A statistical approach for estimating the radiocesium interception potential of soils

The solid-liquid distribution of radiocesium (Cs-137) varies extensively among soil types, A statistical approach was used to relate Cs-137 sorption with soil properties. Eighty-eight pasture soils were sampled in Belgium. All soil samples were characterized for texture, cation exchange capacity, exchangeable K, Ca, and Mg content, pH and organic C content, Soil samples were equilibrated in a 10(-1) M CaCl2 - 5.10(-4) M KCI solution. Adsorption of Cs-137 was measured after 16 h in this medium. The solid-liquid concentration ratio, K-D, of Cs-137 in this specific ionic scenario allows for the calculation of the Radiocesium Interception Potential (RIP), which in turn can be used for KD predictions for a wide range of different ionic scenarios, The RIP varied from 54 to 11 200 mmol(c) kg(-1). Stepwise multiple regression between RIP and soil characteristics yielded statistically significant models (P < 0.001) for samples from each northern and southern Belgium. For northern Belgium the regression model only included texture fractions (R = 0.86, n = 62) whereas for southern Belgium the regressors were texture fractions and soil pH (R = 0.88, n = 26), The correlations between the RIP and soil characteristics depend strongly on the geological origin of the soil.