Mapping of anthropogenic trace elements inputs in agricultural topsoil from Northern France using enrichment factors

• DOI: 10.1016/j.geoderma.2010.04.009
• Publication Date: 2010-05-24T08:29:03Z
• Authors: Hocine Bourennane; Francis Douay; Thibault Sterckeman; Estelle Villanneau; Henri Ciesielski; Dominique King; Denis Baize
• Keywords: [SDV.SA]Life Sciences [q-bio]/Agricultural sciences; [SDV.SA] Life Sciences [q-bio]/Agricultural sciences; 550; accuracy; sequential indicator simulation; uncertainty assessment; trace element; enrichment factor; 04 agricultural and veterinary sciences; 15. Life on land; 01 natural sciences; stochastic simulation; forest soils; arable soils; contamination; 13. Climate action; 0401 agriculture, forestry, and fisheries; precision; spatial-patterns; soil-water content; 0105 earth and related environmental sciences

Abstract Anthropogenic trace element contamination of soils in the densely populated and industrialized Nord-Pas de Calais region (France) was quantified using enrichment factors (EFs). Total concentrations of Al and 18 trace elements (As, Bi, Cd, Co, Cr, Cu, Hg, In, Mn, Mo, Ni, Pb, Sb, Se, Sn, Tl, V, and Zn) were determined in 252 sampling sites both in topsoil and deep horizons of soils developed from sedimentary materials. Thus, at each sampling site an enrichment factor was computed for each of the 18 trace elements as the ratio between total concentrations of the trace element in topsoil and deep horizons in the same soil pit, normalized by concentrations of Al measured in the same horizons and the same pit. To estimate EF values at unsampled locations and then to assess the spatial magnitude of anthropogenic trace element contamination over the whole study area (12 400 km2), variography analysis was used to quantify the spatial structure of each EF. Results have shown that values of 8 EFs are spatially correlated. These EFs were thus mapped using the sequential Gaussian simulation technique. One hundred horizon samples from 50 supplementary sites were analysed for the structured trace elements and Al contents and used for the mapping validation. Results of the validation based on the accuracy plots and goodness statistics (G and D) allowed us to conclude that EF maps exhibited high certainty in regard to the validation data set. In addition, results have shown that E-type estimates were accurate when they are used to estimate the measured EF value of the validation data set. Accordingly, these maps were used to discuss the origin of patches of surface horizon enrichment displaying an exogenous input of the trace elements investigated. The whole study area under discussion is strongly enriched with cadmium, originating from both industrial and agricultural activities. Surface horizons also exhibit locally high levels of Pb and Zn enrichment, mainly around well-known former or current industrial sites. However, Pb and Zn enriched surface horizons are far less widespread than those enriched in Cd. Higher Cu, Bi and Sn EF values were observed in the northern part of the region studied than in the southern part. The dissimilarity underlined a rather diffuse contamination from the urbanized zone in the north to the rural zone in the south of the region. The enrichment of surface horizons by In and Tl seems insignificant. Some unstructured EFs were attributes to point-source enrichments.