ERT-based root water uptake quantification in field-grown wheat under terminal drought

Background and Aims Drought reduces wheat yields, yet field-scale quantification of root water uptake (RWU) remains challenging because below-ground processes are difficult to monitor. This study developed a non-invasive hydrogeophysical framework integrating Electrical Resistivity Tomography (ERT), TDR-based soil monitoring, and depth-aware Random Forest calibration to quantify depth-resolved RWU and evaluate genotype-specific water-use strategies under terminal drought. Methods Time-lapse ERT (44 surveys, ≥ 3 week⁻ 1 ) was combined with TDR sensor measurements of soil water content (n = 278 paired ρ–θ observations) to convert resistivity measurements into depth-resolved RWU estimates across 0.1–1.0 m depth. Five petrophysical models were evaluated using date-grouped fivefold cross-validation, with the depth-aware Random Forest performing best. Three wheat genotypes with contrasting root architectures were monitored under terminal drought (142 mm available water). ERT-derived RWU were analysed alongside stomatal conductance, chlorophyll fluorescence, and grain yield. Results ERT resolved RWU strategies among genotypes. WM-203 exhibited aggressive, coordinated multi-layer water extraction across the soil profile (r = 0.80–0.98), whereas WM-140 showed a delayed uptake strategy characterized by early deep-layer dominance followed by mid- and deep-profile engagement, and IPLR-760 displayed inconsistent uptake with mid-profile hydraulic decoupling. Genotypic RWU rankings were consistent with stomatal conductance and grain yield, spanning from 7.0 t ha⁻ 1 in WM-203 to 1.5 t ha⁻ 1 in IPLR-760 despite comparable total water extraction. Conclusion ERT-based quantification of RWU provides a robust, non-invasive approach for resolving genotype-specific water-use strategies under field conditions. The framework enables characterization of water-use coordination patterns and offers a tool for phenotyping drought-resilient wheat genotypes.