Updated soil to fruit concentration ratios for radiocaesium compiled under the IAEA MODARIA II Programme.

Under the International Atomic Energy Agency (IAEA) Modelling and Data for Radiological Impact Assessments (MODARIA II) Programme, Working Group 4 activities included collating radionuclide transfer data from Japan following the Fukushima Daiichi Nuclear Power Plant accident and separately collating concentration ratio (CR) data for root uptake of radionuclides by crops grown in tropical and arid climates. In this paper, the newly compiled radiocaesium CR data for fruit from Japan, tropical and arid climates have been combined with the data originally compiled for the IAEA Technical Reports Series No. 472 (TRS 472) and additional data identified from the literature to produce an enhanced MODARIA II dataset of fruit radiocaesium CR values. Statistical analysis of the MODARIA II dataset by climate class (based on the Köppen-Geiger climate classification) indicated that the CR values for tropical climates were significantly higher (p< 0.05) than those for arid, temperate and cold climates. Statistical analysis of the MODARIA II dataset by soil group (based on soil texture) indicated that the CR values for coral sand soil (tropical climates only) and organic soil (temperate climates only) were significantly higher (p< 0.05) than those for the clay, loam and sand soil groups. Statistical analysis of the MODARIA II dataset by plant group (based on plant morphology) indicated that the CR values for non-woody trees (tropical climate bias) were significantly higher (p< 0.05) than those for herbaceous plants, shrubs and woody trees. Comparison of the MODARIA II dataset with original TRS 472 values showed only small changes in the fruit radiocaesium CR values for herbaceous plants and shrubs in temperate climates. There was a decrease in the CR values for woody trees in temperate climate across all soil groups. There was also a decrease in the CR values for tropical climates for all comparable soil groups.

Long-term transfer of global fallout ¹³7Cs to cow's milk in Iceland.

The aim of this study was to provide improved information on the long-term transfer of global nuclear weapons (137)Cs fallout to cow's milk in Iceland many years after deposition. The spatial variation in deposition was confirmed to be explained by precipitation. Soil samples showed a significant difference in (137)Cs deposition density between the main agricultural areas, with the South having the highest values, then the West and North and the lowest in the Northeast. There was no significant difference between the effective half-lives in (137)Cs activity concentrations in milk and milk powder from the main dairies in Iceland based on data for milk from 1990 to 2007 and for milk powder from 1986 to 2007. There was, however, a significant difference between the effective half-lives obtained for these two regions, 13.5 years for the Northern and 10.5 years for the Southern regions. These half-lives for global fallout are longer than those previously reported for similar time periods in other Arctic areas. The transfer of (137)Cs to cow's milk was quantified for different agricultural regions using aggregated transfer coefficients (T (ag)) for the period of peak global fallout soil inventory in 1965-1967. The values ranged from 2.8 × 10(-3) to 10.6 × 10(-3) m(2) kg(-1). By 2001-2004, the T (ag) values had only declined, in the main agricultural areas, to 0.6 × 10(-3)-1.0 × 10(-3) m(2) kg(-1). Long-term transfer rates to milk many years after deposition were high in Iceland compared with most other reported data. The transfer is potentially relevant for some of the contaminated areas around the Fukushima Nuclear Power Plant after the accident in March 2011 since limited information is available on uptake from Andosols and associated effective half-lives.

A multi-criteria weight of evidence approach for deriving ecological benchmarks for radioactive substances.

Dose rate benchmarks are required in the tiered approaches used to screen out benign exposure scenarios in radiological ecological risk assessment. Such screening benchmarks, namely the predicted no-effect dose rates (PNEDR), have been derived by applying, as far as possible, the European guidance developed for chemicals. To derive the ecosystem level (or generic) PNEDR, radiotoxicity EDR(10) data (dose rates giving a 10% effect in comparison with the control) were used to fit a species sensitivity distribution (SSD) and estimate the HDR(5) (the hazardous dose rate affecting 5% of species with a 10% effect). Then, a multi-criteria approach was developed to justify using an assessment factor (AF) to apply to the HDR(5) for estimating a PNEDR value. Several different statistical data treatments were considered which all gave reasonably similar results. The suggested generic screening value of 10 microGy h(-1) (incremental dose rate) was derived using the lowest available EDR(10) value per species, an unweighted SSD, and an AF of 2 applied to the estimated HDR(5). Consideration was also given to deriving screening benchmark values for organism groups but this was not thought to be currently appropriate due to few relevant data being currently available.

Protection of the environment from ionising radiation in a regulatory context--an overview of the PROTECT coordinated action project.

The outcome of the PROTECT project (Protection of the Environment from Ionising Radiation in a Regulatory Context) is summarised, focusing on the protection goal and derivation of dose rates which may detrimentally affect wildlife populations. To carry out an impact assessment for radioactive substances, the estimated dose rates produced by assessment tools need to be compared with some form of criteria to judge the level of risk. To do this, appropriate protection goals need to be defined and associated predefined dose rate values, or benchmarks, derived and agreed upon. Previous approaches used to estimate dose rates at which there may be observable changes in populations or individuals are described and discussed, as are more recent derivations of screening benchmarks for use in regulatory frameworks. We have adopted guidance and procedures used for assessment and regulation of other chemical stressors to derive benchmarks. On the basis of consultation with many relevant experts, PROTECT has derived a benchmark screening dose rate, using data on largely reproductive effects to derive species sensitivity distributions, of 10 microGy h(-1) which can be used to identify situations which are below regulatory concern with a high degree of confidence.

Predicting the radiation exposure of terrestrial wildlife in the Chernobyl exclusion zone: an international comparison of approaches.

There is now general acknowledgement that there is a requirement to demonstrate that species other than humans are protected from anthropogenic releases of radioactivity. A number of approaches have been developed for estimating the exposure of wildlife and some of these are being used to conduct regulatory assessments. There is a requirement to compare the outputs of such approaches against available data sets to ensure that they are robust and fit for purpose. In this paper we describe the application of seven approaches for predicting the whole-body ((90)Sr, (137)Cs, (241)Am and Pu isotope) activity concentrations and absorbed dose rates for a range of terrestrial species within the Chernobyl exclusion zone. Predictions are compared against available measurement data, including estimates of external dose rate recorded by thermoluminescent dosimeters attached to rodent species. Potential reasons for differences between predictions between the various approaches and the available data are explored.

An international model validation exercise on radionuclide transfer and doses to freshwater biota.

Under the International Atomic Energy Agency (IAEA)'s EMRAS (Environmental Modelling for Radiation Safety) programme, activity concentrations of (60)Co, (90)Sr, (137)Cs and (3)H in Perch Lake at Atomic Energy of Canada Limited's Chalk River Laboratories site were predicted, in freshwater primary producers, invertebrates, fishes, herpetofauna and mammals using eleven modelling approaches. Comparison of predicted radionuclide concentrations in the different species types with measured values highlighted a number of areas where additional work and understanding is required to improve the predictions of radionuclide transfer. For some species, the differences could be explained by ecological factors such as trophic level or the influence of stable analogues. Model predictions were relatively poor for mammalian species and herpetofauna compared with measured values, partly due to a lack of relevant data. In addition, concentration ratios are sometimes under-predicted when derived from experiments performed under controlled laboratory conditions representative of conditions in other water bodies.

Decision-making in environmental radiation protection: using the ERICA Integrated Approach.

The ERICA Integrated Approach and its associated tool and databases provide a method by which the likely impact of radioactive discharges on the environment can be evaluated, see Fig. 1. The various factors, which should be taken into account when making decisions both during and after an assessment has been made, are discussed for each stage in the assessment process for a hypothetical case study. The assessment will demonstrate the issues associated with the decision-making process at Tiers 1 and 2 within the ERICA Tool and how they may vary. The case study, set in England, evaluates the environmental impact of radioactive substances released under authorisation in response primarily to conservation legislation, because of the need to demonstrate that no adverse impacts will occur on Natura 2000 sites as a result of the release of an authorised substance.

Derivation of transfer parameters for use within the ERICA Tool and the default concentration ratios for terrestrial biota.

An ability to predict radionuclide activity concentrations in biota is a requirement of any method assessing the exposure of biota to ionising radiation. Within the ERICA Tool fresh weight whole-body activity concentrations in organisms are estimated using concentration ratios (the ratio of the activity concentration in the organism to the activity concentration in an environmental media). This paper describes the methodology used to derive the default terrestrial ecosystem concentration ratio database available within the ERICA Tool and provides details of the provenance of each value for terrestrial reference organisms. As the ERICA Tool considers 13 terrestrial reference organisms and the radioisotopes of 31 elements, a total of 403 concentration ratios were required for terrestrial reference organisms. Of these, 129 could be derived from literature review. The approaches taken for selecting the remaining values are described. These included, for example, assuming values for similar reference organisms and/or biogeochemically similar elements, and various simple modelling approaches.

Estimating the exposure of small mammals at three sites within the Chernobyl exclusion zone--a test application of the ERICA Tool.

An essential step in the development of any modelling tool is the validation of its predictions. This paper describes a study conducted within the Chernobyl exclusion zone to acquire data to conduct an independent test of the predictions of the ERICA Tool which is designed for use in assessments of radiological risk to the environment. Small mammals were repeatedly trapped at three woodland sites between early July and mid-August 2005. Thermoluminescent dosimeters mounted on collars were fitted to Apodemus flavicollis, Clethrionomys glareolus and Microtus spp. to provide measurements of external dose rate. A total of 85 TLDs were recovered. All animals from which TLDs were recovered were live-monitored to determine (90)Sr and (137)Cs whole-body activity concentrations. A limited number of animals were also analysed to determine (239,240)Pu activity concentrations. Measurements of whole-body activity concentrations and dose rates recorded by the TLDs were compared to predictions of the ERICA-Tool. The predicted (90)Sr and (137)Cs mean activity concentrations were within an order of magnitude of the observed data means. Whilst there was some variation between sites in the agreement between measurements and predictions this was consistent with what would be expected from the differences in soil types at the sites. Given the uncertainties of conducting a study such as this, the agreement observed between the TLD results and the predicted external dose rates gives confidence to the predictions of the ERICA Tool.

Application of the ERICA Integrated Approach to the Drigg coastal sand dunes.

The EC-funded project 'Environmental Risks from Ionising Contaminants: Assessment and Management' (ERICA) developed an 'Integrated Approach' for assessing the impact of ionising radiation on ecosystems. This paper presents the application of the ERICA Integrated Approach, supported by a software programme (the ERICA Tool) and guidance documentation, to an assessment of the Drigg coastal sand dunes (Cumbria, UK). Targeted sampling provided site-specific data for sand dune biota, including amphibians and reptiles. Radionuclides reported included (90)Sr, (99)Tc, (137)Cs, (238)Pu, (239+240)Pu and (241)Am. Site-specific data were compared to predictions derived using the ERICA Tool. Some under- and over-predictions of biota activity concentrations were identified but can be explained by the specific ecological characteristics and contamination mechanism of the dunes. Overall, the results indicated no significant impact of ionising radiation on the sand dune biota and the Integrated Approach was found to be a flexible and effective means of conducting a radiation impact assessment.

Review of Russian language studies on radionuclide behaviour in agricultural animals: Transfer to animal tissues.

Data on radionuclide transfer to animals from research performed in the former Soviet Union were reviewed to collate transfer coefficient values (Ff) to animal tissues such as liver, kidney and bone, but not muscle which has previously been reported. The derived values were compared with selected data published in the English language literature. The new data are mainly for 90Sr and 137Cs, although some data were also provided for 3H, 54Mn, 59Fe, 60Co, 22Na 65Zn, 131I and U. The Russian language data may provide a basis for better informed evaluation of radiation dose from the consumption of such animal products, which can form important components of the diet in some countries.

Radiostrontium uptake by plants from different soil types in Kazakhstan.

The transfer of 90Sr to a range of different plant species grown on a range of different soil types in Kazakhstan, including three from the Semipalatinsk Test Site (STS), has been measured in a lysimeter experiment. 90Sr uptake by Stipa spp was significantly higher than for other vegetation species. The uptake of 90Sr from chernozem was significantly lower than that from the other soil types which is consistent with other literature. There was a significant negative relationship between 90Sr uptake and calcium, humus and CEC concentration in the soil for Agropyrum spp, Artemisia spp but not for Stipa spp or Bromus spp. The transfer to vegetation from soil has been quantified using the aggregated transfer coefficients for each species. Tag values range from 0.6 to 11.9 m2 kg-1x10(-3) over all measurements. The transfer of 90Sr to plants from the Kazakh soils was low compared to previously reported data and to that given from literature reviews.

The transfer of radiocaesium to ewes through a breeding cycle--an illustration of the pitfalls of the transfer coefficient.

A study to measure the transfer of radiocaesium to adult female sheep through a breeding cycle is described. The transfer of radiocaesium from the diet to muscle (estimated as the equilibrium transfer coefficient) was significantly lower to pregnant, and especially lactating, animals compared to non-lactating and barren animals. High dry matter intake rates were also associated with significantly lower transfer coefficients. Known relationships between dry matter intake rates and protein turnover could credibly explain some of these differences. However, when described as the concentration ratio, radiocaesium transfer to meat was apparently highest during lactation. The apparent difference in results obtained by the two approaches of determining transfer is the consequence of daily dry matter intake being a denominator within the estimation of transfer coefficient. A wider discussion of transfer coefficients and concentration ratios leads us to suggest that the concentration ratio is the more robust and potentially generic parameter.

The transfer of radionuclides from saltmarsh vegetation to sheep tissues and milk.

Radionuclides released into the Irish Sea by the Sellafield reprocessing plant are deposited onto tide-washed pastures along the western coast of the United Kingdom. Many of these pastures are grazed by sheep or cattle. This paper describes a controlled feeding study, in which saltmarsh vegetation harvested from close to the Sellafield plant, was fed to lambs and adult female sheep for a period of 8 weeks. Activity concentrations of (60)Co, (95)Nb, (106)Ru, (134)Cs, (137)Cs, (238)Pu, (239,240)Pu and (241)Am were determined in edible tissues and transfer parameters estimated. The activity concentrations of some of the radionuclides will not have been in equilibrium with those in the diet. Nevertheless, the study was reasonably realistic in terms of agricultural management as the period of the study was similar to that for which lambs graze on the saltmarshes. A field study to determine the activity concentrations of (137)Cs and (239,240)Pu in the milk of ewes grazing a saltmarsh close to Sellafield is also described.

Review of Russian language studies on radionuclide behaviour in agricultural animals: part 1. Gut absorption.

An extensive programme of experiments was conducted in the former USSR on transfer of radionuclides to a wide range of different agricultural animals. Only a few of these studies were made available in the English language literature or taken into account in international reviews of gastrointestinal uptake. The paper gives extended information on Russian research on radionuclide absorption in the gut of farm animals performed in controlled field and laboratory experiments from the 1960s to the current time. The data presented in the paper, together with English language values, will be used to provide recommended values of absorption specifically for farm animals within the revision of the IAEA Handbook of Parameter Values IAEA [International Atomic Energy Agency, 1994. Handbook of Parameter Values for the Prediction of Radionuclide Transfer in Temperate Environments, IAEA technical reports series No. 364. International Atomic Energy Agency, Vienna].

Factors contributing to radiocaesium variability in upland sheep flocks in west Cumbria (United Kingdom).

Following the Chernobyl accident in 1986, restrictions were placed on the movement and slaughter of sheep within upland areas of the UK because radiocaesium activity concentrations in their meat exceeded 1000 Bq kg(-1) fresh weight. Some farms remain under restriction in 2007. From 1991 to 1993 detailed studies were conducted on three sheep farms within the restricted area of west Cumbria to systematically assess the various parameters which may contribute to the observed variability in radiocaesium activity concentrations within sheep flocks. This paper reports the spatial variation in soil and vegetation activity concentrations across the grazed areas at these farms and determines the influence of grazing behaviour on variability in (137)Cs activity concentrations between individual sheep within the flocks. Together with previously reported results, these new data are used to draw conclusions on the factors determining variability within the three flocks. However, the factors are too site specific to be able to generalise the findings to other farms within the restricted areas of the UK.

Review of Russian language studies on radionuclide behaviour in agricultural animals: part 2. Transfer to milk.

An overview of original information available from Russian language papers on radionuclide transfer to milk is provided. Most of the data presented have not been taken into account in international reviews. The transfer coefficient (F(m)) values for radioactive isotopes of strontium, caesium and iodine are in good agreement with those previously published. The Russian language data, often based on experiments with many animals, constitute a considerable increase to the available data for many less well-studied radionuclides. In some instances, the Russian language data suggest changes in recommended values (e.g. Zr and Ru). The information presented here substantially increases the amount of available data on radionuclide transfer to milk and will be included in the current revision of the IAEA TRS Handbook of parameter values for radionuclide transfer.

The transfer of different forms of 35S to goat milk.

Sulphur-35 is released during the routine operation of UK gas-cooled reactors. An experiment to determine the rates of transfer of different forms of (35)S to goat milk is described. Lactating goats received (35)S orally as single administrations of sulphate, L-methionine, or grass contaminated either through root uptake of (35)S as sulphate or through aerial deposition of (35)S as carbonyl sulphide onto the grass. Transfer was higher for (35)S administered as methionine compared with (35)S administered as sulphate. Changes in activity concentrations in milk for all sources of (35)S demonstrated two components of loss. The first component had a half-life of circa 1 d for all sources, the second was longer in goats administered carbonyl sulphide (44 d) than in all of the other treatments (circa 10 d). The rate of transfer of (35)S to milk of a further group of goats receiving (35)S-sulphate daily appeared to reach equilibrium within 30 d. Extrapolation of transfer parameters derived to other dairy ruminants is discussed.

The transfer of 239/240Pu to cow milk.

In summer 1993 we measured the transfer of (239/240)Pu to milk from herbage from a pasture located 5 km from the Chernobyl nuclear power plant. In one treatment cows were allowed to graze freely on the pasture. In a second treatment, cows were fed herbage collected from the pasture in stalls. The milk transfer coefficient; F(m) did not vary significantly between treatments and the mean value of 7.5x10(-6)d l(-1) was higher than previously reported values. Despite many values of F(m) for Pu in the literature we identified few relevant original data sets. Transfer coefficient values for Pu are only appropriate when used in conjunction with a specified time period or an appropriate model which allows for the biological half-life. We recommend for screening purposes an F(m) value of 1x10(-5)d l(-1) for Pu, with an order of magnitude lower value being appropriate for cows which are only exposed for one grazing season.

Improving the quantity, quality and transparency of data used to derive radionuclide transfer parameters for animal products. 2. Cow milk.

Under the International Atomic Energy Agency (IAEA) MODARIA (Modelling and Data for Radiological Impact Assessments) Programme, there has been an initiative to improve the derivation, provenance and transparency of transfer parameter values for radionuclides from feed to animal products that are for human consumption. A description of the revised MODARIA 2016 cow milk dataset is described in this paper. As previously reported for the MODARIA goat milk dataset, quality control has led to the discounting of some references used in IAEA's Technical Report Series (TRS) report 472 (IAEA, 2010). The number of Concentration Ratio (CR) values has been considerably increased by (i) the inclusion of more literature from agricultural studies which particularly enhanced the stable isotope data of both CR and Fm and (ii) by estimating dry matter intake from assumed liveweight. In TRS 472, the data for cow milk were 714 transfer coefficient (Fm) values and 254 CR values describing 31 elements and 26 elements respectively. In the MODARIA 2016 cow milk dataset, Fm and CR values are now reported for 43 elements based upon 825 data values for Fm and 824 for CR. The MODARIA 2016 cow milk dataset Fm values are within an order of magnitude of those reported in TRS 472. Slightly bigger changes are seen in the CR values, but the increase in size of the dataset creates greater confidence in them. Data gaps that still remain are identified for elements with isotopes relevant to radiation protection.