Concurrent variability of soil moisture and apparent electrical conductivity in the proximity of olive trees

Abstract

Detailed knowledge of soil moisture (θ) is crucial for implementing appropriate soil and crop management decisions and assuring crop productivity and soil functioning. θ measurement is challenging in the stony soils of traditional olive orchards where the operationality of common sensors is often compromised and fine scale variability of soil properties is complex as a result of the presence and specific architecture of the tree root system and canopy. In such environments, non-invasive electromagnetic induction (EMI) sensors constitute an alternative for measuring θ. Here we evaluate this alternative by exploring relationships between θ and apparent electrical conductivity (ECa) and contribute to improved understanding of spatio-temporal θ variability in the proximity of olive trees. Continuously measured θ in a trench on one side of an olive tree was complemented with 47 ECa surveys conducted periodically at the opposite side of the tree during a period of 15 months. θ was measured in the trench at five distances from the trunk (0.6–3.8 m) and at five depths (0.1–0.8 m). ECa was measured at the same distances from the trunk. We observed similar temporal patterns of θ and ECa in response to precipitation and evaporation at the site. The proportionality between spatial means of θ and ECa and the similarity in the relationships between spatial mean and standard deviation for each variable revealed concurrent spatio-temporal patterns. Dependence on distance to the trunk was different for θ and ECa, with an apparent erratic response for the former and a positive relationship for the latter. The slope of the linear θ-ECa models showed power law decay with distance to the trunk as a result of the varying soil and tree properties around the tree. These results pave the way for implementing coupled static-mobile ECa monitoring systems and θ estimation in olive orchards from the plot to the field and farm scales.