Application of multivariate statistics in the source identification of heavy-metal pollution in roadside soils of Bursa, Turkey

Bursa located within the western part of Turkey is a rapidly growing and heavily industrialized city, hence a considerable amount of pollutions reaching intolerable levels around the city caused especially by anthropogenic activities. This study aims at finding out causes, intensity, and extent of pollution around the D-90 highway and O-33 motorway. Soil sampling was invoked as the first method to identify and discuss the nature of the pollution. A total of 91 soil samples from 10 different locations were collected away from the roadsides. The samples were analyzed for As, Cd, Cr, Co, Cu, Hg, Mn, Ni, Pb, and Zn. The analytical results were evaluated to understand chemical variability of soil compositions using several multivariate statistical methods such as correlation analysis, principal component analysis (PCA), and cluster analysis (CA). Cluster analysis and PCA clearly reveal three distinct groups of elements as (a) Pb, As, and Cd; (b) Cu, Zn, and Hg; and (c) Ni, Cr, Co, and Mn. Two of these groups (Cd, Pb, and As and Cu, Zn, and Hg) are originated from anthropogenic sources (traffic and industrial activities). Ni, Co, Cr, and Mn elements are reflecting underlying lithology. Soil pollution was investigated with the use of enrichment factor (EF), pollution index (PI), and integrated pollution index (IPI) values. The results consistent with the multivariate statistical evaluations demonstrate that soils along the Bursa highway are seriously polluted by toxic trace elements. Lead, As, Cd, Zn, Hg, and Cu are ascertained to be of anthropogenic origin and derived from traffic and industrial activities.