Investigating Multi-Scale Soil Moisture Dynamics and Hydrological Processes in Silvoarable Agroforestry in Hessen, Germany

Agroforestry represents one of the oldest and most sustainable forms of land use, integrating woody perennials with crops and grasslands to enhance resource efficiency while meeting human requirements for food, timber, and fiber. While its potential to conserve natural resources is well-recognized, a full understanding of the interactions between system components remains limited, particularly regarding the hydrological processes that regulate plant growth, nutrient dynamics, and energy exchange. This study presents ongoing research at a silvoarable experimental site in Gladbacherhof, Hesse, Germany, designed to quantify soil moisture dynamics across multiple spatial scales.

To characterize these dynamics at a high resolution, volumetric soil moisture is monitored using sensors installed along three transects oriented perpendicular to apple tree rows at specific distances of 1, 2.5, 6, and 10.5 meters from the trees and at soil depths of 10, 40, and 60 centimeters. These point-scale observations are complemented by field-scale variability captured through cosmic-ray neutron sensing (CRNS). To bridge these disparate scales, the study employs machine learning approaches—including random forest models, multilayer perceptron neural networks, and deep learning techniques—to derive spatially continuous representations of soil moisture at an intermediate scale.

Furthermore, the acquisition of high-temporal-resolution data allows for the investigation of hydraulic lift, which is inferred from observed nocturnal increases in soil moisture during prolonged dry periods. Building on these findings, a subsequent phase of the experiment will apply water stable isotope techniques to track the specific spatio-temporal patterns of water uptake by trees, grassland, and arable plants. As an ongoing study, this research aims to clarify the key factors and spatial controls governing soil moisture dynamics, ultimately supporting more informed design and management of resilient, multifunctional agroforestry systems.