Searching for heavy metals grouping roadside soils as a function of motorized traffic influence

Ninety two soil samples were collected in four sampling seasons in La Coruna, NW of Spain, and its surroundings to study the impact of vehicular traffic onto the metallic content of roadside soils in a medium-size city. Samples exhibited different levels of exposure and, therefore, sample groups should arise as a function of this anthropogenic impact. Surprisingly, this was not so when all the nine metals initially considered were subjected to multivariate analyses. Then, different unsupervised multivariate statistical techniques were used to discover those elements able to group the samples according to the level of road traffic. Such an objective was achieved only after removing the natural variability. Accordingly, Pb, Cd, Cu and Zn were found to satisfactorily conduct to this end, Pb being the most discriminant one. (C) 1998 Elsevier Science B.V. All rights reserved.