Introduction

The input of fission products into the atmosphere reached its maximum in 1962, due to fusion bomb tests performed by the USSR at Novaya Zemlya and the USA at Johnston and Christmas Island [1]. Concerning the radiological impact, 90Sr was one of the most important fallout radionuclides, due to its long half-life, the high-energy beta radiation of its daughter 90Y, its similar behaviour to calcium in the food chain, and its long-term accumulation in the human body. Aiming at minimization of the internal dose after ingestions via the food chain, research was therefore initiated to identify crop species with low uptake of 90Sr. As a consequence of the Chernobyl accident in 1986, large areas in the former Soviet Union were contaminated with high activities of 137Cs and 90Sr, which made agricultural use impossible. The Chernobyl accident revived the need and interest in research on crops with low 137Cs and 90Sr uptake. In 2005, 444 nuclear power plants in 31 countries with an aggregate electrical power of 370 GWe were in operation, four new reactors went into operation, and 23 new power plants in ten countries were under construction [2]. With respect to the growing importance of nuclear energy [3–5], concepts for the management of contaminated landscapes are required. Different strategies for countermeasures that would allow food production on contaminated areas have been summarized recently by Howard and co-workers within the STRATEGY project [6]. Twenty-nine agricultural countermeasures were selected after critically reviewing of the literature. The cultivation of certain species of crops was not recommended, however, because the authors felt that the available data were not sufficient to allow for an appropriate crop cultivar selection. However, the uptake of 90Sr and 137Cs by different varieties of crops should be studied in more detail because ‘the proper selection of plant varieties is one of the simplest, most effective, and least expensive measures in reducing the contamination of agricultural products’ [7]. Especially when other cultivars of a crop type are used that already had been cultivated in a certain area prior to a nuclear accident, social and economic consequences could be reduced significantly compared to the application of most other types of countermeasures.

Wheat is one of the most important foodstuff. For this reason, its uptake of 90Sr and 137Cs has already been studied in the 1950s (e.g. [8]). Recently, our group published the results of a study on the soil-to-plant transfer of fallout 137Cs in 28 winter wheat cultivars, each of them grown in 1999 at three different locations in Upper Bavaria, Germany [9]. In this study it was shown that the concentration ratio C r (Eq. 1) of 137Cs depends primarily on the sites where the wheat was grown, which can be explained by locally different soil characteristics.

$$ C_{{\text{r}}} {\text{ = }}\frac{{a_{{{\text{grain}}}} }} {{a_{{{\text{soil}}}} }} $$
(1)

where C r is the concentration ratio for a given wheat cultivar, and a grain and a soil are the activity concentrations in grain and soil, both given on a dry-mass basis.

The obtained mean C r values were 0.00049 ± 0.00016 in Gereuth, 0.00075 ± 0.00009 in Pettenbrunn, and 0.00042 ± 0.00008 in Pulling, respectively. The maximum-to-minimum ratio (MMR) of the C r for 137Cs was 3.1, 1.5 and 2.2 at the three sites, respectively. For a better comparison of the cultivars normalized soil-to-grain ratios NCr’s (Eq. 2) and site-averaged normalized soil-to-grain ratios site-averaged normalized concentration ratios (SANCr’s) (Eq. 3) were calculated, to eliminate the influence of the different local soil characteristics.

$$ {\text{NC}}_{{\text{r}}} \;{\text{ = }}\;\frac{{C_{{\text{r}}} }} {{\overline{C} _{{\text{r}}} }}\;{\text{ = }}\;n\frac{{a_{{{\text{grain}}}} }} {{{\sum\nolimits_{i{\text{ = 1}}}^n {a_{{{\text{grain,}}i}} } }}} $$
(2)

where NCr is the normalized soil-to-grain ratio for a given wheat cultivar i at a site and n the total number of cultivars

$$ {\text{SANC}}_{{\text{r}}} {\text{ = }}\frac{{\text{1}}} {{\text{3}}}{\sum\limits_{j{\text{ = 1}}}^{\text{3}} {{\text{NC}}_{{{\text{r,j}}}} } } $$
(3)

where SANCr is the site-averaged normalized soil-to-grain ratio for a given wheat cultivar, and j is the index of the site.

The cultivars Flair (SANCr = 0.74 ± 0.13), Kornett (0.76 ± 0.14), and Previa (0.79 ± 0.08), were found to show SANCr values significantly lower than the mean SANCr of all cultivars. In other words, these species showed a smaller uptake of 137Cs than the other cultivars.

Apart from 137Cs, the strontium isotope 90Sr is the radionuclide of major importance to the radiological long-term hazard of fallout from nuclear accidents, as was pointed out for example by Mück [10]. Therefore, the aim of this work was to determine the soil-to-grain-transfer of 90Sr in the 28 wheat cultivars from the three sites in Bavaria that were described above, and to identify cultivars with significantly lower uptake of 90Sr compared to other winter wheat cultivars. Concerning their possible practical application in food production on contaminated farmland, cultivars with both minimum uptake of 90Sr and of 137Cs would be ideal.

Materials and methods

Details of the sampling strategy and location, soil properties, grain yields and the determination of 137Cs have already been described elsewhere [9]. Briefly, the sampling locations were near the villages Gereuth, Pettenbrunn, and Pulling, which are located within a radius of about 10 km around the city of Freising (Upper Bavaria, Germany). Selected soil parameters of the sampling locations are given in Table 1. As can be seen from Table 1 the soils were well supplied with nutrients [11]. About 1 kg of grains were air-dried, cleaned, sieved, and ashed at a temperature of 400°C. Soil samples (about 2 kg) were dried and sieved (2 mm), four sub-samples of 50 g were used for 90Sr analysis. 137Cs and 40K were determined by gamma spectrometry. After these measurements the samples were ashed at 550°C and leached with 65% nitric acid. 90Y was extracted with tributyl phosphate (TBP) and purified by hydroxide and oxalate precipitation and, if necessary, by the removal of thorium isotopes by means of anion exchange chromatography. Finally, 90Y was determined by proportional or Cherenkov counting. The radiochemical purity of the counting samples was assured by controlling the 90Y decay curves. The chemical yield of Y, determined gravimetrically and by complexometric titration with ethylenediamine tetraacetic acid (EDTA), was in the range of 90%. Minimum detectable activities (MDAs) were calculated according to [12]. Typical MDA’s were in the range of 7–12 mBq per sample. Quality assurance was done by participating national round-robin tests. Calibrations were performed with certified reference standard solutions purchased from the physikalisch–technische bundesanstalt (PTB), Braunschweig, Germany. Details of the radiochemical procedure are described in [13]. Statistical data analysis was performed with the software package STATISTICA 7.1 (Statsoft, Tulsa, USA, 2005)

Table 1 Selected soil parameters of the Ap horizons in Gereuth, Pettenbrunn and Pulling [9]
Full size table

Results

Table 2 shows the 90Sr and 137Cs activity concentrations and the 137Cs/90Sr activity ratios of the Ap soil horizons at the three locations. The 90Sr activity concentrations range from 1.08 to 2.06 Bq kg−1, which corresponds to 314–514 Bq m−2. The corresponding 137Cs/90Sr activity ratios range from 14 to 20 which are typical values for the region near Munich. The small standard deviations of the 90Sr concentrations in the soil indicate a small spatial variability of the local 90Sr fallout at the study sites, as was also found for 137Cs [9]. Table 3 shows the 90Sr activity concentrations of the wheat grains. For better comparability with literature data, Table 3 also includes the corresponding mean Tag values (Bq kg−1 grain/Bq m−2 soil) which are often used in the Russian literature. Mean 90Sr activity concentrations in the grains were 0.163, 0.286 and 0.123 Bq kg−1, respectively (for Gereuth–Pettenbrunn–Pulling; unless otherwise stated, all following data on the three experimental sites will be reported in this order). For all sites the 90Sr activity concentrations were log-normal distributed (Shapiro–Wilkinson). There was no correlation found between the grain yields and the 90Sr activity concentrations in the grains, at any site. Based on the data shown in Tables 2 and 3, concentration ratios C r (Eq. 1) for each variety at each site, and their MMR were calculated (s. Table 4). Mean C r were 0.151 ± 0.029, 0.205 ± 0.035 and 0.060 ± 0.012, respectively. As shown by the MMR, the C r for a given cultivar varied by a factor from 2.46 to 5.00 between the sites. The C r values between the cultivars at a given site varied by factors of 2.57, 1.89 and 2.48, respectively.

Table 2 90Sr and 137Cs [9] activity concentrations (±combined standard uncertainty) and 137Cs/90Sr activity ratios (±1 standard deviation) in the investigated Ap soil horizons
Full size table
Table 3 Activity concentrations (Bq kg− 1 dry mass) of 90Sr (±combined standard uncertainty) in wheat grains
Full size table
Table 4 Concentration ratios C r (Bq kg− 1 dry mass/Bq kg− 1 Ap soil horizon) of 90Sr (±combined standard uncertainty) in wheat grain and MMR between the C r of the three sites
Full size table

In order to eliminate the site-specific influences on the uptake of 90Sr, for each cultivar NCr values (Eq. 2) were calculated for each site, as well as the corresponding SANCr values (Eq. 3). The resulting NCr and SANCr values are shown in Figs. 1 and 2, for all cultivars. In Fig. 2, the culivars were sorted by increasing SANCr values, which ranged from 0.6656 ± 0.0617 to 1.5027 ± 0.1610.

Fig. 1
figure 1

Normalized concentration ratios (NCr) for the soil-to-grain-transfer of 90Sr in Gereuth, Pettenbrunn and Pulling

Full size image
Fig. 2
figure 2

Site averaged normalized concentration ratios (SANCr) for the soil-to-grain-transfer of 90Sr in Gereuth, Pettenbrunn and Pulling

Full size image

Discussion

The C r values for 90Sr obtained in the present work agree very well with the mean recommended value of 0.15 given in [14, 15] for the soil-to-grain transfer of cereals grown on loamy soils, and with the corresponding mean concentration ratios (identical with transfer factors) recently published by the German Federal Ministry for Environment, Nature Conservation and Nuclear Safety (BMU) [16]. From Table 4 it becomes obvious that the uptake of 90Sr by wheat grains mainly depends on the site characteristics, and to a lesser degree on the wheat cultivars. The mean C r values for 90Sr are clearly but negatively associated with the Ca2+ content in the soil (Table 1). Although it was often observed that the uptake of strontium by crops is suppressed by calcium [17], an evaluation of a big pool of data by the BMU revealed a significant negative correlation only for sandy, but not for loamy soils [16].

As mentioned above, research on the uptake of 90Sr by different wheat cultivars was already initiated decades ago. However, results reported in the literature are difficult to compare, due to inconsistent data presentation or different experimental conditions. For example, it was shown recently that the mean transfer factors for radiocaesium and radiostrontium strongly depend on the experimental setups (Table 5) [16]. However, in most cases it was possible to calculate MMR values of different cultivars of wheat grown up under identical conditions. As shown in Table 6, MMR values of different wheat cultivars vary by factors between about 1.6 and 4.2. The MMR values presented in this work for the three sites (2.57, 1.89 and 2.48) agree very well with data already published in the literature. The observed variability of 90Sr uptake by wheat is in the same range as the variability between different types of cereals. This has to be taken into account when comparing the uptake of these radionuclides by different types of crops. In 1965, Lee and Sosulski [18] already concluded that ‘the validation of determining differences between crops on the basis of single varieties to represent each crop may be questionable’. Malikov and co-workers [19] found an MMR value of 2.0, while Green and co-workers [20] found MMR values of 3–4 and pointed out that these factors are comparable to the differences of the same cultivar in different years. Unfortunately, Green’s data only refer to cereals as a group, including wheat, barley and oats.

Table 5 Mean transfer factors for the soil-to-grain transfer of radiostrontium
Full size table
Table 6 Ratios between wheat cultivars with maximum and minimum uptake of radiostrontium
Full size table

Putyatin and co-workers [21] compared the uptake of 137Cs and 90Sr in six cultivars of spring wheat. They identified cultivars with significantly lower uptake and reported MMR values to be 1.6 for both radionuclides. A short summary of the variability of the uptake of 90Sr by wheat grains is given in Table 7. These data again support Green’s statement that the MMR of different cereals in one year are in the same range as the MMR of the same cultivar in different years. Thus, more effort has to be done to quantify this variation, in order to identify suitable wheat cultivars for mid- and long-term application.

Table 7 Variability of the MMR of the uptake of radiostrontium in wheat grains
Full size table

With respect to the reduction of internal doses due to the uptake of 137Cs and 90Sr it would be desirable to identify wheat cultivars with minimum uptake of both 90Sr and 137Cs. Table 8 compares the SANCr for 90Sr obtained in this work with those for 137Cs reported in [9], for each cultivar. Some cultivars showed SANCr values significantly lower than one for 90Sr, others for 137Cs. However, none of the investigated cultivars showed both a significantly lower uptake of 90Sr as well as of 137Cs. In Fig. 3, the SANCr values for 90Sr as determined in this work are plotted against those for 137Cs [9]. In addition we also calculated normalized concentration ratios based on the data published by Malikov and co-workers (Table 3 in [19]) and by Putyatin and co-workers (Table 1 in [21]). These are the only publications known to us which deal with the variety of both 90Sr and 137Cs in different wheat cultivars. Figure 3 supports the hypothesis that there is no correlation between the uptake of 137Cs and 90Sr by wheat grains. This missing correlation is expected because 90Sr, an earth alkali metal, and 137Cs, an alkali metal, show different properties, as far as their fixation to soil particles and the mechanisms governing plant uptake are concerned. Since the uptake of 90Sr and 137Cs is shown to be independent parameters, it should be possible to optimize (decrease) them both by specific breeding. Putyatin and co-workers identified a certain cultivar (Quattro) which had a ‘significantly’ lower uptake of 90Sr as well as 137Cs. This cultivar also fits our criterion (SANCr value +1 standard deviation ≤ 0.9) for ‘significantly lower’ uptake. However, by considering all available data shown in Fig. 3 this appears to be a fortunate coincidence.

Table 8 Site average normalised concentration ratios (SANCr) of 90Sr and 137Cs [9]
Full size table
Fig. 3
figure 3

Comparison of the SANCr of 90Sr and 137Cs obtained in this work, and reported by Schimmack [9], Malikov [19] and Putyatin [21]

Full size image

Conclusion

It was shown that 90Sr is accumulated in grains of wheat with a mean C r value of 0.14, varying by factors up to 2.6 from cultivar to cultivar at a given site, and by factors up to 5.0 for a given cultivar from site to site. No correlation could be found between the uptake of 137Cs and 90Sr. Thus, when cultivating wheat cultivars at highly contaminated sites, a decision might be required on the selection of cultivars with minimum 90Sr uptake or minimum 137Cs uptake. Based on the C r values presented in this work and on those presented by Putyatin and co-workers [21], maximum acceptable 137Cs and 90Sr soil activity concentrations for contaminated farmland can be calculated. For this purpose we used the sanitary standard 10-117-99 of Belarus [22], as an example. For the European Union no applicable regulation exists. Other prescriptive limits for 90Sr and 137Cs in wheat grains used for human nutrition production were not known to us. Using our mean C r values of 0.00055 for 137Cs and 0.14 for 90Sr, wheat grown on a soil contaminated by more than 79 Bq kg−1 90Sr or 164 kBq kg−1 137Cs would show higher activity concentrations than allowed by the sanitary standard 10-117-99 of Belarus. Based on the deduced concentration ratios this corresponds to a 137Cs/90Sr ratio of 2,080. Using the C r values given in [19] this ratio would be 220 (Table 9). Such high ratios are generally not found in soil contaminated by the Chernobyl fallout for which typical ratios significantly below 100 were reported [1, 23–34]. This suggests that investigations on identifying, cultivation and breeding of wheat cultivars with minimum uptake of radionuclides should focus on 90Sr. However, since in other countries other limits have to be taken into account, the national regulations as well as the local 137Cs/90Sr activity ratios have to be considered when selecting the best wheat cultivar.

Table 9 Maxima for the 137Cs/90Sr activity ratio for soils containing the maximum possible 90Sr activity concentrations calculated with different concentration ratios
Full size table