Electrical resistivity tomography as a non-destructive method for mapping root biomass in an orchard

Abstract

Multi-electrode soil electrical resistivity (p) tomography was used for the non-invasive study of tree roots in situ and their spatial distribution in an agricultural soil. The quantitative relations of p and root biometry and the contribution of different root size classes were investigated with two- and three-dimensional 48-electrode tomograms in an orchard in southern Italy on a Typic haploxeralf fine, mixed termic soil. Root biomass density (RD) and root length density (RLD) were measured destructively on coarse (>2 mm diameter) and fine roots, and soil paste electrical conductivity, water content, stone content, texture, organic matter and pH were measured on soil samples taken up to 0.48-m deep. Areas of large ρ values (up to 460 ohm m) were found close to tree trunks and variability in ρ was related to RD (0―0.137 Mg m ―3 ) only; the resistive response was from coarse roots. The effect of other soil variables on ρ was overshadowed by the presence of roots and therefore no significant multivariate relationship was found. A highly significant ρ-RD gamma GLM model used to fit positively skewed data provides a useful framework for regression analysis when p is dominated by roots. Soil electrical resistivity is promising as a proxy for RD in orchards, but not for RLD, and the effect of tree roots on p needs to be taken into account in electrical surveys of soils.