Application of multivariate statistical methods for evaluating soil quality indices in Lordegan semiarid region

Soil quality is affected by the type of land use and management. Evaluation of soil quality indices is a suitable guide for adopting sustainable management practices and preventing soil degradation. The goals of this research were to explain soil variations and to determine the most important soil quality assessment indices. Therefore, totally 50 soil samples were collected from four land uses including forest, pasture, rain fed and irrigated farmlands in Lordegan watershed. Then soil characteristics consisted of soil bulk density, soil texture components, calcium carbonate equivalent, cation exchange capacity, pH, electrical conductivity, organic carbon, total nitrogen, carbon to nitrogen ratio, microbial basal respiration and mean weight diameter of aggregates were measured. Also, average monthly temperature, annual precipitation, elevation and slope of sampling points were determined. The results of principal component analysis illustrated that organic carbon, sand content, annual precipitation and elevation were the most important factors causing soil variations in the area. This method could explain 81% of soil variations in the region by creating 5 principal components. Furthermore, various kind of land uses were distinguished successfully according to a set of selected soil attributes using Discriminant Analysis. The percentage of correct classification in this analysis was 94%. The results of this study emphasize on the applicability of multivariate statistical methods in the assessment of soil quality. The individual examination for each of selected indices indicated the high quality of forest soils compared to the other land uses. However, rain fed farmlands had a poor quality situation. Therefore, achieving sustainable management in the area requires the prevention of land use change and the restoration of degraded lands for improving soil quality.