LAUSR

Abstract Fruit production is relevant to the European agricultural sector. However, orchards in semi-arid areas of southern Europe may contain soils with constraints for tree development. This is the case of soils with high CaCO3 content or limiting layers at variable depth. To assess spatial and in-depth variation of these soil constraints, an apparent electrical conductivity (ECa) survey was conducted in an orchard by using a galvanic contact soil sensor (Veris 3100). Different soil properties were randomly sampled at two depths (topsoil and subsoil) in 20 different sampling points within the plot. ECa raster maps were obtained for shallow (0–30 cm) and deep (0–90 cm) soil profile depths. In addition, an inversion modelling software was used to obtain horizontal ECa slices corresponding to 10 cm thick soil layers from 0–10 cm to 80–90 cm in depth. Concordance analysis of ECa slices allowed the soil profile to be segmented into four homogeneous horizons with different spatial conductivity pattern. Then, a multivariate analysis of variance (MANOVA) was key, i) to better interpret the specific soil properties that mainly contributed to the spatial variation of ECa (CaCO3 and organic matter (OM) contents), and ii) to delimit the soil layer and the specific spatial pattern of ECa that allows potential management areas to be delineated by presenting the same trend in CaCO3 and OM for topsoil and subsoil simultaneously. Moreover, assessing 3D variation of ECa made it possible to identify different soil areas that, linked to previous earthworks to optimize the parcelling of the farm, are the main cause of spatial variability within the orchard.