Determining the Optimal Water Storage Depth of Apple Orchards in the Chinese Loess Plateau: Balancing Soil Desiccation Mitigation, Plant Utilization, and Engineering Construction

ABSTRACT

Deep‐rooted plantations are central to dryland ecological restoration, yet their long‐term cultivation drives widespread deep soil desiccation, threatening ecosystem sustainability. While artificial water replenishment is a key deep soil desiccation mitigation strategy, the depth‐dependent positional effect of stored water on eco‐hydrological processes remains poorly quantified, with no consensus on the optimal replenishment threshold. Here, we used a calibrated and validated HYDRUS‐1D model to simulate soil water dynamics, evapotranspiration partitioning, and deep percolation across eight storage depths in mature apple orchards (a typical deep‐rooted system) on the Chinese Loess Plateau, under extreme dry, normal, and extreme wet precipitation levels. Results revealed consistent depth‐dependent patterns across all scenarios, defining critical thresholds: soil water storage increased with depth, while depths above 5 m were insufficient for sustained water retention. Cumulative transpiration ( T ) exhibited three stages (slow decrease at < 3 m, sharp decline at 3–5 m, plateau at > 5 m), matching root water uptake effectiveness. Cumulative evaporation ( E ) dropped sharply at < 3 m then stabilized, indicating evaporation's maximum influence depth. Evapotranspiration ( ET ) initially declined steeply with depth before leveling off, whereas the T / ET ratio increased above 3 m and decreased at greater depths. Deep percolation remained negligible (< 1 mm year ⁻¹ ) at all depths. A multi‐criteria evaluation identified 5 m as the optimal eco‐hydrological threshold. This study refines soil water reservoir theory and provides transferable guidance for deep soil desiccation restoration in global drylands.