Quantitative clay mineralogy predicts radiocesium bioavailability to ryegrass grown on reconstituted soils.

Current radiocesium (Cs) models to evaluate the risk of Cs transfer from soil to plants are based on the clay and exchangeable potassium (K) contents in soil. These models disregard the mineralogy of the clay fraction and are likely not capable of accurately predicting the Cs transfer factor (TF) in soils of contrasting parent rocks and weathering stages. The objectives of this study were to test that hypothesis and to identify whether quantitative information on mineralogy can improve the predictions. A pot cultivation experiment was set up with clay-sand mixtures in single and double clay doses that were fertilized, spiked with Cs and grown with ryegrass for 30 days. Four clays (illite, biotite, smectite and vermiculite) along with six deposits from clay-rich geological units were compared. The TF generally decreased with increasing clay dose for each of these ten different clay groups, however, the TF varied two orders of magnitude across clay groups and doses. The TF was highest for clays with little Cs specific sites such as bentonite and/or where the exchangeable K content was low compared to the other clays. The TF was well predicted from the soil solution Cs and K concentrations (R = 0.72 for log transformed TF), corroborating earlier findings in natural soils. The TF (log transformed) was statistically unrelated to total phyllosilicate content or 1:1 and 2:1:1 type phyllosilicate content while it significantly decreased with increasing 2:1 phyllosilicate content (R = 0.32). A multiple regression model with four different X-ray diffraction (XRD) based phyllosilicate groups yielded the strongest predictive power (R = 0.74). We conclude that XRD quantification is valuable for describing Cs bioavailability in plant substrates. These findings now await confirmation for natural soils.