Multiple linear modeling between soil properties, magnetic susceptibility and heavy metals in various land uses

Abstract

Land use and intrinsic soil properties play vital role in variability of heavy metals and magnetic susceptibility within a landscape. This study was conducted to use multiple linear regressions modeling to explore the variability of magnetic susceptibility and heavy metals within five different land uses and to find the interaction of these variables with intrinsic soil properties in Ahvaz district, Khuzestan Province, Southern Iran. Five land uses included urban area, agricultural land, steel company, Zargan Power Plant, and Ramin Power Plant. Twenty soil samples from each land use and totally one hundred samples from 0 to 5 cm topsoil were randomly collected. Some chemical and physical soil properties, magnetic susceptibility as well as concentration of some heavy metals were measured. The results showed that significant correlations were found between heavy metals and magnetic susceptibility at low frequency (χlf), showing high capability of magnetic susceptibility for prediction of heavy metals. Calcium carbonate equivalent and soil organic matter (SOM) showed negative correlation with χlf, while SOM and clay content had substantial influences on variability of heavy metals in the studied area. The results of the comparison among land uses showed that Cr, Ni were the highest in agricultural land because of the application of the sewage sludge and fertilizers, while the metals (ZN, Mn, Cu, Co and Fe) and χlf were the highest in steel company due to the industrial activities. Multiple linear regression analysis showed that the combination of soil properties and magnetic susceptibility could explain 94, 62, 69, 53, 81, 77 and 57% of the variability in Fe, Mn, Cu, Zn, Ni, Co and Cr, respectively.