Assessment of physicochemical properties among petroleum contaminated and normal soil: A multivariate statistical analysis

Soil surveying and mapping are vital operations in soil science. They contribute to the evaluation of changes in soil quality between various geographical regions. This study intended to investigate the effect of geographic variability on soil properties gathered from different geographical areas of the Rajasthan State of north-western India, using conventional analytical methods. Normal and petroleum-contaminated soils were collected from four diversified regions of Rajasthan: namely, hyper saline wetland (Sambhar), arid desert region (Jaisalmer), semi-arid region (Ganganagar) and humid irrigated plains (Banswara). The correlations between normal and contaminated soil for these four stations were investigated using multivariate analyses, principal component analysis (PCA) and statistical quality control. The one-way ANOVA test is applied to determine the statistical significance between different sampling stations for both normal and contaminated soil. Since the p-value in both cases is less than 0.05, it reflects a significant difference in physical properties among all four stations. The PCA analysis suggests that the first three components have Eigen values greater than one that can describe 100% of the total variability in soil properties. Heavy metals (Lead, Nickel and Copper) are the primary measures of variability between sample points for contaminated soil. In contrast, physicochemical properties (pH, Conductivity and Sodium) are the primary measures of variability between sample sites for normal soil.