Site-specific natural background concentrations of metals in topsoil from the Trail region, British Columbia, Canada

The natural background level of elements in the surface soil of a given area is largely dependent on the geological setting of the region and the underlying soil parent material. Due to the heterogeneity of soil in various environments, it is essential to determine the site-specific natural background levels of elements to quantify the degree of influence of the anthropogenic source(s) within a given study area. This paper discusses the process involved in establishing the natural background level of trace metals and other elements in topsoil for the area surrounding the Teck Cominco zinc–lead smelter in Trail, British Columbia, Canada. The environmental assessment of soil in the Trail region is difficult due to the natural enrichment of metals in surficial material and topsoil as a result of the influence of mineralized bedrock in the region. Detailed examination of the study area based on the deposition data and previously collected soil data led to the selection of presumably the least anthropogenically influenced background sites in the study area, which also reflect the geological diversity of the region. The range of background was defined using the median plus or minus two median absolute deviations [median±(2 × MAD)], after the outliers were excluded. The outliers were identified using the Tukey boxplot concept and the cause of high metal concentrations in the outlier samples is attributed to their proximity to the historical mining sites in the region. The upper limit of the background range was estimated to be (in mg/kg): As, 11.0; Cd, 0.81; Cu, 38.1; Hg, 0.07; Pb, 27.7; Zn, 152. The resulting values are significantly more conservative estimates than would be obtained by other widely used methods (e.g. the mean plus or minus two standard deviations) due to the robust procedure, which prevents any interference by the extreme values in the collected soil background dataset. The upper background thresholds of metals estimated by the 95th percentile resulted in significantly different estimates, especially for As, Hg, Pb and Zn, from the previously reported values. This highlights the need for revision of the risk-based screening levels by environmental regulations in the study area.