Assessing soil fertilization effects using time-lapse electromagnetic induction

Abstract

Abstract. Adding mineral fertilizers and nutrients is a common practice in conventional farming and is fundamental to maintain optimal yield and crop quality; nitrogen is the most applied fertilizer and is often used excessively, leading to adverse environmental impacts. To assist farmers in optimal fertilization and crop management, non-invasive geophysical methods can provide knowledge about the spatial and temporal distribution of nutrients in the soil. In recent years, electromagnetic induction (EMI) has been widely used for field characterization, to delineate soil units and management zones, or to estimate soil properties and states. Additionally, ground-penetrating radar (GPR) and electrical resistivity tomography (ERT) have been used in local studies to measure changes in soil properties. Unfortunately, the measured geophysical signals are confounded by horizontal and vertical changes in soil conditions and parameters, and the individual contributions of these conditions and parameters are not easy to disentangle. Within fields, and also between fields, fertilization management might vary in space and time, and, therefore, the differences in pore fluid conductivity caused directly by fertilization or indirectly by different crop performance make the interpretation of large-scale geophysical surveys over field borders complicated. To study the direct effect of mineral fertilization on the soil electrical conductivity, a field experiment was performed on 21 bare-soil plots with seven different fertilization treatments. As fertilizers, calcium ammonium nitrate (CAN) and potassium chloride (KCl) were chosen and applied in three dosages. Soil water content, soil temperature, and bulk electrical conductivity were recorded continuously over 450 d. Additionally, 20 EMI, 7 GPR, and 9 ERT surveys were performed, and on days of ERT measurements, soil samples for nitrate and reference soil electrical conductivity measurements were taken. The results showed that (1) the commonly used CAN application dosage did not impact the geophysical signals significantly. (2) EMI and ERT were able to trace back the temporal changes in nitrate concentrations in the soil profile over more than 1 year. (3) Both techniques were not able to trace the nitrate concentrations in the very shallow soil layer of 0–10 cm, irrespective of the low impact of fertilization on the geophysical signal. (4) The results indicated that past fertilization practices cannot be neglected in EMI studies, especially if surveys are performed over large areas with different fertilization practices or on crops grown with different fertilizer demands or uptakes.