Soil bulk electrical resistivity and forage ground cover: nonlinear models in an alfalfa (Medicago sativa L.) case study

Alfalfa is a highly productive and fertility-building forage crop; itsperformance, can be highly variable as influenced by within-field soilspatial variability. Characterising the relations between soil and for-age-variation is important for optimal management. The aim of thiswork was to model the relationship between soil electrical resistivity(ER) and plant productivity in an alfalfa (Medicago sativaL.) field inSouthern Italy. ER mapping was accomplished by a multi-depth auto-matic resistivity profiler. Plant productivity was assessed through nor-malised difference vegetation index (NDVI) at 2 dates. A non-linearrelationship between NDVI and deep soil ER was modelled within theframework of generalised additive models. The best model explained70% of the total variability. Soil profiles at six locations selected alonga gradient of ER showed differences related to texture (ranging fromclay to sandy-clay loam), gravel content (0 to 55%) and to the presenceof a petrocalcic horizon. Our results prove that multi-depth ER can beused to localise permanent soil features that drive plant productivity.