Element transport in aquatic ecosystems – modelling general and element-specific mechanisms

Lena Konovalenko

©Lena Konovalenko, Stockholm University 2014

ISBN 978-91-7649-026-6

Printed in Sweden by US-AB, Stockholm 2014

Distributor: Department of Ecology, Environment and Plant Sciences

Cover illustration: Lena Konovalenko

This thesis is devoted to my family.

“Uncertainty is everywhere and you cannot

escape from it.” Dennis Lindley

Abstract

Radionuclides are widely used in energy production and medical, military and

industrial applications. Thus, understanding the behaviour of radionuclides

which have been or may be released into ecosystems is important for human

and environmental risk assessment. Modelling of radionuclides or their stable

element analogues is the only tool that can predict the consequences of

accidental release.

In this thesis, two dynamic stochastic compartment models for

radionuclide/element transfer in a marine coastal ecosystem and a freshwater

lake were developed and implemented (Paper I and III), in order to model a

hypothetical future release of multiple radionuclides from a nuclear waste

disposal site. Element-specific mechanisms such as element uptake via diet

and adsorption of elements to organic surfaces were connected to ecosystem

carbon models. Element transport in two specific coastal and lake ecosystems

were simulated for 26 and 13 elements, respectively (Papers I and III). Using

the models, the concentration ratios (CR: the ratio of the element or

radionuclide concentration in an organism to the concentration in water) were

estimated for different groups of aquatic organisms. The coastal model was

also compared with a 3D hydrodynamic spatial model (Paper II) for Cs, Ni

and Th, and estimated confidence limits for their modelled CRs. In the absence

of site-specific CR data, being able to estimate a range of CR values with such

models is an alternative to relying on literature CR values that are not always

relevant to the site of interest.

Water chemistry was also found to influence uptake of contaminants by

aquatic organisms. Empirical inverse relationships were derived between CRs

of fish for stable Sr (CRSr) and Cs (CRCs) and water concentrations of their

biochemical analogues Ca and K, respectively (Paper IV), illustrating how

such relationships could be used in the prediction of more site-specific CRCs

and CRSr in fish simply from water chemistry measurements.

Key words: radionuclides, elements, concentration ratio, bioaccumulation,

biomagnification, fish, modelling, aquatic food web, ecosystem, Cs, Sr,

environmental risk assessment.

List of papers

Paper I

Konovalenko L., Bradshaw C., Kautsky U., Kumblad L., 2014. Radionuclide

transfer in marine coastal ecosystems, a modelling study using metabolic

processes and site data. Journal of Environmental Radioactivity, 133, pp. 48-

59

Paper II

Erichsen A.C., Konovalenko L., Møhlenberg F., Closter R., Bradshaw C.,

Aquilonius K., Kautsky U., 2013. Radionuclide Transport and Uptake in

Coastal Aquatic Ecosystems: A Comparison of a 3D Dynamic Model and a

Compartment Model. AMBIO, 42(4), pp. 464-475.

Paper III

Konovalenko L., Andersson E., Bradshaw C., Kautsky U. Transfer of 13

elements in a lake using a process-based ecosystem model. (Manuscript for

Ecological Modelling).

Paper IV

Konovalenko L., Bradshaw C., Andersson E., Lindqvist D., Kautsky U.

Evaluation of factors influencing accumulation of stable Sr and Cs in lake and

coastal fish. (Manuscript for Journal of Environmental Radioactivity).

My contribution to the papers:

Paper I: Main writer of the paper, implemented the marine compartment

model and carried out calculations, reviewed the literature.

Paper II: One of the main writers of the paper together with A.C. Erichsen

and C. Bradshaw, carried out calculations for the compartment model,

reviewed the literature.

Paper III: Main writer of the paper, developed the lake compartment model

and run simulations and evaluated results, reviewed the literature.

Paper IV: Main writer of the paper, analysed results of chemical analysis of

water and fish, produced calculations and regression data analysis, reviewed

the literature.

Table of contents

Introduction .................................................................................................. 1 Objectives of the thesis ............................................................................................ 3

Study regions ............................................................................................... 4

Modelling as a tool in environmental risk assessment ........................ 7 Key parameters in radionuclide transport models .............................................. 8

Distribution coefficient (Kd) ................................................................................ 8 Concentration ratio (CR) ..................................................................................... 9

3-D hydrodynamic transport models ..................................................................... 9 D- Model: hydrodynamic-ecological model ................................................... 10

Compartment model ............................................................................................... 11 K-model: ecological compartment model for coastal environments ........ 12 L-model: ecological compartment model for lakes ...................................... 14

Discussion ................................................................................................... 16

Conclusions ................................................................................................. 22

Future research .......................................................................................... 24

Acknowledgment ....................................................................................... 25

References .................................................................................................. 26

Abbreviations & definitions

CI Confidence Interval

CR Concentration ratio - the ratio of the element or

radionuclide concentration in an organism to the

concentration in surrounding media

D-model Hydrodynamic spatial model for element transfer

in a marine coastal ecosystem (Paper II)

DHI Danish Hydraulic Institute

DIC Dissolved inorganic carbon

DOC Dissolved organic carbon

HD Hydrodynamic model

IAEA International Atomic Energy Agency

K-model Stochastic compartment model for element

transfer in a marine coastal ecosystem (Papers I,

II)

L-model Stochastic compartment model for element

transfer in a fresh water lake (Paper III)

PDF Probability Density Function

PM Particulate matter

POC Particulate organic carbon

SD Standard Deviation

SFR Swedish repository for low and intermediate

radioactive waste

SKB Swedish Nuclear Fuel and Waste Management

Co, Svensk Kärnbränslehantering AB

NPP Nuclear Power Plant

3D 3 Dimensional

1

Introduction

A number of accidents have resulted in radionuclide contamination of the

environment. Due to the accidents in Kyshtym (1957), Three Mile Island

(1979), Chernobyl (1986) and recently Fukushima (2011) there is a need for

better descriptions and understanding of the transfer of radionuclides in food

webs, and the availability of these radionuclides to humans. Many efforts are

currently being made to improve environmental protection capabilities by

comparing and validating dose assessments models for biota that have been

developed for planned, existing and emergency radionuclide releases.

Radiation protection programmes have traditionally focused on the protection

of humans. However, during the last few years there have been more efforts

made to develop better evaluation tools to specifically assess the potential risk

to non-human biota (ICRP 2007; Beresford et al., 2008).

The importance of tools to adequately predict radionuclide concentration in

aquatic biota and their surrounding media has been emphasized by several

authors (Yankovich, 2005; Monte et al., 2009b; IAEA 2012a; Avila et al.,

2013). These predictions are fundamental for internal and external dose

estimates to aquatic biota in models developed for use in environmental

protection. The prediction can either be based upon field measurements or

model assessments. Since field measurements or literature values may not

always be available for the particular ecosystem or radionuclide of interest

(Copplestone et al., 2013; Howard et al., 2013), especially when considering

future hypothetical releases (Kautsky et al., 2013; Lindborg et al., 2013), a

modelling approach to evaluate these concentrations is essential.

Radionuclides are assumed to have the same chemical properties as their

stable isotopes, except for the slight differences resulting from the difference

in mass (Santschi & Honeyman, 1989). In the aquatic environment,

radionuclides are therefore transported and taken up by biota in a similar way

to their stable element analogues. If stable and radioactive isotopes of a

particular element have the same chemical form, they are thought to be

indistinguishable to organisms (Sazykina 2003). Bioaccumulation or sorption

of a certain radionuclide by living organisms is not linked to its radioactivity,

but reflects the difference between the content of the stable chemical element

in the environment and in the organism (Sazykina 2000). Modelling concepts

based on stable elements are therefore applicable to radionuclide modelling.

2

Anthropogenic radionuclides present in the Baltic Sea and its catchment area

have two main sources: 1) During 1950-1980 the United States and the Soviet

Union carried out atmospheric nuclear weapons tests (UNSCEAR, 2000); 2)

The accident at the Chernobyl nuclear power plant in 1986 caused

considerable fallout of Cs-137 and Sr-90 over the Baltic Sea and areas of

eastern Sweden (HELCOM, 2009; Håkanson et al., 1992). However, it should

be remembered that there are other potential sources of a wide range of

different radionuclides.

The future potential release of radionuclides from nuclear facilities is an issue

that will be important for the coming centuries or more. Radionuclides have

military, industrial and medicinal applications as well as being used to produce

electricity through nuclear power production. After use, these radionuclide

sources become radioactive waste that must be stored safely, for example in a

repository, and isolated from the environment as long as they pose a risk for

humans and biota. The amount of radioactive waste is predicted to increase in

the future, as energy needs grow and many countries (e.g. Vietnam, South

Africa, India, China, Finland) are building new nuclear power plants, of which

many will be situated in coastal areas (Fowler and Fisher, 2005). As a

consequence, authorized discharge into the sea will increase, as well as the

probabilities of new nuclear accidents. The whole nuclear fuel cycle, from

radionuclide mining through production and use of the fuel, also includes

possibilities for environmental contamination. The risk of radionuclide

releases to the environment will exist as long as society continues to actively

use anthropogenic radionuclides. Thus, we require knowledge about

radionuclide behaviour in the different types of ecosystems, especially in

aquatic environment, because water helps to transfer and distribute elements

in the ecosystem.

The main focus of this thesis is to describe and model the transfer of

radionuclides in the aquatic environments of a specific marine coastal region,

including brackish water bays and freshwater lakes, located in the Forsmark

area, eastern Sweden. This area is the location of one of Sweden’s nuclear

power plants, as well as one existing and one proposed facility for nuclear

waste storage. The site was chosen for the good availability of ecosystem and

element concentration data from previous and on-going work by the Swedish

Nuclear Fuel and Waste Management Company (SKB). The thesis takes as its

starting point the potential release of radionuclides from nuclear facilities, but

it also advances our knowledge and expertise in ecosystem modelling in

general, and also its application to the transport of non-radioactive elements

in aquatic ecosystems.

3

Objectives of the thesis

The main aim of this thesis is to model radionuclide and element transfer in

aquatic food webs (using generic and element-specific processes) in order to

improve assessment of environmental risk.

The specific objectives are to:

develop ecological element transport models for lake and coastal

ecosystems and simulate concentration ratios (CRs) of biota which

are widely used in risk assessment to estimate element

concentrations in biota from measured water concentrations (Paper I

and III).

compare the compartment model described in Paper I with a 3D

spatial dynamic model (Paper II) and investigate how the two

different models can complement each other and for what situations

the models are most applicable.

evaluate how environmental factors such as water chemistry and

seasonal variation may influence bioaccumulation of Sr and Cs in

fish (Paper IV) and impact on the range of the CR values that are

used in risk assessment.

4

Study regions

The main study area is the Forsmark area in eastern Sweden, including both

coastal brackish water areas (Öregrundsgrepen) and freshwater lakes (Fig. 1).

In June 2009, the Swedish Nuclear Fuel and Waste Management Co. (SKB)

decided to put forward the Forsmark area as the suggested site for the

geological repository of Sweden’s spent nuclear fuel and carried out an

extensive safety assessment analysis (SKB, 2011). The sketch of this planned

repository is illustrated in Fig. 2 (b). The study area is ideal for element

transport modelling, because of the detailed descriptions and large amount of

field data available (Lindborg et al., 2008, 2010; Aquilonius et al., 2010).

The model presented in Paper I was applied to a specific area of

Öregrundsgrepen, one of 28 sub-basins referred to as number 116 (Fig. 1).

This area was also used in simulations by a 3D spatial dynamic model in Paper

II. Basin 116 is located above the existing final repository for radioactive

operational waste (SFR) which is embedded in the bedrock 50 m under the

seabed, and stores low- and intermediate level radioactive waste (Fig. 2, a).

The area is shallow (mean depth c. 10 m) and salinity is low (c. 5 psu); this

brackish water results in a combination of marine and freshwater species being

present in the coastal area. A more detailed description of marine ecosystem

Forsmark area is available in Aquilonius et al. (2010) and Kumblad et al.

(2003, 2006).

In Paper III, Lake Eckarfjärden was studied and for model comparison data

from two other Forsmark lakes were used, Bolundsfjärden and Fiskarfjärden

(Fig. 1). A detailed description of these lakes is available in Andersson (2010).

These same three lakes were also considered in Paper IV, together with Lake

Frisksjön in the Laxemar-Simpevarp area in south-east Sweden (Fig. 1) where

another of Sweden’s nuclear power plants is located. In Paper IV, data from

the coastal areas of Forsmark and Laxemar-Simpevarp were also included in

the analyses.

5

Figure 1. Maps showing the location of the Forsmark and Laxemar-Simpevarp study areas, situated on the Swedish Baltic Sea coast. Top left: the locations of the coastal site sub-basin 116 in Forsmark that was modelled in Papers I and II. Top right: the three lakes in the Forsmark area used for modelling (Paper III) and for fish and water samples (Paper IV), and the coastal area used in Paper IV. Bottom: the lakes and coastal bays where data for Paper IV were collected. NPP − Nuclear Power Plant.

6

Figure 2. (a) Swedish repository for low and intermediate radioactive waste (SFR) and an example of the planned extension for low-level radioactive waste. SFR is located in the vicinity of Forsmark Nuclear Power Plant (NPP) (from Berner et al. 2009). (b) Sketch of the planned future final repository for spent nuclear fuel, 500m below ground in the Forsmark area, Sweden (sketch by SKB).

7

Modelling as a tool in environmental risk assessment

With the help of a modelling approach it is possible to describe and identify

element transport processes (both specific and generic) and make predictions

for risk assessment about transport and cycling of both naturally-occurring

and human-introduced elements, including radionuclides. Many European

countries have implemented models of radionuclide release into the

atmosphere and aquatic environment that are imbedded in Decision Support

Systems such as RODOS (Hofman et al., 2011) and MOIRA (Monte et al.,

2009a). Models are the only tools to predict the consequences of accidental

release and to support decision makers for remediation action.

This thesis considers hypothetical future releases from a high level radioactive

waste repository. For long-term risk assessment of the impact to humans and

biota in case such releases, more than 40 radionuclides need to be considered

(Kautsky et al. 2013). In Sweden, Canada and many other countries it is

obligatory for the waste disposal company to demonstrate that the

environment is adequately protected.

In radioecological modelling, several methodological approaches have been

applied to describe radionuclide migration in aquatic ecosystems. The most

common are compartment models (Paper I, III) and spatial and temporal

hydrodynamic models (Paper II). The structure of a model can be increased

almost indefinitely by incorporating more fluxes (mass/time), amounts (mass)

and concentrations (mass/volume) to and from the defined compartments, but

one of the main tasks is to find the most important processes and optimum

amount of parameters adequate to characterise the system, which are capable

of keeping the model’s robustness. Inter-comparisons of a range of modelling

approaches from different countries (Beresford et al., 2008) have recognised

that variation in estimates of final radiation dose to organisms depends

significantly on the prediction of whole-body activity concentrations, which

are in turn defined by concentration ratios (CRs; see below). The main reason

for some of the observed variability is the inclusion in assessments of

inappropriate CR values, especially when site-specific data are not available

(e.g., CR values are taken from a biogeochemically similar element or a

‘similar’ organism). Instead of using CRs of analogue organisms or elements,

8

which are often misleading, a modelling approach for the prediction of a range

of site-specific CRs for different functional groups of aquatic organisms is

proposed in Papers I-III.

Key parameters in radionuclide transport models

Aquatic systems are influenced by a complex variety of physical, geochemical

and biological processes, which influence the behaviour, transport and fate of

radionuclides released into the water or sediment. Two of the key parameters

that are used extensively in environmental risk assessment analysis to describe

geochemical and biological processes in transport models are distribution

coefficients (Kd) and concentration ratios (CR).

Distribution coefficient (Kd)

Radionuclides in the aquatic environment occur in two phases: dissolved

radioactive substances which are redistributed by the flow field in a

conservative form more or less passively, and particle reactive radionuclides

which tend to attach to particles or suspended material in the water column.

Particle reactive and non-reactive radionuclides/elements are classified

according to a radionuclide/element specific distribution coefficient Kd (m3 kg

dw-1 (dry weight)) which is defined as the ratio of the concentration of

radionuclide/element in solid phase of sediment (Cs) or particulate matter

(CPM) and in liquid phase in water (Cw) in an equilibrium state (Harms et al.,

2003):

𝐾𝑑 =𝐶𝑃𝑀

𝐶𝑤

=(

𝐵𝑞𝑘𝑔 𝑑𝑤

)

(𝐵𝑞𝑚3)

or (

𝑘𝑔 𝑘𝑔 𝑑𝑤

)

(𝑘𝑔𝑚3)

= (𝑚3

𝑘𝑔 𝑑𝑤)

Particle reactive elements with high distribution coefficient (Kd >10 m3 kg-1)

have a tendency to attach to suspended particles or to the sediment strongly

enough to change their dispersion behaviour in the aquatic ecosystem. Non-

reactive elements which have low (Kd ≤10 m3 kg-1) mainly dissolve in the

water and can be transported long distances from the release point with water

currents. The measured Kd values of stable elements for marine and freshwater

ecosystems recommended by IAEA for modelling radionuclide transport can

be found in reports (IAEA 1985a, 2004, 2009, 2010).

The distribution coefficient of suspended particulate matter, Kd PM, in marine

or freshwater environments is usually expressed in the units m3 kg-1, but can

also be converted into m3 gC-1 by multiplying by the ratio of dry weight to

carbon content for particulate matter (PM), e.g. 0.006 kg dw gC-1 (Kumblad

9

et al., 2006). This is more suitable for some models, because biomass is often

presented in mass of carbon since this is a more ecologically relevant unit than

dry weight.

Concentration ratio (CR)

In the latest IAEA reports (IAEA 2004; IAEA 2010) and the online database

www.wildlifetransferdatabase.org the concentration ratio (CR) is defined as

the ratio of the radionuclide concentration in the organism tissue (fresh weight,

fw) from all exposure pathways (including water, sediment and ingestion

pathways) relative to that in water:

CR (L kg-1) = Caquatic biota/ Cw= (Bq kg fw-1)/(Bq L-1),

where Caquatic biota is the concentration per unit mass of organism (kg/kg fw or

Bq/kg fw); Cw is the concentration per unit volume of sea water (kg L-1 or Bq

L-1). The CR definition assumes that the radionuclide in the organism is in

equilibrium with its surrounding media. The time required to achieve such

equilibrium is dependent on both the biological half-life of the radionuclide in

the organism and the radionuclide physical half-life.

The CR can be also presented in the units m3 kg C-1 where the mass of biota

is expressed as the carbon content (e.g. kg C), for example, the models in

Paper I and Paper III and in site assessments carried out by the Swedish

Nuclear Fuel and Waste Management Company (SKB) (e.g., Nordén et al.,

2010; Tröjbom & Nordén, 2010; Avila et al., 2010; Kumblad & Bradshaw,

2008). The units m3 kg C-1 can be useful when the mass balance of aquatic

ecosystems is calculated in carbon and it is assumed that pollutant follows

organic fluxes in the food web. Conversion coefficients for all marine biota in

the study area for converting data from wet weight, dry weight or carbon

weight are published in Kumblad and Bradshaw (2008). In some cases, it is

adequate for modelling purposes to use average CR values and assume

equilibrium over the spatial and temporal scales examined. However, in other

cases it may be more appropriate to use or estimate ranges of CR in space and

time. This is discussed further in all four thesis papers.

3-D hydrodynamic transport models

Three-dimensional hydrodynamic radioecological models (e.g., Paper II) are

often used for short term assessment of accidental release; these models can

model pollution in the ecosystem with great spatial and temporal resolution,

and identify areas of concern such as likely hotspots of contamination. For

instance, models described by Erichsen et al. (2010) or Margvelashvili et al.

10

(2002) can produce the results of simulations in terms of maps of radionuclide

distribution in the water, sediment and biota, but they are computationally

heavy and quite complicated. Also, in order to make model predictions more

accurate, site-specific input parameters are desirable, which can be time

consuming and costly.

Hydrodynamic models for the dispersion of elements (e.g., Paper II) typically

involve two coupled models: a hydrodynamic model (HD model) and a

transport model. The HD models are models that consist of mathematical

equations that are solved at discrete time steps on a regular or irregular grid

that covers the model domain, i.e. the region of interest. HD models are able

to calculate three-dimensional flow fields based on the realistic topography of

the area and realistic forcing functions like wind or water density. In some

cases, time steps of less than one hour are able to resolve a spectrum that

ranges from tidal motions to decadal variability. Numerical transport models

of dissolved elements/radionuclides calculate the advection and the diffusion

of a substance or a tracer. Combined HD and transport models are able to

provide results with a high spatial and temporal resolution (Harms et al.,

2003). They are also able to model transport of contaminants both in the

dissolved and particulate phase. Suspended particles may play an important

role

as the carriers of contaminants over long distances from the release areas with

water currents, or alternatively transport contaminants down to the bottom

where they may be buried in the sediment.

D- Model: hydrodynamic-ecological model

Erichsen et al. (2010) developed the hydrodynamic-ecological-radionuclide

model (referred to here and in Paper II as the D-model) for Öregrundsgrepen,

Sweden. The D-model was used to predict the spread of radionuclides and

subsequent accumulation in sediments and biota following a hypothetical loss

of radionuclides from a repository for nuclear waste at Forsmark nuclear

power plant. The D-model assumes that radionuclides are introduced to the

pore water of the lower sediment layer via groundwater inflow. Seven

radionuclides Cl-35, Cs-135, Nb-94, Ni-59, Ra-226, Th-230 and C-14 were

modelled explicitly. The model coupled ecosystem and radionuclide food web

models with a hydrodynamic model for the area. Conceptually, the ecosystem

model and the radionuclide model were developed based on the general food

web structure developed in earlier modelling studies within the area (Kumblad

& Kautsky, 2004; Kumblad et al., 2006). The results from the hydrodynamic

model provide inputs to the coupled ecosystem model and the results from the

ecological model provide inputs to the radionuclide models. The ecosystem

model is used to illustrate the spatial and temporal variation in important

processes and parameters. In the biotic components, radionuclides are either

11

adsorbed to surfaces or accumulated internally in the organisms.

In autotrophic organisms radionuclide uptake and accumulation are driven by

and scaled to their photosynthetic activity. In heterotrophic organisms

radionuclides are taken up through the food chain by grazing or predation. The

D-model was applied and validated in Paper II, where it was also compared

with a compartment model (L-model, see below).

Compartment model

Compartment models (Papers I, II, III) are used to estimate the impact on the

environment of long-term release when equilibrium between organisms and

the surrounding sea or freshwater can be expected. These models belong to

the category of the so-called “fully mixed” hydrological dispersion models

that have been well described in the scientific literature (e.g., IAEA, 1985b;

1994; Håkanson & Monte 2003). Many of them are physical compartment

models, e.g. models by Avila et al. (2010), Bird et al. (1993), IAEA (2000),

BIOMOVS II (1996a, b), for predicting the radionuclide concentration in

aquatic biota based on the steady-state approach, which assumes a constant

equilibrium between the radioactivity concentration in water and in

organisms, through the concentration ratio (CR). Thus, physical models

estimate elements/radionuclide concentrations in abiotic compartments, and

then the concentrations in aquatic organisms are assessed by multiplying

measured CR with the element/radionuclide concentration in water (or

sediment).

More complicated models are ecological/radioecological compartment

models which include biological processes and are based on the equations for

biomass dynamics of populations or ecological groups of species. The

ecological equations describe the transfer of chemical elements

(radionuclides) from the environment to the food chains of organisms and

back to the environment (ecological cycles of the elements). The compartment

models in this thesis are based on the fluxes of carbon in the ecosystem. One

model of this type (K-model), applicable for the simulation of element

transport in a coastal ecosystem, is discussed in Paper I and further examined

in Paper II. Another one was developed for element transport in a freshwater

lake (L-model) and implemented in Paper III. The compartment models

(Paper I, Paper III) are computationally faster than 3D hydrodynamic models

(Paper II) and can allow estimates for more than for 100 years into the future,

which is important when considering long-lived radionuclides.

12

K-model: ecological compartment model for coastal

environments

Historical model development

Kumblad et al. (2003) first presented an ecosystem model describing the flows

of carbon and C-14 through a coastal food web in the ecosystem above the

underground nuclear waste repository (SFR) in the Forsmark area. From the

carbon flow model, additional parameters and assumptions were made in

order to estimate the flow of C-14 in the same area in 2000 years from the

present (Kumblad et al., 2004a) and for 25 other radionuclides present in the

nuclear waste repository (Kumblad et al., 2006). In this thesis, this multi-

radionuclide model was then implemented in the new software Ecolego

(http://ecolego.facilia.se/), updated with newly available site-specific data and

verified with concentration ratios (CR) for 26 elements (Paper I) and

compared with simulation results of a 3D hydrodynamic transport (Paper II).

The conceptual approach of this model was then used as the basis for a lake

ecosystem model (Paper III).

Model description

Details of the model are described in Paper I, but a short description is given

here to demonstrate the similarities and differences with the other modelling

approaches. The food web model describes the biomass distribution and the

carbon dynamics of the ecosystem. It was conceptualized with eight

compartments, 6 biotic and 2 abiotic (Fig. 3). In the biota compartments,

organisms having the same ecological function and residing in the same

habitat were grouped together. Initial data were determined at species or

family level, either from site-specific in situ measurements or derived from

site-specific biomass data. The carbon flows were constrained by temperature,

light intensity and inorganic carbon available for photosynthesis, and these

parameters were based on field data from the site. The basic biological

processes respiration, consumption, faeces production and excess excretion

are represented in the model (Fig. 3). The main model input parameters were:

biomass of each compartment

net productivity (primary producers) and consumption, respiration

rates and assimilation efficiencies (AEs) of consumers

dietary composition of fish

average radius of species in each functional group and of particulate

matter

These parameters and their distributions were used in probabilistic

simulations. A water exchange mechanism was connected to all pelagic

components of the ecosystem, i.e. phyto- and zooplankton, dissolved

inorganic matter (DIM) or dissolved inorganic carbon (DIC) and particulate

matter (PM) or particulate organic carbon (POC), to be able to model import

13

and export of matter to and from the system. DIC was defined as entering the

food web through photosynthesis (primary production) and to be transferred

to higher trophic levels via grazing and predation according to the food web

structure. The biomass of the compartments (standing stock) was assumed to

remain constant between years.

Radionuclides were then assumed to follow the flow of organic matter in the

food web and radionuclide relocation was regulated by two radionuclide

specific mechanisms: adsorption to organic surfaces and uptake by primary

producers. Radionuclide-specific dynamics depended on distribution

coefficients, Kds, between radionuclides in suspended particulate matter. In

most cases, element site-specific Kd data were applied (Nordén et al., 2010).

For those elements where element-specific site data were lacking, literature

data were used. Kds were converted into the units m3 gC-1, since all organic

transport in the food web model was based on carbon flow.

Uptake of radionuclides by consumers was subdivided into two processes:

adsorption to the organisms’ body surface which are estimated using Kd;

accumulation through the ingestion and assimilation of contaminated

food.

Radionuclides were retained in the organisms, and release took place only

when organisms died, were consumed or were released from organisms with

the faeces.

14

Figure 3. Visualization of element transport and cycling in the coastal marine ecosystem compartment model (K-model) (Paper I, Paper II). The arrows illustrate the element flows, shapes illustrate model compartments.

L-model: ecological compartment model for lakes

The adsorption of elements from water on to the surfaces of organisms and

particles was taken into account, additionally assuming that pollutants follow

proportionally to flows of organic matter via the food web. The model was

partly descriptive, rather than predictive, mainly because input site-specific

data were only available for modelling and there was not enough data for

complex model validation. The element-specific parameters implemented in

the model were Kd PM and Kd sed, which were derived based mainly on measured

site-specific data and accessible for 13 elements from an earlier study

(Tröjbom et al., 2013). The key concepts of the lake transport L-model (Paper

15

III) were identical with the K-model for the coastal area (Paper I; Kumblad et

al., 2003, 2006), though additional processes and compartments that were

specific for the lake were included. The lake ecosystem was subdivided into

16 compartments, 5 abiotic and 11 biotic (Fig. 4). Major differences to the K-

model were to include the process of sedimentation of organic matter and

additionally 4 abiotic compartments: Sediment gyttja, Burial sediment, DOC

and POC. Moreover, two biotic compartments, bacterioplankton and benthic

bacteria, were considered separately in the L-model, while in the K-model

they were taken into account as a part of phytoplankton and benthos,

respectively. Fish were subdivided into three compartments, zooplanktivorous

(Fish-Z), benthivorous (Fish-B) and piscivorous (Fish-P), in contrast to the

coastal K-model where only one general fish compartment was considered

and included all three fish groups.

Figure 4. The carbon budget of Lake Eckarfjärden showing carbon fluxes (kgC/y) and initial biomasses (kgC).

16

Discussion

For long-term risk assessment of impact to humans and biota from a high level

radioactive waste repository in the case of a hypothetical release in future,

multiple radionuclides need to be considered (Kautsky et al., 2013). CR values

are the most commonly used parameters in many estimates of radionuclide

concentrations in organisms. However, a major source of error in the

prediction of these concentrations is the use of inappropriate CR values, taken

from the literature or extrapolated from other sources. Thus modelling

approaches that can estimate a likely range of CRs are a preferable method.

Development of compartment models of element transport for lakes and

marine coastal ecosystems

Estimations of radionuclide redistribution between water, sediment and their

transport with water currents, as well as the uptake by aquatic organisms, are

fundamental for concentration and dose assessments. The studies in this thesis

have contributed to assessment the fate of the radionuclides and elements in

the aquatic environment and their transfer to biota. Prior to this thesis

considerable research has been devoted to modelling and prediction of

radionuclide transport of Cs-137 and Sr-90 (Heling & Bezhenar., 2009;

Lepicard et al., 2004) but less attention has been paid to predicting other

ecologically significant radionuclides such as I-131, Zr-95, Nb-95, Pu-239

and Po-210 in marine and freshwater environments. Only occasionally have

uptake of a few elements by a specific species (Baines et al., 2002; Vives i

Batlle et al., 2008) or transfer via a simple food chain (Smith, 2009) been

considered. However, consideration of a set of radionuclides/elements is

particularly important for evaluating the radioecological and ecological

consequences of accidents in the early critical period of accidental

contamination and for long-term assessment (Kryshev et al., 1999). In this

thesis, the coastal food web model in Paper I modelled and verified the transfer

of 26 elements (Ac, Am, Ag, Ca, Cl, Cm, Cs, Ho, I, Nb, Ni, Np, Pa, Pb, Pd,

Po, Pu, Ra, Se, Sm, Sn, Sr, Tc, Th, U, Zr) and 8 compartments, and the lake

transport model in Paper III modelled 13 elements (Al, Ca, Cd, Cl, Cs, I, Ni,

Nb, Pb, Se, Sr, Th, U) and 16 compartments. Both models were able to

estimate ranges of CR for the radionuclides and organisms that generally

agreed well with the data available for verification, especially those elements

with high Kd. The models also allow extrapolation to future altered ecosystems

(e.g. through land rise or climate change) since they provide a method that can

17

be scaled to ecosystem properties and where element-specific processes are

separated from generic processes which are equally valid for all elements.

The main structural difference between the lake (L-model, Paper III) and

marine (K-model, Paper I) element transport models are that sedimentation

processes of particulate organic matter were taken into account in the lake

model, because the lake is a more isolated water system with lower water

retention time (328 days) and much greater sedimentation rate than coastal sea

area. In the K-model for the coastal area, where water retention time is very

short (1 day), water currents are more important since they transport

particulate organic carbon (POC), phyto- and zooplankton, together with

adsorbed contaminants, in and out of the modelled area.

This short retention time in the model, and the use of measured stable element

CRs for comparison, may have been responsible for the low predicted

zooplankton CRs of low Kd PM radionuclides. The model assumes that

adsorption of elements on zooplankton surfaces from the water happens

immediately and the concentration on their surfaces reaches equilibrium with

the water concentration. Thus for elements with high Kd PM where adsorption

is mainly responsible for total element accumulation of zooplankton and the

dietborne pathway is less important, the calculated CRs for zooplankton were

close to measured CRs (Paper I, Fig. 4). In the model, the clean zooplankton

which had previously fed on uncontaminated food outside the modelled area,

are transported with water currents into the area where radionuclide release

occurs. The retention time in the study basin is only one day, while the lifetime

of zooplankton is around one month; thus, they are unlikely to reach a steady-

state condition for dietborne contaminant exposure. This could be a reason

why predicted CRs were lower than measured CRs for elements with low Kd

PM, where dietary uptake is important. This could be tested by ‘turning off’ the

flux of zooplankton to and from the modelled area and comparing the new

modelled values with measured CRs.

Another difference between the coastal and lake models was that dissolved

organic carbon (DOC), microphytobenthos, and heterotrophic bacteria

compartments were included in the lake L-model. The L-model (Paper III)

estimated that benthic bacteria has the largest surface adsorption of all

functional groups for the studied lake, up to 89% of total surface adsorption,

though their biomass is only 11% of the total biomass in the lake. Because of

the large surface to volume ratio of bacteria, they can absorb a large amount

of pollutants on their surface, especially particle-reactive elements

characterized by high Kd PM. Thus, pollutants once adsorbed on to bacterial

surfaces from the surrounding water or sediment pore water could be

incorporated into the food web via mixotrophic phytoplankton and benthic

fauna. The benthic part of the ecosystem was much more important in element

transfer in the lake ecosystem than the marine ecosystem due to its dominance

18

in terms of biomass and surface area and due to differences in water depth and

water exchange between the two. The fact that essentially the same model

could be applied to two such different ecosystems is evidence of its robustness

and flexibility.

Most radiological models have difficulty in estimating confidence intervals

for their model predictions, because of the absence of a standard methodology

for uncertainty assessment and its application for environmental predictions

(Iosjpe et al. 2006, Mardevich et al. 2014, Periáñez et al. 2015). The models

in Paper I and III are capable of estimating the range of concentration ratios

(CRs) for marine coastal and freshwater biota. The uncertainty intervals in the

predictions in most cases overlapped the uncertainty intervals of the

observations for CRs of zooplankton, grazers of macroalgae and benthos. In

the context of risk assessment of radionuclides centuries or millennia in the

future, this level of agreement is acceptable (Avila et al., 2013). The

development and use of probabilistic radioecological models should be

prioritized in the future, because they are more suitable for the prediction of

the range (confidence intervals) of calculated results than simple deterministic

models which give only average estimated values that can lead to under- or

overestimation of radionuclide transfer. There is also a need for more

standardised criteria for model verification, which is currently lacking. Results

of deterministic models are usually verified with average measured data, or

modelled results are considered acceptable if they are within the error bars of

the measured data (Tateda et al., 2013, Periáñez et al., 2012, 2015). For

probabilistic models, authors usually consider predicted values of the median

(50% percentile) and 95% confidence interval (CI) and check how much these

overlap with measured average and standard deviation (SD) (Kryshev et al.,

1999; IAEA 2012b; IAEA 2012c).

Estimation of concentration ratios (CRs) in aquatic biota

Uptake and elimination of radionuclides, many of which are metals and heavy

metals, are not governed by simple diffusion processes, but are rather a

function of the exposure concentration, the geochemical form, the biology of

the species and physiological mechanisms (Chapman & Adams, 2007). The

process is controlled by complicated biological mechanisms. Most of the

metal transport proteins present in biological cell membranes are involved in

ion regulatory processes and the uptake of essential elements (Simkiss &

Taylor, 1995).

The models in this thesis give a reasonable prediction of CRs for most

functional groups of marine and freshwater organisms for large set of elements

(Paper I and III). At least in the context of risk assessment of radionuclides

centuries or millennia in the future, this level of agreement is acceptable

(Avila et al., 2013). However, the models in Papers I and III estimated CR

19

values for fish that were higher by up to one order of magnitude compared to

measured CRs for many of investigated elements. This is probably largely

because of the assumption that the excretion rate is zero, a simplification made

since it is very difficult to precisely simulate complex accumulation and

elimination mechanisms. Besides, assimilation of metals by fish is determined

both by subcellular metal distribution in the prey, as well as the feeding

process and digestive physiology of the fish (Wang & Rainbow, 2008). The

estimation of the role of trophic transfer in metal bioaccumulation is important

in the context of hazard evaluation. Effects of dietary exposure are metal- and

species-specific, and therefore, are most accurately assessed through studying

specific food-consumer relationships. Thus, in order to produce more realistic

estimates by modelling, it may be necessary to take into account additional

element-specific mechanisms and parameters, e.g. element assimilation

efficiency (AE) and elimination rates for the investigated organisms.

However, an assumption of zero excretion may be appropriate to ensure

conservative estimates of radionuclides in fish that may be subsequently be

eaten by humans.

The atmospheric fallout from the Chernobyl and Fukushima accidents

consisted largely of the radionuclide Cs-137, which was discharged into the

environment and taken up by organisms into the food chain. Thus Cs-137 can

be potentially one of the radionuclides most responsible for internal exposure

from contaminated food for people eating products coming from polluted

sites. Many models (e.g., Kryshev & Ryabov, 2000; Tsumune et al., 2011)

have therefore been developed for predicting transfer of Cs-137 in

ecosystems, and estimating doses and possible adverse effects on human

health and wildlife from Cs-137. Such models are often assumed to work

equally well for other radionuclides. Both of the models developed in this

thesis (K-model and L-model, Paper I and III) reproduced CRs and ranges of

variation for Cs well for most investigated organism groups. Nevertheless, the

difference between the modelled and measured CRs values for some other

elements was more than one order of magnitude. Consequently, it was

demonstrated that a model verified only for one element (e.g., Cs), does not

necessarily guarantee the same satisfactory results for other elements.

Environmental factors influence bioaccumulation of Sr and Cs in fish

The concentration ratio definition assumes that the radionuclide in the

organism is in equilibrium with its surrounding media (IAEA 2004). However,

in reality the relationship between the concentration of an element or

radionuclide in a living organism and the surrounding water is dynamic. Rates

of both uptake and excretion are known to be affected by body size, rate of

change of body size, temperature, salinity, etc. Moreover, seasonal variation

in the biological uptake of elements may be great. Relatively little effort has

been made to quantify the seasonal variation in CRs in aquatic organisms. It

20

is to be expected, consequently, that real differences exist between some CRs,

even for the same element and species, and that the variability in the data

reflects true environmental fluctuations in any area (IAEA 2004).

Paper IV explored variation in the CRs of stable Sr and Cs for freshwater and

brackish-water fish species based on field observations in the Forsmark area,

Sweden. There were significant seasonal variations (factor of 3-7) in the

measured concentrations in lake water for K, Ca, Sr and Cs. Thus, relying on

single measurements or extrapolating from other datasets may result in

misleading CR values. The determination of these element concentrations in

water is important, because it can be crucial for further calculation of CR for

fish. The determined range of CRs for lake fish using the L-model (Paper III)

overlapped with the range of CRs derived from measured data (Paper IV) for

fish muscle tissue: simulated Cs was 4.9 – 840 m3 kg dw-1 and measured Cs

was 1.1 – 23.6 m3 kg dw-1; simulated Sr was 0.019 – 7.8 m3 kg dw-1 and

measured Sr 0.002 – 0.043 m3 kg dw-1. Hence, the lake L-model (Paper III)

was verified for the estimation CR of Sr and Cs for fish. The L-model takes

seasonal variation into account by calculating probability distribution

functions of CR; the estimated range covers this variation.

Inverse relationships between CRs of fish for Sr and Cs and water

concentrations of Ca and K.

Water chemistry may also influence uptake of contaminant by aquatic

organisms. Calcium (Ca) and potassium (K) are biological macroelements,

and also the analogues of strontium and caesium, respectively (Vanderploeg

et al., 1975; Smith et al., 2003, 2009). In Paper IV, in contrast to the other

modelling Papers I, II and III, empirical inverse relationships were derived

between CRs of fish for stable Sr and Cs and water concentrations [Ca] and

[K], respectively. These CRs were based on site-specific measured

concentrations of water and fish from lakes and coastal area in Forsmark, and

these relationships agreed well with previous studies for isotopes Sr-90 and

Cs-137 (Outula et al., 2009; Kryshev, 2006; Smith et al., 2000). Using these

empirical equations from Paper IV, it should be possible to estimate

accumulation of Sr in fish only by measuring the levels of Sr and Ca in the

water (and for Cs in fish using only concentration Cs and K in the water).

Applications and comparison of the different model approaches

The comparison of the compartment K-model described in (Paper I) with the

3D spatial dynamic D-model (Erichsen et al., 2010) exposed strengths and

weaknesses of the two model approaches (Paper II). Using these two models,

the CRs were estimated for Ni, Cs and Th for phytoplankton, zooplankton and

fish. The agreement between empirical data and model predictions was good

for most observations, suggesting that both models are robust. However, the

models showed different levels of complexity with one based on a simple

21

compartment approach to the other making use of transport-diffusion

equations and finite element method.

The model comparison gives the possibility to evaluate: i) which models are

more appropriate for the management of complex environmental problems; ii)

how to provide more effective environmental monitoring; iii) how uncertainty

of model input parameters influences the output results; and iv) what kind of

model is better to apply for specific contamination scenario.

In the case of new accidental releases, dynamic models are essential to capture

the temporal variation in transport and radioactive decay or radionuclides. For

example, the environmental impact of the nuclear accident at Fukushima has

been estimated by applying different radionuclide transport models, in

combination with extensive field sampling. In the course of the Fukushima

nuclear accident the majority of the radionuclides were transported offshore

and deposited in the Pacific Ocean (Steinhauser et al., 2014). Many field

samples of water, sediment and biota were carried out and analysed for

radioisotopes in order to assess risk for aquatic organism and dose for human

via ingested marine food (Buesseler et al., 2012; Tagami & Uchida, 2012;

Takahashi 2014). Three dimensional hydrodynamic transport models have

been used to simulate distribution of Cs-137 (Periáñez et al., 2012; Tsumune

et al., 2014) in the marine coastal water and sediment near Fukushima. A

dynamic compartment food chain transfer model has been developed for

simulation of Cs-137 transfer in the southern Fukushima coastal biota (Tateda

et al., 2013). A dynamic compartment transfer model for Cs-134, Cs-137, I-

131 has been applied to predict concentrations and dose for different aquatic

organism in the Fukushima Dai-Ichi South Channel (Vives i Batlle et al.,

2014).

22

Conclusions

Two models were developed and evaluated: the element transport

compartment K-model (Paper I) for coastal food-webs for 26 elements and the

lake L-model (Paper III) for 13 elements. The uncertainty intervals in the

predictions, in most cases, overlapped the uncertainty intervals of the

observations for CRs of zooplankton, grazers of macroalgae, benthos and fish,

indicating that such a method is suitable for estimating CRs where site data is

not available. These models estimations may be preferable to using CR data

from the literature, which may come from very different environments and is

often only available as averages or best estimates. The distribution coefficients

(Kd PM) suspended particulate matter and upper sediment (Kd sed) played an

important role in modelling and were largely responsible for the different

resulting output values for different elements, because these parameters are

element-specific. The distribution coefficients that were applied with

probability density functions (PDF) in probabilistic simulations also

contributed to the CR ranges obtained. In the absence of site-specific CR data,

being able to estimate a range of possible CR values with such models is more

robust than relying on single CR values from the literature or extrapolating

from data that are not always relevant to the site of interest. CRs modelled in

this way also indirectly take into account temporal variation in environmental

element concentrations, such as those found in Paper IV.

In the case of radioactive waste disposal, it is required that environmental

protection and human safety is demonstrated for at least 10 000 years into the

future. In this situation, biosphere assessment models are necessary to make

quantitative predictions of risk, because this cannot be demonstrated directly.

The marine and lake ecosystem models (Papers I and III) described in this

thesis can be part of the biosphere model which can then be applied to estimate

possible total doses to humans and biota. These food-web models are able to

assess the range of bioaccumulation of radionuclides (or elements) in aquatic

organisms, so they can also be of great use in environmental toxicology,

determining environmental quality criteria and establishing effective

countermeasures for, in particular, metals.

The models in Papers I and III demonstrated the possibility to combine a large

amount of different measured data collected for the same region, but

23

distributed in time and space. Such modelling can help to better understand

and assess the ecological risk associated with dietborne and waterborne

exposure of radionuclides and metals to aquatic organisms.

The K-model (compartment) and D-model (3D dynamic) (Paper II) could be

used to complement each other in different ways: the D-model, with its greater

spatial and temporal resolution, could identify areas of concern, and then this

area could be modelled on a longer time scale with the K-model; the D-model

can serve as a partly independent validation of the K-model.

Paper IV concluded that stable Cs and Sr may be used for the determination

of CR of fish as an analogue for Cs-137 and Sr-90, respectively. Thus, the

results of estimation CR of Sr and Cs for freshwater and marine fish can be

included as additional values for assessment of the range of variation in CRs.

In addition, it demonstrated the strong inverse relationship between CRCs and

water [K] and between CRSr and water [Ca], illustrating how such

relationships could be used in the prediction of more site-specific CRCs and

CRSr in fish simply from water chemistry measurements.

24

Future research

Further work is required to gain a better understanding of the factors that

influence radionuclide/element transfer to biota, particularly for less well-

studied species (e.g. invertebrates, mixotrophic phytoplankton), for less well-

studied radionuclides/elements, also for types of biota that show wide ranges

of radionuclide transfer factors (e.g. macroalgae, phyto- and zooplankton). For

further validation of the developed lake model (Paper III) it is important to

simultaneously collect samples for water and lake organisms from all

functional groups in L. Eckarfjärden, and to chemically analyse a wide set of

elements, in order to determine CRs. The CRs based on site-specific measured

data will be possible to compare with predicted and estimated more precisely

transfer via food chains.

In the aquatic models developed in this thesis (Paper I−III) the simulations

were executed separately for every element, thus it was assumed that one

element concentration did not influence another element’s assimilation or

adsorption by organisms. However, in future models the water chemistry

needs to be taken into account because, for example, concentrations of K (an

analogue of Cs) and Ca (an analogue of Sr) may influence uptake by aquatic

organisms, as demonstrated in Paper IV. It is also known that the analogue of

Ra is Ba, and of Th is Pu, but it would be useful to identify other analogues

and evaluate how the presence of one may influence the uptake of the other

by biota. More processes that affect aquatic organisms uptake and excretion,

such as body size, temperature, light (in the case of algae) and salinity need to

be included in future models.

25

Acknowledgment

This thesis would not have been possible without the help, support, friendship

and patience of my principal supervisor, Associate Professor Clare Bradshaw,

not to mention her advice and unsurpassed knowledge of marine biology. The

good advice and support of my supervisors Dr. Ulrik Kautsky, Dr. Linda

Kumblad and Dr. Peter Saetre have been invaluable on both an academic and

a personal level, for which I am extremely grateful.

I would like to acknowledge, SKB who provided the necessary financial

support for this research. I thank participants of SR-Site SKB project for

useful discussions, particularly Eva Andersson, Karin Aquilonius, Sara

Nordén and Mats Tröjbom. I thank the staff at Facilia AB for their help and

assistance, particularly Rodolfo Avila, Per-Anders Ekström, Kristofer

Stenberg, Per-Gustav Åstrand, Erik Johansson and Robert Broed for their

advice concerning implementation of the compartment model in the software

Ecolego. I am most grateful to Sara Grolander, Jesper Torudd and Idalmis De

la Cruz for providing me with computer files of site-specific data and

references to publications, which have been valuable and reliable.

I would specially like to thank my article co-authors Anders Christian

Erichsen and Flemming Møhlenberg from DHI for their extremely valuable

experiences, support, and insights.

I am indebted to my many colleagues from the Department of Ecology,

Environment and Plant Sciences who supported me, especially Josefin

Segerman, Hanna Oskarsson, Ben Jaeschke, Göran Samuelsson, Nadja

Stadlinger, Charlotte Berkström and Nils Hedberg. Another special thanks

goes to Prof. Nisse Kautsky and Olle Hjerne for useful comments on the

manuscripts.

I would like to extend my sincerest thanks and appreciation to my parents

and sister for their emotional support no matter what path I choose. My sister

Galyna has been my best friend all my life and I love her dearly and thank her

for all her advice and support.

I wish to give my heartfelt thanks to my beloved Andrej, whose

unconditional love, patience, and continual support of my academic

endeavours especially over the last year of my study enabled me to complete

this thesis. Finally, there are my children, who have given me much happiness

and funny moments in my life.

26

References

Andersson E., (2005). Benthic-pelagic microbial interactions and carbon

cycling in clear water Lakes. Ph.D. thesis. Uppsala University, Sweden.

Aquilonius, K., (Ed), (2010). The marine ecosystems at Forsmark and

Laxemar-Simpevarp. SR-Site Biosphere. SKB Report, TR-10-03.

Swedish Nuclear Waste Management Co., Stockholm, Sweden.

Avila, R., Ekström, P. A., Åstrand, P.G., (2010). Landscape dose conversion

factors used in the safety assessment SR-Site. SKB TR-10-06, Swedish

Nuclear Waste Management Co., Stockholm, Sweden.

Avila, R., Kautsky, U., Ekström, P.-A., Åstrand, P.-G., Saetre, P., (2013).

Model of the long-term transport and accumulation of radionuclides in

future landscapes. AMBIO, 42(4): 497-505.

Baines, S.B., Fisher, N.S., Stewart, R., (2002). Assimilation and retention of

selenium and other trace elements from crustacean food by juvenile striped bass (Morone saxatilis). Limonology and Oceanography,

47: 646–655.

Beresford, N.A., Balonov, M., Beaugelin-Seiller, K., Brown, J., Copplestone,

D., Hingston, J.L., Horyna, J., Hosseini, A., Howard, B.J., Kamboj, S.,

Nedveckaite, T., Olyslaegers, G., Sazykina, T., Vives i Batlle, J.,

Yankovich, T.L., Yu, C., 2008. An international comparison of models

and approaches for the estimation of the radiological exposure of non-

human biota. Applied Radiation and Isotopes, 66: 1745-1749.

Berner, B., Drottz-Sjöberg, B.-M., Holm, E. (2009). Inledning. I

Samhällsforskning 2009. Betydelsen för människorna, hembygden och

regionen av ett slutförvar för använt kärnbränsle. Swedish Nuclear

Waste Management Co., Stockholm, Sweden. pp. 4-17.

27

BIOMOVS II, 1996a. Validation test for carbon-14 migration and

accumulation in a Canadian Shield Lake. Technical Report no. 14.

Swedish Radiation Protection Institute, Stockholm, Sweden, p. 40.

BIOMOVS II, 1996b. Assessment of the consequences of the radioactive

contamination of aquatic media and biota. Technical Report no. 10.

Swedish Radiation Protection Institute, Stockholm, Sweden, p. 70.

Bird, G.A., Stephenson, M., Cornett, R.J., (1993). The surface water model

for assessing Canada's nuclear fuel waste disposal concept. Waste Mgmt,

13: 153-170.

Buesseler, K., Jayne, S., Fisher, N., (2012). Fukushima-derived radionuclides

in the ocean and biota off Japan. Proceedings of the National Academy

of Sciences 109(16), 5984-5988

Chapman, P.C., Adams, W.J., (Eds.) (2007). Assessing the Hazard of Metals

and Inorganic Metal Substances in Aquatic and Terrestrial Systems.

CRC Press. 146 pp.

Copplestone, D., Beresford, N.A., Brown, J.E., Yankovich, T., (2013). An

international database of radionuclide concentration ratios for wildlife:

development and uses, Journal of Environmental Radioactivity, 126:

288-298.

Erichsen, A.C., Møhlenberg, F., Closter, R.M., Sandberg, J., (2010). Models

for transport and fate of carbon, nutrients and radionuclides in the

aquatic ecosystem at Öregrundsgrepen. DHI, SKB R-10-10. Swedish

Nuclear Waste Management Co., Stockholm, Sweden.

Fowler, S.W., Fisher, N.S., (2005). Chapter 6 Radionuclides in the biosphere,

in: Hugh, D.L. (Ed.), Radioactivity in the Environment. Elsevier, pp.

167-203.

Harms, H. I., Karcher, M.J., Burchard, H., (2003). Modelling radioactivity in

the marine environment: The application of hydrodynamic circulation

models for simulating oceanic dispersion of radioactivity. Chapter 3,

pp.55-85. Modelling radioactivity in the environment, E. Marian Scott

(Editor), © 2003 Elsevier Science Ltd.

HELCOM 2009. Radioactivity in the Baltic Sea 1999-2006, Baltic Sea

environment proceedings, No 117: 47-49.

28

Heling, R., Bezhenar, R., (2009). Modification of the dynamic radionuclide

uptake model BURN by salinity driven transfer parameters for the

marine foodweband its integration in POSEIDON-R. Radioprotection

44(5): 741–746.

Hofman, D., Monte, L., Boyer, P., Brittain, J., Donchyts, G., Gallego, E.,

Gheorghiu, D., Håkanson, L., Heling, R., Kerekes, A., Kocsy, G.,

Lepicard, S., Slavik, O., Slavnicu, D., Smith, J., Zheleznyak, M., (2011). Computerised Decision Support Systems for the management of

freshwater radioecological emergencies: assessment of the state-of-the-

art with respect to the experiences and needs of end-users. Journal of

environmental radioactivity, 102: 119–27.

Howard, B. J., Beresford, N. a, Copplestone, D., Telleria, D., Proehl, G.,

Fesenko, S., Jeffree, R. A., Yankovich, T. L., Brown, J. E., Higley,

K.,Johansen, M P, Mulye, H., Vandenhove, H., Gashchak, S., Wood,

M. D., Takata, H., Andersson, P., Dale, P., Ryan, J., Bollhöfer, A.,

Doering, C., Wells, C. (2013). The IAEA handbook on radionuclide

transfer to wildlife. Journal of Environmental Radioactivity, 121: 55–74.

Håkanson, L., Monte, L. (2003) Radioactivity in lakes and rivers. In Scott,

E.M. (ed.), Radioactivity in the Environment. Elsevier, pp.147-200.

Håkanson, L., Andersson, T., Nilsson, Å., (1992). Radioactive Caesium in

Fish in Swedish Lakes 1986-1988. General Pattern Related to Fallout

and Lake Characteristics Tord Andersson & Ake Nilsson, Journal of

environmental radioactivity, 15: 207–229.

IAEA 1985a. Sediment Kds and concentration factors for radionuclides in the

marine environment. Technical Report Series No. 247., p.76.

International Atomic Energy Agency, Vienna, Austria.

IAEA 1985b. Hydrological Dispersion of Radioactive Material in Relation to

Nuclear Power Plant Siting. A Safety Guide. Safety Series, 50-SG-S6.

International Atomic Energy Agency, Vienna, Austria.

IAEA 1994. Handbook of Parameters Values for the Prediction of

Radionuclide transfer in temperate Environments. Technical Reports

series, 364. International Atomic Energy Agency, Vienna.

IAEA 2000. Modelling of the transfer of radiocaesium from deposition to lake

ecosystems. Report of the VAMP Aquatic Working Group. IAEA-

TECDOC-1143. International Atomic Energy Agency, Vienna, Austria.

29

IAEA 2004. Sediment distribution coefficients and concentration factors for

biota in the marine environment. Technical reports series no. 422.

International Atomic Energy Agency, Vienna, Austria.

IAEA 2009. Calmon, P., Fesenko, S., Voigt, G., & Linsley, G. (2009).

Quantification of radionuclide transfer in terrestrial and freshwater

environments for radiological assessments. Report IAEA-TECDOC-

1616. International Atomic Energy Agency, Vienna.

IAEA 2010a. Handbook of parameter values for the prediction of radionuclide

transfer in terrestrial and freshwater environments. Technical reports

series no 472. International Atomic Energy Agency, Vienna, Austria.

IAEA 2012b. Environmental Modelling For Radiation Safety (EMRAS).

Summary Report of the Results of the EMRAS Programme (2003–

2007). IAEA-TECDOC-1678. © International Atomic Energy Agency,

Vienna.

IAEA 2012c.Testing of Models for Predicting the Behaviour of Radionuclides

in Freshwater Systems and Coastal Areas. Report of the Aquatic

Working Group of EMRAS Theme 1. IAEA-TECDOC-1678.

International Atomic Energy Agency, Vienna, Austria. http://www-

ns.iaea.org/downloads/rw/projects/emras/final-reports/aquatic-tecdoc-

final.pdf

IAEA 2012b. Modelling radiation exposure and radionuclide transfer for non-

human species: report of the Biota Working Group of EMRAS Theme

3. Vienna, International Atomic Energy Agency, 244pp. http://www-

ns.iaea.org/downloads/rw/projects/emras/final-reports/biota-final.pdf

ICRP 2007. Recommendations of the international commission on

radiological protection. Ann. ICRP 37, 2e3. Publication 103.

Iosjpe, M. (2006). Environmental modelling: Modified approach for

compartmental models. in: P.P. Povinec, J.A. Sanchez-Cabeza (Eds.),

Radionuclides in the Environment. Radioactivity in the Environment,

Series Ed, 8, M.S. Baxter, pp. 463–476

Kautsky, U., Lindborg, T., Valentin, J. (2013). Humans and ecosystems over

the coming millennia: overview of a biosphere assessment of radioactive

waste disposal in Sweden. AMBIO, 42(4): 383-392.

Kryshev, A. I. (2006). 90Sr in fish: a review of data and possible model

approach. The Science of the total environment, 370, 182–9.

30

Kryshev, A. I., & Ryabov, I. N. (2000). A dynamic model of Cs-137

accumulation by fish of different age classes. Journal of Environmental

Radioactivity, 50, 221–233.

Kryshev, I.I., Sazykina, T.G., Hoffman, F.O., Thiessen, K.M., Blaylock, B.G.,

Feng, Y., Galeriu, D., Heling, R., Kryshev, A.I., Kononovich, A.L.,

Watkins, B., (1999). Assessment of the consequences of the radioactive

contamination of aquatic media and biota for the Chernobyl NPP cooling

pond: model testing using Chernobyl. Journal of Environmental

Radioactivity, 42: 143–156.

Kumblad L., Kautsky U., (2004a). Effects of land‐rise on the development of

a coastal ecosystem of the Baltic Sea and its implications on the long‐term fate of C‐14 discharges. Hydrobiologia, 514: 185‐196.

Kumblad, L., Kautsky, U., (2004b). Models for transport and fate of carbon,

nutrients and point source released radionuclides to an aquatic

ecosystem. SKB TR-04-13, Swedish Nuclear Waste Management Co.,

Stockholm, Sweden.

Kumblad, L., (2004). Radionuclides in the Baltic Sea. Ecosystem models and

experiments on transport and fate. Doctoral Thesis, Department of

System Ecology, Stockholm University.

Kumblad, L., Bradshaw, C., (2008). Element composition of biota, water and

sediment in the Forsmark area, Baltic Sea. Concentrations,

bioconcentration factors and partitioning coefficients (Kd) of 48

elements. SKB TR-08-09, Swedish Nuclear Waste Management Co.,

Stockholm, Sweden.

Kumblad, L., Gilek, M., Næslund, B., Kautsky, U., (2003). An ecosystem

model of the environmental transport and fate of C-14 in a bay of the

Baltic Sea, Sweden. Ecological Modelling, 166: 193-210.

Kumblad, L., Kautsky, U., Næslund B., (2006). Transport and fate of

radionuclides in aquatic environments: the use of ecosystem modelling

for exposure assessments of nuclear facilities. Journal of Environmental

Radioactivity, 87: 107-129.

Lepicard, S., Heling, R., Maderich, V., (2004). POSEIDON/RODOSmodels

for radiological assessment of marine environment after accidental

31

releases: application to coastal areas of the Baltic, Black and North Seas.

Journal of Environmental Radioactivity, 72(1-2):153-161.

Lindborg, T., (ed.), (2010). Landscape Forsmark - data. methodology and

results for SR-Site. SKB Report, TR-10-05. Swedish Nuclear Waste

Management Co., Stockholm, Sweden.

Lindborg, T., (ed.) (2008). Surface system Forsmark Site descriptive

modeling SDM-Site Forsmark. SKB Report, TR-08-11. Swedish

Nuclear Waste Management Co., Stockholm, Sweden.

Margvelashvili, N., Maderich, V., Yuschenko, S., Zheleznyak, M., (2002). 3-

D numerical modelling of mud and radionuclide transport in the

Chernobyl Cooling Pond and Dnieper-Boog Estuary. Proceedings in

Marine Science. Fine Sediment Dynamics in the Marine Environment,

5: 595–609.

Maderich, V., Bezhenar, R., Heling, R., Jung, K.T., Myoung, J.G., Cho, Y.-

K., Qiao, F., Robertson, L., (2014). Regional long-term model of

radioactivity dispersion and fate in the Northwestern Pacific and

adjacent seas: application to the Fukushima Dai-ichi accident. Journal

of Environmental Radioactivity, 131: 4–18

Meyer J.S. Ed., (2005).Toxicity of Dietborne Metals to Aquatic Organisms.

Publisher: Pensacola, FL: Society of Environmental Toxicology and

Chemistry.

Monte, L., Brittain, J.E., Gallego, E., Håkanson, L., Hofman, D., Jiménez, A.,

(2009a). MOIRA-PLUS: a decision support system for the management

of complex fresh water ecosystems contaminated by radionuclides and

heavy metals. Computers & Geosciences, 35: 880-896.

Monte, L., Periañez, R., Boyer, P., Smith, J.T., Brittain, J.E., (2009b). The role

of physical processes controlling the behaviour of radionuclide

contaminants in the aquatic environment: a review of state-of-the-art

modelling approaches. Journal of Environmental Radioactivity, 100:

779–784.

Nordén, S., Avila, R., de la Cruz, I., Stenberg, K., Grolander, S., (2010).

Element-specific and constant parameters used for dose calculations in

SR-Site. SKB Report, TR-10-07. Swedish Nuclear Waste Management

Co., Stockholm, Sweden.

32

Outola, I., Saxén, R.L., Heinävaara, S., (2009). Transfer of 90Sr into fish in

Finnish lakes. Journal of Environmental Radioactivity, 100: 657–64.

Periáñez, R., Suh, K.-S., & Min, B.-I. (2012). Local scale marine modelling

of Fukushima releases. Assessment of water and sediment contamination

and sensitivity to water circulation description. Marine Pollution

Bulletin, 64(11), 2333–9.

Periáñez, R., Bezhenar, R., Iosjpe, M., Maderich, V., Nies, H., Osvath, I.,

Outola I., de With, G. (2015). A comparison of marine radionuclide

dispersion models for the Baltic Sea in the frame of IAEA MODARIA

program. Journal of Environmental Radioactivity, 139: 66–77.

Periáñez, R., Suh, K.-S., & Min, B.-I. (2012). Local scale marine modelling

of Fukushima releases. Assessment of water and sediment contamination

and sensitivity to water circulation description. Marine Pollution

Bulletin, 64(11), 2333–9.

Santschi, P.H., Honeyman, B.D., (1989). Radionuclides in aquatic

environments. Radiation Physics and Chemistry, 34: 213-240.

Sazykina, T., (2003). Radioactivity in aquatic biota. Modelling radioactivity

in the environment. pp. 201-222, E. Marian Scott (Ed.). © 2003 Elsevier

Science Ltd.

Sazykina, T.G., (2000). ECOMOD: an ecological approach to radioecological

modelling. Journal of Environmental Radioactivity, 50: 207-220.

Simkiss, K., Taylor, M.G., (1995). Transport of metals across membrenes. In:

Tessier, A., Turner D.R. (Eds.), Metal Speciation and Bioavailability in

Aquatic Systems, UK: John Wiley & Sons, Wallingford, Chichester. 44

pp.

SKB 2011. Long-term safety for the final repository for spent nuclear fuel at

Forsmark. Main report of the SR-Site projectSKB Report, TR-11-01.

Swedish Nuclear Waste Management Co., Stockholm, Sweden.

Steinhauser, G., Brandl, A., & Johnson, T. E. (2013). Comparison of the

Chernobyl and Fukushima nuclear accidents: A review of the

environmental impacts. The Science of the Total Environment, 470-

471C, 800–817.

33

Smith, C.N., Clarke, S., McDonald, P., Goshawk, J.A., Jones, S.R., (2000).

Reconstructing historical radionuclide concentrations along the east

coast of Ireland using a compartmental model. Science of the Total

Environment, 254: 17-30.

Smith, J.T., Kudelsky, A.V., Ryabov, I.N., Hadderingh, R.H., & Bulgakov,

A.A. (2003). Application of potassium chloride to a chernobyl-

contaminated lake: Modelling the dynamics of radiocaesium in an

aquatic ecosystem and decontamination of fish. The Science of the

Total Environment, 305, 217–227.

Smith, J.T., Sasina, N.V., Kryshev, A.I., Belova, N.V., Kudelsky, A.V.,

(2009). A review and test of predictive models for the bioaccumulation

of radiostrontium in fish. Journal of Environmental Radioactivity, 100:

950–954.

Tagami, K. & Uchida, S. (2012) Marine and freshwater concentration ratios

(CRwo-water): review of Japanese data. Journal of Environmental

Radioactivity 126, 420–6.

Takahashi, S., (Ed.), (2014). Radiation Monitoring and Dose Estimation of

the Fukushima Nuclear Accident. (S. Takahashi, Ed.). Tokyo: Springer

Japan.

Tateda, Y., Tsumune, D., & Tsubono, T. (2013). Simulation of radioactive

cesium transfer in the southern Fukushima coastal biota using a

dynamic food chain transfer model. Journal of Environmental

Radioactivity, 124, 1–12.

Tröjbom, M., Nordén, S., (2010). Chemistry data from surface ecosystems in

Forsmark and Laxemar-Simpevarp. Site specific data used for

estimation of CR and Kd values in SR-Site. SKB R-10-28, Swedish

Nuclear Waste Management Co., Stockholm, Sweden. Svensk

Kärnbränslehantering AB.

Tröjbom M., Grolander S., Rensfeldt V., Nordén S., (2013). Kd and CR used

for transport calculations in the biosphere in SR-PSU. SKB Report, R-

13-01, Swedish Nuclear Waste Management Co., Stockholm, Sweden.

Svensk Kärnbränslehantering AB, pp.363.

Tsumune, D., Tsubono, T., Aoyama, M., & Hirose, K. (2011). Distribution

of oceanic 137 Cs from the Fukushima Dai-ichi Nuclear Power Plant

simulated numerically by a regional ocean model. Journal of

Environmental Radioactivity, 3, 1–9.

34

Wang, W.-X., Rainbow, P.S. (2008). Comparative approaches to understand

metal bioaccumulation in aquatic animals. Comparative biochemistry

and physiology. Toxicology & pharmacology: CBP, 148: 315-23.

Vives i Batlle, J., Wilson, R.C., Watts, S.J., Jones, S.R., McDonald, P.,

Vives-Lynch, S. (2008). Dynamic model for the assessment of

radiological exposure to marine biota. Journal of environmental

radioactivity, 99: 1711-30.

Vives I Batlle, J., Aono, T., Brown, J. E., Hosseini, A., Garnier-Laplace, J.,

Sazykina, T., Steenhuisen, F., Strand, P. (2014). The impact of the

Fukushima nuclear accident on marine biota: retrospective assessment

of the first year and perspectives. The Science of the Total

Environment, 487, 143–53

Vanderploeg, H.A., Parzyck, D.C., Wilcox, W.H., Kerchner, J.R., Kaye,

S.V., 1975. Bioaccumulation Factors for Radionuclides in Freshwater

Biota. ORNL-5002. Oak Ridge National Laboratory, Oak Ridge,

Tennessee.

Yankovich, T.L., (2005). EMRAS Biota Working Group (BWG): The Perch

Lake Freshwater Scenario Description (Revision 2). ETB-05–86 (Rev.

2), Atomic Energy Canada Limited, Chalk River, Canada. Available

from: /http:// www-ns.iaea.org/projects/emras/emras-biota-wg.htm