No abstract

Five peat soils and a mineral soil were artificially contaminated with 137Cs. Soil solution activity and radio-lability of 137Cs were monitored over 709 days to quantify progressive 137Cs fixation. The peat soils fixed large amounts of 137Cs, but less than the mineral soil did. Distribution coefficients (&, cm3 g-') ranged from 30 to 5000 at the end of equilibration. A labile 137Cs distribution coefficient, Kdlr was estimated by a method involving solid e solution equilibration in dilute solution.In a separate study several concentrations of KC1 were added to soils in increasing concentration both before and after the addition of 137Cs. Differences in apparent adsorption strength of radiocaesium indicated that K+ induced the collapse of expanded mineral interlayers, thereby trapping ions. It seemed that 137Cs adsorbs at sites in the small micaceous clay fraction of the peat soils.The different rates of 137Cs adsorption and fixation in the peat and mineral soils, in which the rate of access of 137Cs to fixation sites in peat soils is less, seems to have been caused partly by lack of K, and partly by the scarcity of fixation sites.

A model is presented that dynamically estimates the radiocesium activity in herbage from readily available soil parameters. Three key properties underlying the bioavailability of radiocesium in soils were estimated in the model: the labile radiocesium distribution coefficient (k dl ), the solution K + concentration ([m K ]), and the radiocesium concentration factor (CF, Bq kg -1 plant/Bq dm -3 soil solution). These were determined as functions of the soil clay content and exchangeable K status. The effect of time on radiocesium fixation was described by two first-order decay equations. The model was initially parametrized using radiocesium uptake data from a ryegrass pot trial. Without further parameterization, the model was then tested for a wide range of soil and crop combinations using a database of published and unpublished information from a variety of sources and covering contamination time periods of 0.5-11 years. Model predictions of activity concentrations in crops were in generally good agreement with observed values.