Comparison of cadmium extractability from soils by commonly used single extraction protocols

This study compares Cd extractability using 13 commonly used extraction procedures for 28 soils, varying in composition (0–88% sand; 0.9–12.8% organic matter; pH 5.2–8.2) and contamination level (0.2–41.4mg kg− 1 Cd). The procedures used were: soil solution extraction using rhizon soil moisture samplers, 0.01M CaCl2, 0.1M Ca(NO3)2, 0.1M NaNO3, 1M NH4NO3, 1M NH4OAc, 1M MgCl2, 0.11M HOAc, 0.1M HCl, 0.5M HNO3, 0.02M EDTA + 0.5M NH4OAc + 0.5M HOAc (pH 4.65), 0.005M DTPA + 0.01M CaCl2 + 0.1M TEA (pH 7.3) and aqua regia digestion. Single extractions were compared and clustered based on Pearson correlation. Relative Cd extractability, expressed as percentages of pseudo-total content (aqua regia), were compared between soils with pH below and above 7. Finally, multivariate regression functions for Cd extractability as a function of soil parameters were constructed. Distinctive subgroups of related extraction procedures could be distinguished using cluster analysis. A first major class of significantly related procedures consisted out of the weaker extractants and contained Rhizon extraction in addition to NaNO3, CaCl2, NH4NO3, Ca(NO3)2. The second major class contained the pseudo-total content (aqua regia), more aggressive chelate based (DTPA, EDTA) and acid based (HOAc, HCl, HNO3) procedures, and stronger extractions used to estimate exchangeable Cd in the soil (NH4OAC, MgCl2). Differences in extractability between acidic and alkaline soils were observed in the weaker extractants whereas no or little influence of soil pH on the extractability was observed when the more aggressive extractants were used. Cd extractability by the weaker extractants was found to be especially affected by soil pH, total Cd content and cation exchange capacity. Regression curves were constructed based on these three parameters. Finally, observed extraction values for the various procedures in 11 unpolluted control soils were included for future reference.