Geostatistical analysis of heavy metals in a one-hectare plot under natural vegetation in a serpentine area

The objective of this study was to examine the spatial variability of selec ted heavy metals in a soil developed over serpentine. Both total and EDTA-e xtractable Fe, Mn, Cr, Ni, Cu, Zn and Co were determined in 53 samples, col lected from the topsoil of a 1-ha forested plot. Naturally occurring soil C r and Ni concentrations were much higher than critical limits for safety. E xperimental semivariograms were computed and modelled by a nugget component plus a structure with autocorrelation ranges varying between 25 and 90 in. EDTA-extractable heavy metal contents exhibit a different spatial variatio n pattern from that of total contents, although Ni and Cu semivariograms pr esent some similarities. The joint spatial variation for pairs of variables with significant correlation was also investigated. The nugget variances i n the cross-semivariograms were not very different from those of individual semivariograms, suggesting heterogeneity within the shortest sampling inte rval. Semivariograms provided a clear description of the spatial structure of heavy metals and some insight into possible processes affecting their di stribution. Kriging maps allowed the identification of small regions with d istinct metal concentrations and confirmed the suitability of geostatistics for investigating processes controlling heavy metal variation. Isotopic co kriging performed better than kriging, but the gain for mapping purposes wa s limited.