O efeito da concentração de solutos na relação umidade do solo–resistividade elétrica em solos tropicais

Abstract

Different agricultural practices such as soil acidity correction, fertilization and fertirrigation or soil contamination with salts and metals alter the concentration of solutes in the soil. Electrical Resistivity Tomography (ERT) has been used to determine the soil electrical resistivity (ρ) and with it estimate the soil water content (θ). In addition to θ, ρ is influenced by variations in solute concentration, temperature, soil texture, soil organic carbon, soil porosity, pH and soil bulk density. The estimation of θ through ρ has been demonstrated in several studies, however, the relationship between θ-ρ has not been studied for soils with different solute concentrations yet, which may compromise the estimation of θ through ρ. The hypothesis of this work is that with the increase of soil solute concentration, the solute concentration starts to have more influence on ρ than θ, making more difficult to monitor θ, causing its overestimation. Thereby, the objectives of this study were: (1) evaluate the influence of solute concentration in the soil solution under the θ-ρ relationship; (2) identify which tool estimates θ more accurately; and (3) identify the variables with greater sensibility to ρ. Samples with preserved structure of three soils - Typic Dystrustept, Rhodic Hapludult and Rhodic Hapludox – were submitted to 5 treatments, 1 was the control treatment and 4 treatments were with increase of electrical conductivity (σ) to 10, 15, 25 and 50 dSm-1 , in 4 repetitions. Initially, the samples were saturated with solutions of 55, 110, 275 and 550g of NaCl L -1 . Subsequently were taken readings of temperature, σ and θ with a capacitance sensor, ρ with a resistimeter using an arrangement of 4 electrodes, and θ through the standard method. The data of θ, ρ and σ were adjusted to mathematical models, like potential, logarithmic, polynomial, exponential and linear, and then evaluated according to R 2 , root mean square error (RMSE) and AIC (Akaike Information Criterion). The potential model was the one that best described the θ-ρ relationship for soils with solute concentrations until 25 dSm-1 . Above this value, the polynomial and logarithmic started to have better performance. The ρ was able to better separate treatments in saturated soil condition than σ, and variables related to solute concentration were the most sensitive to ρ. Still, the concentration of solutes influenced ρ more than θ. Therefore, ρ is able to estimate θ more accurately in soils with different solute concentrations and was able to differentiate soil solute concentrations better than σ , when θ was above 0.4m3m-3.