Investigating the soil-plant continuum of maize crops using ground penetrating radar

The soil-plant continuum of agricultural crops is regulating key processes that affect plant performance and agricultural productivity. As climate change impacts agricultural systems, understanding these processes will become increasingly important, especially when increasing yield productivity, while minimizing the environmental footprint are key aspects. Quantifying the impact of climate change and management practices on crop growth requires understanding about the dynamics of the root systems of crops. Ground penetrating radar (GPR) combined with root imaging and modeling techniques offers a unique opportunity to study these dynamics in function of soil, climate and management. As a first step, this study examined the relationship between root development and soil dielectric permittivity variability using root images and 200 MHz time-lapse horizontal crosshole GPR at two field minirhizotron (MR) facilities in Selhausen, Germany. The data was acquired over three maize growing seasons, in 7-m long rhizotubes at six different depths, ranging between 0.1 m - 1.2 m and for three different plots with varying agricultural treatments. We calculated trend-corrected spatial permittivity deviations to isolate root-related effects by removing static and dynamic influences caused by soil heterogeneity and changing weather conditions. This permittivity deviation increased during the growing season, correlating with root presence. Cross-correlation analysis between permittivity variability and root volume fraction yielded in coefficients of determination above 0.5 for half of the data pairs. From this study some questions remained unanswered, such as identifying individual roots or quantifying the influence of roots and above-ground shoot on the GPR signal. Subsequently, synthetic forward modeling was conducted using the data acquisition of the previous study as a template and the open-source electromagnetic simulation software gprMax. GPR traces were modeled and analyzed for scenarios with varying soil-plant continuum compositions, including soil, roots, and above-ground shoots in two- or three dimensions. The models incorporated realistic root contributions based on trench wall counts. We found that the presence of roots, which resulted in a permittivity increase on one hand, had a higher influence on the GPR signal than the above-ground shoot and on the other hand the roots affected the first arrival time and amplitudes of the GPR signal. Hence more sophisticated analysis techniques such as full-waveform inversion are necessary. Furthermore, we introduced an approach to derive the soil water content within the soil-plant continuum, where the CRIM petrophysical model was extended with the root phase. This showed that neglecting the root phase leads to overestimation of soil water contents.