The importance of radioecology to the discussion of radioactive contamination is discussed here. This paper discusses the history of radioecology, alongside the most recent developments in the science. It describes the need for more environmental data, and explains where the gaps in current knowledge lie. The calculation of radiation dose to wildlife along with the complications in performing such calculations is discussed. The paper also tackles the difficult question of the implications of radioecology on nuclear waste management and site decommissioning policies. From the beginnings of the science to today, radioecology is poised to be an important field of study as humans continue to rely on ionizing radiation to improve their lives.
“Radioecology is the study of behavior and effects of radioactive elements in the environment. This can be broken into three subdivisions: radionuclide movement within ecological systems and accumulation within specific ecosystems components such as soil, air, water, and biota; the effects of ionizing radiation on individual species, populations, communities, and ecosystems; and the use of radionuclides and ionizing radiation in studies of structure and function of ecosystems and their component subsystems” [
Radiation protection has historically focused on humans as the object of protection standards. In 1977, the International Commission on Radiological Protection (ICRP) stated that “Radiation protection has historically been solely focused on human protection, with the reasoning that… the level of safety required for the [radiation] protection of all human individuals is thought likely to be adequate to protect other species, although not necessarily individual members of these species. The Commission therefore believes that if man is adequately protected then other living things are also likely to be sufficiently protected” [
In 2008, ICRP recommendations shifted toward a new paradigm, one in which wildlife populations are considered to be their own protection endpoint. This decision has raised numerous challenges for the scientific community, and efforts are currently in progress to deepen the understanding of how radiation impacts non-human biota (NHB). With this new paradigm, evaluating radiation dose to biota in response to planned, existing, or emergency situations is of significant concern. As doseresponse relationships are not currently well understood for NHB, it is imperative to make an accurate determination of dose rates at which NHB are exposed.
This paper aims to place radioecology in the context of radiological contamination, with the intent to inform readers about this field of study and some of the ongoing work to improve radioprotection of the environment.
There is confusion over appropriate endpoints of protection. For humans, we are interested in reducing the lifetime cancer risk to as low as reasonably achievable, a stochastic endpoint, as our radiation protection standards virtually rule out deterministic effects. The dose-response relationship for humans is assumed to be linear, with no threshold, meaning that any increase in dose is an increase in risk. The dose response relationship for NHB is unknown in many cases.
There is also a scarcity of long-term multigenerational studies for NHB. One study shows that bank voles living in the Chernobyl exclusion zone had levels chromosomal aberrations in bone marrow that remained constant with each generation, despite decreasing levels of dose. In addition, the percent mortality of their embryos increased with time [
Finally, there is a growing abundance of data to challenge established paradigms. Several studies done by Møller and Mousseau in the Chernobyl region suggest that there may be population level effects at radiation doses below that which were previously assumed safe [4,5]. These studies challenge our belief that low levels of radiation (e.g., a few multiples of background dose rates) are essentially harmless to NHB. They highlight the need for a coordinated research effort to both qualify the types of effects and quantify their level of damage. There is ongoing research that is attempting to assist in the evaluation of radiation dose received by NHB.
It is generally taken as true that site specific analytical results are more representative of a location than generic data. Studies of analogues may be used for site analysis, but the information should be limited to parameter fitting [
An analysis of the source term data for the biosphere sub-model of Yucca Mountain by Higley et al. [
a peer reviewed article, 210 were from institutional publications, 140 had no listed reference, and 49 were derived during the creation of the model by the authors [
There are ways to fill data gaps for site assessments. Collection of samples of opportunity from locations representative of the site can fill gaps for specific locations [10,11]. Collection efforts, along with, rapid processing and analysis can provide data that is accurate for the site and applicable.
When transfer (e.g., concentration ratio) data is collected, there are no guidelines for filling in data gaps. What is collected is at the discretion of the team of researchers completing the study. The possibilities for research are open ended, but tend to be focused in specific directions towards certain elements and species.
There are two ways to consider how data is available, the first is element specific. IAEA Technical Document 1616 [
The second methodology for transfer is by animal, but this too has data gaps. The number of listed references for human protein and milk sources was categorized in IAEA Technical Document 1616 [
The amount of data for plants and animals that are not art of the western food chain is even sparser. A review of pasture grazers by Brown et al. [
From both conceptual methodologies, the IAEA considers the amount of data for cesium and strontium to be acceptable and nearly adequate for uranium, radium, manganese and cobalt. These six elements have more than 500 listed references available; the other elements considered have not been studied to the same extent and do not have enough data points to be considered sufficiently categorized.
Several approaches have been developed, tested and compared for the computation of radiation dose in non-human biota. The ICRP has developed a comprehensive approach that includes the use of Reference Animals and Plants (RAPs) for the assessment of dose to NHB.
ICRP’s current approach to dosimetry calculations for NHB relies on simplified organism geometry. Organisms are modeled as ellipsoids made of homogeneous ICRU four-component soft tissue [
Voxel models are three dimensional models created from radiological imaging modalities (e.g. computed tomography and magnetic resonance imaging). These models allow radioecologists to calculated organ-specific DCFs from heterogeneously distributed radionuclides. Voxel models are more robust and easily defendable than the previously used ellipsoidal models.
a 3-dimensional model is created. 3D Doctor allows for the export of a boundary file, which specifies the coordinates of each segmented tissue on a three dimensional matrix grid. This file is imported into a program created by the Human Monitoring Laboratory called Voxelizer [
decision-making process, such that environmental remediation, removal of wildlife, or other actions deemed necessary can be taken.
Performance assessments comprise the formal means by which the long-term safety of nuclear waste repositories is evaluated. In its peer review of the biosphere modeling program of the US Department of Energy’s Yucca Mountain site characterization project, the International Atomic Energy Agency states that an essential component of a performance assessment is the development and application of a methodology for assessing the potential impact of any future releases of radionuclides that may reach the surface environment, i.e. the biosphere [
Broadly speaking, the modeling process for radioactive waste internment or site decommissioning involves 1) construction of a conceptual model that describes the system and includes each of the important processes and their couplings, 2) translation of the conceptual model into a mathematical model and coding it into a computer program, 3) verification of the numerical correctness of the code, and 4) validation of the code’s applicability to the repository system to assess its predictive capabilities [
The last two components, release of radionuclides to the biosphere and their subsequent uptake by biota, are most relevant to the field of radioecology. Biosphere models must consider 1) radionuclide transport through many food chain pathways, such as deposition on soil and vegetation via irrigation water; 2) crop interception and retention; 3) radionuclide accumulation in soils as a result of long-term deposition by irrigation; 4) radionuclide leaching from the soil and retention mechanisms in root zones; 5) re-suspension of contaminated soil onto vegetation; 6) soil-to-plant uptake via roots; 7) transfer of radionuclides from feed to animal products; and 8) food ingestion rates of humans [
Although they are highly complex, these models by necessity include hypotheses, assumptions, and simplifications. Biosphere models are generally based on empirically determined bio-concentration factors, which predict radionuclide concentrations in plants and animals based on their concentrations in the environment (soil, water) or an animal’s diet. Concentration ratio (CR) is the ratio of radionuclide concentration in plant or animal tissue to the radionuclide’s concentration in the surrounding medium (soil, water). Transfer coefficients, also known as transfer factors, relate radionuclide concentrations in an animal’s diet to radionuclide concentrations in foods produced from the animal.
The most minimal acceptable model of radionuclide uptake by vegetation is illustrated by Robertson et al. in their review of plant and animal transfer factors used in performance assessment models [
where:
Cdci(Tyr) = concentration of radionuclide i on plant type c at harvest from deposition processes for a one-year period (Bq/kg wet weight)
Tyr = one-year exposure period (y)
Rid = constant dry deposition rate of radionuclide i (Bq/m2y)
Riw = constant wet deposition rate of radionuclide i (Bq/m2y)
Γdc = interception fraction from airborne dry deposition for plant type c (dimensionless)
Γwc = interception fraction for airborne wet deposition to plant type c (dimensionless)
Cci(Tyr) = average concentration of radionuclide i in farmland soil for crop type c for the current one-year period (Bq/m2)
RFc = re-suspension factor for crop soil (m−1)
Vdi = deposition velocity of radionuclide i (m/s)
Γic = interception fraction for irrigation deposition to plant type c (dimensionless), generally equal to Γwc
Mc = fraction of the year irrigation takes place for plant type c TVc = translocation factor for plant type c (dimensionless)
Bc = total standing biomass for plant type c (kg wet weight/m2)
λei = effective loss rate constant from plant surfaces representing weathering and radioactive decay for radionuclide i (y−1)
λei = λwi + λI
λwi = weathering rate constant for crops for radionuclide i (y−1)
Tgc = crop growing period for plant type c (days)
3.15E7 = units conversion factor (sec/y)
2.74E−3 = units conversion factor (y/d)
Equation (2) illustrates how concentration of a radionuclide in a plant due to root uptake is calculated:
where:
Crci(Tyr) = concentration of radionuclide i in crop type c from root uptake pathways for a 1-year period (Bq/kg wet plant)
Bvci = concentration ratio for plant type c (Bq/kg dry plant per Bq/kg dry soil)
fc = dry-to-wet ratio for plant type c (kg dry plant/kg wet plant)
P = areal soil density (kg dry soil/m2)
The total radionuclide concentration in the plant at the time of harvest is the sum of radionuclide contributions by deposition and root uptake is described by Equation (3):
where:
Chci(Tyr) is the concentration of radionuclide i in plant type c at harvest for a one-year period (Bq/kg wet plant), and other terms are as previously defined.
A cursory review of the equations for plant radionuclide uptake reveals the complexities inherent to evaluating radionuclide concentrations in crops. The equations describing radionuclide concentrations in animals used for foodstuffs are no less complex. Transfer factors are often element and species specific. For example, a transfer coefficient calculated for 99Tc cannot be used interchangeably with one calculated for 137Cs, as these elements have different chemical and physiological properties. One cannot assume that a transfer coefficient calculated for 90Sr in beef is applicable to 90Sr in chicken. Calculation of transfer coefficients must also take into consideration each of the following variables: mode of absorption (gastrointestinal, inhalation, etc.); homeostatic control mechanisms that may cause variation in transfer coefficients over a wide range of conditions (for example, if calcium and strontium are under homeostatic control with regard to their concentrations in milk, estimates of strontium intake by the animal may not allow accurate determination of strontium concentration in the milk); equilibration of radionuclides in animal tissues; the effects of a radionuclide’s chemical form, generally associated with solubility, on rates of absorption; isotopic effects of radionuclides lacking stable isotope carriers (such as technetium and plutonium); interference of dietary components such as fiber; age of the animal; variations in geography; and other variables such as soil ingestion by the animal [
In the absence of solid experimental, observational, or theoretical support, performance assessment model parameters are generally selected to yield conservative results. Erring on the side of caution is preferable to underestimating dose to biota; however, this approach almost certainly overestimates risk in some cases. With regard to waste repository performance assessments, regulatory decisions and allocation of resources should rely on experimentally determined model input values whenever possible. Recent investigations have revealed significant data gaps in the parameters used for modeling radionuclide uptake by both plants and animals [9,24,25]. Concentration ratios and transfer coefficients for 137Cs and 90Sr have been experimentally determined for many plant and animal species, due to the predominance of these isotopes in releases from atmospheric testing and events at Chernobyl in 1986 and the Fukushima Daiichi plant in 2011. However, data for determining biosphere concentrations of many other components of high-level nuclear waste are either incomplete or lack empirical support.
A great deal of work has been done over the past ten years to establish site-specific concentration ratios and transfer factors for many of the long-lived radionuclides identified in the YMPPA. Napier et al. performed soil-toplant uptake studies for 99Tc, 238Pu, and 241Am using soils and groundwater collected from sites in the northwest, southeast, and southwest United States [
neric CRs typically used in performance assessments. Both studies underscore the requirement of site-specific data for meaningful dose analysis near waste repositories and decommissioned sites.
36Cl is a major isotope of concern in the YMPPA, and a close examination of CR and TF literature for 36Cl illustrates some of the problems associated with many model input values currently in use. This isotope is a neutron activation product of 35Cl, which is present in small quantities in graphite, cladding, nuclear fuel, and other sources. It is persistent, with a half-life of just over 300,000 years, and it is highly mobile in the environment as chloride anion. In a comprehensive literature review for this isotope, no studies were found that addressed foliar interception of 36Cl by crops irrigated with groundwater containing it [
The compendia of generic concentration ratios and transfer factors currently in use have significant gaps and inadequacies. For many of the longest-lived components of nuclear waste, very little data exists. Generic CR and TF values that are available for these radionuclides are frequently characterized by high uncertainties and dubious pedigrees. Experimentally determined site-specific CR and TF values for several sites in the US have been published recently. While their uncertainties are generally lower, they are by definition best suited for site-specific modeling. Significant differences in uptake result from physical and chemical differences among soils and water sources. Dose assessments performed with modeling software—whether they are for waste repositories, decommissioned sites, or emergency response—should rely on high-quality and experimentally determined input values. Much work remains to be done to increase the quantity and quality of CR and TF data used in modeling applications. This can only be accomplished with robust radionuclide uptake studies.
Ultimately, to ensure protection of the ecosystems which receive anthropogenic radioactivity releases, we must have the tools to determine what doses are being delivered and the knowledge of what doses produce the endpoints of concern.
How well can we determine absorbed dose to NHB that exist in spatially and temporally varying radiation fields? The endpoints of concern are a matter of two variables: how sensitive is a given organism, and what level of deleterious effects is acceptable? The most sensitive taxa (e.g. mammals, birds) have many regulations in place that focus on preventing individual level effects. Sessile organisms are among the easiest organisms for which to calculate dose. However, they are also generally far more radio-resistant than other organisms, with the exception of trees (see
This introduces challenges that to date have been addressed to some degree. The simplest of these operate
simply via CRs as discussed earlier, to convert a soil measurement into estimated contamination levels of all organisms in the vicinity. This simplified approach is easily extended to apply to temporal variation, as organisms are simply assumed to instantaneously reach the published CRs. With the orders of magnitude variation in CRs and the need for site-specific calculations, this approach leaves such a great deal of uncertainty that translates into far more stringent regulations on environmental contamination than may truly be necessary to protect the ecosystem in question.
The next higher approach is to apply a network model for the food web and various medium-to-organism transfers. This is ideal for temporal variability as the transfer model already incorporates rate terms. The POSEIDONR model [
Presently, our regulatory restrictions are based more loosely on population level effects, which introduce considerable leeway in the level of spatial variability detail we need. Rather than needing to describe movement patterns of individuals, we simply need a spatial occupation factor description. This is much easier to determine via routine surveys of the number of organisms present at each site, whereas individual level protection necessitates individual level movement description (e.g. radio tag studies). Determining external doses through such an approach is entirely feasible, but internal dose assessment is more challenging. CRs have large uncertainties but are easy to apply, which is why they remain a part of practices today. Food-web transfer models with spatial patterns for each species in the web would necessitate models that track spatially distinct communities that may be grazed by more mobile predators. However, with data on diet, transfer factors, and biological half-lives such an approach is computationally possible. Doing so for an individual-level protection scheme would not, but a combination of these approaches (individual level for species of concern, population level for other food web components) may very well be with sufficient movement pattern description, be in the form of a function or as a randomly sampled collection of real-life recorded spatial movements.
Our measure of dose to wildlife remains fairly simplistic. In humans, we have substantial medical data upon which to weight exposures to different organs as well as weight different types and energies of radiation. Human protection focuses on stochastic effects (mainly cancer) associated with acute exposures. NHB protection depends on chronic exposures that produce largely deterministic effects. While there has been much written on the topic, we do not yet have a consensus on the appropriate NHB weighting factors to use for different radiation types and energies [
The field of radioecology is rapidly expanding, its relevance to humans becoming more apparent as humans continue to rely on the use of ionizing radiation to improve their lives. From handling spent fuel or decommissioning of former nuclear power plants to medical isotopes and industrial radiography sources, we have a multitude of potential environmental consequences to consider from these uses. Radioecology aims to ensure we reap all the uses we’ve found for these technologies without sacrificing related ecosystems in the process.