The variability of standard artificial soils: cadmium and phenanthrene sorption measured by a batch equilibrium method.

Artificial soil (AS) is used in soil ecotoxicology as a test medium or reference matrix. AS is prepared according to standard OECD/ISO protocols and components of local sources are usually used by laboratories. This may result in significant inter-laboratory variations in AS properties and, consequently, in the fate and bioavailability of tested chemicals. In order to reveal the extent and sources of variations, the batch equilibrium method was applied to measure the sorption of 2 model compounds (phenanthrene and cadmium) to 21 artificial soils from different laboratories. The distribution coefficients (K) of phenanthrene and cadmium varied over one order of magnitude: from 5.3 to 61.5L/kg for phenanthrene and from 17.9 to 190L/kg for cadmium. Variations in phenanthrene sorption could not be reliably explained by measured soil properties; not even by the total organic carbon (TOC) content which was expected. Cadmium logK values significantly correlated with cation exchange capacity (CEC), pH and pH, with Pearson correlation coefficients of 0.62, 0.80, and 0.79, respectively. CEC and pH together were able to explain 72% of cadmium logK variability in the following model: logK=0.29pH+0.0032 CEC -0.53. Similarly, 66% of cadmium logK variability could be explained by CEC and pH in the model: logKd=0.27pH+0.0028 CEC -0.23. Variable cadmium sorption in differing ASs could be partially treated with these models. However, considering the unpredictable variability of phenanthrene sorption, a more reliable solution for reducing the variability of ASs from different laboratories would be better harmonization of AS preparation and composition.