellence fund (project number: K444/2022)
• Deutsche Forschungsgemeinschaft (INST 186/1360-1 FUGG, project number 437452730)
• Projekt DEAL (Open access funding)

Citation

@article{Demir2026Flexible,
  author = {Demir, Gökben and Markwitz, Christian and Dubbert, David and Emad, Anas and Knohl, Alexander and Dubbert, Maren},
  title = {Flexible water uptake across soil depths buffers transpiration in winter cereals during dry spells},
  journal = {Agricultural and Forest Meteorology},
  year = {2026},
  doi = {10.1016/j.agrformet.2026.111515},
  url = {https://doi.org/10.1016/j.agrformet.2026.111515}
}