Radionuclide Uptake from Soil to Plants: Influence of Soil Classification

Abstract

The entry of trace contaminants, which are present in the terrestrial environment, into human food chains is controlled in the long term by their uptake by plant roots. The radionuclides released into the environment can give rise to human exposure by the transport through the atmosphere, through aquatic systems, or through soil subcompartments. Soil-to-plant factor is one of the important parameters to be used in transfer models for predicting the concentration of radionuclides in crops/plants and for estimating dose impacts to man. Existing databases are limited to experimental values in a restricted number of soil systems and are largely comprised of temperate environment data. In general, transfer factors show a large degree of variation dependent upon several factors such as soil type, species of plants, and other environmental conditions. Soil-to-plant transfers of the radionuclides varied considerably with the times of deposition also. In general, the variations with the deposition times were by a factor of up to 10. This chapter describes a review of published data on the transfer factor (TF) and classification of soils based on TFs.

1 Introduction

Radionuclides exist in the environment either naturally or artificially. The natural and artificial radionuclides that are present in the environment are the main sources of radiation exposure for human beings (UNSCEAR 2000). It has been estimated that on average, 79 % of the radiation to which humans are exposed is from natural sources, 19 % from medical application, and the remaining 2 % from fallout of weapons testing and the nuclear power industry (UNSCEAR 2000). However, most of the public concern over radiation from radionuclides has been due to the global fallout from atmospheric nuclear weapons testing and the operation of nuclear facilities. Both of these activities have introduced a substantial amount of man-made radionuclides into the environment and have caused radionuclide contamination of large areas of land worldwide. The natural radioactive background originates from uranium and thorium series, from potassium-40, and from the interaction of cosmic radiation with matter, while man-made sources include various applications of radionuclides in medicine, industries, consumer products, and nuclear weapons tests. Weathering of the earth’s crust is the ultimate mechanism for the release of primordial radionuclides into the soil, which constitutes the principal source of natural background radiation. The principal natural chains are those originating, respectively, from ²³²Th, ²³⁸U, and ²³⁵U. ²³²Th is one of the natural thorium isotopes (together with, e.g., ²³¹Th and ²³⁴Th that are, respectively, daughters of ²³⁵U and ²³⁸U) with a half-life of 1.4 × 1010 years. The final stable product of its decay chain is ²⁰⁸Pb. Thorium is present at different concentrations in the earth crust. The most abundant in the earth crust is ²³⁸U followed by a very weak presence of ²³⁵U and ²³⁴U (Mortvedt 1994). Uranium is found in almost all soils and rocks at different concentrations, depending on the nature and typology of soil (Blanco Rodriguez et al. 2006; Baeza et al. 2005). Artificial radioactivity is present in the environment as a consequence of different anthropogenic sources. It is possible to enlist some of these, ranging from nuclear weapons to radioactive material leaching from nuclear power plants and waste products of general and medical industry. Artificial radioactivity derives from man activity and is generated in different applications, ranging from nuclear installations, nuclear accident, or normal radionuclides applications used in industry, e.g., medical machinery and different kinds of electronic devices. The primary man-made products are ⁹⁰Sr and ¹³⁷Cs (Abbazov et al. 1978; Absalom et al. 1999, 2001; Askbrant and Sandalls 1998). Radionuclides of long half-lives, such as ⁹⁰Sr and ¹³⁷Cs, are of special importance from the point of human health, since these nuclides can enter the human body via the food chain and increase the radiation burden for many years. The understanding of mechanisms that affect the radionuclide uptake by plant species which grow under field conditions becomes a subject of increasing interest.

Radioactive contamination of the human environment became a reality on July 16, 1945, when the first fission was tested near the town of Alamogordo in New Mexico, USA. However, large-scale production of anthropogenic radionuclides began when the first nuclear reactor started operation in Chicago on December 2, 1942 (Aarkrog 1994). Prior to the release of large quantities of radioactive fission and activation products into the environment during the atmospheric weapons tests 1945–1963, there was limited scientific interest in the environmental radioactivity as the world inventory of artificial radionuclides was very small (Shaw 2007). Atmospheric testing of nuclear weapons was carried out in two main series from 1952 to 1958 and from 1961 to 1962. The USSR, USA, and UK signed the partial atmospheric test ban treaty in 1963, and thereafter, only a limited number of atmospheric tests were carried out by France, China, and India (MacKenzie 2000).

Contamination of soils with typical fission product radionuclides, such as ¹³⁷Cs and ⁹⁰Sr, has persisted for far longer than what was originally expected. Radionuclides in soil are taken up by plants, thereby becoming available for further redistribution within food chains. Radionuclides in the environment can, therefore, eventually be passed on to human beings through food chains and so may represent an environmental threat to the health of populations (UNSCEAR 2000). Although atmospheric testing of nuclear bombs has been banned globally, in the foreseeable future, the nuclear power industry will continue to make an increasing contribution to power consumption, a trend which could help to reduce the global warming process due to the consumption of fossil fuel. As accidental and routine releases of radionuclides from the nuclear industry are inevitable and can cause local, regional, and even global environmental contamination, remediation of soils contaminated with radionuclides is becoming an increasingly important aspect of radiological protection.

Artificial radionuclides have entered the human environment as a consequence of atmospheric nuclear weapons tests. Global fallout of fission products in the 1950s and 1960s involved deposition of ¹³⁷Cs (t_{1/2 phys}: 30.17 years), ⁹⁰Sr (28.5 years), ⁸⁹Sr (50.5 days), ³H (12.3 years), ⁵⁴Mn (312 days), ⁶⁵Zn (244 days), ⁹⁵Zr (64 days), ^103/106Ru (39.4 days/368 days), ¹²⁹I (1.6 × 10⁷ years), and ¹⁴⁴Ce (284.8 days). At present, the most important sources are routine releases from nuclear power (NPP) and reprocessing plants, which can include ³H, ¹⁴C, ⁶⁰Co (t_{1/2 phys}: 5.27 years), ⁵⁴Mn, ⁸⁹Sr, ⁹⁰Sr, ⁹⁵Zr, ⁸⁵Kr (10.8 years), ²³⁸Pu (87.7 years), ^239/240Pu (24,000 years/6,600 years), ¹²⁹I, ⁹⁵Nb (35 days), ^103/106Ru, ^110mAg (249.9 days), ¹²⁵Sb (2.77 years), ^134/137Cs (2.06 years/30.17 years), ¹⁴⁰Ba (12.8 days), and ¹⁴⁴Ce (284.8 days).

On a global scale, terrestrial ecosystems have received the largest inputs of anthropogenic radionuclides from nuclear weapons tests and from the accident at the Chernobyl nuclear power plant (UNSCEAR 2000) (Table 1).

Table 1 Major global radionuclide releases in PBq (10¹⁰ Bq) (UNSCEAR 2000)

Figure 1 illustrates the principal sources of artificial radionuclides released to the environment and the complex pathways by which these sources are redistributed and ultimately impact on organisms. Radionuclides released to the atmosphere and marine waters can be transported over large distances and eventually find their way to the terrestrial environment, often with surprising rapidity. Once terrestrial ecosystems become contaminated through such routes, the residence times and environmental impacts of individual radionuclides within specific ecological compartments is a function of the physical half-life of the radionuclide, the chemistry of the element to which it belongs, and the nature of the compartment itself. Long-lived radionuclides pose environmental hazards on different timescales. Hence, ¹³⁷Cs and ⁹⁰Sr (half-lives = 30 and 28 years, respectively) deposited on land surfaces around the world at the peak of atmospheric weapons testing in 1963 are still present in the environment and are likely to remain detectable for another 150 years.

Fig. 1

Pathways leading to redistribution throughout the environment of artificial radionuclides from nuclear-related facilities

The primary application of soil-to-plant transfer factors is in food chain models used for calculating radiological consequences from routine or accidental release of radioactive substances into the environment. These models usually are designed to give conservative assessments (Hoffman et al. 1984; Peterson 1995). For this purpose, the use of parameters that are as simple as possible is desirable to keep model complexity down. The wealth of soil-to-plant transfer factor data now available at least for temperate environments seems to be adequate to derive probability distributions from which representative values for use with screening models can be derived (IAEA 1994; Sheppard and Evenden 1997; Nisbet 2000).

In radioecological research, on the other hand, interest shifts more and more to elucidating the chemical, biological, and physical mechanisms governing the root uptake and translocation of radionuclides from soils. The best example may be the pioneering work of Cremers and coworkers (Cremers et al. 1988; Valcke and Cremers 1994) on the chemistry of radiocesium in soils. Recent understanding of processes involved in soil–plant transfer offers the perspective for developing mechanistic models, but such models, although deemed required (Nisbet 2000), are not yet available. Focusing on radiocesium and radiostrontium, this paper intends to summarize present knowledge on processes which influence root uptake of radionuclides (as well as of other trace metals) by plants and to identify areas where our present understanding is still limited. Although important, species variations in soil–plant transfer are not discussed, since comprehensive compilations and reviews are available (e.g., Andersen 1967; Frissel 1992).

Eventually, the most important conceptual limitation of the transfer factor approach is that it does not take into account competition between ions. Soil-to-plant transfer factors are most often measured for trace substances whose behavior in the soil–plant system largely depends on the concentrations of macronutrients present. For example, an activity concentration in a soil solution of 1 Bq l⁻¹ of ⁹⁰Sr or ¹³⁷Cs corresponds to ca. 2 × 10⁻¹⁵ M l⁻¹, whereas median concentrations of Ca, K, and Mg in soil solution are in the order of 1 mM l⁻¹ (Robson and Pitman 1983). A number of substances naturally present in soils have been found to influence the uptake of radionuclides and heavy metals by plants, though not always beneficially (Wallace 1989; Desmet et al. 1991; Lorenz et al. 1994a, b). For radioactive cesium and strontium, these competitive effects form the basis for countermeasures at the soil–plant level after a nuclear accident (Howard and Desmet 1993). As discussed in detail in the following paragraphs, the concentration of a trace substance accumulating in plants may not primarily depend on its absolute concentration in the soil–plant system but on the concentration ratio of other micro- and macronutrients.

In fallout situation, the human population can be exposed to external and internal radiation by different pathways (Fig. 2). As generally found, the transfer to a certain crop will be much less if fallout takes place between two growth seasons than during a growth season. The direct transfer from fallout to a crop exposed to the atmosphere during a growth season will always be much higher than the indirect transfer from a contaminated ground surface, which in turn will be higher than the transfer by root uptake from a contaminated soil or plow layer.

Fig. 2

Main pathways for radionuclides to man

There is increasing interest in radiological assessments of discharges of naturally occurring radionuclides into the terrestrial environment. Such assessments often require the use of predictive models because measurements in environmental materials generally contain a contribution from the natural background. The transfer of artificial radionuclides along terrestrial food chains has been studied extensively over the last 30 years, with understandable emphasis on radiocesium since 1986 (Bunzl and Kracke 1989; Chiu et al. 1999). Naturally occurring radionuclides have not been studied to the same extent as their artificial counterparts, but some comprehensive investigations have been carried out in various parts of the world. There is increasing interest in radiological assessments of the discharges of naturally occurring radionuclides into the terrestrial environment, in terms of current releases both from industrial sites and from the presence of historical contamination in the soil.

The natural and artificial radionuclides that are present in the environment are the main sources of radiation exposure for human beings and constitute the background radiation level (Eisenbud 1973). Contamination of the environment has been a frequent legacy of industrialization, and recognition of the adverse health and environmental and economic effects of this contamination resulted in legislation for its minimization and monitoring. Thus, determining the distribution of these radionuclides is necessary for assessing the effects of radiation exposure (Eisenbud 1973). There is now increasing pressure to develop effective technologies not just to minimize and monitor but to decontaminate ecosystem compartments such as soils that have become contaminated; the nuclear industry is similar to other industries in these respects (Eisenbud 1973).

Radioactivity from the Chernobyl accident affected food production systems throughout Europe. Most affected were the Scandinavian countries where activity concentrations in reindeer, goat’s milk, sheep, game animals, and freshwater fish were above the intervention levels, which are still subject to restrictions. Radiocesium and radiostrontium were still present in the global environment at relatively low concentrations prior to the Chernobyl accident as a result of atmospheric nuclear weapons and discharges from natural facilities to a lesser extent. The global fallout of ¹³⁷Cs resulting from the atmospheric nuclear weapons testing during the 1960s has been estimated at 2.8 kBq⁻¹m² in the Northern Hemisphere and 2.2 kBq⁻¹ m² in central Sweden (Rosèn et al. 1999). However, from Chernobyl accident, they were 1.8 and 1.4 kBq⁻¹ m², respectively (Rosèn et al. 1999). The total amount of ¹³⁷Cs deposited over Sweden has been estimated to be 4.25 × 1015 Bq (UNSCEAR 2000). In some places, the Chernobyl fallout of ¹³⁷Cs in Sweden was up to 100 times higher than the global bomb fallout (Rosèn et al. 1999). The greater activity of ¹³⁷Cs in the initial fallout, twice the activity of ¹³⁴Cs and its longer physical half-life, makes ¹³⁷Cs the most important of the two cesium isotopes (Rosen et al. 1999).

About 1,300 different radionuclides exist partly originating from natural sources and partly anthropogenically produced. Natural radionuclides can be generated by the activation of stable isotopes via cosmic radiation, for example, ³H (12.3 years of half-life), ⁷Be (53.3 days), ¹⁴C (5,730 years), and ³⁵S (87.5 days) known as cosmogenic radionuclides, which are originated during the creation of the universe. The latter are known as primordial radionuclides and include ⁴⁰K (1.3 × 109 years) and the isotopes of uranium and thorium that give rise to various daughter nuclides including ²²⁶Ra (1,600 years), ²²²Rn (3.8 days), and ²¹⁰Pb (22.3 years) (Strebl et al. 2007). Artificial radionuclides have entered the human environment as a consequence of atmospheric nuclear weapons tests and also by nuclear accidents. Global fallout of fission products in the 1950s and 1960s involved deposition of ¹³⁷Cs (30.17 years), ⁹⁰Sr (28.5 years), ⁸⁹Sr (50.5 days), ³H, ⁵⁴Mn (312 days), ⁶⁵Zn (244 days), ⁹⁵Zr (64 days), ^103/106Ru (39.47368 days), ¹²⁹I (1.6 × 107 years), and ¹⁴⁴Ce (284.8 days). At present, the most important sources are routine releases from nuclear power plants (NPP) and reprocessing plants. Very high emissions of these isotopes can occur in very rare circumstances of nuclear power plant accidents such as occurred at Chernobyl (Ukraine) in 1986. This accident caused deposition of substantial quantities of radionuclides over a large part of Europe, and in Sweden, four counties were mainly affected (Rosén et al. 1999). Hence, it increased the ^134/137Cs and ⁹⁰Sr soil inventories of Sweden and many other European countries considerably (Strebl et al. 2007). In the USA, many nuclear plant accidents were occurred and caused fatalities and property damages. One of them is the accident happened in Idaho Falls, Idaho, January 3, 1961, where the damage costs around 22 million dollar.

In Japan, the Fukushima I nuclear accidents occurred after a 9.0 magnitude Tōhoku earthquake and subsequent tsunami on March 11, 2011, only 14 days before the reactor was to be shut down. The earthquake triggered a scram shutdown of the three active reactors at the Fukushima I Nuclear Power Plant (Fukushima Dai-Ichi). The ensuing tsunami inundated the site, stopped the Fukushima I backup diesel generators, and caused a station blackout. The subsequent lack of cooling led to explosions and meltdowns at the Fukushima I facility, with problems at three of the six reactors and in one of the six spent fuel pools. It caused fatalities and properties damage and evacuation of millions of people.

Irradiation of humans can occur via external and internal exposure to radionuclides (Voigt et al. 2007). Doses to humans are estimated by considering ingestion of radionuclides in drinking water and food, external irradiation from radionuclides in soil, and inhalation of radionuclides on airborne dust particles, and these accounts for a substantial part of the average radiation doses received by various organs of the human body (Khan et al. 2010; Voigt et al. 2007).

Plants acquire man-made and naturally occurring radionuclides via their roots or leaves, and animals acquire them through consumption of plants, phosphate-based mineral food supplements, and soils. These radionuclides are ultimately transferred to man by eating animal meat or milk or directly from plants by using them as food. Radionuclides ingested in food and, to a lesser extent, water account for a substantial part of the average radiation doses received by various organs of the human body, especially the skeleton. Certainly, radionuclides in an environment will persist with more or less predictable residence times and with associated radiation exposures in proportion to the type and activities of radionuclides present. Radioactivity from the Chernobyl accident affected food production systems throughout Europe.

Plant development stage and seasonality are the varying response to radioactive contamination of vegetation when the contamination occurs. Furthermore, the migration and accumulation of radionuclides in the soil–plant system is a complex phenomenon, involving processes such as leaching, capillary rise, runoff, sorption, root uptake, and resuspension into the atmosphere (Ehlken and Kircher 2002; Frissel et al. 1990; Fortunati et al. 2004; Haylock 1999; Higley and Bytwerk 2007; Kabate-Pendias 2004). The plant uptake of the main natural radionuclides, uranium, thorium, potassium, and radium, is naturally due to the plant’s need of nutrients. Through the mineral uptake process, the plant transfers natural radioactive substances, which accumulate in the plant vital portions, including the edible ones. Radionuclides are taken up by plants by some of the same mechanisms (Baeza et al. 2004; Frissel 1990; Duskesas 2009; Bunzl 1997; Bynzl et al. 2000; Choi et al. 1998; Djingova et al. 2005; Djuric et al. 1996; Djingova and Kuleff 2002) as plant nutrients due to their similar chemical and physical characteristics. In this way, radionuclides enter human food chain through consumption and lead to a long-term internal exposure to human body (Pulhani et al. 2005). This chapter describes a review of published data on the transfer factor and classification of soil based on TFs.

2 Transfer Factors

Radionuclide uptake by plants from contaminated soil represents a key step of radionuclide input into human food chain; this phenomenon is described by soil–plant transfer factor that is defined as the ratio of plant-specific activity to soil-specific activity. Plants are the primary recipients of radioactive contamination to the food chain following atmospheric releases of radionuclides. The transfer factor (TF) is a value used in evaluation studies on impact of routine or accidental releases of radionuclide on the environment for most important agricultural products known. For other areas and especially the developing countries, TFs are less known. The soil-to-plant transfer factor is regarded as one of the most important parameters in environmental safety assessment needed for nuclear facilities. This parameter is necessary for environmental transfer models which are useful in the prediction of the radionuclide concentrations in agriculture crops for estimating dose intake by man.

Plants are the primary recipients of radioactive contamination to the food chain from the abiotic environment through the uptake of radioactive debris from the atmosphere by aboveground parts of plants and a sorption of debris from the soil by the root system of plants (Aarkrog 1994). Experience since the Chernobyl accident indicates that the intake of radioactivity through food is an important source of the dose to the population in the Nordic countries such as Sweden. In the long term, the problems will be connected to the contamination of various sensitive ecosystems. Plant development stage and seasonality are the varying response to radioactive contamination of vegetation when the contamination occurs. Short-lived radionuclides (as ¹³¹I) and those that enter the food chain by direct contamination (e.g., ¹³⁷Cs) are important in this connection (Aarkorg 1994).

Radionuclides from natural decay chains despite short physical half-lives can be of high radiological importance because they are continuously produced and thus remain in the environment at a constant level (Streble et al. 2007). Besides the total amount deposited, the radiological importance of artificial radionuclides is determined by their radioactivity and radiation type (alpha, beta, or gamma), their bioavailability, and behavior within the food chain (Streble et al. 2007). Some radioactive isotopes like those of iodine cannot be discriminated from the stable forms and readily enter living systems.

The soil–plant–human is the principal pathway being studied for the transfer of radionuclides to human beings (IAEA 1982, 1994, 1999). Soil-to-plant transfer is one of the major environmental pathways leading to human ingestion of radionuclides. Thus, since the 1950s and 1960s, many efforts have been undertaken to predict and quantify radionuclide root uptake and to implement suitable models in radioecological/radiological models. The transfer factor concept is the simplest model for the quantification and prediction of crop contamination with radionuclides. It is usual to adhere to strict protocols when obtaining data to calculate transfer factors, as originally specified by the International Union of Radioecology (IUR 1989; IAEA 2004, 2006, 2009). The radionuclide concentration in soil is always determined on a dry weight basis down to a depth of 20 cm for all crops except pasture grass (10 cm). When determining radionuclide concentrations in plant material, the edible parts of crops (grain, tubers, fruits, edible leaves, etc.) have been most often investigated although, in the case of grazed and seminatural ecosystems, the vegetative parts of plants which provide nutrition for animals are just as important.

Soil–plant–human is recognized as one of the major pathways for the transfer of radionuclides to human beings. Knowledge, description, and modeling of radionuclide transfer in food chains are one of the key topics in radiation protection and radioecology (Voigt et al. 2007). Although the majority of research to date has focused on the contamination of food products and its prediction, an attempt is made here to draw together the more ecological aspects such as the distribution and fluxes or radionuclides in the different compartments. In response, countermeasures to reduce especially radiocesium, radioiodine, and radiostrontium in animal products have been studied, tested, and implemented, including those for animals living in seminatural environments. However, to apply countermeasures most effectively, the behavior and transfer of radionuclides in animals need to be understood and properly modeled (Voigt et al. 2007).

An important assumption in using the TF is that it is independent from the absolute radionuclide concentration in the soil. This assumption does not always appear to hold true under real-world conditions (Bunzl et al. 2000). Other simplifications in the transfer factor concept include (1) the artificial definition of the rooting zone, (2) the lack of discrimination between radionuclide pools of different availability in soil, (3) the fact that the TF does not really describe the process of root uptake but merely provides a concentration ratio including, for example, mass loading of plant surfaces with contaminated soil particles, and (4) the omission of any plant physiological parameters. The influence of these and additional simplifications discussed above is reflected by the huge variability of transfer factor values obtained under field and experimental conditions, which, for many radionuclides, exceeds three orders of magnitude (Coughtrey et al. 1983; IAEA 1994).

Another comprehensive data set published can be found in Gerzabek et al. (1998). A careful evaluation of the available Cs and Sr transfer data was published by Nisbet et al. (1999) who investigated the effects of aging, pH, organic matter, and exchangeable potassium (or calcium) on the soil-to-plant transfer of Cs and Sr. Post-Chernobyl studies confirmed previous work showing the influence of soil properties (clay mineral content and exchangeable potassium concentration) on radiocesium uptake by the food chain (Smith et al. 2005). Important factors affecting the transfer of radiocesium to crops/plants are the distribution of the root system in the soil profile as well as the soil pH and nutrient status (Rosén et al. 1999).

Primordial radionuclides (²³⁸U, ²³⁵U, ²³²Th, and ⁴⁰K) were formed by the same stellar processes which formed the other heavy elements of the earth (Eisenbud 1973). They are long-lived species that have been present on the earth since its formation 4.5 × 10⁹ years ago (MacKenzie 2000). The radionuclides ²³⁸U, ²³²Th, and their decay products and ⁴⁰K are natural terrestrial radionuclides and significantly form the major part of the natural radiation dose (UNSCEAR 2000). Due to its low natural abundance, ²³⁵U and its decay products do not form a significant part of natural radiation exposure. The primordial radionuclides ²³⁸U, ²³⁵U, and ²³²Th are the parent members of the three natural radioactive decay series, and their main members along their half-lives and principal decay modes are shown below (MacKenzie 2000): Radiation doses from primordial radionuclides are primarily from external gamma radiation, ingesting, and inhalation. Both external and internal doses can vary significantly according to differences in the geology of a region.

The soil-to-plant transfer factor (TF), the ratio of the concentration of radioactivity in the crop to the radioactivity per unit mass (sometimes surface area is used) of the soil, is a value used in evaluation studies on the impact of releases of radionuclides on the environment. The calculated dry weight radioactivity values in the soil and plant samples were used to derive transfer factors from each of the soils into each crop. The definition of a transfer factor is given by the following equation (IAEA 2004):

$${\text{TF}} = \frac{{{\text{Concentration}}\,{\text{of}}\,{\text{radionuclide}}\,{\text{in}}\,{\text{plant}}\,\left( {{\text{Bq}}\,{\text{kg}}^{ - 1} \,{\text{dry}}\,{\text{crop}}\,{\text{mass}}} \right)}}{{{\text{Concentration}}\,{\text{of}}\,{\text{radionuclide}}\,{\text{in}}\,{\text{soil}}\,\left( {{\text{Bq}}\,{\text{kg}}^{ - 1} \,{\text{dry}}\,{\text{soil}}\,{\text{mass}}\,{\text{in}}\,{\text{upper}}\,20\,{\text{cm}}} \right)}}$$

For grass, the soil depth considered is 10 cm.

Radioactivity in soil is defined here as the average activity concentration in the top 20 cm. This is an accepted international compromise arising from alternate measures that are often based on deposition per unit area assuming atmospheric fallout. Transfer factor values in excess of one imply active bioaccumulation of activity. Values less than one imply either strong binding of the radioactivity to the soil or that the plant is not accumulating that material. For most of Europe and the USA, the TFs for most important agricultural products are known. For other areas and especially the developing countries, TFs are not so readily available (IAEA 1994, 2010).

The transfer factor (TF) is a value used in evaluation studies on the impact of routine or accidental releases of radionuclide on the environment. This parameter is necessary for environmental transfer models that are useful in the prediction of the radionuclide concentrations in agricultural crops for estimating dose intake by man. The main factors that determine the variability of TFs are the type of radionuclide, type of crop, type of soil (soil characteristics), and stable element concentration. Minor factors are differences in crop varieties, differences in agricultural management (fertilization), and differences in the weather. The term weather is used deliberately, because what is meant is not the overall climate but the day-to-day differences of the weather which are believed to be the main reason for the great spread of TFs. As the transfer factors can differ between areas due to different climates, soil types, and vegetation, local transfer factors should be observed.

The transfer of artificial radionuclides along terrestrial food chains has been studied extensively over the last 30 years, with understandable emphasis on radiocesium since 1986. Naturally occurring radionuclides have not been studied to the same extent as their artificial counterparts, but some comprehensive investigations have been carried out in various parts of the world. Many studies have been carried out to determine TFs for most important agricultural products (Nisbet and Woodman 2000; Ng et al. 1982; Muck 1997; Masconzoni 1989; Martinez-Aguirre 1996; Livens et al. 1991; IUR 1992; Jacobson and Overstreet 1998; Deb et al. 2004; Frissel et al. 1990, 2002; Pulhani et al. 2005; Popplewell et al. 1984; Pietrazak-flis and Suplinska 1995; Mollah et al. 1998, 2004; Mollah and Begum 2001). Several projects were run by the International Atomic Energy Agency (IAEA) to determine TF mainly for ⁹⁰Sr and ¹³⁷Cs (IAEA 2010). These data have been used extensively in radiological assessment models. Natural environmental radioactivity arises mainly from primordial radionuclides, such as ⁴⁰K, and the radionuclides from ²³²Th and ²³⁸U series, and their decay products are considered to be the main contributor to internal radiation dose. Several studies on transfer of natural radionuclides from soil to plant have been carried out in different regions in the world (Shappard and Evenden 1988, 1990; Vera Tome et al. 2003; Hanlon 1994; Vandenhove et al. 2009; Yunoki et al. 1993; Yamamoto et al. 1996). However, there seem to be few data on transfer of natural radionuclides from soil to plant in semiarid environments. Therefore, the present study aimed to determine TF for natural radionuclides in semiarid settings to some agricultural crops under natural field conditions.

Transfer factor depends upon many factors such as electrical conductivity (EC), PH, and bicarbonate contents of soil. Radionuclide plant–soil ratio is affected by many factors that control plant uptake. These factors are as follows:

• 1. Physicochemical form of radionuclide.

• 2. Plant species and internal translocation mechanisms within the plant.

• 3. Soil characteristics.

• 4. Fertilizers and agricultural chemicals.

• 5. Chelating agents.

• 6. Distribution of radionuclides in soil.

The physicochemical form of the radionuclide strongly affects its retention by the soil particles and its availability for uptake by plants. The soil type affects strongly the behavior of radionuclides in soil and soil retention characteristics (Skarlou et al. 1996; Remeikis et al. 1957; Riise et al. 1990; Abbazov et al. 1978; Baeza and Gullen 2006; Champlin and Eichholz 1967). Sandy soils do not have the retention capacity of clay soils. Clay soils are composed of smaller particle sizes with larger surface area and negative charge surfaces. The soil’s pH value affects the plants uptake. In alkaline soils (high pH), insoluble precipitates may be formed with carbonate, hydroxyl, phosphate, or sulfide ions. These insoluble precipitates reduce the availability of radionuclides for plants. In acid soils (low pH), hydrogen replaces the adsorbed cations which become more available to plants. In highly acidic soils (pH < 5.5), some trace elements (particularly iron and manganese) may become toxic to plant growth. Fertilizers are chemical compounds added to increase the soil fertility and enhance plant production. They strongly affect both the stable element concentration and soil acidity (Fort 1978; Jang et al. 2004; Juo and Barber 1970; Kuhn et al. 1984; Massas et al. 2002). The effect of limestone (CaCO₃) addition appears to raise soil pH, increase exchangeable calcium concentration, and decrease the uptake of strontium. This is possibly due to the decreased solubility of SrCO₃ in alkaline conditions. Fertilizer with nitrogen in the nitrate form (potassium or calcium nitrates) and phosphate fertilizers may decrease soil acidity (Foth 1978). Organic fertilizer affects the ion exchange capacity, pH, stable element content of soil, and soil retention properties (Champlin and Eichholz 1967; Routson and Cataldo 1978).

Chelating agents are organic compounds which increase the ion mobility and reduce soil retention. This increases the plant uptake. Moreover, these agents enhance the translocation ability within the plant itself (Champlin and Eichholz 1967). In some situations of plant nutrient deficiencies, they are useful because they decrease soil retention, therefore increasing plant uptake (Pickering et al. 1966). Their effectiveness depends upon soil properties (particularly soil pH), chemical form of the radionuclide, and the nature and concentration of chelating agent.

Radiological assessments are always based on a prediction (or reconstruction) of radionuclide transport patterns. Recently, a number of high-quality publications have been produced for many of the transfer parameter values which merit consideration. The primary application of soil-to-plant transfer factors is in food chain models used for calculating radiological consequences from routine or accidental release of radioactive substances into the environment. These models usually are designed to give conservative assessments. For this purpose, the use of parameters that are as simple as possible is desirable to keep model complexity down. The wealth of soil-to-plant transfer factor data now available at least for temperate environments seems to be adequate to derive probability distributions from which representative values for use with screening models can be derived (IAEA 1994; Nisbet and Woodman 2000).

For many years, the IAEA has been publishing documents aimed at the support of the assessment of the radiation impacts on both human beings and the environment. Two major supporting data documents are the Technical Report Series No. 247 (TRS 247) “Sediment K _d s and Concentration Factors for Radionuclides in the Marine Environment” and TRS No. 364 “Handbook of parameter values for the prediction of radionuclide transfer in temperate environments” (IAEA 1994). The main goal of the Technical Report Series No. 364 “Handbook of parameter values for the prediction of radionuclide transfer in temperate environments” was to provide realistic transfer parameters for radiological assessments. Together, these documents covered available by that time transfer parameters for marine, freshwater, and terrestrial environments, and over many years, they have been key references for assessors, providing environmental impact assessments. In 2000, the IAEA initiated a revision of TRS 247 resulted in the publication of Technical Report Series No. 422 “Sediment Distribution Coefficient and Concentration Factors for Biota in the Marine Environment” (IAEA 2004). However, TRS 422 has proved to be one of the most important sources of information used in assessing the impact of radionuclides on the marine environment and can be widely used for the further development of generic model for to assess the impact of radioactive discharges on marine ecosystems. The document contains revised sediment Kds for marine ecosystems and concentration factor (CF) values for some marine biota species. In addition, TRS 422 contains CFs for a limited number of elements for marine mammals not included in TRS 247. In 2003, in the framework of the IAEA EMRAS ("Environmental Modelling for Radiation Safety”) project, the IAEA also started a revision of TRS 364. The developments of the above documents were supported by several CRPs, namely “Radionuclide Transfer from Air, Soil, and Fresh Water to the Food chain of Man in Tropical and Subtropical Environments,” “The classification of soil systems on the basis of transfer factors of radionuclides from soil to reference plants,” and “Radiochemical, chemical and physical characterization of radioactive particles in the environment” (IAEA 2006).

Investigations on the factors influencing the TF of radiocesium have been reported in an enormous number of studies, especially after the Chernobyl accident. Many of the results have been summarized in Frissel (1992), IAEA (1994), and IUR (1992), Nisbet et al. (1999). From this experimental material, many conclusions have been reached about the influence of some soil, climatic, and agricultural factors on the TF of radiocesium. However, a lot of contradictory facts have accumulated as well, indicating that more needs to be known about the dependence of TF on all soil properties and their combinations and on the regional conditions and seasonal changes which might cause variation in the TF. Therefore, in spite of the numerous existing data during the last decade, the FAO/IAEA initiated two successive coordinated research programs (“Radionuclide transfer from air, soil and freshwater to the food chain of man in tropical and subtropical environments” and “Classification of soil systems on the basis of transfer factors of radionuclides from soils to reference plants”) aiming to propose a classification of soils on the basis of reference values for the transfer of radiocesium and radiostrontium. The preliminary results are reported in Frissel et al. (2002) and Skarlou et al. (2003).

3 Factors Influencing Transfer Factors

The TF values of artificial and natural radionuclides vary enormously. The main factors that cause this variability for any particular radionuclide are the type of crop and type of soil. The length of time the radionuclide has been in the soil is also important. Other factors are crop variety, agricultural practice (especially fertilization), and differences in the weather during the growing season (Fig. 3).

Fig. 3

Factors affecting TF values

Soil properties that are likely to affect TF values include mineralogical and granulometric composition, organic matter content pH, and fertility. For cations such as Cs⁺ and Sr²⁺, cation exchange capacity and the nature of exchangeable bases are important.

The TF for Cs depends not only on the soil properties but also on the time that the radionuclide has been present in the soil as the availability of Cs decreases with time. A variety of processes are involved including fixation with soil minerals, incorporation by microorganisms, and migration within the rooting zone. Consequently, a reference value is contact time dependent. The TF for Cs in non-equilibrium conditions may be up to a factor 10 higher than for equilibrium conditions. Equilibrium for Cs may take some 5 years to attain, sometimes longer although the rate of decrease in TF is greatest during the first three months. The availability of Sr also decreases with time, but the effect is much less pronounced than with Cs.

Aside from the inherent characteristics of a crop to take up cations, TFs are a function of the physiological state of the plant which is affected by soil fertility (hence agricultural practice), the weather, and its stage of development. In the IAEA CRP (Frisel et al. 2002), this variability is reduced by considering only the radionuclide content of the harvested part of the plant, although for crops such as grass or chard, which are cut several times a season, the TF remains a function of the stage of physiological development.

Soil parameters that exert a major impact on radionuclide mobility are, according to Bunzl (1997), (1) the composition of the soil solution (pH, concentration of inorganic ions, redox potential, and concentration of organic substances), (2) physical and chemical soil properties (species/characteristics and contents of clay minerals, oxides and organic matter, and surface and charges of particles), (3) microorganisms and fungi (mycorrhiza), and (4) temperature. Most radionuclides in soil are present in cationic forms. In general, therefore, low pH values, low clay contents, and low cation exchange capacities lead to an increase in radionuclide mobility within the soil profile and favor plant uptake of radionuclides. Changes in soil organic matter content can yield different effects, depending on the ability of the respective radionuclide to form organic complexes or not (Frissel et al. 1990).

For radionuclides present in terrestrial ecosystem, information about their migration in soils is crucial, since this process controls their long-term behavior in the environment and their uptake by flora and fauna including food chains, but also their potential as groundwater contaminant (MacKenzie 2000). In soil, the type of radionuclide element determines only their potential ability to migrate through the profile. In soil profile, it is believed that the initial movement of deposited radioactive elements or radionuclides from the soil surface is relatively rapid as a result of infiltration process. And, the radionuclide migration intensity is observed as a result of simultaneously occurring processes: transport caused by rainfall infiltration (convective transport), dispersion caused by spatial variations of convective velocities, transfer on colloidal and fine-dispersed particles (diffusion process), migration along the plant root system, and others. In addition to these abiotic processes, soil fauna may contribute to the transport of radionuclides in soils, but their action under general conditions results in a dispersion-like translocation. Hence, the vertical migration of the radionuclides is much slower as a significant amount of the fallout of the majority is sorbed to the soil matrix. However, for some radionuclides such as radiocesium, their mobility might be decreasing over period of time after fallout due to the presence of slow sorption reactions in soil. Thus, the physical and chemical behaviors of the fallout (radionuclides) influence its migration in the soil. The concentration (amount) and distribution of radioactive elements in the root zone of plants also affects the transfer of radioactivity to the food chain and the transport of radioactive elements from contaminated soil to plants (Eisenbud 1973).The factors that determine the variation of radionuclide activity with depth in soil profile are rainfall intensity, pH, organic matter content, soil moisture content, soil texture, soil structure, infiltration rate, sorption characteristics of the particular radionuclide, land use, and management practices (Rosen et al. 1999). Due to the inherent complexity and spatial variability of the soil–plant system, the uptake of radionuclides in vegetation from soil is difficult to quantify. The transfer factor (TF) is a useful parameter, which is usually used for evaluating the impact of releases of radionuclides on the environment (IAEA 1994). The TF depends on the vegetation type, soil properties, and the type of radionuclides (Eisenbud 1973). For example, Velasco (2008) established an association of transfer factors of radiocesium and radiostrontium with soil organic matter, an association which was absent in the data analyzed by Nisbet and Woodman (2000). The enormous number of observations that have been accumulated during the last decades demonstrates that for a number of long-lived radionuclides, soil-to-plant transfer factors show variations which may exceed three orders of magnitude (Coughtrey and Thorne 1982; Frissel 1992). For radiocesium uptake from agricultural soils, transfer factors show ranges of up to three orders of magnitude even for individual soil–crop combinations (Nisbet and Woodman 2000).

The reason for the high variability of soil-to-plant transfer factors is obvious. This macroscopic parameter integrates a number of soil’s chemical, biological, hydrological, and physical processes and plant’s physiological processes, each of which shows its own variability and in addition may be influenced by external factors such as climate and human agricultural practices. Evaluation of the influence of these processes has been attempted by statistical inference from soil-to-plant transfer factor databases (Van Bergeijk et al. 1992; Sheppard and Evenden 1997; Nisbet and Woodman 2000) but with moderate success.

3.1 Soils

Many types of soil exist (UNFAO 1998) because of the wide range of biological, geological, and climatological conditions. Major classes in which soils are grouped on the basis of texture are gravel, sand, silt, and clay. In addition, the term loam is widely used for certain combinations of sand, silt, and clay. In soil classification for agricultural purposes, loam indicates a certain distribution of particle sizes. Soils of dry regions normally have pH values of about 7, and the usual order of abundance of exchangeable cations in the soil is Ca > Mg > I > Na. Soils of humid regions usually have pH values below 7, and under conditions of extreme leaching, hydrogen may represent 90 % or more of the total exchangeable cations. The usual order of abundance of exchangeable cations in humid regions is H = Ca > Mg > K > Na.

Table 2 summarizes the effects of key soil parameters on the mobility of some elements with important radionuclides, and the specific behavior of these elements is described below.

Table 2 Influence of soil characteristics on the behavior of different radionuclides in soils (compiled from Bunzl 1987; Coughtrey et al. 1983; IAEA 1994; Mortvedt 1994)

3.2 Radionuclide Dependence

Table 3 shows the TF AM and GM values for the principal radionuclides without differentiating between plant and soil types. The TF values show a very wide variability. Differences in the order of a factor of 2 were found for Co and Zn, but as high as a factor of 5 for Cs and Ra. For GM values, the following sequence was found: TFZn > TFSr > TFCs ≅ TFCo > TFRa ≅ TFU > TFPb ≅ TFI.

Table 3 Principal radionuclides in tropical and subtropical database (IAEA 1994)

3.3 Influence of Plant Group

The distribution of TF for some plant groups was reviewed (IAEA 2010). In Fig. 4, the values of TF obtained for three plant/plant part groups are shown, distinguishing the data obtained for different soil types. For rice, in spite of the TF GM being lower in clay soil, no significant differences in TF distribution were observed for the different soil types. Leafy vegetable TF values from clay soil demonstrate a larger transfer factor than loam and sand soils. The main conclusions obtained are as follows:

Fig. 4

Box chart representation of TF for Cs for different plant groups. GM and AM indicate geometric and arithmetic mean values, respectively

• A wide TF value variability was found for all radionuclides when plant and soil groups were not distinguished. The difference in the order of magnitude of TF value ranged from 2 (Co and Zn) to 5 (Cs and Ra).

• When different plant group/plant part combinations are considered, TF value variability is markedly lower.

• It was found that for many combinations, TF values are reasonably fitted by a lognormal probability distribution.

• In most plant groups, Zn and Sr have the highest TF values. Ra and U have low TF values, and the TF is lowest for Th. A relatively high TF was found for Ra in grasses. TFs for Cs have intermediate values.

• Leguminous and leafy vegetables have the highest TF values, while for grains, particularly rice, transfer factors are lower.

• Soil type seems not to have a decisive influence on TF values. Only the Cs TF value in tubers grown in sand soil has a noticeably higher value than TF values obtained from clay and loam.

3.4 Comparison of TF Values Derived from Pot, Lysimeter, and Field Experiments

In some evaluations of soil–plant transfer data for agricultural crops, pot experiments turned out to yield different average values when compared with data observed under field conditions. According to Frissel et al. (1990), the reasons for this observation are as follows.

• A more intensive exploration of the contaminated substrate by plant roots due to the limited space available in plant pots.

• In pot experiments, soils are mostly contaminated artificially using radionuclides in easily soluble form (dissolved in aqueous solutions). Binding to the soil substrate can increase with time (aging effects), until equilibrium conditions are reached.

• To avoid plant drought stress, plant pots have to be irrigated regularly as the limited soil volume and high root density result in rapid water loss by evapotranspiration.

These differences can be reduced to a minimum by choosing an experimental design which allows enough space per plant for the root system and some equilibration time after artificial contamination. The differences between the data are within the normal range of variability, and a systematic bias to higher TF values for pot experiments was not observed.

The IAEA, together with the United Nations Food and Agriculture Organization (UNFAO 1998) and the International Union of Radiologists (IUR), conducted a Coordinated Research Project (CRP) on Transfer of Radionuclides from Air, Soil and Freshwater to the Food chain of Man in Tropical and Subtropical Environments. This produced a set of values for key transfer parameters of radionuclides between the various components of tropical and subtropical ecosystems that can be used in dose assessment models. It concentrated on what are considered as the key parameters in assessment models—radionuclide transfer from soil to plant and from freshwater to fish. A data bank was developed from transfer factors for radionuclides, principally ¹³⁷Cs and ⁹⁰Sr, from soil to cereals, fodder crops including grass, legumes, root crops, green vegetables, and plantation crops. Account was taken of soil properties, nature of the contamination (artificial, weapons testing fallout, Chernobyl fallout, and so on), and the type of experiment (field, pot, or lysimeter) that generated the values. For soil-to-plant transfer, the following conclusions were drawn:

• there are no systematic differences between soil-to-plant transfer factors in temperate, subtropical, and tropical environments;

• the effect on transfer of (a) soil pH, (b) nutrient status of the soil, and (c) time elapsed since the soil was contaminated with radionuclides is generally independent of the climatic zone;

• there exist, however, ecosystems with a relatively high or low uptake (by a factor of 10 or even 100 higher or lower than average values);

• a higher or lower uptake condition is nuclide specific; an ecosystem may show a relatively high or low uptake for a particular radionuclide and not at all for other radionuclides;

• a higher or lower uptake condition is not crop specific. If an ecosystem shows a relatively high or low uptake for one crop, all crops show this behavior qualitatively.

The earlier CRP Transfer of Radionuclides from Air, Soil and Freshwater to the food chain of Man in Tropical and Subtropical Environments concluded that the generic values as published in the IUR/IAEA Handbook of Parameter Values for the Prediction of Radionuclide Transfer in Temperate Environments (IAEA 1994) may be used as a first approximation in assessment studies, but that site-specific deviations of a factor of 10 or more must be expected. For more precise assessment studies, an investigation to identify the conditions causing local deviations is necessary.

The investigators reported almost 3,000 TF values, mostly for radioisotopes of Cs and Sr but also of Mn, Zn, Po, Pb, Th, U, and K, obtained in some 25 soil lower-level units from 16 different soil groups using 25 crops. Geometrical means were used because data in the literature consistently show lognormal distributions. It is stressed that the mean values consider all observations even in cases where it might seem advisable to reject certain observations. This appendix also contains data for other nuclides and includes information about soil properties and crops. The data sheets from which they were calculated are available at ftp://iaea.org/dist/rifa-trc/Crete/RCM/Frissel/raw-data/.

Greenhouse experiments were conducted to investigate the dependence of ⁵⁴Mn, ⁶⁰Co, ⁸⁵Sr, and ¹³⁷Cs transfer from sandy soil to soybean plants on the growth stage when a radioactive deposition occurs. A solution containing ⁵⁴Mn, ⁶⁰Co, ⁸⁵Sr, and ¹³⁷Cs was applied onto the soil surfaces in the lysimeters at six different times—2 days before sowing and 13, 40, 61, 82, and 96 days after sowing. Soil-to-plant transfer was quantified with a transfer factor (m² kg^–1 dry) specified for the deposition time. The transfer factor values of ⁵⁴Mn, ⁶⁰Co, ⁸⁵Sr, and ¹³⁷Cs for the seeds were in the range of 1.5 × 10⁻³–1.0 × 10⁻², 4.7 × 10⁻⁴–3.2 × 10⁻³, 5.7 × 10⁻⁴–1.0 × 10⁻², and 3.0 × 10⁻⁵–2.7 × 10⁻⁴, respectively, for different deposition times. The corresponding values for the leaves were 6.4 × 10⁻³–3.2 × 10⁻², 4.3 × 10⁻⁴–2.0 × 10⁻³, 5.1 × 10⁻³–5.3 × 10⁻², and 9.2 × 10⁻⁵–1.9 × 10⁻⁴, respectively. The values for the seeds were on the whole highest following the middle growth stage deposition. After the presowing deposition, the transfer factor values of ⁵⁴Mn, ⁶⁰Co, and ¹³⁷Cs for the seeds decreased annually, so those in the fourth year were 53, 75, and 34 % of those in the first year, respectively. The present results may be useful for predicting the radionuclide concentrations in soybean plants due to their root uptake following an acute soil deposition during the vegetation period and for validating a relevant model.

4 Classification of Soils Based on TFs

4.1 Soil Classification

Soil classification involves a range of criteria, but pedological considerations dominate. It assesses the essential properties of the soil itself including the processes of formation, distribution, mineralogical composition, the organic matter content, and the texture. There are 30 soil groups:

Acrisols; Albeluvisols; Alisols; Andosols; Anthrosols; Arenosols; Calcisols; Cambisols; Chernozems; Cryosols; Durisols; Ferralsols; Fluvisols; Gleysols; Gypsisols; Histosols; Kastanozems; Leptosols; Lixisols; Luvisols; Nitisols; Phaeozems; Planosols; Plinthisols; Podzols; Regosols; Solanchaks; Solonetz; Umbrisols; and Vertisols. They are further subdivided into 121 lower-level units, several of which may occur in the same soil group so that 509 group/unit combinations are listed in (UNFAO 1998). It is at the unit level that criteria include some of agricultural significances such eutric (fertile defined as >50 % base saturation), dystric (unfertile, <50 % base saturation), and calcaric (calcareous). However, soils influenced by human activity may be placed in the Anthrosol group. Given the secondary level of significance allocated to agricultural properties, it cannot be assumed that this classification will be the most appropriate for the prediction of TFs, but clearly, it must be taken into consideration.

4.2 Weather

Weather data were not systematically collected although most investigators record some details in their reports. There is no obvious procedure with which to assess the effect of weather on the TFs measured in such a relatively small data set. Djingova et al. (2005) found that TFs to winter cabbage were higher than those to summer cabbage, presumably because it takes longer for the winter crop to reach maturity, an indirect effect of weather. Schuller et al. (2002) noted that TFs to chard, which is cut several times per season, declined through the year although it is not clear whether or not this should be interpreted as a weather effect.

4.3 Crop Variety

Skarlou et al. (2003) found that TFs to cabbage variety Brunswick from soils were 2.5, 1.3, 1.3, and 2.2 times those to variety Kozanko, and corresponding figures for sweet corn cultivars Vilmorin and Elite were 1.3, 0.5, 2.8, and 2.2, so in these cases, the effect was relatively small.

4.4 The Influence of Soil Properties on Cs Reference Transfer Factors

Table 4 shows reference TFs derived from the IUR database and values obtained in the earlier CRP (Frissel et al. 2002). This combined data set is larger than that obtained in new CRP, so any classification developed here should therefore not conflict with this table. Unfortunately, with a few exceptions, soil groups as defined by the FAO classification are not listed in the IUR database and most investigators are no longer available. With the earlier CRP, the situation is better, but not optimal. Also, in most cases, the IUR database does not include exchangeable K. Only, the soil texture is almost always reported, as is the notation, P, for peat soils, i.e., histosols in the FAO classification. Further, it would be desirable to estimate the influence on TF of different conditions such as the time that radionuclide was in contact with the soil, fertilization, irrigation, soil management, and the weather, but these factors were not systematically recorded.

Table 4 Reference transfer factors of Cs for cereals under equilibrium conditions (Bq kg⁻¹ dry crop)/(Bq kg⁻¹ soil in the upper 20 cm of soil)

4.5 Contact Time and Fixation

Fixation of Cs is a serious problem. Fixation is the partly irreversible, slow adsorption of Cs, especially to clay minerals. Usually, fixation is most important in the first few years after contamination of the soil, but the amount that will be fixed differs from soil to soil. The process may continue for a long time, but the kinetics of the process have not been characterized numerically. One would hardly expect to see fixation effects on soils contaminated after the Chernobyl accident which occurred 13–15 years before the measurements were taken, and the results of Sanzharova and Prister are broadly consistent with this view. Similar changes occur in other data sets (Fig. 5).

Fig. 5

Influence of the time elapsed since the Cs contamination of soil on the TF for cereals. Legend TFyamamoto = FAO/IAEA/IUR Workgroup collected data by Yamasaki, TFhaak = IUR data bank (available at ftp://iaea.org/dist/rifa/Crete/RCM/Frissel). TF in (Bq kg⁻¹ crop)/(B kg⁻¹ soil)

4.6 Time Trends

Radionuclides can be released and deposited onto soils at any time of the year. The physiological activities of plants, the developmental stages of their organs, and the availability of radionuclides for root uptake change with time (Choi et al. 2005, 2009). In the case of an acute release during the vegetation period, therefore, soil-to-plant transfer of radionuclides may greatly depend on the time of their deposition. Accordingly, time-dependent deposition values of a soil-to-plant transfer factor would be useful for estimating the root uptake from an acute vegetation–period deposition (Choi et al. 1998, 2009).

Transfer factors of radiocesium and radiostrontium in lysimeter and field experiments were frequently observed to decrease slowly with time for some years after contamination of the soils (Squire and Middleton 1966; Noordijk et al. 1992; Nisbet and Shaw 1994). Commonly, this time dependency is attributed to a slow irreversible fixation of the radionuclides to the soil matrix (IAEA 1994). Following the work of Cremers and coworkers (Cremers et al. 1988), a long-term decrease in radiocesium uptake by plants most often is attributed to its sorption and fixation to clay minerals (Shand et al. 1994; Hird et al. 1996). From the long-lived radionuclides, cesium (¹³⁴Cs and ¹³⁷Cs) and strontium (⁸⁹Sr and ⁹⁰Sr) isotopes burden the environment for greater time period.

Soil-to-plant transfer of various radionuclides is known to be affected by soil properties, plant species and variety, climatic condition, and cultural practices (Papanikolaou 1972). Variation of the concentration of radionuclides on the soil surface depends mainly on its mineralogical composition, its chemical and physical properties, meteorological conditions, and the possible transfer of material to deeper soil layers (Missaelidis et al. 1987; Vosniakos et al. 1998). The possibility of fixation of Cs isotopes by geological material and soil has been the subject of previous studies (Sikalidis et al. 1988). The mechanism of fixation depends strongly on the mineral composition of the soil. The existence of ¹³⁷Cs in the soil is important because of its possible transfer to the cultivated plants and eventually to animals and humans. It also increases the direct-exposure doses received by humans from terrestrial natural radioisotopes by 10 % (Kritidis and Kollas 1992). The transfer of Cs is also increased with increasing organic matter content.

4.7 Classification by Cs TFs

As stated earlier, any new classification should not conflict with Table 4. They include all Cs TF values for soil units of the CRP which correspond with groups 1 and 2 of Table 5. Maximum and minimum values for expected TF values were fitted by eye. They form the core of this classification. For soils in group 1, TF seems not to be related on texture, but in group 2, TFs, which are considerably higher, seem to be related to the texture. This is in agreement with the groupings in Table 5, but their nutrient status was a leading criterion, whereas here soil units are used. These considerations lead to the following classification of soils for the uptake of Cs by cereals (Table 5).

Table 5 Classification of soils by Cs TFs

As discussed earlier, this classification is imprecise because many of the TF measurements included values obtained before complete equilibrium of Cs distribution in the soil had been reached and only limited data are available for many soil groups. Consequently, the lower limits of expected TFs for groups 2 and 3 are the same and overlap the range of group 1. However, for planning purposes, it would be prudent to use the upper limits in which case there are reasonably clear distinctions. Therefore, these results give promise that soils could be classified in this way, but very many additional data are required for confirmation.

4.8 The Reference TF for Strontium

In addition to pedological considerations, properties such as organic matter, soil acidity, calcium carbonate content, and texture are used to define many soil groups and lower-level units. All these properties influence the uptake of Sr, so it should be possible to group soil units on the basis of reference Sr TF values. Figure 6 shows the TFs as a function of the exchangeable Ca.

Fig. 6

TFs as a function of the exchangeable Ca

The scheme in Table 5 uses clay, sand, and peat contents as the main criteria. Therefore, Fig. 7 distinguishes TF values on this basis. For clay/loam and sand histosols, it gives a reasonable fit. The vertisol does not follow the general trend, but it concerns one only. Fig. 7 is not very conclusive because it does not contain enough observations. IUR data do not include exchangeable Ca; therefore, only the mean value of the TFs for each soil texture is given. The agreement for clay/loam and sand soils is good, but it is poor for peat soils (not shown).

Fig. 7

Relation between soil texture and Sr TF

4.9 Classification by Sr TFs

The foregoing considerations lead to the following classification of soil for the uptake of Sr (Table 6).

Table 6 Classification of soils based on Sr TF

There were enough data sets to give a fair insight into the behavior of Cs for some soil units (fluvisols, luvisols, chernozems, and podzols), but others were only represented by one set of data. It is possible that one set of data is not representative. It was assumed that all lower-level units belong to the same soil group, but there is no proof. From some units, there was no representative at all. With soil group 1 (Cs classification), the majority of data refer to old contaminations, while for soil group 2, the data refer to recent contaminations; this could have unbalanced the choices made. The Sr classification is based on very limited data, and in addition, the measurement of Sr in some cases contained a large degree of uncertainty. The range of TFs is rather wide; the higher the values occurring with the most acid and dystric classes, the lower the values occurring with the calcaric and eutric classes. At the moment, there is not sufficient information to go into more detail. Within soil group 2 (Cs classification), there are two eutric and one dystric units listed. The TF of the latter unit is lower than that of the former two, but one cannot derive a rule from three sets of observations. One might think of a statistical analysis within a soil group, but lack of quantitative data on the individual factors affecting TF values precludes this. Despite this, the CRP made a successful beginning with the classification, and it is worthwhile to consider possible applications.

A classification of soil ecosystems might be a way to reduce uncertainties due to the enormous range of uptake parameters. Our contribution is to produce data on transfer factors of ¹³⁴Cs from soil to reference plants in a range of Bangladeshi soil systems to characterize systems in which TFs might differ substantially from what would be regarded as normal. The comparison between the results is obtained for TF of ¹³⁷Cs for wheat (grain and straw) during the 4 years of the experiment. The results can be summarized as follows:

• The TF observed for wheat (grain and straw) grown in soil type 1 was higher than for soil type 2.

• The TF for straw is higher than for grain as expected (e.g., Nisbet and Show 1994). The trends in the TF change with the time after contaminations are the same.

• The TF decreases with the time as expected for plants grown on soil type 1. Obviously, fixation of radiocesium takes place during the first year. According to the data obtained by this study, the TF values remain practically constant for the next 2 years within the biological variations. This coincides with the opinion of Frissel et al. (2002) that 1 year might be generally enough time for ¹³⁷Cs soil fixation, although for “safety” reasons, 4 years are to be considered.

• TF for plants on soil type 2, however, increases after the first year to return to the first-year results after the second year and the third year. Similar effects were established in Massas et al. (2002) for other crops. Obviously, for this type of soil, equilibrium for Cs is reached after a longer period, although the higher elite content suggests quicker fixation.

• Although the general expectation is that TF values follow lognormal distribution (e.g., Sheppard and Evenden 1990), the investigation of the skewness (a), defined as a = √2(ln E- lnM) (M median, E average), proved that the distribution of the TF in the present study tends to be normal (the value of a is <0.3).

• The present TF value for soil type 1 is very near to the reference TF values, published by IAEA (1994), Frissel et al. (2002). For soil type 2, however, the TF is within the IAEA interval and just below the lower limit of the range of Frissel et al. (2002). It is quite obvious that further refinement of the soil classification should be done using soil types and parameters because according to the preliminary classification in Frissel et al. (2002), both the investigated soils fall within the same group on the basis of nutrient status. Evidence in this direction is the variation of the TF values (more than 10-fold) in both IAEA (1994), Frissel et al. (2002).

• The TFs of potassium calculated for both the investigated soils show practically identical values, which obviously is due to the nutrient function of this element, and therefore, its uptake by the plant is not seriously influenced by the soil properties.

The transfer of radionuclides (¹³⁷Cs, ⁴⁰K, ²²⁶Ra, ⁹⁰Sr, ^239+240Pu, and ²⁴¹Am) had also been analyzed in previous work (Baeza et al. 2005, 2006). Compared to the uranium and thorium TF values, those of ¹³⁷Cs were generally higher, those of ⁴⁰K, ²²⁶Ra, and ⁹⁰Sr were similar or a little higher, and those of ^239+240Pu and ²⁴¹Am were similar (Fig. 8).

Fig. 8

Median and range of available transfer factors (TF) for artificial and natural radionuclides for several species of mushroom (Baeza et al. 2004)

Given the medians and ranges of the available transfer for mushrooms, the efficiencies of the radionuclide transfers were ranked as follows:

$$\begin{aligned} &^{228,230,232} {\text{Th}} \approx {}^{40}{\text{K}} \ge {}^{137}{\text{Cs}} \ge {}^{234,238}{\text{U}} \\ & \approx {}^{226}{\text{Ra}} \ge {}^{90}{\text{Sr}} \ge {}^{239 + 240}{\text{Pu}} \approx {}^{241}{\text{Am}}. \\ \end{aligned}$$

The migration and distribution of radiocesium in the soil profile varies depending on soil properties such as soil texture, organic matter content and pH and on climatic conditions, land use, and management practices. Important factors affecting the transfer of radiocesium to crops/plants are the distribution of the root system in the soil profile as well as the soil pH and nutrient status (IAEA 2010).

Both soil and plant samples of nine different plant species grown in soils from southeastern China contaminated with uranium mine tailings were analyzed (Chen et al. 2005) for the plant uptake and translocation of ²³⁸U, ²²⁶Ra, and ²³²Th. Substantial differences were observed in the soil–plant transfer factor (TF) among these radionuclides and plant species. Lupine (Lupinus albus) exhibited the highest uptake of ²³⁸U (TF value of 3.7 × 10⁻²), while Chinese mustard (Brassica chinensis) had the least (0.5 × 10⁻²). However, in the case of ²²⁶Ra and ²³²Th, the highest TFs were observed for white clover (Trifolium pratense) (3.4 × 10⁻²) and ryegrass (Lolium perenne) (2.1 × 10⁻³), respectively. In general, the TFs across all plant species for the three radionuclides were in the following order: ²³⁸U ~ ²²⁶Ra > ²³²Th.

5 Conclusions

Chemical properties, especially similarities with nutrient elements, determine the extent to which anthropogenic radionuclides will become involved in terrestrial nutrient cycling. Long-lived radioisotopes of cesium and strontium have received the greatest attention due to their similarity with K and Ca and hence bioavailability and their level of global deposition. Due to the focus of most studies on the human food chain, the best understood processes are those important for the soil-to-plant and plant-to-animal transfer. Due to the inherent complexity and spatial variability of the soil-plant system, the uptake of radionuclides in vegetation from soil is difficult to quantify. The transfer factor (TF) is a useful parameter, which is usually used for evaluating the impact of releases of radionuclides into the environment. The TF depends on the vegetation type, soil properties, and the type of radionuclides. The factors that determine the variation of radionuclide activity with depth in soil profile are rainfall intensity, pH, organic matter content, soil moisture content, soil texture, soil structure, infiltration rate, sorption characteristics of the particular radionuclide, land use, and management practices. Generally, the soil-to-plant transfer of radionuclides depends on soil type, pH, solid/liquid distribution coefficient, exchangeable K⁺, and organic matter contents. Absalom et al. (1999, 2001) presented a model that predicts the soil-to-plant TF of radionuclides (clay content, organic carbon content, exchangeable potassium, and pH). The Absalom model has been tested in Europe with successful prediction of the fate of Chernobyl and weapons fallout of ¹³⁷Cs. However, testing and validation of this model for the tropical food chains in many countries in South Asia is very limited. As countries in the South Asian region like Bangladesh is expanding applications of nuclear technology, a comparable model is required to predict the impact of deposited radionuclides based on the regional parameters derived for wet–dry tropical environments. In order to apply the Absalom model and/or to modify the model, regional databases for model validation need to be developed for the tropical environments. There are two promising alternative approaches to classifying soils with regard to radionuclide behavior. One uses correlation and factor analysis to identify the most important soil properties controlling TFs which can then be used predictively. The other is a semiempirical procedure based on the assumption that ion exchange capacity, pH, and organic matter content control ion availability in the soil.