Chemical characterization and spatial distribution of PAHs and heavy hydrocarbons in rural sites of Campania Region, South Italy

In this paper, the behaviour and distribution patterns of heavy hydrocarbons and several polycyclic aromatic hydrocarbon (PAH) priority pollutants, as listed by the US Environmental Protection Agency, were evaluated in 891 soil samples. The samples were collected in three expected polluted rural sites in Campania (southern Italy) as part of the LIFE11 ECOREMED project, funded by the European Commission, to test innovative agriculture-based soil restoration techniques. These sites have been selected because they have been used for the temporary storage of urban and building waste (Teverola), subject to illicit dumping of unknown material (Trentola-Ducenta), or suspected to be polluted by metals due to agricultural practices (Giugliano). Chemical analysis of soil samples allowed the baseline pollution levels to be determined prior to any intervention. It was found that these areas can be considered contaminated for residential use, in accordance with Italian environmental law (Law Decree 152/2006). Statistical analysis applied to the data proved that average mean concentrations of heavy hydrocarbons could be as high as 140 mg/kg of dry soil with peaks of 700 mg/kg of dry soil, for the Trentola-Ducenta site; the median concentration of analytical results for hydrocarbon (HC) concentration for the Trentola-Ducenta and Giugliano sites was 63 and 73.4 mg/kg dry soil, respectively; for Teverola, the median level was 35 mg/kg dry soil. Some PAHs (usually benzo(a)pyrene) also exceeded the maximum allowed level in all sites. From the principal component analysis applied to PAH concentrations, it emerged that pollutants can be supposed to derive from a single source for the three sites. Diagnostic ratios calculated to determine possible PAH sources suggest petroleum combustion or disposal practice. Our sampling protocol also showed large dishomogeneity in soil pollutant spatial distribution, even at a scale as small as 3.3 m, indicating that variability could emerge at very short spatial scales.