Plant uptake of elements in soil and pore water: field observations versus model assumptions.

Contaminant concentrations in various edible plant parts transfer hazardous substances from polluted areas to animals and humans. Thus, the accurate prediction of plant uptake of elements is of significant importance. The processes involved contain many interacting factors and are, as such, complex. In contrast, the most common way to currently quantify element transfer from soils into plants is relatively simple, using an empirical soil-to-plant transfer factor (TF). This practice is based on theoretical assumptions that have been previously shown to not generally be valid. Using field data on concentrations of 61 basic elements in spring barley, soil and pore water at four agricultural sites in mid-eastern Sweden, we quantify element-specific TFs. Our aim is to investigate to which extent observed element-specific uptake is consistent with TF model assumptions and to which extent TF's can be used to predict observed differences in concentrations between different plant parts (root, stem and ear). Results show that for most elements, plant-ear concentrations are not linearly related to bulk soil concentrations, which is congruent with previous studies. This behaviour violates a basic TF model assumption of linearity. However, substantially better linear correlations are found when weighted average element concentrations in whole plants are used for TF estimation. The highest number of linearly-behaving elements was found when relating average plant concentrations to soil pore-water concentrations. In contrast to other elements, essential elements (micronutrients and macronutrients) exhibited relatively small differences in concentration between different plant parts. Generally, the TF model was shown to work reasonably well for micronutrients, whereas it did not for macronutrients. The results also suggest that plant uptake of elements from sources other than the soil compartment (e.g. from air) may be non-negligible.

Inferences about radionuclide mobility in soils based on the solid/liquid partition coefficients and soil properties.

To assist transport modeling in assessments of the radiological impact of a geological repository for radioactive wastes, the mobility of various elements was studied in arable and wetland soils in the Forsmark region, Sweden. Pore water and total element contents were determined for five types of unconsolidated deposits (regolith), spanning a wide range of soil properties with respect to pH and organic matter content. Two soil depths were sampled to capture element mobility in regolith layers affected and unaffected by soil-forming processes. The solid/liquid partition coefficients (K d values) for most elements varied significantly among regolith types. For most elements, the observed variations in K d values could be explained by variations in soil properties. For many elements, mobility increased with decreasing soil pH. The results provide a significant addition of data on radionuclide retention in soils, taking account of soil properties and processes.