Soil-to-plant transfer factors for phytomanagement

The transfer factor (TF) depends on radionuclide, crop or vegetation type, soil type, soil characteristics, climate conditions, the type of contamination and the type of experiment. Therefore, TFs taken from literature should be considered with care if these TFs are used in radiological evaluations. Since TFs were mostly studied in context of the ingestion radiation dose to men, most TFs found in literature are for edible crops and grasslands. For the analysis of TFs, crops are often grouped. Cereals, tubers, leafy green vegetables, brassicas, root vegetables, legumes, onions and grassland are among the most important groups. Rooting depth, water uptake, exudates which complex radionuclides or alter the root environment (e.g. changes on pH), plant nutrient requirements (e.g. plants with high K contents have generally a high Cs-uptake), ... are among the crop characteristics affecting uptake. Soil type (often broadly classified as sand, loam, clay and organic) and its associated physical and chemical characteristics can affect the transfer of radionuclides to crops both directly and indirectly. There are almost no empirical modelling approaches existing which predict the activity concentrations in vegetation based on soil properties. An exception is the model by Absalom et al. (1999) which predicts the Cs concentration in crops (grass, cereals, some vegetables) considered pH, organic matter content clay and silt content and exchangeable Ca and K status. Regional differences in climate can affect growing season period, crop yields and crop husbandry practices, such as irrigation. Precipitation and temperature are the climate factors which have the most influence on agriculture. The type of the contamination will influence the radionuclide TF to the crop. The chemical form of a radionuclide will determine its availability. The time between the contamination of the soil and the harvesting of the (mature) crop, i.e. the age of the contamination, affects the TFs. This is certainly important when one intends to extrapolate data obtained from potted soil, lysimeter or field experiments (via own research or literature data) to in situ conditions. The type of the experiment certainly affects the TF values obtained. Modes TFs can be obtained is via potted soil, lysimeter or field experiments or via in situ investigation. Generally speaking, TFs obtained in potted soil experiments are higher than values obtained under field or in situ conditions. No wide-scale empirical models that predict soil-to-plant transfer of radionuclides are currently available for use in arable systems. Consequently, transfer must be predicted from existing experimental and field data. To date, the largest compilation of data on TFs has been carried out by the International Union of Radio ecologists (IUR, 1992). More than 7000 records of TFs for 21 radionuclides have been recorded. The largest numbers recorded are for radiocaesium and radiostrontium. The IUR database contains information on a limited number of soil characteristics, principally a broad texture class definition, pH and organic matter content. In 1989, multiple linear regression was used to derive best estimate TFs for radiocaesium and radiostrontium for 4 soil types - sand, loam, clay, peat - and five edible crops - cereals, tubers, root vegetables, legumes and green vegetables. The best estimate TFs for peat soils were based on no data or very few data. These data are still commonly used in radiological assessments. Nisbet et al. (1999) have made an extensive compilation of radiocaesium transfer factors following a critical review of published and unpublished literature. Best estimate TFs are presented in Table, this table also show the IUR (1992) recommended TFs. A wide range of physical and chemical characteristics of soils can affect the transfer to radiocaesium. The parameters considered in the study by Nisbet et al. (1999) were pH, organic matter content, clay and silt content, exchangeable K and Ca. Statistical analysis showed that none of the soil parameters considered could be reliably used to predict TFs for either radiocaesium or radiostrontium. This was due to the inherent variability of data and insufficient numbers of re-cords for several soil-crop combinations. However, despite the absence of any robust mathematical relationship, low levels of exchangeable potassium (i.e.< 0.5 meq per 100 g soil) were shown to result in significantly higher TFs for radiocaesium for several crop types. High ammonium levels tend to increase the TFs. Absalom et al. (1999) successfully predicted the Cs concentration in crops (grass, cereals, some vegetables) follow-ing multicomponent regression considering the total deposition, the percentage clay and the fraction of soil exchange-able K. Three key properties underlying the bioavailability of radiocaesium in soils were estimated in the model: the la-bile solid-liquid distribution coefficient (Kd defined as the ratio of the radionuclide in the solid soil fraction to the radionuclide concentration in the soil solution: Bq kg-1/Bq dm-3), the soil solution K concentration and the radiocaesium concentration factor. The Cs-TFs for cereals, green vegetables, brassicas, root vegetables and legumes tend to be highest for organic soils and sand and lowest for loam and clay: maximal differences between soil types for these crops are generally small. (Authors)