Is the transfer factor a relevant tool to assess the soil-to-plant transfer of radionuclides under field conditions?

The radiological impact of radionuclides released to the terrestrial environment is usually predicted with mathematical models in which the transfer of radionuclides from soil to the plant is described with the transfer factor (TF). This paper questions the validity of the protocols proposed by the International Atomic Energy Agency to measure TF in the field and in greenhouses conditions. We grew maize (Zea mays L.) both in the field after a surface application of radionuclides ((54)Mn, (57)Co, (65)Zn, and (134)Cs) and in a greenhouse with the same soil that has received the same fertilization and that had been previously sieved and homogeneously labeled with the same radionuclides before being repacked in pots. The analysis of the displacement of radionuclides in the field soil profile showed a higher concentration of the surface-applied radionuclides in the preferential flow path (PFP) in comparison to the soil matrix indicating that they infiltrated heterogeneously in the soil profile due to the structure-induced non-uniform water flow. A significantly higher recovery of (57)Co and (134)Cs was observed in the plants grown in the field soil, whereas no differences in the recovery of (54)Mn and (65)Zn between the two experiments were detected. These results suggest that (i) under field conditions the soil-to-plant transfer of radionuclides that co-exist as stable elements present at low concentrations in the soil and in the plant is higher than that measured under greenhouse conditions and (ii) the implicit assumption made when calculating the TF (that radionuclides are homogeneously distributed in the soil profile) is not valid, thereby preventing the calculation of an average concentration to obtain the TF parameter.