1

Novel Methods for Integrated Risk Assessment ofCumulative Stressors in the Environment

NOMIRACLE

Integrated Project to Call FP6-2003-Global-2

Priority 1.1.6.3 of the Sixth Framework Programme: ‘Global Change and Ecosystems’Topic VII.1.1.a (‘Development of risk assessment methodologies’)

Co-ordinator

Hans LøkkeNational Environmental Research Institute (NERI), Denmark

2

List of ParticipantsN

o.

Acronym Institution Nation

SME

Person in charge

Gen

der

1 NERI National Environmental Research Institute Denmark Hans Løkke M2 NERC Natural Environment Research Council UK David Spurgeon M3 UFZ UFZ Centre for Environmental Research Germany Gerrit Schüürmann M4 DESUN University of Nijmegen Netherlands Ad Ragas M5 DSTA University Piemonte Orientale Italy Aldo Viarengo M6 VU Vrije Universiteit, Amsterdam Netherlands Kees van Gestel M7 NIPH National Institute for Public Health Czech Republic Milon Tichy M8 UWC Cardiff University UK Stephen Stürzenbaum M9 UCAM Cambridge University UK Jules Griffin M10 UJAG Jagiellonian University Poland Ryszard Laskowski M11 EKUT University of Tübingen Germany Heinz Köhler M12 WU Wageningen University Netherlands Jan Kammenga M13 DBUA University of Aveiro Portugal Amadeu Soares M14 UA University of Antwerp Belgium Wim De Coen M15 WRcNSF WRc-NSF Ltd UK X Jason Weeks M16 LEMTEC LemnaTec Germany X Matthias Eberius M17 USALZ Salzburg University Austria Albert Duschl M18 JRC Directorate General Joint Research Centre,

European CommissionItaly David Pennington M

19 SYKE Finnish Environment Institute Finland Timo Assmuth M20 APINI Kaunas University of Technology Lithuania Jurgis Staniskis M21 ALTERRA Alterra Netherlands Jack Faber M22 EAWAG Swiss Institute of Environmental Science

and TechnologySwitzerland Kai-Uwe Goss M

23 RIVM National Institute of Public Health and theEnvironment

Netherlands Dik van de Meent M

24 LIMCO LimCo International Germany X Almut Gerhardt F25 RWTHA Aachen University of Technology Germany Ingolf Schuphan M26 ECT ECT Oekotoxikologie GmbH Germany X Thomas Knacker M27 UNIMIB Consorzio Interuniversitario Scienze del

Mare – Università di Milano BicoccaItaly Marco Vighi M

28 ENVI Environment Park SPA Italy X Federica Peirano F29 EPFL Ecole Polytechnique Fédérale de Lausanne Switzerland Jérome Payet M30 ULANC Lancaster University UK Hao Zhang F31 ITM Stockholm University Sweden Michael McLachlan M32 DIA DIALOGIK Germany Ortwin Renn M33 URV Universitat Rovira i Virgili Spain Francesc Giralt M34 LHASA LHASA Ltd UK X Philip Judson M35 LMC Bourgas "Prof. As. Zlatarov" University Bulgaria Ovanes Mekenyan M36 CSIC Consejo Superior de Investigaciones

CientificasSpain Damia Barcelo M

37 USOUTH University of Southampton UK Jean Lou Dorne M38 SYMLOG Institut SYMLOG de France France Claire Mays F

3

Table of Contents

page

B.1 Scientific and technological objectives of the project and state of the art 6

B.2 Relevance to the objectives of the Sub-Priority “Global Change and Ecosystems” 13

B.3 Potential Impact 14

B.3.1 Contributions to standards/policies/regulations 17

B.4 Outline implementation plan 18

B.4.1 Research, technological development and innovation activities 20

Research Pillar 1 – Risk Scenarios 20

Research Pillar 2 – Sound Exposure 24

Research Pillar 3 – Effect Assessment 33

Research Pillar 4 – Risk Assessment 39

Management – Pillar 5 47

B.4.2 Demonstration activities 52

B.4.3 Training activities 52

B.4.4 Management activities 56

B.5 Description of the Consortium 56

Experience and role of partners 56

B.6 Description of project management 64

B.7 Project resources 70

B.8 Detailed implementation plan – first 18 months 75

Work package list (18 month plan) 91

Deliverables list (18 month plan) 93

Work Package 1.1 description:Establishment of data background for scenario selection (18 month plan)

100

Work package 1.2 description:Scenario selection and ranking (18 month plan)

103

4

Work Package 2.1 description:Matrix-compound interaction (18 month plan)

105

Work Package 2.2 description:Available exposure (18 month plan)

107

Work Package 2.3 description:Metabolic fate (18 month plan)

109

Work Package 2.4 description:Region-specific environmental fate (18 month plan)

112

Work package 3.1 description:Interactive toxicological effects in diverse biological systems (18 month plan)

114

Work package 3.2 description:Combined effects of natural stressors and chemicals (18 month plan)

117

Work package 3.3 description:Toxicokinetic modelling (first 18 month plan)

120

Work package 3.4 description:Molecular mechanisms of mixture toxicity (18 month plan)

123

Work package 4.1 description:New concepts and techniques for probabilistic risk assessment (18 month plan)

126

Work package 4.2 description:Explicit modelling of exposure and risk in space and time (18 month plan)

129

Work package 4.3 description:Dealing with multiple and complex risks in a management context (18 month plan)

132

Work package 4.4 description:Risk presentation and visualisation (18 month plan)

135

Work Package 5.1 description:General Project Management (18 month plan)

138

Work Package 5.2 description:Data management (18 month plan)

140

Work Package 5.3 description:Training and Demonstration (18 month plan)

142

Work package 5.4 description:Dissemination and Exploitation (18 month plan)

143

5

B.9 Ethical issues: using animals for science 145

B.10 Gender issues 146

B.10.1 Gender Action Plan 146

B.10.2 Gender issues 147

Annexes

Annex 1 – References used in Part B

Annex 2 – Extended description of Partner capabilities

6

B.1 Scientific and technological objectives of the project and state of the art

Science & technology objectives

1. To develop new methods for assessing the cumulative risks from combined exposures to severalstressors including mixtures of chemical and physical/biological agents

2. To achieve more effective integration of the risk analysis of environmental and human healtheffects

3. To improve our understanding of complex exposure situations and develop adequate tools forsound exposure assessment

4. To develop a research framework for the description and interpretation of cumulative exposureand effect

5. To quantify, characterise and reduce uncertainty in current risk assessment methodologies, e.g.by improvement of the scientific basis for setting safety factors

6. To develop assessment methods which take into account geographical, ecological, social andcultural differences in risk concepts and risk perceptions across Europe

7. To improve the provisions for the application of the precautionary principle and to promote itsoperational integration with evidence-based assessment methodologies

State of the art

The assessment of risks from chemicals to humans and the environment in the European Union hasresulted in a number of comprehensive regulatory frameworks for various classes of compounds.For pesticides the European Community has developed Directive 91/414/EEC (EC 1991), definingstrict rules for authorising the use and application. The Directive requires extensive risk assessmentsfor effects on health and the environment to be carried out, before a product can be placed on themarket and used. For biocides, Directive 98/8/EC (EC 1998) on the placing on the market ofbiocidal products was adopted in 1998 where the regulation for pesticides served as a model; forpharmaceuticals a Directive is under development. In parallel, a new Directive is being developedfor industrial and other new and existing chemicals. The proposed new EU chemicals strategy andthe REACH (Registration, Evaluation and Authorization of CHemicals) system will unify andamend these pieces of legislation (CEC 2001, 2003a).The Technical Guidance Documents (TGD)for risk assessment in these contexts have been recently updated (EC, 2003) and represent apragmatic support tool for regulatory purposes. Recently, the European Environment and HealthStrategy (SCALE) addressed the shortcomings of the current methods (CEC 2003a).

Many acute environment and health related problems have been solved, but much remains to bedone, in particular with respect to the health implication of chronic exposures, as reported byorganisations such as the European Environmental Agency, WHO and a number of nationalorganisations. They indicate that the interaction between environment and health is far morecomplex than is commonly understood. In particular, little attention has been paid to the interactionof different pollutants in the human body as well as in the environment. Even low level exposureover a period of time to a complex cocktail of pollutants in air, water, food and consumer productsis likely to contribute significantly to the health status of European citizens. It is estimated thataround 25-33% of the burden of disease in industrialised countries can be attributed toenvironmental factors, with the bulk of this affecting children and vulnerable groups (Smith et al.1999). The majority of Europeans also perceives the magnitude of the problem: in a recent survey,some 89% are worried about the potential impact of the environment on their health (Eurobarometer2002). Furthermore, new technologies, changing lifestyles, work and life patterns, present new andsometimes unexpected impacts on the environment and its influence on health. Within this projectwe will improve both human and environmental risk assessment procedures by addressing a seriesof major shortcomings that exist within the current approaches, namely that:

7

i) they are based on direct effects of single compounds or productsii) they apply uncertainty factors which are not strictly based on scientific principlesiii) they do not account for multiple stressors and indirect effects in a dynamic and heterogeneousenvironmentiv) they typically do not account for cumulative (integrated over time, space, substances) effects,andv) they do not allow for site specific and other spatially detailed evaluations

Although it is generally acknowledged that chemical, biological, radiological, and other physicaland even psychological stressors can cause a variety of human health or ecological health effects,assessing the risks associated with them is considerably more complex methodologically andcomputationally than current risk assessment practices. Given these lacunas there is an urgent needfor ”cumulative risk assessment” which can be defined as “an analysis, characterisation, andpossible quantification of the combined risks to health or the environment from multiple agents orstressors” (US-EPA 2003). Development of a framework for such complex risk assessments willgreatly improve understanding of the effects of cumulative exposures occurring under the variety offield conditions within Europe and will provide a better scientific basis for forecasting risks andassociated uncertainties. The understanding of the complexity of cumulative risks is a prerequisitefor development of more efficient guidelines to provide data for future regulation of chemicals, onone hand taking into account the proportionality of different risks, and on the other hand meet theneed for improving environment and human health.

In the integration of human and environmental risk assessment, a powerful and promising strategyis to follow the line of a bioinformatics approach to ecological systems, where the high degree ofinternal complexity is accepted as an inherent property, and consequently the state of the systemmust be analysed in terms of possibly several thousand measurable variables. The current projectadvocates such an approach based on a robust and computational reliable framework that answersthe call for cumulative risk assessment and for reducing uncertainty. To formulate this framework,it is essential that the underlying processes and mechanisms are understood, as it is only byknowledge of how chemicals interact with the abiotic environment and within organisms thatmodels to describe such effects can be developed and applied. By adopting this approach,NOMIRACLE will place Europe as the world leader in considering the complexity of interactionsbetween populations inhabiting the real world.

In the following, the state of the art concerning each science & technology objective is presentedtogether with the expected improvements provided by the NOMIRACLE project.

Ad 1 To develop new methods for assessing the cumulative risks from combined exposures toseveral stressors including mixtures of chemical and physical/biological agents

State of the art:The shift from a single-compound to a cumulative approach has radical consequences for theassessment of exposure, effects and risks. In exposure assessment, current procedures tend to focuson the fate of chemicals in relatively homogeneous environments (e.g., in EUSES) and simple rulesare applied to estimate the amount of chemicals that reaches the receptor (McKone & Ryan 1989).However, in a truly cumulative approach it is realised that human or ecological receptors are notexposed to individual substances in a relatively homogeneous environment, but to toxic mixturesand other natural stress factors (e.g., severe drought or extreme temperatures) in a heterogeneousenvironment. These factors have not been accounted for in current risk assessment procedures(Heugens et al. 2001). Also, space and time play a crucial role because the spatial and temporalpatterns of the stressors and the receptors determine the ultimate effect.

8

In effect assessment, a substantial effort has been made to label the effect of mixtures as synergistic,antagonistic, or additive. Although these concepts can be communicated relatively easily, they havecontributed fairly little to our understanding of joint toxic action. Part of the difficulty is that theunderlying toxicological concepts are weakly defined and that the use of simplified mathematicalcharacterisations of synergism and antagonism reveal either interaction or no-interaction and carrylarge uncertainties.

Currently, the division of the exposure estimate with an effect measure, as a proxy for dose-response relationship (e.g. PEC/PNEC-ratio), results in a characterisation of risk (Van Leeuwen1995). This procedure can only be followed in cumulative risk assessment when the effects of thestressors can be expressed in comparable endpoints, either in biological terms or in terms ofvaluation.

This project adopts a unique spatial- and receptor-oriented approach for assessing the integratedexposure to multiple stressors. Using promising research directions, such as toxicokinetic andtoxicogenomic studies, we will develop biologically based models of joint toxicity to replacecurrent interaction labels for mixture risk assessment procedures (Groten et al. 2001, Hertzberg &MacDonell 2002). By doing so it will be possible to assess the overall effect of combined exposureto mixtures and (a) biotic stressors, and subsequently to develop integrated assessment methods,such as advocated by Van Straalen (2003). Mixture toxicity studies in combination with biotic andabiotic stressors will be based on a statistically robust framework for assessing combined effectswhich was developed, e.g. within the EU funded MIXTOX project ENV4-CT97-0507 (Jonker et al.in press a,b,c).

Expected Improvements:

• Methods for assessing potential effects and for characterising cumulative risk will be developedtaking into account the real characteristics of potentially exposed human and ecologicalreceptors (sensitivity, vulnerability, value as natural resource, realistic probability of exposure,etc).

• A unique aspect of NOMIRACLE is the conjunction of toxicity studies with mechanisticresearch at the chemical, biochemical, and genomic level, which will give understanding of theunderlying mechanisms of combinatorial effects as, e.g., advocated by Van Straalen (2003).

• Development of integrative endpoints that can be used for characterisation of cumulative risks.• Novel spatial- and receptor-oriented approaches for assessing the integrated exposure to

multiple stressors.

Ad 2 To achieve more effective integration of the risk analysis of environmental and humanhealth effects

State of the art:A key dimension in integration of risk assessment is that between human and non-human receptors(WHO 2001). This is a central element in the evolving EU Environmental Health Strategy focusedon chemicals (CEC 2003); it is also part of the new EU Chemicals Policy and the REACH system(CEC 2003). The development of science toward cross-cutting theories and methods has aided suchintegration e.g. through molecular biology and genetics. In environmental science, a reflection ofthis strive is the ecosystem health concept (e.g. Rapport et al. 1998). In exposure assessment betterinteraction of chemical fate data and models for the various receptors is needed. In toxicologicalrisk assessment the traditional area of integration is extrapolation from animal models to humans. Akey here is comparative tissue dosimetry utilising biokinetic and toxicodynamic models and data(Welsch et al. 1995, Anderson & Dennison 2001, Walton et al. 2001). In ecotoxicology the focus

9

has been on species sensitivity distributions (Forbes et al. 2001, Pennington 2003) but human-nonhuman integration will require also other inputs from comparative and evolutionary toxicology.Relatively little research has been made in human-nonhuman integration for substances withspecific sites (Kalberlah et al. 2002) or modes of action (Barton & Clewell 2000, Bogdanffy et al.2001), except for drugs. Mixture effects complicate this; options to address them include modellingof interactions (Frederick et al 2000) and integrative biomarkers (e.g. Gibson & Starr 1988). Someof the limitations to integration of human and nonhuman assessment are fundamental, sometechnical; some are related to differing ecology and exposure, some to physiology and metabolism(e.g. Kalberlah et al 2002). It is often stressed that human health unlike ecological assessmentstarget individual level risks. Regardless of differences in receptor significance or in exposure andeffects mechanisms, advanced methods in the human health area e.g. to distinguish the contributionof risk factors in multi-stressor settings offer tools also for ecological risk assessment; and viceversa. A better understanding of the shared and specific biological and other processes in thevarious receptors is highly needed, and will be developed in the NOMIRACLE project at variouslevels from molecular to ecosystems.

Expected Improvements:

• Statements about the validity of the systems for toxicity testing in human toxicology andecotoxicology especially based on advances in the understanding of molecular and cellularmechanisms and biomarkers of toxicity, providing a better transferability betweenecotoxicology and human toxicology

• Since this project is the first one that investigates ecotoxicological and human toxicologicaltest principles simultaneously, a comparison will result in a better understanding of theoutputs, which is expected to help interpret the results and transferability between both.

• Direct comparability between effects of chemicals (under different conditions) on theenvironment and the human health.

• The uncertainty analyses will help elucidate the relative share of the human and nonhumancomponents in overall uncertainty, and the options for integration across receptors andsectors

Ad 3 To improve our understanding of complex exposure situations and develop adequatetools for sound exposure assessment

State of the art:Within the current regulatory framework of risk assessment (EC TGD 2003), the regional and conti-nental exposure assessment of chemical substances is usually based on predictions from genericmultimedia fate models such as EUSES (EC 1996, den Hollander et al. 2003). Whilst multimediamodels are suitable instruments for exposure assessment, their application is currently hampered byserious shortcomings, particularly as regards modern bioactive agents such as biocides, pesticidesand pharmaceuticals. Firstly, the release patterns and their variations in time and space dependingon the uses of these agents are not well taken into account. Secondly, the model predictions arebased on phase partitioning algorithms that work well only for non-polar and hydrophobicchemicals. Thirdly, the models lack a functionality to relate total compound concentrations to thematrix-specific effective concentrations that govern the resultant risk potential, as has beenemphasised in recent publications (ECETOC 2002 and 2003). Quantifying available exposure isalso a problem for soil-, sediment- and cell-based laboratory systems, and this problem affectsexposure-concentration curves and the determination of endpoints for environmental and humantoxicology accordingly. Fourthly, generic multimedia models do not account for indoor exposurethat is of primary importance for humans, and do not address emissions from consumer productsand confounding factors such as human time-activity patterns (e.g. ECETOC 2001). Fiftly, the

10

exposure predictions are frequently highly uncertain due to the paucity of reliable methods ofpredicting environmental degradation rates quantitatively (Sabljic & Peijnenburg 2001), and theyusually do not address the fate of metabolites formed under natural conditions (Fenner et al. 2002).Sixthly, the models currently used to support risk assessment in the regulatory context are generic innature and have not been parameterised to account properly for spatial and temporal variationacross the European continent (Huijbregts et al. 2003).

Expected Improvements:

• Fate and exposure models that account properly for the compound-matrix interaction andresultant phase partitioning of modern bioactive agents

• Methodologies for quantifying the matrix-specific available exposure of xenobiotics thatare relevant for toxic and ecotoxic effects, covering both the field and laboratory biotestsystems

• Tools for improved prediction of degradation rates and pathways of organic chemicals• Non-generic models that can be tailored to the complex exposure situations encountered in

the European Union• Indoor exposure patterns and assessment methods for homes, kindergartens and public

building across Europe

Ad 4 To develop a research framework for the description and interpretation of combinedexposure effects that leads to the identification of biomarkers of cumulative exposure andeffect

State of the art:It has been suggested that current reference models for mixture toxicity (concentration addition,independent action) could incorrectly predict combined effects in at least 25% of cases (Hertzbergand McDonell, 2002). Additionally, while the models are only descriptive it is impossible toaddress the uncertainty in when, how and why they may fail. This represents an unacceptable levelof uncertainty in our ability to conduct predictive risk assessments for chemical mixtures. To createa robust and scientifically valid approach for risk assessment of complex exposures, there is a needto go beyond descriptive approaches and analyse the mechanisms and pathways that linkcumulative exposures to eventual effects (Eggen et al. 2003, Hertzberg and McDonell, 2002, VanStraalen, 2003. Such mechanistic understanding should also enable risk assessment to include 1) theeffects of environmental stressors which interact to change an organisms sensitivity to chemicalexposure and 2) the effects of chemicals on the tolerance limits of the organism to natural stressors(e.g. Holmstrup et al. 2004; Heugen et al. 2003).

Adopting this mechanistic approach will require a coalescence of state of the art methods intoxicokinetics and the molecular biosciences. NOMIRACLE will use these approaches toinvestigate interaction mechanisms at the physiological level. Conducting this work in diversespecies will highlight species specific and common biomarkers for used in ecological monitoring ofcumulative exposure and resulting effect. These will be applicable as effects based monitoring toolsfor direct use in cases where the current state of knowledge leads to uncertainties in risk prediction.Further, as NOMIRACLE include mammalian cells and rodents among the systems investigatedthese indicators could be applicable to the epidemiological assessment of human population health.

11

Expected Improvements:

• Development of a biological-systems orientated approach to understanding the occurrenceconsequences of cumulative stress effects in the environment by linking molecular genetic andwhole organism approaches.

• Understanding of the physiological mechanisms through which effects of combined exposuresbecome manifest, allowing better estimates of when predictive models are reliable and whenthey may fail.

• Highlighting potential species specific and common biomarkers that can be used in monitoringof cumulative stress effects on environmental and human health.

Ad 5 To quantify, characterise and reduce uncertainty in current risk assessmentmethodologies, e.g. by improvement of the scientific basis for setting safety factors

State of the art:In risk assessment, and particularly in cumulative risk assessment, levels of uncertainty are typicallyhigh, mainly due to the complexity of the problems involved and our limited knowledge of theunderlying phenomena (Hellström 1996). This uncertainty often remains obscured by the use ofarbitrary deterministic safety factors and default assumptions. The growing awareness that ignoringuncertainty can result in conservative or erroneous risk estimates (and consequently in a waste ofresources) has resulted in a shift from deterministic towards probabilistic risk assessment (US-EPA1997), also in EU e.g. for pesticides, but as yet in a rather rudimentary and non-integrated fashion.This shift requires novel concepts and techniques to characterise, quantify, reduce and deal withuncertainty in a risk management context. NOMIRACLE aims to develop PRA techniques that arescientifically sound and practicable for cumulative risk assessment.

An area of risk assessment where the role of uncertainty is particularly profound is theapplication of default safety factors to extrapolate laboratory toxicity date to human and ecologicalendpoints. The default factors currently used lack a sound scientific base and ignore uncertainty(Dourson & Stara 1983; Chapman et al. 1998). The processes covered by the default safety factors(e.g., inter-individual and inter-species differences in toxic effects) are rapidly being unraveled bymolecular and genetic studies in (eco)toxicology (Renwick et al. 2000, Wild et al. 2002, Clewell etal. 2002, Snell et al. 2003). It is becoming increasingly clear that toxicological processes aregoverned by a limited number of mechanistic descriptors such as molecular characteristics, geneticpredisposition and the toxic mode of action. The NOMIRACLE project aims to incorporate thesenew scientific insights to derive scientifically sound safety factors for human and ecological riskassessment that explicitly account for the uncertainties involved in the extrapolation process.

Expected improvement:

• Development of new concepts and techniques to characterise, quantify, reduce and deal withuncertainty that are scientifically sound and practicable for cumulative risk assessment

• Improved safety factors for human and ecological risk assessment based on new scientificinsights in the underlying toxicological processes that explicitly account for uncertainty

Ad 6 To develop assessment methods which take into account geographical, ecological, socialand cultural differences in risk concepts and risk perceptions across Europe

State of the art:One of the main problems for the site-specific application of predictive models for exposureassessment is the variability of environmental and land use characteristics. The implementation of avariety of models for exposure and effect prediction in conjunction with Geographical Information

12

Systems (GIS) would allow to handle spatial and temporal variability and to make reliablepredictions at different scales. Most of the model implementation to date has been done at aregional scale using simple index approaches or models that neglect site-specific factors and theirvariability. Within the project, scenarios and models for release, fate and exposure assessment atdifferent scales will be adapted and linked. The NOMIRACLE project will provide guidance tohelp identify what resolution and data are appropriate and when they are needed. The objective is torefine and integrate the techniques for more realistic PEC calculations for different Europeanregions and land uses (FOCUS, 2002).As for effect assessment, a PNEC is calculated from data obtained on simplified trophic chains (e.g.algae, Daphnia and fish) that do not represent natural biological communities. This approach doesnot fulfil the requirements of the Water Framework Directive (WFD) for an assessment of the bio-ecological quality of water bodies. A procedure capable to take into account ecological differencesin aquatic and terrestrial ecosystems could be based on the Species Sensitivity Distribution (SSD)concept (Posthuma et al., 2002) combining this with mechanisms and ecogeography for spatialresolution and realism in assessment. Moreover, the impact on natural populations and communitiesshould take into account resilience and recovery capability but also the possibility for theirbreakdown or decline (e.g. lagged).Finally, in order to provide information suitable to assess the social and economic impacts ofmultiple environmental stresses, there is the need to describe (and, as far as possible, to quantify)the relevance and value of the potentially endangered system (strategic natural resources, protectedareas, etc.)

Expected Improvements:

• The methods developed for assessing exposure and for characterising risk will be integrated in acomprehensive methodology with the development of suitable models and software forassessing location-specific risk, for integrating it into a GIS and for producing(eco)toxicological risk maps.

• The methodology will allow the production of maps of predicted exposures (PECs) andestimates of other validated metrics better reflecting actual exposure, of ecosystemcharacteristics and vulnerability and of (eco)toxicological risk at different scales (from the localscale, i.e. specific terrestrial and aquatic ecosystems, to the regional scale, i.e. different typicalEcoregions). The method should take into account different European environmental and landuse characteristics and, therefore, should be valid across the European Union.

Ad 7 To improve the provisions for the application of the precautionary principle and topromote its operational integration with evidence-based assessment methodologies

State of the art:The debate on how to evaluate and manage risks focuses on three major strategies (Stirling 1999):(a) evidence-based approaches, including numerical thresholds (e.g. NOEL), (b) reduction activitiesderived from the application of the precautionary principle (e.g. ALARA, BACT, containment, orconstant monitoring), and (c) standards derived from discursive processes such as roundtables,deliberative rule-making, mediation or citizen panels. This project will particularly deal withstrategy b). A precautious approach to risk management is required as a paradigmatic change inresponding to environmental and health risks caused by both chemicals and other stressors(Harremoes et al. 2001). With the communication on precaution in the 2000, the EU has taken thelead in precautionary approaches to risk (e.g., Tickner & Raffensberger 2001). However, it is stillunclear how to implement the principle in various cases of chemicals management, and what itsrelationships are with evidence-based risk assessment, both current and novel. This will be aparticular challenge in developing assessment approaches for multi-dimensional and uncertain risks

13

from multiple stressors and to multiple receptors and generally in developing more detailed and yetmore inclusive responses to risk. The precautionary principle essentially involves proactive riskmanagement based on weak evidence e.g. on dose-response-relationships and action mechanismsbut strong indices on either hazard criteria or widespread exposure. The typical characteristics thattrigger precautionary actions, ubiquity, persistence, bioaccumulation, irreversibility of effect andintensive psychosomatic impacts (Renn and Klinke 2002), have not been systematically assessed orintegrated in a predictable and consistent framework of risk management. Applying the principleacross-the-board implies the danger of impeding necessary innovations since there is always achance for unforeseen negative impacts. Suggestions for making prudent use of the precautionaryprinciple include investigating the magnitudes and qualities of risks and associating uncertainties(Finkel 1995) and exploring risk reduction opportunities and their consequences. A precautionaryapproach also allows for the inclusion of social, psychological and cultural aspects of riskexperience. If society responses to risk situations are included in the analysis, health-related,environmental and socio-cultural criteria can be combined for the risk evaluation process (Stirling1999; Klinke and Renn 2001). This is pointed out by the analyses of the impacts of REACH (DGEnterprise 2003a,b,c; JRC 2003a,b; RPA 2003) but needs to be studied for specific classes ofchemicals as well (e.g. Römbke et al. 2001).

Expected improvements:

• The project will improve the knowledge base and methodologies for efficient implementation ofthe precautionary principle in managing risks from chemicals and other stressors through multi-disciplinary studies of the key cognitive, knowledge-related and social issues in risk assessment.

• These studies will elucidate ways to integrate the precautionary principle with detailed scientificrisk assessments, depending on the decision situation (e.g. the chemical product, receptor andregion). The work will focus on the use of scientific information in integrated assessment toprovide policy-relevant advice and on related processes of inference and deliberation.

• This R&D is expected to have significant value for the development and implementation ofintegrated risk assessment and for risk management, be it predominantly science-based orprecautionary, in a variety of contexts, primarily in the project domains but also more generally.

• The R&D in this area will, by elucidating risk views and knowledge and inference inassessment, also serve to integrate the project.

B.2 Relevance to the objectives of the Sub-Priority“Global Change and Ecosystems”

The programme of activity offered by the Sub-Priority “Global Change and Ecosystems” willstrengthen the necessary scientific knowledge for the future orientation of the ST strategy and the6th Environmental Action programme; it will also provide the socio-economic tools and assessmentsand the overall management practises. Furthermore it will ensure their implementation at theenlarged EU level and, when relevant, at the world level. Risk assessment of chemicals is a centralissue within this framework. Despite the implementation of a series of measures, chemicals stillcause pressure on natural environmental resources and environmental health. Further, chemicals area major factor in a general disaffection with technology and science that potentially threatenseconomic and social development in both the Member and Accession Countries, and at the worldlevel. The NOMIRACLE Consortium will conduct a comprehensive and coherent researchprogramme providing the scientific basis for a coming shift in paradigm for risk assessment of

14

chemicals taking into account realistic exposure and the cumulative risk following exposure toseveral stressors.

Contribution to the general objectives of area VII “Complementary research”NOMIRACLE will focus on the development of advanced methodologies for risk assessment ofchemicals aiming at integration of methods for assessment of environmental and human health. Theproject will develop new tools for risk assessment aiming at a general improvement of theenvironmental quality in Europe. A part of the activities will be pre-normative research onimproving the current state of measurements and testing, e.g. for exposure assessment, for testing ofcombined stressors, and for indicators for monitoring of populations of humans (indoor andoutdoor) and species in the environment.

Contribution to the topic VII.1 “Development of advanced methodologies for risk assessment”NOMIRACLE will contribute to the overall aim to strengthen and advance risk assessmentknowledge and practises with particular emphasis on cumulative risk assessment, taking intoaccount recent trends in science. The Consortium combines 38 partners from 17 countries,including 11 EU and 4 NAS countries as well as Switzerland. The topic VII.1 will be addressed in 4Research Pillars on respectively “Risk Scenarios”, “Sound Exposure”, “Effect Assessment”, and“Risk Assessment”. All pillars integrate human health and environmental quality, and knowledgefor integrated risk assessment will effectively be transferred between the experts of the pillars. Forthe assessment of human health both indoor and outdoor exposure will be dealt with. The researchwill produce the base for establishment of an integrated risk assessment scheme based ongeographic information that takes into account the diverse geographical, ecological, social andcultural differences in Europe. By integrating socio-economic studies, NOMIRACLE will analysethe use of the precautionary principle in relation to the findings, and address the question of riskcommunication with relation to risk assessment practise.

Contribution to topic VII.1.a “Development of risk assessment methodologies”NOMIRACLE will focus on a conceptual change in risk assessment with emphasis of assessment ofthe effects of combined exposures by developing new models and test systems, integrated forenvironmental and human health, however based on existing methods and models wheneverpossible. The stressors will include chemical mixtures, pathogens, climatic stressors, and otherenvironmental stressors such as anoxia and acidification. New developments are needed for realisticexposure and for measurement of combined exposures. The project will address the aim “to developmethodologies for assessing the risk of substances and molecules designed for provoking specificinteractions with biological structures”. The work will include pesticides, biocides andpharmaceuticals, and seek to provide for a future scientific basis for general harmonisation ofprotocols for all types of chemicals. NOMIRACLE will cover the aquatic and terrestrialenvironment (but not marine environments) and deal with exposure through water and air, includingindoor exposure for humans. In dealing with chemical mixtures, the development of methodologiesfor assessing the risk of exposure to chemicals in products will have a high priority.

B.3 Potential Impact

In Europe, health effects are related to environmental factors, such as respiratory diseases, asthmaand allergies that are associated with pollution. Additionally, environmental pollution continues tostress ecosystems in spite of existing regulations. Stress on human health and ecosystems isincreasingly related to complex chemical mixtures from multiple, dispersed sources.NOMIRACLE will mobilise expertise and resources on a sufficient scale to give Europe a leadingscientific position in relation to addressing such issues.

Working with the JRC and other Directorate Generals, NOMIRACLE will provide inputs tothe European Environmental and Health Strategy (the SCALE initiative) recently launched by the

15

Commission. NOMIRACLE will help fill the knowledge gap on the link between environment andhealth, in a first phase focusing on molecules designed for provoking specific interactions withbiological structures.

This IP will provide increased knowledge and improvements to existing tools in relation tothe transfer of pollutants between different environmental compartments and on the impact ofcumulative stressors, including combinations of chemicals as well as different physical stressors.In particular, while fitting into the sustainable development framework of European policy, thefindings of NOMIRACLE will contribute to improved methodologies in support of future revisionsof, for example, the Plant Protection Directive (91/414/EEC), the Biocide Directive (98/8/EEC), thedirective for pharmaceuticals that is underway, as well as revisions to the Sewage sludge disposalDirective.

In direct collaboration with partners from the Commission's JRC, NOMIRACLE will helpsupport the Commission's thematic Strategy on the Sustainable Use of Pesticides, Strategy for SoilProtection, and Strategy for Waste Reduction and Recycling, by providing novel insights related tothe fate of pollutants, species exposure, and cumulative effects attributable to multiple stressors.

The NOMIRACLE Consortium is highly competent in relevant areas, counting leadingscientists within human toxicology and epidemiology, aquatic and terrestrial ecotoxicology,chemistry, biochemistry, toxicogenomics, physics, mathematical modelling, geographicinformatics, and socio-economic science, as well as in the broader context of life cycle assessmentand risk assessment.

To ensure efficient dissemination of the results and worldwide co-ordination, a close co-operation will be established with an independent Advisory Board of Stakeholders from, forexample, government, industry, and non-governmental organisations. These stakeholders are keyplayers in the development of improving risk assessment approaches for environmental and humanhealth protection.

The Advisory Group and all interested stakeholders will be invited to the events planned bythe research Consortium such as workshops, seminars and conferences. NOMIRACLE willestablish a homepage, and a dissemination plan will be prepared to commit the partners todisseminate the results of NOMIRACLE to national, local, and regional authorities, the public,industry, academia (scientific publications), as well as international and non-governmentalorganisations. Consortium partners within the Commission's JRC are expected to play a vital role inthis context, in addition to interactively developing methods in support of EU-level policy.

NOMIRACLE will build on the current state of knowledge as obtained by preceding EUprojects such as MIXTOX (ENV4-CT97-0507) and will co-ordinate the work on pharmaceuticalswith the consortium ERAPharm, dedicated to the work programme of topic VII.1.1b of this FP6subprogramme. The project will be co-ordinated with the FP6 IP ALARM (Assessing Large-scaleenvironmental Risks with tested Methods), sharing geographical and other data, which can find usein both projects.

NOMIRACLE will enhance the scientific and applied competitiveness of all partnersinvolved. In particular, knowledge will be transferred between research institutions and SMEs, andbetween experts in the domains of human health and environmental quality. Partners from differentregions of Europe will be able to directly share in improving the state of the art as produced in theproject. The Consortium will employ at least 40 PhD students, who will receive supervision andtraining from leading scientists within the scope of NOMIRACLE and from collaborations amongstthe partners.

Socio-economic impact assessmentThe socio-economic impact assessment issue is approached from two complementary angles:• The impacts of the project on socio-economic processes and factors• The studies and assessments of socio-economic impacts of chemicals policy and related policies

and procedures

16

The socio-economic impacts of the project itself are manifold, corresponding to the variety of thefunctions of dissemination and exploitation (cf. the plan on these). Some impacts will be realizedimmediately, e.g. through the empirical and theoretical methods produced, the assessment toolsdeveloped, and the dialogues initiated about assessment policies, while other results will have amore gradual and indirect impact on society that still may be of paramount long-term importance.This is true especially of the scientific results providing improved understanding and concepts.

The better grasp of how to integrate precautionary approaches with science-based assessment andmanagement, expected to emerge from the project e.g. through the studies in the policy aspects ofrisk assessment, will be important from both economic points of view, e.g. by clarifying thenecessary but not excessive requirements for assessment and reasonable conditions for innovation,and from a policy and social points of view by elucidating the processes and factors in building trustin the chemical risk management area.

For the important issue of testing strategies, including the objective of reducing animal testing, theproject has, generally speaking, a two-fold contribution. First, new approaches to and methods fortoxicity testing are expected to be identified and developed, capturing more extensively, efficientlyand accurately crucial effects of the chemicals and stressors studied when in vivo testing is needed,as is the case for many endpoints that are difficult to predict or ascertain based only on theory,QSARs and in vitro tests. Secondly, the integrative uncertainty analyses including consideration ofthe information needs in a risk management decision context will aid to define the levels and kindsof toxicity test data needed; some of it may e.g. be diminished when a more precautionary approachcan be shown to be feasible and justified. Both within the project itself and in a broader evaluationbased on its results, improved testing strategies will thus be devised that fulfil both the ethicalrequirements (cf. B.9) and the need for realistic information for risk assessment and management.

The regional dimension in risks and in assessing and managing them are among the key socio-economic aspects. This entails both the consideration of the natural (including human demographic)variations between regions of Europe, as well as the policy-related issues and implications of howrisks from chemicals and accompanying stressors are perceived and responded to across Europe andhow this interacts with cultural differences. The project will make contributions in both areas, in thefirst by a significantly improved spatial resolution, analysis and presentation of risks thatincorporates the variability in environmental conditions and processes, and in the second by studiesof risk views particularly in case regions and of the general policy issues within integration in riskassessment across geographic scales and across related levels or regimes of administration.

Socio-economic aspects of chemicals policies will be studied in relation to risk perception, riskknowledge, policy links of assessment, and risk communication. However, within this project it isneither possible nor meaningful to attempt e.g. a follow-up or expansion of the analyses by theCommission (DG-Enterprise 2003a,b,c, DG-Environment 2003) and elsewhere (JRC 2003a,b, RPA2003) of the socio-economic consequences of REACH. Instead, the studies in this area will targetissues directly related to risk assessment and to the production and use of scientific information,particularly in the areas of integration (multiple stressors, receptors and regions) and for the classesof stressors subject to closest study. The project will herein address the socio-economic dimensionby a decision, policy and communication analytical approach emphasising risk views anduncertainty management instead of e.g. risk-benefit economics, apart from value-of-informationanalyses.

All in all, it is expected that the project will have a significant impact on the socio-economic area,and will contribute to a comprehensive, balanced and broadly acceptable approach to risks that isprotective and supportive of both environmental quality and human health as well as social andeconomic values in a long-term sustainable perspective. Finally, it should be pointed out that an

17

important contribution to this goal will be made already by identifying and explicating the issues,conditions and constraints, including controversies and impediments, in science and technology andin other areas that are relevant for its fulfilment, even if not attempting to solve them, as this is thedomain of policy and decision makers, not researchers (even decision or policy scientists).

B.3.1 Contributions to standards/policies/regulations

As indicated in the description of the overall strategic and socio-economic impacts and in moredetail in the work descriptions, the project will offer many contributions, direct and indirect, to thedevelopment and implementation of EU policies and regulations within its scientific scope. It willhelp resolve how best to refine the risk assessment procedures based on the present and alreadyproposed regulations (especially under REACH but also other relevant areas). Still moreimportantly, the project will provide knowledge and methods for assessment of agents, cofactorsand risks that are intractable by present methods; this is expected to be crucial for the futuredynamic development of EU risk management policies and regulations. In connection with this, thelinkages between policy and these novel needs, factors and concepts in risk assessment will beilluminated.

In terms of collaboration and co-ordination between branches of administration (both at theCommunity level and at national levels), the integrated treatment of ecosystem and human healthwill make an important contribution; in addition, in addressing the context and conduct of riskassessment, also other branches, particularly that within enterprise, will be included, due also to thedeveloping role of industry in risk assessment. As to chemicals regulations, the focus on mixtureeffects assessment will by definition help unify the presently separate areas of policies andregulations (such as between different categories of chemicals) especially in areas of greatest gapsand needs.

No unequivocal and detailed standards can be produced for risk assessment, due to the inherentcomplexity and multi-dimensionality of risks and to the dependence of assessment on its contextand purpose. Likewise, harmonisation of risk assessment is only possible to a limited degree, toretain and promote the attention of variation e.g. based on particular sector, regional or otherreasons. However, at a more general level, commonly applicable frameworks can and will beproduced, and integrated and harmonised information useful for the varied assessments will beprovided. This will serve a sensible level of harmonisation that allows e.g. both common EUpolicies and the necessary subsidiarity in identifying and responding to risks.

18

B.4 Outline implementation plan

Description of science and technology approach and ways to achieve theNOMIRACLE objectives - General approach

The risk assessment models and methods for chemicals currently in use within the European Union(e.g. European Communities, 2003; Annex VI of Directive 91/414/EEC) are based on a series ofdefault assumptions and deterministic assessment factors which tend to result in relativelyconservative predictions of risk. It is a major challenge to introduce more detail into these riskassessment procedures so that better informed management decisions become possible. Dealingwith mixture toxicity and multiple stress is one of the areas where risk assessment of chemicals canand should be considerably improved. Other important areas include spatial differentiation,temporal variation, inter-individual variability (in exposure, toxicokinetics and dynamics),uncertainty and theoretical issues such as the quantification of critical stress levels in situationswhere cumulative stressors act in combination.

The research of NOMIRACLE will in particular focus on pesticides, biocides andpharmaceuticals and their effects in conjunction with other important natural or anthropogenicstressors such as industrial chemicals. The assessment scheme that will be developed will beequally applicable to each of these types of substances. The scheme for industrial chemicals willdiffer, however also the applicability in assessment of industrial chemicals of the new methods to bedeveloped will be considered. Realistic exposure scenarios will be evaluated with respect tochemical mixtures and cumulative stressor combinations likely to occur.

It is clear that the number of combinations of stressors is practically infinite, and therefore it isnot possible to cover this to the full extent. However, in this project a shortcut is suggested wherefocus is put on the most frequent chemical mixtures present in the regions of Europe, incombination with other major stresses such as extreme climate, acidification, eutrophication,particles etc. By this exercise it is ensured that relevance and realism is given high priority.

The project will develop methods focusing on the effects of long-term exposure, and willinvestigate the possibility to make regional risk assessment based on “stress maps” including thesemajor background environmental factors acting on humans and the environment. In the end, e.g.regional safety factors will be assessed for individual chemicals based on sound exposure takinginto account the real world exposure.

19

Research Pillar

Pillar 1: Risk

scenarios

Pillar 2: Sound

exposure

Pillar 3:Effect

assessment

Pillar 4:Risk

assessment

WP 1.1Data

backgroundfor scenario

selection

WP 2.1Matrix-

compoundinteraction

WP 1.2Scenarioselection

and ranking

WP 2.2Availableexposure

WP 2.3Metabolic fate

WP 2.4Regionspecific

environmentalfate

WP 3.1Interactive toxicology in diversebiological systems

WP 4.1New concepts for probabilistic risk assessment

WP 4.2Explicit modellingof exposure and

risk in spaceand time

WP 4.4Risk

presentationand

visualisation

WP 3.2Combined effects

of naturalstressors and

chemicals

WP 3.3Toxicokinetic

modelling

WP 3.4Molecular

mechanismsof mixture

toxicity

WP 5.3 Training and Demonstration

WP 5.2 Data Management

WP 5.4 Dissemination and ExploitationWP

5.1

G

ener

al P

roje

ct M

anag

emen

t

WP 4.3Dealing withmultiple and

complex risks

Pill

ar 5

: M

anag

emen

t

Figure B.4-1: NOMIRACLE activities and their components (Pert diagram).

The NOMIRACLE approach is summarised in Figure B.4-1 and consists of 4 main ResearchPillars (RP) each containing a number of Work Packages (WP). Briefly, in RP 1 the most importantand relevant stressors occurring across Europe and EU accession countries will be identified inorder to find the most relevant scenarios of cumulative stressors to be studied. These “potential riskscenarios” will feed into RP 2, RP 3 and RP 4, and dictate which particular combinations ofstressors and risk scenarios will be studied in detail.

RP 2 will address the European-wide fate of environmental contaminants and its dependenceon compound properties as well as external factors such as emission sources, climate parameters,their matrix-specific availability, and in particular the human-specific exposure pattern.

The goal of RP 3 is to generate data to be used in development of generic rules for theassessment of combined exposure effects that are underpinned by mechanistic understanding. RP 3will investigate interactive effects in species ranging from plants to bacteria to vertebrates to fish tohumans. Comparative analysis will be used to identify the generic physiological changes that dictatethe consequences of exposure to toxicant mixtures and combinations of chemicals withenvironmental/pathological stresses identified as important during RP 1.

The main aim of RP 4 is to develop novel methods for integrated risk assessment that enablethe optimum use of available information in the decision-making processes, thus ensuring anefficient use of valuable resources. This aim will be realised by integration of the results ofResearch Pillars 1, 2 and 3 within a probabilistic and spatially explicit modelling framework.

The crosscutting management pillar (Pillar 5) will secure the overall co-ordination of theproject via the work of the projects secretariat and a Management Board (described in section B6).A sound data management strategy will be developed in WP 5.2 securing that data generated during

20

the project will be accessible to all partners in pre-defined formats to optimise the use of data.Training of personnel, students and end-users will be ongoing during the whole project period inorder to have the adequate skills available for the project. Dissemination of the results of the projectwill have high priority in NOMIRACLE, and a special WP 5.4 is dedicated to this task withfeedback between end-users, stake holders and the relevant WPs.

NOMIRACLE is a 5-year integrated project where each RP will be running with varyingactivity levels during the whole project period. RP 1 is a comparatively small component (in termsof budget) of the project, and the main deliverables of RP 1 will be finalised during the first 2 yearsof the project. RP 2 and RP 3 contain the largest body of work, and will run for most of the projectperiod. The main work of RP4 will be carried out in the last three years of the project, but withsome activities running during the whole period. A detailed overview of the manpower allocated toeach component is described in section B.7.

B.4.1 Research, technological development and innovation activities

In this section the components of the research and innovation activities are described including themanagement, training and dissemination activities. To visualise the work effort allocated to eachcomponent, tables for each WP are showing who is undertaking the work (shown as the acronym ofthe institution), and how many manmonths are allocated to each activity. In this section, manmonthsduring the whole project period are indicated.

Research Pillar 1 - Risk Scenarios (Sub-coordinator: Hanne Bach, NERI)

WP 1.1Acronym ALTERRA APINI DESUN JRC NERI UFZ UNIMIBPersonmonths 10 4 0.5 16 23 6.5 9

WP 1.2Acronym NERC DESUN ENVI JRC NERI SYKE UNIMIB URVPersonmonths 6 2 8 2 24 6 5 4

Aims and rationaleThe aim of this pillar is to develop new methods, which can guarantee the best possible selectionand design of risk assessment scenarios, reflecting the complexity of the entire risk assessmentmethodology. The scenario selection is a highly important task for the fundamental set up andfocuses of developing integrated risk assessment methods. The interactions between mixtures ofchemicals in the environment and combinations of stress and fate conditions for understanding eco-toxicity produce a nearly infinite set of possible risk assessment scenarios. It is neither feasible norrelevant to describe the effects of every scenario and a procedure is therefore needed to find the bestscenarios. Uncertainty in the scenario selection procedure carries the risk for identification of sub-optimal scenarios as false optimal for further investigations. An uncertainty analysis focusing onreal risk estimates for humans and environment needs to face this problem as it will otherwiseinduce false confidence by having an incomplete uncertainty picture. The development ofprocedures for scenario selection problem will therefore be taken into account as a part of thescientific challenge of improving the risk assessment methodology.

RP 1 will develop a procedure, which can support the activities in the other research pillarsby identifying a set of the most relevant scenarios with highest risk potential, based upon priorknowledge. Data collection and assessment will take place in WP 1.1. This involves release ratesand position for the chemical substances of concern and essential environmental and human factorsfor fate toxicity and other relevant stressors including assessment of data quality and uncertainty.The objective in WP 1.2 is to develop a multi-criteria procedure, which can identify a set ofscenarios having the highest risk potential taking into account cumulative assessment of mixture

21

toxicity. The methodological development includes uncertainty analysis of the scenario ranking.This task will apply a series of empirical data mainly from field scale and risk assessment results inconjunction with the other research pillars.

WP 1.1 Establishment of data background for scenario selection (Leader: Arwyn Jones, JRC)The purpose of this WP is to develop the data background including uncertainty assessment for thefollowing activities: (a) A basis for scenario selection and related uncertainty estimates in WP 1.2.(b) Input to the exposure assessment in WP 2.4 (in form of emission inventories, environmentalcharacterisation and non-chemical stressors). (c) Supporting the uncertainty assessment in ResearchPillar 4 (d) Supporting the risk mapping and other means of risk visualisation in RP 4. This includesthe following types of data: (1) Substance related data relevant for integrating exposure and effectsof human and ecological character. (2) Product related data of a substance’s composition andrelease rate including links to human activity. (3) Climatic, physical and ecological conditions (e.g.temperature, precipitation and eutrophication) relevant for characterising stressors for cumulativerisk assessment. (4) Stressors for cumulative risk assessment linked to human activity (e.g.wastewater, land use and ozone). (5) Landscape classification and land use. (6) Presence andvulnerability of receptors.

Substance related data includes physical chemical properties including measures forpersistence in different environmental compartments, toxicological and eco-toxicological propertiesand information about existing toxicological and eco-toxicological hazard listings. JRC/ECB willhelp to identify appropriate test substances for the NOMIRACLE project by providing physico-chemical, toxicological, ecotoxicological and other relevant data by using the ECB databases, e.g.IU-CLID.

Product related data including emission will be established for both indoor and outdoorconditions. Socio-economic and human population specific (sensitivity of sub-groups) factors needto be registered as conditions that can change the outcomes of exposure. Exposure and release ratesof chemicals in products will be included by combining emission factors and specific content ofchemicals in products.

Maps on physio-geograpical settings, topography and climatic databases are available at EUas well as regional and local level. A series of very relevant characterizing parameters can begathered from these data sets, reflecting environmental properties of aquatic and terrestrialecosystems.

Stressors for cumulative risk assessment linked to human activity covers broad spectra ofdifferent data sources at different levels of spatial aggregation (site, suburb, county, country,watershed, etc.). Demographic data such as population density, social structure and housing will begathered. These data are important both in relation to potential stressors and as input to emissionsestimates. Air pollution originating from energy consumption of heating and transportation andother human activities are important stressors. Transportation in terms of intensity and structure willalso be important as input for stressors both for human health (e.g. noise) and ecological health (e.g.fragmentation of habitats). Induced oxygen deficit in surface and marine waters, as a consequenceof wastewater and eutrophication are other examples of relevant stressors linked to human activity.This also includes critical aspects of the ecosystem sensitivity related to different types of habitats,especially regarding Natura 2000 sites. This procedure is in accordance with a recently developedapproach for site-specific ecological risk assessment applied in the Netherlands (Rutgers et al.2001). The approach has been submitted for incorporation and further development of astandardised pan-European procedure for ERA, within the framework of the Concerted ActionCLARINET (Crommentuijn et al. 2001).

Aspects of land use and designated areas will also be developed for risk assessment. Forexample areas for the extraction of drinking water, nature conservation areas and agricultural areas,or even urban areas ideally should all be assessed differently in terms of risk assessment because theessential parts or functions of the ecosystem to fulfil the specific goals for land use are different.

22

Presence and vulnerability of receptors is important in terms of demographic data onpopulation density, social structure and vulnerable nature areas, which are important in relation toexposure of relevant human and ecological receptors. Socio-economic and human populationspecific (sensitivity of sub-groups) factors need to be registered as conditions may influence theeffect of exposure.

The uncertainty inherent in the data background will be investigated and quantified so muchas possible in order to secure valid conditions for the uncertainty analysis in WP 1.2.

WP 1.2 Scenario selection and ranking (Leader: Peter B. Sørensen, NERI)The objectives of this WP is to develop the multi-criteria procedure needed to identify scenarioswith the highest risk potential taking into account new concepts of handling mixture toxicity andcumulative exposure and effect assessment, which is primarily based on data from WP 1.1. Ascenario in this context, is a set up of conditions for more detailed investigations in terms of eitherexperimental work or modelling. This includes elements as (1) Emission patterns; (2) Compositionof mixtures for mixture toxicity assessment; (3) Stress factors for cumulative risk assessment and(4) Characteristics and vulnerability of the biological community and ecosystem. A scenario may belocated physically as a position in GIS or just reflect realistic conditions without assigning to ageographical reference.

This WP will yield input for the activity in RP 2 and RP 3 and the risk visualisation in WP4.4 and the uncertainty analysis in general in RP 4. A set of scenarios will be selected as possessingthe highest risk potential for cumulative assessment of mixture toxicity. The work will beundertaken under the condition of high complexity and limited knowledge, so the selection willidentify a “best choice” set of scenarios including uncertainty analysis of the selection procedure.The uncertainty analysis will attempt to predict the certainty for which one can claim that the “inreality most relevant scenario” is included in the selected set of best choice scenarios. Theuncertainty estimates will be used in the uncertainty evaluation in RP 4 and also as an instrument inthe refinement of the scenario selection procedure in this WP. A spin off will be a method toidentify substances of potential interest for risk assessment but where further testing is inhibited bya low analytical-chemical capability for exposure measurements. This can set up future guidelinesfor development of analytical measuring methods in general to support risk assessment.

The criterion setting is the link between the data background from WP 1.1 and the selectionprocedure. The input to the selection procedure is composed of many criteria because many aspectsneed to be included simultaneously in the selection of scenarios especially for cumulative riskassessment of mixture toxicity. The criteria can be considered as indicators for potential risk morethan input for exact risk assessment procedures. They are reflecting the data background directly ina one to one relationship or they are formed by interpretation of data as done by Roelf et al. (2003).Some of the criteria will be useful for assisting later assessment initiatives in the project in otherresearch pillars. Any criteria will be related either to a chemical substance or to a geographicalposition forming two different decision matrixes as shown in the following Fig. B.4-2:

Figure B.4-2: Criteria for selection of scenarios will be related either to a chemical substance or toa geographical position forming two different decision matrixes.

Decision matrix: substances

substances Criteriasubstance 1substance 2---

Decision matrix: sites

Areas CriteriaArea 1Area 2---

Linked byinformation

aboutproducts

23

The decision matrix for the substances lines up all relevant criteria for toxicity, fate, relationship toproducts and human or ecological receptors. In the decision matrix for the sites every area isdescribed by a set of criteria including product usage, climate, physical and ecological conditionsand human activities. The products will in most cases be linked to human activity parameters, asinformation about product consumption seldom can be associated with specific locations (sites).

The methodological approach in this work package will include a variety of differentconcepts. Many criteria have to be used in the selection procedure as outlined above. However, theinteractions between these criteria are not well known at this stage. Several methods for screeningof chemicals in relation to human and/or environmental health have been undertaken (e.g. EU,2001, EU, 1999, Hansen et. al, 1999 and Swanson et. al, 1997). The rationales behind thesemethods will be included as support for the method development of a more integrated selectionprocedure including mixture toxicity and other stressors and based on a more solid theoreticalbackground of multi-criteria methodologies. The number of criteria can be reduced when they aretheoretically related to each other by using aggregation models or the existing screeningmethodologies mentioned. Examples of aggregation models are Utility Functions, Fuzzy Logicmodels or models based on the compartment paradigm as suggested by Pennington (2001) andGouin (2000).

The reduced sub set of criteria will subsequently be analysed using multi-criteria methods.The Partial Order Theory is relevant for this problem and has been utilised and developed forranking chemical substances in relation to potential risk (e.g. Brüggemann and Bartel 1999; Lercheet al. 2002). This is a very general and flexible method, which facilitates the handling ofuncertainties due to the simultaneous use of multiple criteria without having a full systemicknowledge about their interaction. This type of uncertainty can be quantified via a statistictreatment of the ambiguities in the ranking order between different criteria (Lerche and Sørensen,2003). In this way, it is possible to form Bayesian type ranking estimates yielding a probabilityspace for possible rankings. Other multi-criteria methods will also be reviewed like e.g. the specificPromethee Method (Brans et al. 1986) and the very general methodology “Rough Sets Theory”(Greco et al. 2001).

Several methods exist for quantification of modelling uncertainty as a consequence of inputparameter uncertainty. Several of these methods, mainly based on the Monte Carlo type principle,will be adopted in this project in order to analyse the uncertainty in the scenario selection proceduredue to uncertainty of the criteria value setting. A challenge using these methods is to quantify inputvalue variability and possible inter-correlation between different input parameters. Thus theuncertainty estimates in relation to the input parameter uncertainty will be supported andaccomplished using expert knowledge and theoretical constrains.

A selection procedure based on multiple criteria will always be based on some assumeddegree of importance between the single criterion in order to come up with a fixed selection result.Thus, the selection will not be completely firm if the relative importance between the criteria is notwell established. A solid quantification of the uncertainty due to uncertain criteria weightings ispossible when the predicted selection can be validated against real data. This can only be done in acomplete manner for relatively simple models and not for the entire multi-criterion proceduredeveloped in this WP. Sub parts of the selection procedure, however, can be validated usingmonitoring data sets. This type of validation has been done for pesticides by Kolbin et al. (1998),Krüger et al.(1998), Sørensen et al., (2003a), Sørensen et al., (2003b) and Worral and Thomsen(2004) and for dioxins by Margni et al. (2003). All these validation activities have been undertakenin relation to exposure and not by validation of toxicity. However, the multi-criterion approach inthis WP will open for new possibilities for accomplishment of correlation analysis, in whereecological monitoring data can be related to fate and emission criteria, simultaneously. This type ofanalysis will be conclusive not only for this work package but also throughout the project and thusforming a novel link to field scale data.

When a set of scenarios is selected in this work package it will enter as input to the otherresearch pillars and thus be subject to a much more detailed analysis. This will yield an excellent

24

opportunity for reconsideration of the conditions under which the scenario was selected. If a factoron one hand is shown to be important in the detailed risk assessment but on the other hand wasignored as a criterion in the scenario selection procedure then the scenario selection can be claimedto be associated with uncertainty due to ignorance. There are therefore good reasons for evaluationof the scenario selection procedure during progress of the project as detailed investigation resultsarrives from the other research pillars.

Research Pillar 2 - Sound Exposure(Sub-coordinator: Gerrit Schüürmann, UFZ, Germany)

WP 2.1Acronym UFZ EAWAGPersonmonths 87 36

WP 2.2Acronym NERI CSIC ITM JRC UFZ ULANCPersonmonths 43 65 48 6 42 36

WP 2.3Acronym LMC ECT LHASA UFZ URVPersonmonths 72 66 40 88 16

WP 2.4Acronym DESUN ITM JRC RIVM URVPersonmonths 48 48 37 4 38

Aims and RationaleThis research pillar addresses the processes and mechanisms that determine the effective exposureof chemical substances under realistic conditions in the environment.

The major deliverable is a suite of methods that allows quantification the multiple exposureof environmental chemicals and its matrix-specific fraction available for interaction with biota. Thisconcerns both the actual exposure in biotest systems that include sorption-relevant compartmentssuch as soil, sediment and cell material, and the effective exposure in waters, soils and sedimentsunder field conditions. In particular, both region-specific background contamination and site-specific hot spots will be accounted for in a multimedia-level approach, considering realisticoutdoor and indoor conditions in cooperation with RP 1. The envisaged methodology will offer atiered approach, providing screening-level estimates based on only little information, and a strategyto identify major information gaps and resultant management actions in order to arrive at increa-singly sophisticated characterisations of the effective exposure.

The combination with information about the hazardous effects of non-chemical and che-mical stressors as derived in RP 3 will allow a sound assessment of the resultant cumulative risk,which is the subject of RP 4. Major features of the overall scheme are the characterisation of theeffective exposure under in situ conditions, the explicit treatment of uncertainty, spatial andtemporal variability, and the combined assessment of the aggregate exposure of a human or animalcommunity towards multiple stressors via their relevant routes, pathways and sources.

State of the Art and Major Current ShortcomingsThe current European regulatory standard for evaluating the environmental fate of xenobiotics is themultimedia model EUSES (EC 1996, van de Meent 1993, den Hollander et al. 2003). WhilstEUSES addresses all major uptake pathways including bioaccumulation, its modelling approach

25

leans heavily on the assumption that intermedia partitioning is described well by hydrophobicitywith octanol as surrogate for organic compartments, which is not adequate for more polarcompounds. Moreover, EUSES lacks a functionality to relate total compound concentrations tomatrix-specific effective concentrations that are governing the resultant risk potential, and propermodels for predicting abiotic and biotic degradation rates as well as prevalent metabolites frommolecular structure. Also, the employment of generic compartments without a spatial resolution ofclimate and media-specific parameters as well as the inability to address temporal variations put se-rious limitations to the current EU risk assessment of toxic substances.

A further shortcoming concerns current strategies to derive exposure-concentration profilesof xenobiotics in biotest systems that involve soil or sediment compartments. In such systems,sorption to abiotic compartments competes with uptake into the test organism, such that theeffective exposure depends heavily on matrix and milieu parameters. Accordingly, the derivation ofsound exposure-effect relationships and resultant toxic potency parameters such as EC50 values ishampered by the lack of knowledge of the actual exposure under in situ conditions.

In order to overcome these shortcomings, a concerted research activity across different pro-cesses and scales is required. First, a proper methodology needs to be developed to parameterise thematrix-compound interaction for more complex chemical structures of emerging compound classessuch as biocides, pesticides and pharmaceuticals. Secondly, a new approach is needed to determinethe effective concentrations of environmental compounds under in situ conditions that govern theextent of an hazardous impact on biota. This includes the need to establish a sound link betweenoutdoor exposure and exposure in laboratory systems, and to address the specific features of indoorexposure that is of primary importance for humans. Third, realistic degradation rates are needed toevaluate field half-lives of contaminants, and knowledge of prevalent metabolites is required toperform a comprehensive evaluation of the exposure profile associated with xenobiotics. Fourth, amultimedia fate model addressing temporal and spatial variation is required that includes mecha-nistically sound modules for the compound-matrix interactions, and allows a region-specific charac-terisation of the European background exposure based on emission data and use pattern scenarios.Accordingly, the research pillar contains the following four work packages:

WP 2.1 Matrix-compound interaction (Leader: Gerrit Schüürmann, UFZ)At present, the equilibrium partitioning of xenobiotics between fluid and condensed phases is oftendescribed through relationships employing the octanol/water partition coefficient (Kow). For morecomplex chemical structures such as modern biologically active agents, polar interactions betweenhydrogen-bond donor and acceptor sites become important, which however can not be capturedproperly with the simple octanol phase as surrogate for organic compartments. A solution to thisdilemma is given by employing the Abraham descriptors developed for linear free energy solvationrelationships (Abraham 1993, Goss & Schwarzenbach 2001) that offer a mechanistically sound wayto model the equilibrium partitioning between chemical compounds and abiotic or biotic matrixes.

The goal is to develop methods to predict the phase partitioning of organic compounds fromphase-specific parameters and molecular descriptors. To this end, a two-fold approach combiningexperimental and theoretical work is envisaged: Firstly, model systems will be built to measurecompound partitioning between environmentally relevant phases (water, soil, sediment; EAWAG).Here, the method of choice is inverse gas chromatography (IGC), provided that the equilibrationtime needed for thermodynamic partitioning is not too long. If the latter condition can not befulfilled, an alternative option are standard batch equilibrium experiments combined with a head-space analysis. Secondly, semiempirical quantum chemical methods will be used to develop localreactivity descriptors based on perturbational molecular orbital (MO) theory, exploiting both am-plitudes (wavefunction contributions) and energies of MO wavefunctions constructed as linearcombinations of atomic orbital (LCAO) wavefunctions (Karelson et al. 1996, Schüürmann 1998 &2004; UFZ). Here, particular emphasis will be devoted to encode H-bond donor and H-bondacceptor strengths, since so far no context-dependent MO-based H-bonding descriptors are

26

available. In addition, the recently developed electropological hydrogen bonding approach (Kier &Hall 1999) will be explored for its potential suitability to yield molecular descriptors that mimichydrogen bonding features (UFZ).

Like with abiotic compartments, a sound description of the compound affinity for biologicalmembranes requires particular attention when dealing with multifunctional chemical structures,since biological membranes contain a variety of lipid proteins and carbohydrates, withphospholipids constituting major structural components. Due to their amphiphilic character, theydetermine the essential physical functionality of the membrane, especially the permeability and thusthe exchange of material between cell and the surrounding. According to recent findings, themembrane-water partition coefficient (Kmw) offers a promising alternative to the Kow to bettermimic the biopartitioning of organic compounds, particularly when ionisation plays a role (Escheret al. 2002). Possible methods to quantify Kmw include the determination of retention capacityfactors on a immobilized artificial membrane (IAM) stationary phase (Lundahl & Beigi 1997,Taillardat-Bertschinger et al. 2002), and batch equilibration experiments with small unilamellarphospholipid vesicles (e.g phosphatidyl choline, PC), where the partitioning of a substance betweenthe water phase and liposomes can be detected by its depletion in the batch solution. Here, eitherliqiuid-liquid extraction, equilibrium dialysis, solid phase extraction (SPE) with an appropriatesorbent material or solid-phase microextraction (SPME) with a polymer-coated quartz-glass fibrecan be utilised to determine the remaining amount ("free concentration") of test substance in thevesicle suspension (cf. Dulfer & Govers 1995, Escher et al. 2000, Vaes et al. 1997). Starting pointwill be the further development of the SPME-based detection that appears advantageous because ofits simplicity and flexibility, and the employment of liquid chromatographic methods based on IAMcolumns as promising tool for high-throughput measurements and comparative studies within sub-stance classes (UFZ).

For a judiciously selected test set of compounds that covers all major matrix-compoundinteraction types, laboratory experiments with relevant European soil types (as identified in WP 1.1)will be performed to derive the relevant matrix-specific phase parameters (EAWAG), and themeasurement of membrane-water partition coefficients leads to a respective parameterisation of themembrane-water partition coefficients (Kmw), which is more directly related to the hydrophobicity-driven compound partitioning into biological compartments (UFZ). Computational chemistry(Schüürmann 2004) will be used to develop models to predict the compound-specific Abrahamdescriptors as well as Kmw from molecular structure (UFZ), thus allowing a predictive application ofthe new equilibrium partitioning parameters in the context of methods to characterize the effectivein situ exposure in laboratory systems as well as in outdoor compartments, and as module of anintegrated method suite for multimedia fate modelling and cumulative risk assessment.

WP 2.2 Available exposure (Leader: Philipp Mayer, NERI)This WP aims at the direct analytical determination of available exposure, which is complementaryto the prediction of available exposure from experimental or calculated parameters as outlined inWP 2.1, and will provide a novel experimental methodology to be used as part of a tiered decisionsupport system to assess the effective exposure of contaminants.

A major drawback of current exposure determination schemes is the problem of how to cha-racterise the compound portion available for interaction with biota under matrix-specific in situconditions. In this respect, substantial progress has been made through the introduction of newanalytical methods (Diaz-Cruz et al. 2003, Lopez de Alda et al. 2003) and of new diffusivesampling methods for both organic pollutants and metals (Cornelissen et al. 2001, Mayer et al.2003, Zhang et al. 2001). Such sampling methods can be operated in three different regimesfacilitating three complementary strategies – all with the overall goal to determine the exposure thatis effective under in situ conditions:(1) The first strategy is to add a strong sorbent to the sample that constantly traps the “available

quantity” as it is released by the sample matrix. This strategy was demonstrated to have a highpredictive power for biodegradation and uptake of organic compounds into soil and sediment

27

organisms (e.g. Ten Hulscher et al 2003, Cuypers et al 2002), whereas it has been less appliedwithin an environmental toxicology context.

(2) The second strategy applies "equilibrium sampling devices (ESDs)" to sense the activity of thecontaminants (or equivalent quantities such as chemical potential, fugacity or freely dissolvedporewater concentration). This strategy is very powerful within a partitioning context, simplybecause it measures those entities that drive partitioning. The predictive power of such methodswas recently demonstrated for the uptake of a wide range of hydrophobic organic chemicals intoa sediment dwelling worm (e.g. Kraaij et al 2003).

(3) The third strategy has been developed for the sampling of cationic metals, and is called "dif-fusive gradient in thin films (DGT)". The sampler is brought into direct contact with forinstance soil, where it continuously traps the "diffusive flux" towards the sampler. Thepredictive power of this approach has been demonstrated for the uptake of copper into plants(Zhang et al. 2001).

The three sampling strategies will be adapted for the model substances and type of soils andsediments under investigation in NOMIRACLE. As outlined in RP 1, soil and sediment typesrelevant for the European environment will be identified in WP 1.1, and a targeted selection ofproject-wide test compounds will be performed in WP 1.2, in cooperation with QSAR methodsapplied in WPs 2.1, 2.3 and 3.3 (LMC, UFZ). The experimental work comprises analysis of fieldsamples including in situ studies, and laboratory biotest systems involving soil or sediment asemployed in RP 3, WPs 3.1 and 3.3, with the purpose to translate exposure between laboratory andthe field. The three exposure parameters will also be related to measured tissue levels in laboratoryexperiments of RP 3, and their suitability to determine “available exposure” for different types oforganisms will in this manner be tested. (CSIC, ITM, NERI, ULANC).

This work will be extended to an evaluation of methods of predicting available exposure. Thefield measurements of available exposure will be compared with exposure predicted by the sorptionalgorithms developed in WP 2.1 and the other algorithms used in the multimedia fate models (WP2.4). This comparison will form the basis for revised predictive algorithms, which will beincorporated into the chemical fate and exposure models (WP 2.4), enabling these models to predictavailable exposure.

Note that the actual bioavailability of test compounds provides a similar problem in in vitrosystems, due to binding to serum proteins, loss processes or simply due to uptake into cells andtissue or (non) specific binding to other components in the test or culture medium. Thus, even forthe same test and chemical, effect concentrations may vary significantly because of different testconditions (Gülden et al. 1994, Gülden and Seibert 1997, Vaes et al. 1997, Nagel et al. 1998). Thisproblem of bioavailability in in vitro systems has been addressed in the recently submitted STREPproposal SPEACH (Priority 8.1, policy-oriented research) coordinated by NOMIRACLE partnerSchüürmann (UFZ), and a respective cooperation between SPEACH and NOMIRACLE would beagreed upon approval of both proposals, in order to optimise the exploitation of the envisagedresources.

As regards indoor-outdoor relationships of environmental contaminants, air-borne and water-borne pollutants and their ingestive and inhalative pathways into human are of particular concern.Starting from the environmental fate, the aim is to find out clusters and patterns of combinedexposure which characterize specific and typical health-relevant situations (UFZ, JRC). Sinceeffective or available exposure is a concept that integrates both environmental and humantoxicology test systems, the above-mentioned diffusive sampling techniques will be also employedfor the biotest systems in the human health context, in order to enable the derivation of soundexposure-effect relationships. In addition, the suitability of the three environmental exposureparameters as discussed above will be tested with regards to transport and uptake processes that arerelevant for human exposure.

Considering the inhalative pathway, the indoor environment plays the most important role.Here, exposure to diverse chemicals has increased due the combined effects of energy-savingmeasures and the increased use of chemicals in consumer products. As a consequence, the typical

28

indoor VOC spectrum has changed, which requires new analytical techniques that go beyond thetraditional approaches (Seifert 1990, Molhave 1991). Recent findings indicate that the seasonalvariability is an additional confounding factor that has to be taken into account (Rehwagen et al.2003). The goal is to analyse indoor exposure profiles with respect to possible lead compounds thatrepresent certain types of emission sources and activity patterns, which in turn can be linked to thehuman health status. In addition to permanent (continuous) or intermittent indoor sources, outdoorexposure is also relevant due to infiltration into the indoor environment, which is a further pathwaythat requires judicious consideration.

This work package will also assume responsibility for coordinating the trace chemical ana-lysis within the whole NOMIRACLE Consortium. It is essential to ensure that the data collectedwithin the different work packages are of high quality and can be compared with each other. To thisend, this work package will test and select the best analytical methods for the purposes inNOMIRACLE, and distribute these methods to the other work packages conducting chemicalanalysis. In addition, WP 2.2 will serve as an information resource for analytical issues and alsoorganize and conduct intercalibration exercises to ensure that the other laboratories are successfullyimplementing the methods. This work would profit greatly from close links to the EU projectERAPHARM on pharmaceuticals coordinated by NOMIRACLE partner Knacker (ECT), and to theEU project SPEACH on alternative ecotoxicological methods coordinated by NOMIRACLE partnerSchüürmann (UFZ), provided these project become approved.

WP 2.3 Metabolic fate (Leader: Ovanes Mekenyan, LMC)Besides phase partitioning, degradation rates are major determinants of the environmental fate ofcompounds. Again, currently available models (Sabljic & Peijnenburg 2001) were developedmainly for more simple compounds, and with respect to microbial biodegradation most of the un-derlying experimental data were generated in qualitative form (e.g. ready vs. not ready biodegra-dable) mostly according to OECD guideline 301 C or D (OECD 1994). The methods representstandardised artificial laboratory conditions simulating waste water conditions with differentamounts of the test compounds and inoculum. As a consequence, the prediction of outdoor systemhalf-lives is hampered by the lack of quantitative or at least semiquantitative biodegradation ratesthat apply to realistic conditions. Available data sets are typically related to the hydrosphere and notto soils or sediments, where sorption, ageing, sequestration, and cross coupling may affect thebioavailability and metabolism.

To overcome these shortcomings, experimental biodegradation in typical European soils asidentified by WP 1.2 will be studied for a judiciously selected test set of compounds (provided byWP 1.1 in cooperation with QSAR predictions from WP 2.1 and 3.3), focusing on substances withmore complex chemical structures such as biocides, pharmaceuticals, and pesticides (UFZ). Theseresults will be compared to results of the OECD 301 tests (ECT).

Note that for the derivation of a mapping between OECD 301 results and degradation in soil,sediment and in water, there is a need to establish the relevant correlating variables for thesoil/sediment (e.g., soil/sediment chemistry and soil/sediment structure, water and pore water che-mistry), in addition to microbial population characteristics (in soil, sediment and water), tem-perature, and appropriate physicochemical parameters or structural information for the chemicals ofconcern. Soil and sediment characteristics vary regionally and even locally; therefore, one mustestablish a means of classifying soil and sediment parameters that are likely to impactenvironmental degradation. For soils, the soil classification database of the European Soil Bureau(ESB) will form the starting point, and if the need arises to improve the existing ESB classification,fuzzy ARTMAP classifiers and dynamic radial basis functions networks (Carpenter et al. 1992,Giralt et al. 2000, Rallo et al. 2002; URV) will be applied.

Like soils, sediments also exhibit significant variability with respect to their structure, che-mical composition as well as their indigenous microorganism populations. Therefore, theNOMIRACLE team will have to develop a practical database of the appropriate sediment proper-ties, for a reasonable number of locations along with representative degradation data for selected

29

chemicals, to enable QSPR development for sediment degradation rate parameters (URV). Thedatabases of soil and sediment parameters that are relevant to chemical and biochemical de-gradability will be incorporated into a common GIS system (spatial maps) used by theNOMIRACLE research team, will also be highly useful for fate and transport modeling efforts inWP 2.4 and 4.2.

As regards the aqueous phase, a respective experimental study will be performed (ECT),including a systematic search for additional data from literature and competent authorities. Then, anexpert system based on fuzzy methods will be developed (URV) that establishes – as well aspossible – a mapping between OECD 301 results in water to degradation in soil, taking into accountthe above-mentioned confounding factors that determine the soil type (URV).

Whilst the degradation rates in all major compartments are needed for a proper prediction ofthe compound fate in a multimedia environment, such an approach would still ignore the potentialimpact associated with the generation of transformation products. At the same time, algorithmshave been developed in recent years that allow, in principle, the consideration of transformationproducts in the context of multimedia fate modelling and risk assessment (Cahill et al. 2003, Fenneret al. 2000 & 2002, Quartier & Müller-Herold 2000) as discussed in WP 2.4. Accordingly, there is aneed of information about major metabolic pathways and associated transformation products.

To address this need, the NOMIRACLE team will focus on the further development ofknowledge-based expert systems such as PASS-BioTransfo (Borodina et al. 2003), METEOR(Greene et al. 1999, Judson & Vessey 2003, Judson et al. 2003, Button et al. 2003; LHASA) andthe rule-based metabolite generator system CATABOL (Jaworska et al. 2002; LMC), thus enablingthe derivation of new probabilistic estimates to assess the reliability of microbially mediatedmetabolic transformations in water, soils and sediments.At present, the CATABOL system mimics the microbial transformations in two test protocols,OECD 301C (MITI) and OECD 301B (Ready Sturm). The set of principal transformations isderived analysing both parent chemicals and their biodegradation products. It is based on a trainingset of observed catabolic pathways for more than 250 organic compounds from monographs(Gibson 1984, Schwarzenbach et al. 1993, Wackett & Hershberger 2001) and the University ofMinnesota Biocatalysis / Biodegradation Database (UM-BBD, http://umbbd.ahc.umn.edu/) (Ellis etal. 1999, 2000, 2001). Articles devoted to microbial degradation of specific classes of chemicalssuch as halogenacetic acids, terpenes, linear alkylbenzene sulfonate surfactants, bisphenols, etc.were also used (Casellas et al. 1997, Cook and Hrsak 2000, Ellis et al. 2001, Lobos et al. 1992, vander Werf et al. 1999).

LHASA have expertise in the use of non-numerical reasoning methods to rank events (in thiscase biodegradation reactions) according to how likely they are and the combination of thesemethods with those used in CATABOL will lead to new advances in computer methods forreasoning under uncertainty.

At present non-numerical reasoning methods are used in the METEOR system for theprediction of mammalian metabolism but there will be new challenges in applying them to thecomplex field of biodegradation. METEOR predictions about the more and less likely metabolicfates of chemicals in a mammalian system are based on analogy with similar chemicals, assessmentof likely absorption and distribution of the chemical (which may determine whether the chemicalreaches a particular enzyme in a particular organ), and potential excretion of the chemical or someof its metabolites. Some of these factors are less important or quite different for biodegradationprocesses – for example, because of the great microbial variety in a typical soil sample manyenzymes are potentially available without the need for further distribution, and active excretion isnot relevant. On the other hand, other factors may be important, such as the presence or absence ofoxygen, pH, and temperature (which can be presumed constant in a mammalian system). Theproposed work will be further informed by spin-off from a current collaborative project betweenLHASA and Ellis and Wackett at the University of Minnesota, and an internal project at LHASA,both of which are exploring different aspects of the problem of modelling biodegradation.

30

The reasoning methods developed at LHASA are designed to support reasoning about diverseinformation and problems, using quantitative and qualitative, numerical and non-numerical data.However, in METEOR, numerical expressions of probability are not used. CATABOL usesarithmetical techniques to rank its predictions. By bringing together the two systems and the twoapproaches we will, for the first time, implement a model in a practical context that combinesnumerical and non-numerical information about probability and brings together aspects offrequentistic and epistemic probability (i.e. probability based on statistical analysis and probabilitybased on past experience). It is expected that the high-level reasoning methods we develop will beof wide use both elsewhere in the proposed project and outside it, as well as within the specialisedareas of metabolism and biodegradation.

For the photolysis in air, computational chemistry will be used to derive a mechanisticallysound model to predict photolytic degradation rates from molecular structure (UFZ), employingmolecular descriptors based on quantum chemical models (Karelson et al. 1996, Schüürmann2004). The newly derived algorithms will allow more realistic predictions of degradation ratesunder outdoor conditions as needed in multimedia fate modelling (WP 2.4), and allow theconsideration of the formation of metabolites with modified half-lives and toxicity.

WP 2.4 Region-specific environmental fate (Leader: Mark Huijbregts, DESUN)Multimedia fate and exposure models provide information on the likely average concentrations ofcontaminants in environmental media (air, water, soil, sediment, and vegetation), taking intoaccount competing pathways for intermediate transport and loss by advection and degradationreactions. Two major weaknesses of currently used environmental fate and exposure models, likeEUSES (EC 1996, van de Meent 1993), can be identified:1. Current models are predominantly based on partitioning, bioavailibility and degradation

mechanisms that were derived primarily for the historically used hydrophobic substances,turning them less useful for the presently used more polar biocides and pharmaceuticals.

2. Current models tend to deliver time- and space averaged output for generic environments,disregarding the great temporal and spatial variations in concentrations that are often observedin reality. The advantage of the current models is simplicity. Models with much greater spatialand temporal detail tend to lack this important advantage.

In order to extend the applicability range of the models to compound classes with morecomplex and typically less hydrophobic chemical structures such as pharmaceuticals, pesticides andbiocides, the models will be improved to account for novel algorithms for partitioning,bioavailibility and degradation as derived in WP 2.1, WP 2.2 and WP 2.3. Since abiotic or biotictransformation may lead to metabolites with a separate physicochemical and toxicological profile, afurther key activity is to build in the functionality of handling both the initially emitted compoundand its metabolites (Cahill et al. 2003, Fenner et al. 2000 & 2002, Quartier & Müller-Herold 2000).To this end, interfaces will be built that allow to exploit the knowledge about prevalent metabolitesfrom rule-based systems (Behrendt et al. 1999, Button et al. 2003, Jaworska et al. 2002) asdeveloped in WP 2.3.

A further step towards realistic exposure is reached by including of the appropriate spatialand temporal resolution in fate and exposure modelling. To achieve this, the relative importance ofspatial and temporal specification as compared to the uncertainty in substance-specific parameterswill be explored for regionalised European-scale multimedia fate models with various spatial andtemporal resolutions of environmental characteristics. Novel probabilistic modelling techniques willbe employed for this purpose, based on cognitive neural networks and network analysis algorithmsthat are optimized for pattern recognition, feature extraction, community identification,classification and adaptive learning (Buchanan and Smith 1988, Kasabov & Kozma, 1998,Kreinovich et al. 2004, Guimera et al. 2003, Giralt et al. 2004). The required level of spatial andtemporal detail will follow from a combination of (a) uncertainty in substance properties, (b)transport potential of the substance and (c) spatial and temporal variability in emission profiles.Compounds will be selected in close cooperation with WP1.2. As regards the relevant compound-

31

specific physicochemical properties and emission profiles and environmental parameters such assoil types, waters and climate, a coordinated activity with WP 1.1 is foreseen to collect relevant(field) data of the European environment for the NOMIRACLE Consortium.

Output of the models for specific substances will be evaluated by comparison with fieldmeasurements, thus testing the validity of the models to (i) predict and compare environmentalconcentrations of parent compounds and possible metabolites on both a steady-state and temporalbasis and to (ii) identify key fate processes and wildlife/human exposure pathways. Whereasregionalised continental scale models have been previously developed for describing fate in theEuropean region, this would be one of the first attempts to evaluate the predicted environmentalconcentrations of these models against measured concentrations.

In Table B.4-1, the major research themes of the four workpackages as well as their major inter-actions within RP 2 as well as with RP1, RP 3 and RP 4 are summarized.

32

Table B.4-1: Major research themes and interaction patterns of the RP 2 work packages.

Major research themes Major interaction

WP 2.1: Matrix-compound interaction

• Phase partitioning soil/membrane• QSAR prediction of phase partitioning

Input from:• RP 1 (reference soils/compounds)Output to:• WP 2.2 (experimental results)• WP 2.3 (experiment + modeling results)• WP 2.4 (phase partitioning algorithms)

WP 2.2: Available exposure

• Analytical competence for bioactiveagents

• Matrix-specific available exposure• Available exposure in biotest systems• Indoor/outdoor relationships• European indoor exposure profiles

Input from• RP 1 (reference soils/compounds)• RP 3 (compound doses in organisms)• WP 2.1 (phase partitioning)Output to• RP 3 (analytical competence for biotests)• WP 2.4 (matrix-specific available

exposure, indoor exposure)

WP 2.3: Metabolic fate

• Biodegradation in soil/water/sediment• System factors determining biodegrada-

tion• Probabilistic prediction of metabolic

rates and pathways• Quantum chemical prediction of photoly-

sis rates

Input from• RP 1 (reference soils/compounds)• RP 3 (experimental metabolic pathways;

bacteria toxicity)• WP 2.1 (phase partitioning)Output to• RP3 (metabolic pathways)• WP 2.4 (metabolic pathways)

WP 2.4: Region-specific environmental fate

• Spatially explicit soil parameterization atEuropean scale

• Probabilistic optimization of scenario-specific spatial and temporal resolution

• Model evaluation with field data• Sensitivity profile of input/output

relationships

Input from• RP1 (risk scenarios; reference

soils/compounds)• WP 2.1 (phase partitioning algorithms)• WP 2.2 (matrix-dependent availability,

indoor exposure profiles)• WP 2.3 (prevalent metabolites)Output to• RP 4 (Spatially and temporally explicit

available exposure)

33

Research Pillar 3 - Effect Assessment (Sub-coordinator: David Spurgeon,NERC)

WP 3.1Acronym UFZ NERC VU NERI DSTA WU DBUAPersonmonths 60 36 18 12 6 12 18Acronym EKUT NIPH USALZ RWTHA LemTec WRcNSF LIMCOPersonmonths 20 36 36 12 12 30 30

WP 3.2Acronym NERI UFZ WU DBUA UJAG EKUT LIMCOPersonmonths 60 40 12 30 48 18 18Acronym USALZPersonmonths 12

WP 3.3Acronym VU NERC DSTA UJAG UA LMC WRcNSFPersonmonths 48 24 12 36 12 36 18Acronym UFZPersonmonths 12

WP 3.4Acronym DSTA NERC UFZ WU UWC UCAM EKUTPersonmonths 30 36 30 24 36 36 22Acronym UAPersonmonths 36

Aims and rationaleIn natural ecosystems, organisms are frequently exposed to mixtures of chemicals/metabolites andnon-chemical stressors (US EPA, 2003, Heugens et al. 2001). Recognising this, a central challengein delivering the objectives of NOMIRACLE is estimating the interactive effects of single andcombined chemical and non-chemical stressors. As this applies both to human health and ecologicalassessment, RP 3 will aim to integrate environmental and human health effect assessment formixtures. This will be done by adopting a mechanistic-based approach to identify conserved causeand effect relationships in environmental species and human model systems. To achieve suchintegration, partners in RP 3 will undertake a set of mechanistic and experimental studies, with theresults compiled in a single relational data store. This will be used in RP 4 to generate rules for riskassessment of combined effects. By working with researchers in RP 2, the role of chemicalproperties in defining the nature of interactions will be elucidated. These rules can then be used incumulative risk assessments conducted for exposure scenarios relevant to Europe.

Background, the state of the art and the NOMIRACLE approachBecause of the potential importance of chemical mixture effects in biological systems,(eco)toxicologists have developed a number of approaches to mixture toxicity assessment. Amongthese approaches, one paradigm that is based on two underpinning concepts has found wideacceptance (Eggen et al. 2004, Cassee et al. 1998). First, if chemicals have the same mode ofaction, their combined toxicities can be described by the concentration addition model; second, ifthey have different modes of action, their combined effects can be described by the independentaction model. The broad applicability of these two reference models for (eco)toxicologicalassessment of simple and complex mixtures has been demonstrated for both similarly acting(Altenburger et al. 2000, Deneer et al. 1998, Faust et al. 2001, Hermans, et al. 1985) andindependently acting compounds (Backhaus et al. 2000, Deneer et al. 1988, Faust et al. 2003) in arange of species.

34

Although the concentration addition and independent action reference models have thepotential to describe the combined toxicity of many chemical combinations, there are cases inwhich these approaches have been found to fail. These include not only where the well known, butcomparatively rare, conditions of synergism (Johnson et al., 1994, Meled et al. 1998, Forget et al.1999) and antagonism (Van Gestel and Hensbergen, 1997, Posthuma et al. 1997) occur, but alsowhere there are more subtle conditions such as dose level deviations in which combined effectsdiffer from predicted as dose level varies (e.g. synergism at low dose, antagonism at higher dose)(Gennings et al, 2002, Jonker, 2003) and dose ratio dependent deviations (e.g. extent of synergismor antagonism depends on which chemical is dominant) (Van Den Hurk et al., 1998, Jonker et al.2004, in press, Carter and Gennings, 1994). To date the two reference models have been validatedonly for chemicals acting by precisely the same mechanism (e.g. 15 oxidative uncouplers), or forsets of chemicals acting by strictly different mechanisms. The challenge that will be tackeled inNOMIRACLE is assess multifunctional compounds, where interaction could occur synergisticallyor antagonistically for different dose levels and dose ratios.

In current risk assessment mixture effects are at best taken into account using the twoexisting reference models (additive, independent) and at worst merely by including a simple safetyfactors for single compounds. Recognising a more scientific approach to mixture toxicity analysiswas needed, the EU funded MIXTOX project (ENV4-CT97-0507), co-ordinated by NOMIRACLEpartner WU, developed and validated a framework for the detection and statistical interpretation ofthe full range of possible chemical mixture interactions. The approach uses an experimental designwhich, by testing individual compounds and mixtures of different dose levels and dose ratiossimultaneously, provides maximum statistical certainty in identifying the shape of toxicity responsesurfaces (Jonker, 2003, Jonker et al. 2004, in press). The analysis framework can be applied both incases where mode of action is known and importantly also where such information is not available,or where the known primary mode of action is not relevant to the species under study (e.g. herbicideeffects on animals). Using this approach, four biologically relevant deviation patterns from bothreference models can be identified by means of likelihood analysis: no deviation, absolute deviation(synergism/antagonism), dose level and dose ratio dependent deviations (see above for description).As well as exposure to combinations of toxic chemicals, populations living in the real world arealso subject to non-chemical stressors. These can interact either directly with chemicals throughchanging bioavailability or indirectly through changing organism biology (e.g. Friis, et al, 2004,Herbarth et al. 2002, Heugen et al, 2003, Malcolm et al. 2003). While there is some information onthe mechanisms behind the interactions within chemicals mixtures taken from pharmacology (Yang,1994) little is known about how chemical and non-chemical stressors may interact. For example,non-chemical stressors may cause significant changes in the physiology of organisms that may ormay not reveal themselves as effects at the individual level, depending on the plasticity of thespecies regulatory mechanisms (Parker et al. 1999). The growing acceptance that (eco)toxicology ismerely a branch of the wider field of stress biology (Van Straalen, 2003) offers the potential tounify approaches for mixture toxicity (multiple chemical) and multiple stressor (chemical/non-chemical) effect assessment. In NOMIRACLE multiple stressor analysis will be developed in linewith the novel approach for mixture toxicity developed in MIXTOX (Jonker, 2003, Jonker et al.,2004). As NOMIRACLE proceeds, use of this quantitative approach will enable predictionsregarding the combined effects of an ever increasing number of stressors. As complexity builds, ascientific basis for forecasting risks to organisms of cumulative exposures occurring under thevariety of field conditions within Europe will be formulated. At this point work will move into afinal field validation phase that will involve a coalescence of partners from all pillars and the finaldefinition of a unified approach.

Completion of the exposures and response modelling for mixture toxicity (WP 3.1) andmultiple stressors (WP 3.2) will provide data that can be used to establish probabilistic rules thatcan forecast the magnitude of interactions between multiple stressors. To allow these rules to beused for predictions in general stress biology, detailed mechanistic understanding is needed(Hertzberg and McDonell, 2002). Such work should aim to identify generic traits, shared between

35

systems and taxa that define the nature and extent of interactions for mixtures of stressors. Toachieve this improved understanding, as well as simply quantifying multiple stressor effects,NOMIRACLE will also directly investigate the controlling mechanisms in diverse species. Thiswill include investigation of three levels of interaction (as adapted from Calamari and Alabaster,1980 by Van Gestel and Hensbergen 1997 and Posthuma, et al. 1997) namely; those occurring inthe exposure medium (environmental availability) in RP 2, those affecting the physiologicalprocesses of uptake (bioavailability) and elimination in WP 3.3 and excretion and finally those atreceptor and target sites in WP 3.4. These analyses will concentrate on studies in animals,including mammalian cells and rodents, in order to increase the relevance of the data for humanhealth.

WP 3.1 Interactive toxicological effects in diverse biological systems (Leader: AlmutGerhardt, LIMCO)WP 3.1 will take a stepwise approach to mixture toxicity assessment. As detailed analysis of allmixtures combination is unfeasible, an initial analysis of the single compound and mixture toxicityscenarios identified in WP 1.2 will be made using rapid screening bioassays. This will allow us toselect for detailed analysis 1) a set of single compounds with different modes of action causingvarying effects in the different screening assays, 2) a set of chemical mixtures that show nodeviation from the reference (additive, independent) models and 3) a set of mixture that showconsistent deviations from the reference models. Detailed investigation of the most relevantscenarios identified in WP 1.2 will be undertaken using chronic exposures for the environmentallyrelevant species most likely to be affected. This data will then be used to define how conservedthese response profiles are amongst species, life-stages and endpoints.

The rapid screening assays that will be used are:

• the Vibrio fischeri Microtox system (Doherty, 2001) (WRcNSF);• a rapid benthic invertebrate exposure with Tubifex (Tichy et al. 2002) and/or Chironomidae

(Gerhardt & Janssens de Bisthoven 1995, Gerhardt & Schmidt 2002) (NIPH, LIMCO);• the early life stage fish test (OECD, 1992) (LIMCO);• a single celled ciliate protozoan Tetrahymena pyriformis which act as a model for the effect of

pollutants on cells of the human airway (Massolo et al. 2002, Muller and Herbarth, 1994,Netzeva et al. 2003) (UFZ);

• screens for human health effects based on existing and novel human cell lines (e.g. immunecells) (Bommel et al. 2000, 2003) (UFZ, USALZ).

The chronic exposures will be conducted for species that 1) fill significant positions in aquatic andterrestrial food webs, 2) come from different taxonomic groups and have different biologicalcomplexity and (3) have different habitats and requirements. Species selected are:

• aquatic algae and the higher plant Lemna (OECD, 2000a) (LemTec/RWTHA);• Daphnia magna (Barata et al. 2002, OCED, 1995) (DBUA, UA);• Chironomidae (OECD, 2001);• the fish Danio rerio (Gerhardt et al., 2002, in press, Nagel and Isberner, 1998) (EKUT, UFZ,

LIMCO);• the marine mollusc Mytilus galloprovincialis (Panfoli et al. 2000) (DSTA)• the terrestrial oligochaetes Eisenia fetida and Lumbricus rubellus (Kula & Larink 1997,

Spurgeon et al. 2003); (NERC, NERI);• the terrestrial collembolan Folsomia candida (ISO, 1999, Holmstrup and Krogh, 1996, Smit et

al. 1997a,) (VU, NERI);

36

• the higher plants Trifolium pratense, Lolium perenne and Sinapis alba (OECD 2000b, Sverdrupet al. 2002) (NERI, NERC);

• the nematode Caenorhabditis elegans (Kammenga et al. 1997, Jonker et al, 2004) (WU); aterrestrial carabid (Lagisz et al. 2003, Laskowski, 1997) (UJAG);

• the slime mold Dictyostelium discoideum (Falugi et al. 2002)(DSTA) and finally;• the mammalian models, mouse (Hahn et al., 2003, Grunewald et al. 2003) (Mus musculus) and

rat (Rattus norvegicus) (Griffin et al. 2000) (NERC, USALZ).

Exposures will be conducted by experts skilled in the handling of the test organism/system rangingfrom bacteria to higher vertebrates according to routine in-house protocols and approved standards.SME partner Lemna-Tec will act as a technology platform provider to assist all partners to furtheroptimise data collection from each test. During these exposures, a series of responses (e.g. gene andprotein expression, metabolic dysfunction, immune dysfunction, cellular changes, weight change,proliferation, reproduction, behaviour, growth and viability) will be measured. Measurement ofthese endpoints will enable identification of biological dysfunction at all levels of organisation,encompassing not only acute, but also significant chronic effects (Calow and Forbes, 2003,Kammenga and Laskowski, 2000, Van Straalen, 2003). The mixture experiments will be undertakenand analysed using the novel framework developed during the MIXTOX project (Jonker, 2003,Jonker et al., 2004, in press) and outlined above. To allow this approach to be used in house by allpartners, a guidance document will be written and a software version of the existing analysisprogram refined and distributed. Use of this framework will allow modelling of response surfacesfor simple and later complex mixtures using an optimal design and minimum number of testanimals, thus reducing the number of test organisms that need to be sacrificed. When deviations incombined toxicity from additive and independent effects are found, the mechanisms underpinningthese will be elucidated by toxicodynamic (WP 3.3) and molecular (WP 3.4) approaches.Completion of screening and detailed studies of simple chemical mixtures will provide us with thedata required to develop a theoretical basis for predicting combined effects. This theoretical basiswill then be extended to address complex mixtures. These predictions will be validated usingexperimental work that will study interactions for a progressively increasing number of compoundsto ensure continued mechanistic understanding of the interactions. Analysis of complex mixtureshas already been conducted in the MIXTOX project, and has demonstrated the ability of the datainterpretation approach to identify interactions occurring between multiple chemicals (Jonker,2003). To deliver this improved mechanistic understanding, in NOMIRACLE we will extend thisapproach to address the exposure, effect and resultant risk for organisms of cumulative chemicals.At this point work will move into a final field validation phase.

WP 3.2 Combined effects of natural stressors and chemicals (Leader: Martin Holmstrup,NERI)Given the likely importance of multiple stressor exposures, this area has not, to date, been a priorityarea of research (Calow and Forbes, 2003, Eggen et al. 2004, Holmstrup et al. 2000; Højer et al.2001). To address this gap in understanding, work package 3.2 in NOMIRACLE will place classical(mixture) toxicity studies within the context of the diverse environmental conditions that existacross Europe. The environmental and population specific factors considered will be selected in theprioritisation and scooping process undertaken in RP 1. Examples already under investigation bypartners include, diet quality (all partners as applicable), population vulnerability due to geneticerosion as a result of ecological bottlenecks or habitat fragmentation (Lopes et al. 2004, VanStraalen and Timmermans, 2002) (UJAG), prevalence of pathogens/allergens (Fritz and Herbarth,2004, Rolle-Kampcyyk et al. 2002) (USALZ, UFZ, UJAG), UV exposure (Hatch and Blaustein,2003) (DBUA, NERI), extreme temperatures (Heugen et al. 2003, Holmstrup et al. 1998) (allpartners as applicable), acidification (Herrmann et al. 1993) (all partners as applicable), soil

37

moisture content (drought/water logging) (Friis et al. 2004) (NERI, UJAG, VU) andanoxia/eutropication (Ivora et al. 2002) (DBUA).RP 1 will identify which cumulative stress combinations should be assessed for which types of testorganism in order to maximise the relevance of the assessments. Thus, the influence of pathogensmay be most relevant in humans/mammals, the influence of drought most relevant for soilorganisms etc. The ultimate goal of WP 3.2 is to have a broad understanding of the effects ofrelevant cumulative stressor combinations (chemicals combined with environmental stress) on awide range of organisms and human immuno-responses to cover most environments and humancommunities of the European continent.

The approach to identifying the interactive effects of the environmental stressor will mirrorthat used for contaminant mixtures in WP 3.1. For each factor, the response profile for eachenvironmental stressor will first be described in the most environmentally relevant set of theorganisms listed in WP 3.1. These responses will be collated, thereby providing spin-offinformation concerning the limits of the tolerance of diverse taxa to environmental variation andchange. Once the stress response profiles for the single stressors have been established, thisinformation will be used to design multiple stressor studies that can be analysed using the novelframework for chemical mixtures outlined above. Likelihood analysis will be applied in order toidentify if there are deviations (absolute synergism/antagonism, stress level and stress ratio) fromeither reference model (additive, independent). Using both reference models will allowdetermination of which of these acts as the better predictor of each multiple stressor effect.In the later stages of the project a set of field validation studies are planned to verify the interactionsdemonstrated in the laboratory-based studies. These will focus on different types of emissionscenario including point-source/end of pipe, landscape, catchment, regional and national scales incombination with relevant multiple environmental stressor combinations identified in WP 1.2.These field studies will be a crosscutting issue encompassing the disciplines of all four researchpillars. These field studies will also form an important part of the demonstration activities plannedin the last phase of the NOMIRACLE project. In parallel with ecological based field studies, wewill undertake work that enables us to understand the epidemiological significance of combinedstressor interaction identified in human model systems. For this, the results of epidemiologicalstudies that have been done or are ongoing will be sorted and used to provide informationconcerning the effects of the substances and environmental stress scenarios on the population atlarge. This data will be mined and used to confirm or renounce finding concerning the possibleeffects derived on the basis of laboratory studies. This process will provide essential validation ofthe potential of combined exposure as a risk factor for public health.

WP 3.3 Toxicokinetic modelling (Leader: C. Van Gestel, VU)This work package is one of two within NOMIRACLE that will improve mechanistic understandingof how single chemicals, mixtures and multiple stressors affect species with different physiologies.In it the role of uptake and elimination in governing the effects of such exposures will beinvestigated. The approach uses chemical analysis to determine toxicokinetics under each exposurescenario. This work will directly address interactions affecting the physiological processes ofuptake and elimination for which there are numerous examples concerning modulation ofabsorption, metabolism, localisation and excretion of one compound by another (Haddad et al.2000, Van Den Hurk et al. 1998).

A suitable approach to evaluate the relationship between external exposure, organism ortarget tissue dose, and biological outcome is through toxicokinetic models. As these will need todescribe each species in terms of their physiology, biochemistry and ecology, a range of approacheswill be needed to encompass the taxonomic variety included in NOMIRACLE. Methods developedwill range from simple one and two compartment models for aquatic (UA), soil dwelling (VU,NERC) and soil surface dwelling (UJAG) invertebrates; to complex models that describe thephysiological processes of vertebrates (VU, NERC). Fortunately past work has identified potential

38

prototypes for all of these (Andersen 1995, Belfroid et al. 1994, Lien et al. 2001, Mackay & Fraser2000, Widianarko et al. 2001). After an initial review of these methods, the best approaches will bedeveloped into fully operational models (co-ordinated by VU in co-operation with WP 4.1). Thesemodels can then be used to assess the contribution of toxicokinetic modulation to the toxicity ofspecies to mixtures and multiple stressors.

With suitable models in place for each species, toxicokinetic assessments will be made inrelevant species. Colleagues involved in WP 2.2 will help identify suitable methods for the analysisof test chemicals in substrates and organisms and SME partner WRcNSF will provide advice andtechnical supervision to assist partners in further refining these methods. To ensure properimplementation of each analytical method, a series of research training visits will be initiatedbetween the partners in WP 3.3 (particularly WRcNSF) and WP 2.2. Using existing and modifiedprotocols, toxicokinetic assessments will first be made in all species for two compounds selectedfrom the priority list. This data will be used to ensure optimisation of each toxicokinetic model (incollaboration with partners in WP 4.1). As will be the case throughout the project, data concerningthese parameters for compounds (both singly and in combination) will be recorded and used in riskassessment models that consider bioaccumulation (in WP 4.1 and 4.2). Following this refinementstep, assessments of toxicokinetic parameters in mixture experiments can begin. Work will start byfirst assessing toxicokinetic parameters in simple mixtures of chemicals with the same mode ofaction that show no deviations from concentration addition in all tested species. Next, mixtures withdifferent mode of action that fit the independent model will be assessed, before moving on finally toinvestigate the role played by toxicokinetics in simple and complex mixtures that show consistentdeviation from either of the two reference models. As the role of metabolism in prioritisingchemicals will be important both in terms of kinetic parameters, the deactivation of some parentcompound and the production of toxic metabolites for other this will also be modelled in detail. Topredict metabolism and toxicity for compounds, the tissue metabolism simulator (TIMES) systemwill be applied to predict metabolic activation of chemicals (Mekenyan et al., 2004a, 2004b). Thesystem uses a heuristic algorithm to generate plausible metabolic maps from a comprehensivelibrary of biotransformations and abiotic reactions and estimates for system-specific transformationprobabilities. To further the mechanistic understanding and prediction of hazardous effects onhuman health caused by indoor exposure, targeted QSAR investigations of the toxic effects of VOCchemicals on ciliates, immunocompetent and other human cells will be undertaken. This work willinclude effect profile analyses across non-human species, and aims at building a mechanisticmapping between epidemiological human health status and biotest system response patterns. Usingthe range of data collected and by working with QSAR developers in RP 2, a predictive approachthat describes kinetic changes at different dose level and dose ratio combinations will be sought.

WP 3.4 Molecular mechanisms of mixture toxicity (Leader: Aldo Viarengo, DSTA)Even when interaction between chemicals at the levels of bioavailability (WP 2.1) andtoxicokinetics (WP 3.3) are accounted for, there are still likely to be a set of unexplained deviationsin combined effects from concentration addition and independent action due to interactions at thesite of toxicity. To elucidate the conserved and distinct molecular changes that underpin thesystemic response of species to chemical mixtures (WP 3.1) and multiple stressors (WP 3.2), WP3.4 will use a dual approach that combines comprehensive and targeted methods. Thecomprehensive analyses will exploit global gene, protein and metabolite analysis technologies. Atthe genomic level, gene screening, discovery and mapping methods (UWC, UA)(Stürzenbaum et al.1998; De Coen and Janssen 1997, Moens et al. 2003) will be used. Additionally, transcriptomicstudies will be made using established commercial and custom printed microarrays. These include a14,000 gene zebra fish cDNA array (UFZ), a 16,000 gene oligonucleotide nematode array (WU,UWC), an 8,000 gene earthworm cDNA array (Sturzenbaum et al., 2003) (UWC), a 2500 geneDaphnia magna cDNA array (Soetaert et al. 2003), a 2,000 gene cDNA mussel and a 2000 genecDNA Dictyostelium arrays (DSTA). Proteomic analyses will investigate protein expression withintargeted sub-proteome and in limited cases also for global profiles (Pennington & Dunn 2001;

39

Hogstrand et al. 2002, Vido et al. 2001). Both 2-D electrophoresis and chromatography basedseparations (UA, NERC, DSTA) will be used with mass spectroscopy for protein identification(DSTA, UA, UCAM, NERC). To investigate the metabolic consequences of changes in geneexpression and resulting functional protein complement, large-scale analysis of metabolites will beconducted using 1H nuclear magnetic resonance spectroscopy (NMR), GC-MS and LC-MS basedmetabolomics (UCAM, NERC) (Bundy et al. 2002; Nicholson et al. 2002; Griffin et al. 2001).To supplement the use of post-genomic screening techniques, detailed biochemical and moleculargenetic analyses will also be used to further investigate interaction mechanisms. These will includemeasurements for cholinesterase (DelOmo et al. 1996, Amaroli et al. 2003) (DSTA, NERC),cytochrome P450 (Viarengo et al. 1997) (NERC, DSTA), glutathione-s-transferases (Saint-Denis etal. 1998) (NERC), stress proteins (Kohler et al. 1998, 2001, Triebskorn et al. 2002) (EKUT),antioxidant enzymes (Regoli et al. 1998) (DSTA), metabolic enzymes (Long et al. 2003, De Coenand Janssen, 1997) (NERC, UA) and energy budgets via cellular energy allocation (UA) (De Coenand Janssen, 2003a) and organ pathology (EKUT) (Hinton and Lauren, 1990, Triebskorn et al,2002). These approaches (e.g. cellular energy allocation) offer the possibility to link short termbiomarkers with population level responses through the DEBtox model (De Coen and Janssen,2003b, Kooijman, 1993). On completion of analyses, pattern recognition techniques (UCAM,NERC) (Lindon et al. 2001; Raamsdonk et al. 2001) can be used to overlay species responsesprofiles to identify if modes of action and interactions are unique to species or common betweentaxa. This approach has the potential to highlight potential species specific and common biomarkersthat can be used in ecological monitoring of cumulative stress effects. Further because analyses areconcentrated into a series of animal phyla, including mammalian cells and rodents these indicatorscould be applicable to the epidemiological assessment of human population health. Suchbiomarkers will be validated in the field studies conducted as a cross cutting initiative inNOMIRACLE, thereby providing a set of indicators of chemical mixture and multiple stressorexposure for broad application in environmental and human health monitoring.

Research Pillar 4 –Risk Assessment (Sub-coordinator: Ad Ragas, DESUN)

WP 4.1Acronym VU DESUN RIVM USOUTH EPFLPersonmonths 48 52 4 36 18

WP 4.2Acronym DESUN URV UFZ ALTERRAPersonmonths 46 42 49 22

WP 4.3Acronym SYKE DIA JRCPersonmonths 54 21 18

WP 4.4Acronym ALTERRA UFZ URV UNIMIB NERI JRC DESUNPersonmonths 8 7.5 6 4 4 3 2

Aims and rationaleThe main aim of RP 4 is to develop novel methods for integrated risk assessment that makeoptimum use of available data and models, ensuring an efficient use of valuable resources. This aimwill be accomplished by the integration of the results of RP 1, RP 2 and RP 3 within a probabilistic

40

and spatially explicit modelling framework that is tailored to support risk management decisions.The new concepts and techniques developed in RP 4 will be applied to produce probabilisticestimates and characterisations of risk for the scenarios identified in RP 1.

For a selected number of critical environmental functions, endpoints and agents (RP 1),emission and environmental data gathered in RP 1 will be used to produce probabilistic(cumulative) exposure estimates (RP 2 and RP 4). Comparison of these estimates with the effectdata on mixture toxicity and multiple stressors gathered in RP 3 will result in refined probabilisticestimates of cumulative risk including evaluations of qualitative dimensions of risk and indices ofsignificance. The dependencies between management and assessment approaches will be identified,and options outlined to deal with these risk estimates in a multi stakeholder setting, i.e., relating torisk perception, risk communication and a prudent application of the precautionary principle.The activities in RP 4 are organised in four work packages. WP 4.1 aims at developing novelconcepts and techniques to quantify uncertainty in different stages of the integrated risk assessmentprocess (exposure assessment, effects assessment, risk characterisation and risk management).Starting point of WP 4.1 is the notion that uncertainty is a measure of information quality that playsan important role in risk management decisions. WP 4.2 addresses modelling techniques that canintegrate exposure and risk over space and time. Contrary to traditional risk assessment practicesthat tend to concentrate on the stressor, the approach taken here focuses on the receptor as the maintarget of the modelling effort, as the receptor integrates (the effects of) stressors as it moves throughspace and time. WP 4.3 aims to analyse cognitive, social and contextual aspects of integrated riskassessment in order to improve the overall knowledge base for dealing with multiple and complexrisks, uncertainty and ambiguity. Finally, the presentation and visualisation of the cumulative risksthat have been quantified in WP 1.2 (assessment of potential cumulative risks) and WP 4.2 (refinedcumulative risk assessment) will be the subject of WP 4.4.

WP 4.1 New concepts and techniques for probabilistic risk assessment (Leader: Ad Ragas,DESUN)The growing awareness that deterministic risk assessment procedures can result in conservative orerroneous risk estimates (and consequently in a waste of resources) has resulted in a shift towardsprobabilistic risk assessment (PRA; Ragas 2000). However, most techniques currently used in PRAhave considerable shortcomings from a scientific as well as a management perspective. Examplesare a lack of differentiation between various types of output variance (i.e., true uncertainty andinterindividual, spatial and temporal variability), the considerable amount of subjectivity involvedin the choice of the probability distributions, the sub-optimal use of existing data, and the complexinterpretation of the probabilistic output. WP 4.1 aims to develop new concepts and PRA techniquesthat are scientifically sound and practicable for management purposes. Examples are nested andMarkov Chain-based Monte Carlo simulation (Cullen & Frey 1999, Johansson & Jonsson 2002),Bayesian methods including dynamic Bayesian belief networks and nonparametric hierarchicalBayesian analysis (Bates et al. 2003, Varis & Kuikka 1999, Arjas & Andreev 2000), distribution-free techniques (Ferson et al. 1998) and techniques to quantify the added value of new information(Hammitt & Shlyakhter 1999).

Some concepts and techniques that will be developed in WP 4.1 are restricted to a certainphase of the risk assessment process (e.g., derivation of a probabilistic NEC for mixtures), whereasothers are more generally applicable. Two general techniques explored in WP 4.1 are (1) theseparation of uncertainty and variability by means of nested Monte Carlo simulation, and (2)quantification of the added value of new information (VOI) for risk management decisions (Dakins1999) (DESUN). The former technique will initially be applied to an integrated human exposuremodel that describes the uptake of contaminants from relevant environmental exposure pathways(i.e., air, drinking water, swimming water, food, soil and dust). In a later stage of the project, thistechnique will be used in ecological risk assessment, i.e., to separate the influence of intraspeciesand interspecies variability from true uncertainty in estimation of the NECeco. The VOI techniquewill initially be applied to quantify the reduction in uncertainty that can be realised by performing

41

extra toxicity tests before deriving a NECeco. The costs of an extra ecotoxicity test will be weighedagainst the possible benefits of a less stringent NECeco due to reduced uncertainty. In a later stage ofthe project, this VOI technique will also be used in other areas of the risk assessment process.In the area of effect assessment, a new method will be developed to derive a probabilistic NEC forsimple and complex mixtures based on the Dynamic Energy Budget theory of Kooijman (2001)(VU). This theory quantifies the effects of compounds by influencing the resource allocation withinorganisms to various endpoints, i.e., feeding, maintenance, development, growth, reproduction andaging processes. Co-limitation by food and other factors that modulate toxic effects are included inthe model. ISO and OECD have recently accepted the current model for the analysis of data fromstandardised ecotoxicity tests (Kooijman et al. 2004). In the NOMIRACLE project, the method willbe extended for the assessment of simple and complex mixtures. Profile likelihood and Monte Carlomethods will be used to study the confidence intervals of the NEC-estimates.

Another important activity in WP 4.1 is the derivation of new probabilistic assessment oruncertainty factors (UFs) for the extrapolation of laboratory toxicity data to relevant human andecological endpoints (Amler et al. 2003, Pelekis et al. 2003, Roelofs et al. 2003). This will beachieved by meta-analysis of toxicity data, i.e., those gathered in RP 3 and those stored in existinghuman and ecotoxicological databases. The UFs currently used in human and ecological effectassessment have evolved along comparable, but separate lines. Although there are some cleardifferences between human and ecological effect assessment, the awareness is growing that theunderlying toxicological principles are to a large extent comparable and governed by a limitednumber of mechanistic descriptors, e.g., substance parameters, the genetic predisposition ofreceptors and the toxicological mode of action. In WP 4.1, new probabilistic UFs will be derivedseparately for human and ecological endpoints, but the underlying analytical framework used forthe meta-analyses will be harmonised to the extent possible as regards the relevant mechanisticdescriptors. Furthermore, there will be an exchange and evaluation of toxicological data thatdescribe comparable phenomena (e.g., data on interspecies differences are relevant for extrapolationfrom test animals to humans and for deriving the NECeco from a limited number of single speciestests).

Derivation of UFs for human risk assessment (USOUTH) will concentrate on thequantification of human variability in pharmacokinetics and pharmacodynamics for individualcompounds and chemical mixtures that are handled by major polymorphic pathways (CYP2C9,CYP2C19, CYP2D6, NAT, glutathione-S-transferases, sulphation) (Dorne et al. 2002, 2003ab,2004ab). Meta-analyses and quantification of interspecies differences will also be performed as acomparison of pharmacokinetic and pharmacodynamic differences between humans and test species(rat, mouse, rabbit, dog including neonatal animals) (Walton et al. 2001ab, 2004).

For ecological risk assessment, probabilistic UFs will be derived for extrapolation of (1)acute to chronic endpoints and (2) the median value of a species sensitivity distribution to theNECeco (Roelofs et al. 2003, Pennington et al. 2003c). The database used for these analyses(Wintersen et al. 2002) already contains extensive data on pesticides and will be supplemented withdata on pharmaceuticals and other substances, partly gathered in RP 3. The dose response surfaceson mixture toxicity and multiple stress situations provided by RP 3 will be analyzed to identifyadherence and deviation from concentration and effect addition mixture models. Based on theseanalyses, UFs will be derived to describe the possibility and magnitude of particular deviationswhen chemicals have similar and dissimilar modes of action (DESUN, RIVM).

Finally, probabilistic indicators will be developed that enable comparative risk assessment(CRA) due to different stressors, e.g., toxic stress, eutrophication and acidification (EPFL). CRA isbased on the premise that not all environmental problems pose the same degree of risk to humanand ecosystem health, and that all environmental problems cannot be addressed fully at the sametime. In this study, methods developed in Life Cycle Assessment (LCA; Goedkoop & Spriensma1999, Jolliet et al. 2003, Pennington et al. 2003ab) and in the EU FP5 OMNITOX project (Payet etal. 2003; Larsen et al. 2003) will be extended to integrated risk assessment of mixtures and multiplestressors.

42

WP 4.2 Explicit modelling of exposure and risk in space and time (Leader: Uwe Schlink,UFZ)It is common practice in risk assessment of chemicals to use a spatial and temporal average of themeasured concentration(s) to estimate exposure and risk. This approach eliminates all informationabout spatial and temporal patterns of the contamination and of the exposed receptor. It is thereforea rather crude method that may result in over- or underestimation of the actual risk. The main aimof WP 4.2 is to develop new methods and models that explicitly address the temporal and spatialdimensions of cumulative risks, both for human and ecological receptors. The structure of WP 4.2 isillustrated in Figure B.4-3.

Vulnerability analysis ofecological receptors

Human health data onexposure and risk clusters

from WP1.1

Identification of ecologicalparameters/receptors

Spatial and statistical analysis

Identification of causalrelationships

Small-Scale Random Walk Models

Large-Scale Random Walk Models

Temporal exposuremodel and estimates

Aggregation ofspatial data

Routines forup-scaling

Output of Risk Maps to WP4.4

WP4.2

Definition of vulnerableecological receptors in

WP1.1

Figure B.4-3: Flow chart representing the structure of Work Package 4.2.

The temporal dimension of exposure and risk is addressed in a modelling study that focuses on thetemporal variation of human risks for selected inhalative chemical compartments in indoor air basedon data gathered in WP 2.2 (UFZ). For that purpose, principal vector analysis (PVA; Bright et al.1999, Johnson et al. 2002) will be applied. This will result in the identification of characteristicexposure spectra in indoor air as a risk for human health. Combining the exposure with activitypatterns of the inhabitants (e.g., renovation and ventilation; Rehwagen et al. 2003, Schlink et al.2003) a model will be developed for the prediction of long-term risks based on short-termmeasurements.

43

The spatial dimension of exposure and risk is addressed by developing geo-referencedrandom walk models for human and ecological receptors (DESUN). These models will includecritical pathways that are representative for selected stressors and vulnerable environmentalfunctions identified in WP 1.2. In these models, an individual receptor (human or mobile organism)is represented by a set of algorithms that describe the processes relevant for exposure and riskassessment, i.e., movement, dietary composition, food consumption rate, inhalation, migratorybehavior and interactions with other individuals, species and pathogens. The movement algorithmallows the receptor to move over a GIS map, encountering and accumulating different contaminantsand stressors over space and time (Hope 2000, Linkov et al. 2002, Topping & Odderskaer 2004,Rafoss 2003, Woodbury 2003). The receptor thus becomes a spatial and temporal integrator ofstressors. Human and ecological random walk models will be developed along similar lines with anemphasis on ecological models during the first phase of WP 4.2 and on human models during thesecond phase.

Selection of critical ecological pathways and parameters to be included in the random walkmodels will be based on the ecological receptors (identified in WP 1.1 and prioritised by scenarioranking in WP 1.2) on the basis of a vulnerability analysis using multi-criteria analysis (MCA;Faber et al. 2003). This will be in line with recent innovative pilot studies and relevant criticallimits will be derived (De Bruin et al. 1999). The analysis will be further developed to facilitateprobabilistic use of species data in spatial modelling or the development of “virtual species”representing critical target species for modelling on the basis of underlying variability of ecologicaltraits in real species (ALTERRA). Further development of MCA techniques for this purpose will beundertaken in cooperation with the experts of RP 1.

The ecological random walk models will describe exposure and risks at the scale of thehabitat or home range of organisms, and will focus on incorporating spatial variations ofcontaminants in soils and relate these to the habitat use by organisms. Knowledge gathered in theEU INTERREG BERISP project (conditionally approved; development of a decision supportsystem for spatially explicit risk analysis for ecological receptors) will support the development andvalidation of ecological random walk models in WP 4.2 (ALTERRA).

For the development of human random walk models it is crucial to identify the mostimportant causal parameters that determine human exposure and risk. This is no easy task, sincehuman health is affected by a variety of environmental factors including genetics, diet, activitypatterns, and environment-dependent factors such as proximity to hazardous waste sites, air andwater quality, etc. The identification of causal parameters is a research effort in its own right thatwill not only support model development, but will also aid in the development of a non-biasedmethod for identifying at-risk populations and potential hotspots. The methodology will enableevaluation of the effect of spatial scales on observing clusters of risk, cancer mortality or exposure(at predefined ranges). This approach should also provide a framework for developing regionalsurveillance systems.

Analysis of spatial patterns and cause-effect relationships will encompass the use ofcorrelations, regressions, cluster analysis, and artificial neural networks (ANN). ANN have beenused successfully, e.g., to develop predictor variables in the diagnosis of myocardial infarction, fortoxicities of complex mixtures and for modelling and predicting exposure (Tu 1996, Buscema 1997,Gagne et al. 1997, Schlink et al. 2003). This technique is still relatively unexplored in risk analysisand the identification of risk patterns in space and time. Kohonen’s self-organizing maps (Kohonen1982) will be used to facilitate a visual identification of relationships among data and identifypotential exposure and risk hotspots. Fuzzy ARTMAP neural networks will be used to analyzenoisy and incomplete data sets (Espinoza 2001).

To isolate potential predictors (stressors and other relevant causal factors), databaseinformation collected in WP 1.1 will be utilised and converted, as necessary, to GIS compatibledatabases (including census tract information). The GIS environment will be used to analyze scale-related associations between various environmental, social, demographic factors and risk data.Variables such as toxic releases, air quality, demographics, watershed quality, industry

44

distributions, education, income, health coverage, hazardous waste sites, various exposuremeasures, individual activity patterns and other pertinent databases will be analyzed at various scalelevels. Formal statistical tests will be used to assess where exposures and risks are significantlyhigher than average and to examine the strength of variable relationships with the impact measures.Another important issue addressed in WP 4.2 is the aggregation of spatial data in relation to theaccuracy and interpretation of the final spatial output (Haining 1981). Application of Bayesianmodels (UFZ) with conditional autoregressive terms (CAR) will result in interpolated, smoothed,and aggregated risk data (Besag and Kooperberg 1995; Sun et al. 1998). Particular problems thatwill be addressed in WP 4.2 are the specification of the smoothing parameter (Clayton et al. 1993)and the effect of risk attenuation that occurs at aggregated data of a heterogeneous population(Schlink 2002). Besides individual chemical compounds and mixtures, an adjustment will be madefor risks due to further stressors, such as socio-economic factors and climate (Schlink et al. 2002).This activity will be co-ordinated with the studies on mixture toxicity and other stressors in RP 3.

A final issue addressed in WP 4.2 is the consideration of risks at different spatial scales.Data analyses at different scales can determine the relationship between risk clusters and facilitateits geographical identification (Fayyad 1996). For example, in ecological risk assessment, thehabitat of the receptor naturally provides a typical length that can serve as a separator defining smalland large spatial scales (Landis 2003). Properties of ecological risks at different scales will bedescribed and analyzed in order to identify relevant processes that govern risks at different spatialscales and to develop routines for up-scaling (ALTERRA). Also, the analysis of human exposureand risk clusters will be conducted at different spatial scales to determine the relationship betweencluster size and its potential identification relative to geographical scale and the identifying factors.The use of different scales will demonstrate the relative limitations of using large-area scales toidentify impact clusters. In this way, the homogeneity of associations across various geographicscales will be analyzed.

WP 4.3 Dealing with multiple and complex risks in a management context (Leader: TimoAssmuth, SYKE)Risk assessment and management involves the integration of factual assessments and valuejudgments. These overlap and interact causing subjective reasoning also in so-called scientific factsand claims, and challenging the traditional separations between science, 'scientific' assessment andmanagement (see Putnam 2002). Especially when there is uncertainty and ambiguity involved,estimates of cumulative risk become blurred or fuzzy, and the value-fact borderline becomes vague.Examples are scientists that make subjective assumptions about model structures and parameters,and stakeholders that take advantage of uncertainties for their own interests. Communication needsto be improved not only between various stakeholder groups, but also between scientists andassessors of various backgrounds. This requires systematic study of risk perception and cognition,of knowledge-related processes and factors in responses to risks and uncertainties, of views of thequalities and significance of risks, of the multi-actor and multi-level communication in assessmentand management contexts, and of the relationships between science, assessment and managementpolicy. Such studies are important for the development of new assessment and managementstrategies involving e.g. balanced combinations between detailed and simplified approaches,notably those based on the precautionary principle (Pidgeon 1998). In this WP a coherent set ofmental and social aspects of risk will be studied as dictated by the particular contexts and overallfoci and approaches of the project.

The overall aim of this WP 4.3 is to improve the knowledge base for dealing with multipleand complex risks, uncertainties and ambiguities by studying cognitive and knowledge-related,social and contextual aspects of integrated risk assessment, and by providing new interdisciplinary,reflexive and pluralistic approaches to addressing these aspects. Emphasis will be placed on risksassociated with specific multi-stressor activities and on the uses and limits of knowledge inintegration of the precautionary principle with in-depth evidence-based assessments. The WP willin particular address epistemological and policy issues in steering, conducting, developing and

45

evaluating integrated risk assessments in the relevant key areas of EU chemicals policy. Integrationand its relations with differentiation will be investigated along several dimensions, including agents(chemical mixtures, multiple sources, chemicals and other agents), time scales (alsointergenerational) and spatial scales (mainly EU, national and regional level, in interaction with WP4.2), receptors (human and non-human, age groups), endpoints and consequences, other riskattributes (e.g. associated benefits), stages of risk formation (exposure and effect), stages of activity(research or testing – assessment – management – monitoring), management sectors, and actors(experts, regulators, regulated). Given the many dimensions of integration, selected combinationswill be treated more closely in cases.

The key operational objectives and work areas are (1) to study risk perception and cognitionand particularly their patterns and influences in integrative risk assessment, (2) to examine strategicissues in the assessment-management interfaces including precautionary approaches and productionand uses of knowledge to manage associated uncertainties, (3) to analyze expert and stakeholdercommunication about multiple risks and uncertainties, (4) within all these areas, to developmethodologies for multi-dimensional analysis of risks in a management context, and (5) to analyzeand support the dissemination and exploitation of results. These objectives involve both scientificand applied activities. The focus will be on knowledge and its opposites uncertainty, ambiguity andignorance in various domains of activity (research, assessment, policy decisions); this focus willserve to tie the work together internally and, for broad external utility, to direct these activities tothe most meaningful and decisive questions in the various settings given.

Risk perception and cognition related to integrated assessment will be studied among keyactors at European level (SYKE, with partners), based on earlier work (Renn 1998a, Assmuth &Hildén 2002, cf. Carthy et al. 2002). The key general topics include dimensions of risk, roles andkinds of knowledge, modes of thinking and plurality of views (e.g., Funtowicz & Ravetz 1992).Expressed views of risk comparisons will be analyzed, emphasizing multiple stressors and receptorsand taking into account the implications of the above other dimensions and of uncertainty for riskcomparisons (Finkel 1992, 1995). This will be aided by meta-analyses of documented studies andsurveys. The cognitive aspects in the representations and processing of risks will be examined bytheoretical models of risks as socially amplified constructs (e.g., Pidgeon et al. 2003), andempirically by soliciting expert and stakeholder opinions among the consortium and affiliatedexperts and stakeholders including the Competent Authorities and industry responsible forassessment (SYKE). Knowledge about risks and uncertainties will be evaluated in connection withframing issues, extending the value-of-information analyses in WP 4.1 to account for qualitativeand procedural aspects and multidimensionality of risks and for higher-level uncertainties (JRC).The work will be tied to the other WPs to devise conceptual models of risks and inference inassessment, especially to risk and scenario identification and multi-criteria analyses in RP 1.

Risk management strategies that focus on methods for dealing with uncertainty andambiguity (routine, risk-based, precautionary, discourse-based, preventive) will be studied,including analysis of regulatory frameworks and management performance (DIA). Policy issueswill be studied in integration across stressors, receptors, regions and actors (e.g., Renn 2001).Interactions of management and assessment will be analyzed in regulatory procedures for casechemicals (cf. Assmuth et al. 2000), in connection with quantitative environmental risk criteria andgoals, and in environmental health (e.g., Jalonen, accepted) (SYKE). In particular, strategic aspectsin integrating evidence-based and precautionary assessment for case chemical categories will bestudied, including inputs from and back to research, testing and monitoring; this policy-levelanalysis will link with and complement the VOI analyses in WP 4.1. The strategic issues in risk anduncertainty analysis under the REACH system will be an important case (all partners). Thesestudies will also address options for risk prevention e.g. through alternative products and their prosand cons such as counter-veiling risks of alternatives but account also for indirect and processimpacts such as benefits from learning, participation and trust-building. Multi-objective approacheswill be used as traditional risk-benefit analyses are not well suited to deal with multidimensionalityand ambiguity of risks and with multiple goals (e.g., Voulvoulis et al. 2002). Methods and guidance

46

will be developed for addressing multiple risks and uncertainties (JRC), particularly withinscientific advice for integrative chemical policies (cf. Funtowicz et al. 1999, Craye 2003).Risk communication will be studied on a multi-actor and multi-dimensional communicationparadigm in relation to discourses of concepts and approaches in integrative assessment (SYKEwith partners), focusing on inter-disciplinary and expert-stakeholder communication underuncertainty and controversy (cf. Dreyer 1997) and examining links with assessment models (Renn1998b). Communication within risk assessment under EU's new Chemicals Policy and the REACHsystem, related key specific regulations such as the Biocide Directive, and the EU EnvironmentalHealth Strategy development and implementation will be used as cases. Options and obstacles forcross-disciplinary and interactive communication will be identified (cf. Breakwell 2000), includinglanguage and sector barriers. Frameworks for new approaches to communication of risk assessmentwill then be developed (cf. Assmuth 2003) and tested in cases using e.g. dialogue techniques andvisualizations to frame and identify issues (DIA). The latter methodologies will be explored incollaboration with the work on presentation of risks in WP 4.4, complementing this with analyses ofand methods for communication of the social and controversial aspects of risk that will be done inclose collaboration with the applied dissemination and exploitation activities in WP 5.4.

The WP will provide results for dissemination and exploitation processes also by studyingthem and the assessment-management links; efficient exploitation will thus be ensured. The WPwill involve researcher and expert training, and also contribute to training in other ways, e.g., byfocusing on cognition, communication and assessment as learning processes.

WP 4.4 Risk presentation and visualisation (Leader: Joost Lahr, ALTERRA)In NOMIRACLE, risk estimates are produced during different phases of the project (i.e., in WP 1.2,WP 4.1 and WP 4.2). These risk estimates relate to different endpoints (humans, ecosystems,specific species), different spatial scale levels (EU, regional, local), different levels of detail(‘potential cumulative risks’ in WP 1.2 and ‘refined cumulative’ risks in WP 4.2), and differentlevels of accuracy. It will be necessary to integrate and visualise these risks in GeographicInformation Systems (GIS) before they can be presented and communicated to the scientificcommunity, policy makers, stakeholders and the general public. The main objective of WP 4.4therefore is to develop and demonstrate the most appropriate and/or novel techniques forpresentation and visualization of cumulative risks and of environmental and human health riskscombined. The work in WP 4.4 will provide tools to make risk assessment results accessible forfurther dissemination (see WP 5.4). Where WP 4.4 concentrates on the presentation techniques thatmay help to increase the perception of cumulative risks by the end-users, WP 5.4 concentrates onthe most efficient ways (e.g., brochures, internet, workshops, etc.) to communicate these results.WP 4.4 has a highly integrative character throughout the NOMIRACLE project and stronglydepends on the availability and suitability of data and output produced by the consortium membersin other RPs and work packages. The participants in WP 4.4 are all represented in WP 1.1 (databackground) as well.

The work will be divided into two stages. The first stage is to establish ways to produce‘potential cumulative risk maps’, among others for chemical mixtures, based on the type of datagathered in WP 1.1 and the scenario ranking procedure of WP 1.2 (ALTERRA, NERI, UFZ,UNIMIB, URV, JRC). The methods must be suitable to construct GIS-based risk maps for the EUand selected regions (depending on data availability in WP 1.1) that integrate the cumulative risksof exposure to chemicals with a specific mode of action (e.g., pesticides, pharmaceuticals andbiocides) in combination with other relevant stressors (identified in WP 1.2). These initialpresentation and visualization methods will in turn be used as input for WP 4.3, i.e., the examples(maps etc.) produced are evaluated with respect to their suitability to inform key actors andstakeholders and it will be investigated in WP 4.3 in what type of cumulative risks/parameters thesegroups are most interested.

During the second stage of WP 4.4, the ‘potential cumulative risk maps’ and othervisualization products will be updated with information and new scientific insights from the other

47

research pillars and work packages (ALTERRA, NERI, JRC, UFZ, UNIMIB, URV, DESUN). Forexample, the results of WP 2.4 (sound exposure modelling) will be used to update the predictions ofexposure through various environmental media involved. The results of RP 3 (advanced effectassessment) and WP 4.1 (uncertainty factors) will be applied to update effects of cumulativestressors predicted during the initial stage. Uncertainties in the updated risk estimates can bequantified and made visible at the basis of the PRA techniques developed in WP 4.1. Scalingroutines for risks and aggregation methods developed in WP 4.2 can be employed to aggregatespatial data and describe risks at different spatial scale levels. These work packages contribute tomore scientifically sound and accurate predictions of the risks of cumulative stressors. Importantly,the results of the risk perception and communication studies under WP 4.3 will be used to finalisethe presentation and visualization methods that were developed during the first stage of WP 4.4 insuch a way that cumulative risks are made readily perceivable for end-users and decision making isfacilitated. WP 4.4 will also be constantly in touch with WP 5.4 to ensure user-friendlycommunication of results with end-users and other interested parties. The examples of risk mapsproduced during this second stage of WP 4.4 will include actual cumulative risks and risks for alimited number of future scenarios, e.g., dealing with the effects of climate change.

Management - Pillar 5 (Project co-ordinator: Hans Løkke, NERI)

The management activities are organised in a cross-cutting pillar consisting of four work packagesas depicted in Figure B.4-1.

WP 5.1Acronym NERI UFZ NERC DESUNPerson months 31.5 1.5 1.5 1.5

WP 5.2Acronym JRC NERIPerson months 14 1

WP 5.3Acronym APINI DBUA DESUNPerson months 8 1 0.5Acronym DIA ENVI JRCPerson months 0.5 2.5 2Acronym NERI SYKE UFZPerson months 3 1.5 1

All university partnerscontribute with at least 0.5month training activity

WP 5.4Acronym SYMLOG NERI

Person months 7.5 3

All other partners participate in this activitywith input for scientific and populardissemination

WP 5.1 General Project Management (Leader: Hans Løkke, NERI)

This activity is described in detail in section B.6. The WP 5.1 includes information on the activitiesof the Project Co-ordinator, the Management Board, the Project Secretariat, the Advisory Board, themanagement at Research Pillar and at WP level, and the General Assembly. It contains a plan formanagement of knowledge, of intellectual property and of other innovation-related activities arisingin the project.

48

WP 5.2 Data Management (Leader: Arwyn Jones, JRC)

Aims and rationaleEnvironmental data are often irreplaceable; they are nearly always unique, if only in the timing ofcollection. They can also be extremely expensive to collect and are usually of interest to a numberof research groups. For these reasons the NOMIRACLE management team attaches greatimportance to ensuring that maximum benefits are derived from the data sets that will be acquiredduring the course of this project. The commitment of the NOMIRACLE management team to thecollation of data in retrieval formats recognises that scientists producing high quality work oftenconsider the end point of their research to be the production of publications, whether in openliterature or in the form of policy advice. In a truly collaborative and cooperative project such asNOMIRACLE, the retention of data by partners could represent a barrier to the integration of theefforts of the research teams involved. In NOMIRACLE much of the innovation is likely to be adirect result of this integration. For this reason it is vital that all research teams involved have theability to exchange data files in a format that is interpretable to all parties. Thus, while data willindeed be manipulated by the researcher to provide material for publication, these data should beseen as a valuable resource in their own right. Properly managed and preserved, they can providenot only an ability to share information between partners, but also a resource that can be used byfuture researchers, and exploited commercially or educationally. Such further uses, often neverenvisaged in the first instance, will make an additional contribution to NOMIRACLES objectives.

For these reasons, WP 5.2 will provide the NOMIRACLE projects with the mechanisms andtools to provide distributed access to data holdings and information services for participatinginstitutions and organizations. An extension of the goal to provide efficient access to data forinternal use is the dissemination of products to the wider community (e.g. policy makers, generalpublic, industry, etc.). WP 5.2 will also address this issue through collaboration with the WPs onTraining and Dissemination (WP 5.3 and WP 5.4).

Establishment of data management systemThe objective of the NOMIRACLE Data Management system is to bring together knowledge fromthe different research pillars of the NOMIRACLE project (i.e. data, information and services fromall WPs) and provide access to this knowledge base through a single, user friendly interface. Giventhe complexity and variety of data formats required to implement the NOMIRACLE project,responsibility for the management of core data sets and outputs will be given to the individual workpackages. For this reason, NOMIRACLE will not operate a central data repository that stores alldata used by the project. Rather, NOMIRACLE will function as a series of linked nodes, each nodehosting the relevant data for which they are responsible. However, NOMIRACLE has an obligationto exchange information and data between WPs and, to an extent, the wider community. Tofacilitate the dissemination of information and to ensure that the deliverables of each work packageare available to the partners of the project, a distributed data management system will be developedas a tool to manage the recording of information and as a mechanism to facilitate access.

To this end, the NOMIRACLE Data Management System will operate a directory systemthat will locate data and resources, which will provide direct access to data and indicate issuesrelating to the quality of the data.

In summary, this activity will aim to gather the necessary information on the nature and typeof core data sets utilised and deliverables generated by the NOMIRACLE project in order to design,develop, test and implement an internet based, platform independent, meta-data / cataloguingsystem that will be used to record and allow access to NOMIRACLE data sources.

WP 5.2 will commence by making an inventory of the data requirements and deliverables ofthe NOMIRACLE project. This inventory will be carried out initially via a survey of the projectdeliverables and a questionnaire that will be distributed among the NOMIRACLE partners. Thequestionnaire will request partners to indicate data sources, data type, availability, etc. at the

49

participating institutes. Partners will be asked to indicate which data are official deliverables of theproject and are used as inputs by other WPs.

Following this information gathering exercise, a draft data management system will bedesigned based on clearly defined needs of the users and project managers. It is proposed that thedata management system will be operated as an on-line system based on current Internet protocolsand meta-data standards. A draft design plan will be circulated among the project participants forcomments and suggestions. The finalised design concept will then be developed as prototypesystem and requested functionality demonstrated to users.

Within the data management system, a number of key issues can be pre-defined. The systemwill provide an Internet based, platform independent user-friendly interface to the informationstored by the NOMIRACLE Consortium. To satisfy these requirements, it is obvious that a Web-oriented solution is the most fitting. This means that all operations of the data management systemwill be effected through a standard web browser.

An appropriate meta-data or cataloguing template will be defined to record the necessaryinformation on the data. The precise structure and content of the catalogue will be defined followingthe results of the questionnaire. A meta-data gateway will be developed to search the catalogue forinformation on data held at various NOMIRACLE partner sites. The outcome of the search will be alist of relevant data sets as links in a HTML/XML format. Clicking on any of the links will take theuser to the relevant HTML view of the data set record in the catalogue. A query of the DataManagement System will allow searches on any text in a catalogue record, and set bounds on thetime and space coverage of the dataset.

Where required, the data management user interface will provide access to NOMIRACLEdeliverables sources. The format of export data sets will be defined at the outset of the project. It isforeseen that File Transfer Protocol (FTP) will be used by users to download data. As part of thedesign document, the necessity of operating a NOMIRACLE FTP server will be investigated.

A crucial aspect of any data management system is the traceability and reliability of theinformation being stored in the scheme. The NOMIRACLE data management system will provideusers with tools for recording the history of logged data sets and for identifying issues relating tothe quality of the data from downstream users (i.e. downstream use of data by the various WP ofNOMIRACLE project).

The implementation phase of the NOMIRACLE Data Management system is foreseen in threephases:

1. Following the acceptance of the design document, a prototype system will be developed underphase 1 to test the required functionality and operation. The prototype will be demonstrated tothe NOMIRACLE Consortium for feedback and possible modifications.

2. An internal testing phase prior to operational launch of the system3. A full implementation of the system; this phase will be carried out with a view of launching an

operational data management system for NOMIRACLE by month 10 of the project.

Risks to the successful execution of the implementation plan arise from the lack of timely responsefrom the Consortium for data management requirements.

On completion of the operational implementation of the Data Management System, WP 5.2will co-ordinate the data input of the catalogue records and maintain the data management system inline with the progress of the project.

The homepage of NOMIRACLE Web site will have a link to the Data Management Systeminterface. In addition, a seamless link will connect the data management interface to theNOMIRACLE homepage at all times.

To support users, WP 5.2 will maintain a NOMIRACLE Data Management System HelpDesk to assist users of the system.

After 6 months of operation, a review of the usage and functionality of the NOMIRACLEData Management System will be carried out. The opinion of WP leaders and NOMIRACLE

50

project managers will be canvassed. Any revisions to the Data Management System can then beimplemented.

The work package will continue for the duration of the NOMIRACLE project.

WP 5.3 Training and Demonstration (Leader: Amadeu Soares, DBUA )

These activities are described in detail below in sections B.4.2 and B.4.3.

WP 5.4 Dissemination and Exploitation (Leader: Claire Mays, SYMLOG)

NOMIRACLE will produce a great sum of information that may be disseminated to and exploitedby a variety of end-users. Target groups are:• The scientific community• Practising professionals (consultants, risk and uncertainty analysts, risk and safety managers and

planners) in academia and/or in industry, including SMEs• Regulators and policy makers• NGOs and networks• The general public.

WP 5.4 targets quality control of dissemination products and user-friendly communication.

The objectives of WP 5.4 are to:• Contribute to the internal and external dissemination plans• Co-ordinate or perform their implementation.

At all times, an end-user focus is preferred. Feedback to and among R&D WP is organised.

The Dissemination and Exploitation WP 5.4 assure five major functions:

1. Develop and maintain a calendar of expected dissemination products2. Contribute to integration of parallel R&D outputs across the IP (e.g., through a restricted access

web site); compile and input feedback from end-users.

3. Co-ordinate internal peer review of dissemination products.

4. Co-ordinate the dissemination of R&D outputs (data, models, reports, articles, conferencepresentations, PhD theses; novel laboratory, field and office methods for science-basedcumulative risk assessment and management).

5. Ensure the accessibility of NOMIRACLE outputs through multiple media (electronic, print,including popular science outlets) and in regard to the future "mechanism for data sharing,improved data availability, accessibility, comparability and enhanced exchange of information"currently under development by a WG under SCALE "European Environment & HealthStrategy"; c.f.; Brussels Consultative Forum, 12/2003).

The leader of WP 5.4 (SYMLOG) has responsibility for developing procedures and templates forreporting outputs and inputs of various types, and for co-ordinating their use. She acts as a publicaffairs officer for the IP, centralising inquiries and feedback from end-users that may flow into theproject at various access points, and directing information outwards. (This will often imply interfacewith WP 5.1 and WP 5.2 concerned with internet service.) The WP leader also directs attention tospecial communication needs, compiling guidance regarding end-user sub-populationsdistinguished by sector, region, language, gender, vulnerability, etc. The actual production ofdissemination materials is the responsibility of participating scientists. Partner institutes in WP 5.4contribute their in-house human and production resources for publications, graphics, etc. The WP

51

leader consults to these scientific and technical personnel in particular regarding disseminationproducts targeting the general public.

Dissemination and exploitation imply critical interface with several other WPs, in particularwithin RP 4 (Risk assessment) and the WP 5.3 on Training and Demonstration Activities. Theinterface may be characterised in general as reciprocal and recursive feedback relationships. Theseare briefly outlined below:

WP 4.3 Dealing with multiple and complex risks in a management contextInputs to WP 5.4 will include:• Empirical information on pertinent dimensions of public risk perception that may impact on the

characterisation of special end-user communication needs (e.g., gender and/or role sensitivity torisk, as in the case of the public sub-populations of women, or mothers of young children; thesegroups are salient in the context of SCALE focus on children's health which is to be addressedby NOMIRACLE RP 4)

• Identification of critical actors in the national and international risk management systems, towhom dissemination should be directed

• Guidance for avoiding or eliminating obstacles to communication linked to the complex cross-disciplinary nature of the risk assessment field.

In turn, WP 5.4 outputs to WP 4.3 will include (from Month 7, subsequent to publication of e.g.,first Newsletter in Month 6):• Identification of stakeholder categories expressing an active demand for information• Compilation of stakeholder demands on dissemination, providing insight on risk perceptions

and discourse on cumulative risk assessment• Compilation of user feedback on the uncertainty aspects of disseminated materials.

WP 4.4 Risk presentations and visualisationInputs to WP 5.4 include:• Techniques for presentation and visualisation of cumulative risk• Information on non-spatial cumulative risk, i.e., identification of vulnerable sub-populations

determined by activity patterns and socio-economic factors, to aid in the characterisation ofspecial end-user communication needs.

In turn, WP 5.4 outputs to WP 4.3 include (from Month 7, subsequent to publication of e.g., firstNewsletter in Month 6, or exploitation of risk maps from Month 18):• Compilation of user feedback on the presentation and visualisation techniques.

WP 5.3 Training and Demonstration ActivitiesA reciprocal link is maintained regarding user demand and feedback. The Dissemination WP alsoserves as a vector for announcement of the demonstration activities and training courses. WP 5.4offers strong support for the final IP dissemination workshop, based on user feedback gatheredthroughout the project.

The interfaces above highlight the end-user focus of the Dissemination and Exploitationwork package, and in particular the attention accorded to institutional and public stakeholders. WP5.4 does not aim to create a pan-European clearinghouse for risk assessment information directed atall potential users. The work package has the specific role of disseminating NOMIRACLE outputs(taking the form of scientific risk assessment, and, in some cases, uncertainty information andguidance for interpretation when feasible.) However, by attentively compiling user feedback to themultiple contents and formats of dissemination, WP 5.4 builds experience in serving the userssituated outside the traditional R&D target communities of scientists and professionals.

In this light, the early tasks of WP 5.4 will include development of:

52

1. IP internal (including peer review) and external reporting procedures, based ono an informal survey of partner institutions to identify their dissemination and

exploitation experience and guidance, ando "European Environment and Health Strategy" WG guidelines when available

2. Templates in the specific goal of integrating user feedback, e.g.,o Web-based internal reporting by all RP of user needs, perceptions, or feedback

encountered in the course of§ NOMIRACLE R&D activities§ demonstration activities, training, and dissemination

o Web-based input by§ visitors to NOMIRACLE public website§ users of other disseminated material§ participants in demonstration activities or training

3. Procedures to analyse and redirect all such reports to the WPs in which they could be useful

In later phases of the project, the following tasks will be performed, through a typical public affairsapproach (informal interviews, observations, and contacts) as well as through centralisation andanalysis of observations received from other parts of the project, reported by the proceduresmentioned above:

4. Stakeholder mapping, focusing on key actors and representatives of key civil society groups,including emergent ones

5. Identification of issues and needs in user groups through e.g., consultation with theNOMIRACLE Advisory Board of Stakeholders

6. Identification of relevant dissemination outlets including conferences and meetings;development of a staged strategy for deciding when to invest resources in attending suchmeetings.

Recall that IP R&D scientists have direct responsibility for developing dissemination products. WP5.4 will assure specifically:

7. Newsletter publication (from Month 6)8. Web site postings (in co-ordination with WP 5.1 and WP 5.2)

B.4.2 Demonstration activities

In the last phase (year 5) of the project the novel risk assessment framework evolved during theNOMIRACLE project will be demonstrated at an international workshop arranged for all relevantstake holders: National and governmental EPAs, Industry, Academia and NGOs. Case studies willbe demonstrated to illustrate how the new NOMIRACLE risk assessment toolbox can be exploited.This demonstration will show the integrated results of the project from the production of “potentialrisk identification maps” and the scenario ranking procedure to the final modelling of risks and riskvisualisation techniques.

B.4.3 Training activities

Training is considered a vital activity in the spreading of excellence and implementation ofNOMIRACLE aims. In addition to the several training contributions and activities that are madewithin each particular workpackage, the primary importance that the NOMIRACLE Consortiumconfers to the training and exchange of scientific personnel and students is demonstrated by the

53

organization of a special and dedicated work package (WP 5.3, lead by Amadeu Soares, DBUA).The consortia provides an excellent vehicle for the advanced training of researchers (includingpostgraduate students -both MSc and PhD- and post-doctorals) and other key staff such as researchmanagers, industrial executives and potential users. Accordingly, considerable efforts have beendevoted to designing training courses that will offer targeted education programmes about topicalresearch areas of the project (see below), and thus providing an excellent opportunity andcontributing to the professional development of the persons concerned. Scientists and studentsinvolved in this project will have the opportunity to be trained in various methodologies used bydifferent researchers that would not otherwise be possible in a smaller project. Staff exchanges willbe planned during the kick-off meeting, with the likely focus for their activities being the variousrequirements for data and analytical standardization, technical support, student guidance, analysisand the dissemination of results. In addition, the Consortium has expertise in the organization oftraining activities orientated to research managers, industrial executives (via the participation ofSMEs) and other potential end-users (e.g. national/regional regulators, consultants).

Kick-off workshop and courses for partnersThe increase of the interdisciplinarity between different scientific communities/disciplines is ofutmost importance in a large Integrated Project such as NOMIRACLE where there is a risk thatcollaboration between scientists of different scientific background fail due to unawareness of theproblems, aims and scope of other disciplines. In order to prevent such a development we willarrange a number of intensive courses in which participants will give lectures and arrange seminarsfor other non-peer project participants. For example, the ecotoxicologists involved in RP3 willarrange short introductory courses for participants with other scientific backgrounds (e.g.environmental chemistry, risk modelling etc.). These courses will be compulsory for allparticipants, and are to be held at the NOMIRACLE kick-off workshop. These courses have twomain purposes. Firstly, this training will increase each participant’s insight into the overall projectand putting into perspective how his/her own research can supplement and interact with otherresearch activities of the project. Thus, increasing each participant’s understanding ofNOMIRACLE’s scientific disciplines will help in streamlining the concerted scientific efforts.Secondly, these courses are likely to have social value, increasing the interactions betweenscientists and securing integration of the project. It is well known that stimulating discussions aremuch more likely to take place when people meet and come to know each other. This has highpriority in NOMIRACLE, and it is important to have this training activity placed as early aspossible in the project period.

NOMIRACLE Postgraduate SchoolSeveral consortium partners have extensive experience in running training courses at both post-graduate level and for the implementation of new methods and tools in a regulatory framework,involving policy-makers. Regarding formal European education at the University level, it isanticipated that the Bologna Agreement for undergraduate education will result in theimplementation of a postgraduate education that will start at the Master level. These Masters willincorporate research and will be the door towards a doctoral degree. To be successful they will haveto involve at least three European Universities since this will open the door to the Erasmus MundusEU program. Thus, the success of postgraduate research and education will be founded onsuccessful Master programs, i.e., programs that get support for students, share courses, fostermobility of students and professors, have EU visibility, etc. Within the partnership ofNOMIRACLE an EU Master programme will be created, leading towards a doctoral degree,forming the NOMIRACLE Postgraduate School. In addition to the seed funding secured throughthis IP, the NOMIRACLE Consortium will actively seek funds from elsewhere, using appropriatealternative EU 6thFP instruments (e.g. the Marie Curie Program, the ERASMUS Mundus Program,the Alban Program) and National Post-graduate Grant Programs, to encourage postgraduatestudents and post-doctoral scientists to be involved and collaborate in the NOMIRACLE activities.

54

In fact, one of the aims of the external funding is to open the NOMIRACLE training network to awider and international audience, i.e. other interested research teams from universities and researchcenters that do not participate in the NOMIRACLE proposal or from NOMIRACLE partnerinstitutions that participate in other approved IPs, strengthening the NOMIRACLE training networkand assuring also in this way the success of our integrated work at all levels within and beyondNOMIRACLE.

One of the activities will be through a collaborative association by integrating the expertisefrom various disciplines into a network to establish a program for the training of specialists at anearly career stage, in the form of an international PhD school. It is anticipated that at least ca. 20-30postgraduate students will be involved and educated within the project, in the first phase ofNOMIRACLE. These postgraduate students will be enrolled in the Master and PhD schools at theirrespective universities. However, the NOMIRACLE Postgraduate School will add to these nationalschools with cutting edge science generated during the project period. To implement theNOMIRACLE graduate school activities as presented above, a series of short term advanced studiescourses will be organised during the initial phase of NOMIRACLE, and widely announced to thescientific community in Europe and elsewhere. The participation of females in these courses will beencouraged. Examples of some topics for training courses that may be organised by theNOMIRACLE Consortium are presented in Table B.4-2 below. From those, at least 6 will beoffered during the first 18 months of the project. In addition, letters of intend/memorandums ofagreement will be signed by several University partners of NOMIRACLE up to month 6 of theproject, so that the NOMIRACLE Postgraduate school can be implemented, through common M.Scand PhD programmes, and announced by month 17, at the latest.

Other trainingSummer Schools and workshops will be organized by NOMIRACLE. Summer schools will bededicated to different research areas including state-of-the-art scientific questions and technologies.The Summer School will typically last for one week of intensive training with a seminar programfrom invited speakers (who will be asked to stay for the whole time) as well as seminars from thetrainees. Tutorials in small groups will take place in which the trainees are actively involved andhave to contribute. Potential topics for the Summer Schools are also indicated on table B.4-2 below.At least two Summer Schools will be organized during the first 18 months of the project.

The several workshops that will be organized during this IP, some of which will have theparticipation of several stakeholders (i.e. policy-makers, governmental officials, policy-makers,NGOs and other relevant stakeholders) are also an excellent way of promoting the concepts andtools developed in NOMIRACLE. For example, targeted technical workshops, designed to provideinformation exchange and to promote collaboration between existing EU members and newmembers/candidates will be organized. This, in particular the activities targeted to Eastern Europeancountries, will be closely coordinated with the current on-going actions of JRC, to avoid anyunnecessary duplication of efforts.

A training program for authorities and industry of Eastern European countries will bedeveloped on the basis of the specific needs of those countries concerned. This task will beundertaken by APINI (Lithuania). Whilst APINI will organize and actively take part in the training,different project members will be giving lectures at these courses. One or two-day training courseswill be organized in several rounds, but primarily in the later phases of the project. Free materialsused for training will be made available to participants, who in their turn then could furtherdisseminate the knowledge in their respective countries.

ResourcesIn addition to the internal resources that some of the partners dedicate to their training mission andongoing training activities, and to the efforts that the Consortium will dedicate to the seek ofexternal funding (see above), the initial NOMIRACLE training activities will be supported in partby a grant from the IP budget. This funding may be used to meet host institution costs, travel and

55

subsistence costs. A proportion of places on every course and training activity will be reserved forparticipants outside the Consortium, including from Developing countries and Eastern European /new accession countries. The criteria for selection of participants for the several training activitieswill be drawn up by the Management Board, but they will focus around the needs and aims ofNOMIRACLE. The set of criteria, application forms and other relevant information will beconveniently announced on the website, through all participating partners own networks,international scientific societies, etc, assuring the widest dissemination possible. For example, linkswill be sought between NOMIRACLE and current developments in the FP6 Marie Curie Actionsand their various support types, plus the new ERA-MORE Researcher Mobility Service Network tobe launched in May 2004, including the concrete mechanism of the "Researcher's mobility portal"where researchers and organizations may register and announce (also with a map-basedfunctionality) their training activities. Positive action will be taken to assure a proper genderbalance in all training activities.

Table B.4-2: Possible topics for Training Courses and Summer Schools organised by theNOMIRACLE Consortium for targeted groups inside and outside the project. At least six TrainingCourses and two Summer Schools will be offered during the first 18 months of the project. The listof potential responsible partners is non-exclusive, and only major responsible partners are listed forthe easiness of reading.

Topic Target Groups Potential ResponsiblePartners

Multicriteria ranking fordecision making

Risk managers (SME, industry,NGO),risk regulators (government)

NERI, DIA

Environmental analytics PhD students, SME CSIC, ITMAvailable exposuredetermination

PhD students, young researchers,experienced researchers

DBUA, ULANC

Computational toxicology PhD students, young researchers,governmental regulators, industrialexecutives

LMC, UFZ

Multimedia fate modelling PhD students, young researchers,risk assessors (SME, government)

DESUN, RIVM, URV

Mixture toxicity PhD students, young researchers,experienced researchers

WU, VU

Applied molecular biology(geneomics, proteomics,metabolomics)

PhD students, young researchers,experienced researchers

DSTA, UFZ

Probabilistic risk modelling PhD students, young researchers,Risk assessors (SME),risk regulators (government), riskmanagers (industry)

DESUN, URV

Indoor/outdoor relationshipsof exposure

PhD students, young researchers,experienced researchers

UFZ, USALZ

Superposition of chemical andnon-chemical stress

PhD students, young researchers,risk assessors (SME, industry,government)

NERI, UJAG

Cumulative risk management Risk assessors (SME),risk regulators (government), riskmanagers (industry),PhD students

SYKE, DIA

56

B.4.4 Management activities

Management activities are organised in Pillar 5 consisting of four work packages as shown insection 4.1 above. Work package 5.1, General Project Management, is described in detail insection B.6 (description of project management) and therefore not treated here.

B.5 Description of the Consortium

An overview of the partners in NOMIRACLE is shown at the front of this research proposal. Thecomposition of the Consortium reflects the need of research skills to accomplish the goals ofNOMIRACLE. Thus, we needed excellent partners within the fields of, database management,geographic informatics, scenario development and analysis, environmental chemistry, QSAR,ecotoxicology, human toxicology and epidemiology, risk assessment, probabilistic risk assessment,mathematical modelling, risk management, risk communication, and socio-economics.

Factors determining the choice of partners have included first of all excellence in theparticular field of research, but also regional European considerations, stakeholder interests, and thesecuring of complementary research skills have played an important role. Thus, the Consortium hasparticipation from 18 EU and new accession countries. Partners are covering the range ofdisciplines and interests needed to fulfil the aims without any significant overlap of research fieldsor tasks they are performing in the project (see detailed descriptions of partner capabilities inAnnex 2). However, many of the partners involved have a mutual history of productivecollaboration, which will help secure a fruitful co-operation in NOMIRACLE.

High level of commitmentThe research of NOMIRACLE is within the focal point of the research priorities of the partnerswhose capabilities are described in Annex 2. There will therefore be a high level of commitment notthe least due to the high priority of genuine science in the project. All partners have the necessaryequipment and research facilities available for carrying out the described science. We also expect tohave a high number of Master’s students linked to the various research projects because theresearch will be at the cutting edge, and therefore well suited for numerous student projects. Theproject will also benefit from these satellite projects.

SME participationSix of the 38 partners are SMEs showing the high involvement of industry in the project. TheseSMEs will carry out tasks, for which they are especially well qualified, or where they can process ahigh number of analyses, screening tests etc. The SMEs of NOMIRACLE are also devoted todevelopment of methods, and thus will almost certainly benefit from the project in terms ofeconomic and technological developments.

Experience and role of partners

NOMIRACLE Risk scenarios Pillar (RP 1)

Partner 1 – Head of Risk Scenarios Pillar - National Environmental Research Institute(NERI)The focus of the research is mathematical models and predictions under uncertainty. NERI isworking with a wide range of applied mathematics in order to handle exposure (fate) modelling,ranking under complex circumstances and data interpretation. NERI is developing databases for theDanish EPA and stochastic models, which can gain benefit from large data series.

57

Partner 18 – Directorate General Joint Research Centre (JRC)As a Directorate General of the European Commission, the EC-JRC functions as a reference centreof science and technology for the Union. JRC has developed over the years skills and tools toprovide neutral and Europe-wide expertise in the environmental field. The IES has considerableexpertise in the study of the mechanisms by which contaminants are released into the environment,their transformation and impact on environment and climate. In particular, the IES-Soil and WasteUnit carries out research in support of EU policies on the impact of pollutant emissions and wastes.

Partner 20 – Kaunas University of Technology (APINI)The Institute is involved in research, purpose oriented training and technical assistance toenterprises, and education. Among the research areas of the Institute are the following: chemicalscontrol and management, environmental impact assessment, environmental efficiency, modellingand management of surface water quality, simulation of various pollution load scenarios,development and implementation of the concepts and techniques of pollution prevention/ wasteminimisation/ cleaner production.

Partner 4 – University of Nijmegen (DESUN)The Department of Environmental Studies (DESUN) of the Faculty of Science, Mathematics andComputing Science co-ordinates a Bachelor and Master Programme in Environmental Science andperforms scientific research within the domain of integrated environmental science. The mainresearch expertises are ecological restoration of river basins and modelling of environmental risksposed by chemical contaminants and other stressors.

Partner 21 – Alterra (ALTERRA)Alterra engages in strategic and applied research to support design processes, policymaking andmanagement at the local, national and international level. This includes not only innovative,interdisciplinary research on complex problems relating to rural areas, but also the production ofreadily applicable knowledge and expertise enabling rapid and adequate solutions to practicalproblems.

Partner 28 – Environment Park SPA (ENVI)EnviMod, thanks to sophisticated and powerful software tools and to expertise, aims at providingenvironmental consultancy in different thematic areas such as Human Health Risk Assessment,Ecological Risk Assessment and environmental modelling.

Partner 27 – Universita di Milano Bicocca (UNIMIB)The Ecotoxicological Group at UNIMIB has a long experience in field monitoring and modelling ofenvironmental distribution and fate of organic chemicals, environmental risk assessment, aquatictoxicology for single or mixtures of environmental pollutants. The group is well equipped withfacilities for analytical work and modelling.

Partner 19 – Finnish Environment Institute (SYKE)SYKE is the leading institute within the environmental administration of Finland. The key tasks ofSYKE include multidisciplinary research in environmental changes and their causes and insolutions to environmental problems, and environmental monitoring and assessment.

Partner 2 – Natural Environment Research Council (NERC)The institute encompasses expertise in ecology, hydrology, virology and environmentalmicrobiology. CEH is part of the Natural Environment Research Council (NERC), which is theUK’s leading body for research, survey, monitoring and training in the environmental sciences.CEH has an international reputation for scientific excellence and high quality research and provides

58

a combination of research capabilities directed towards understanding and managing terrestrial andfreshwater environments within the UK and overseas.

NOMIRACLE Sound exposure Pillar (RP 2)

Partner 3 – Head of Sound Exposure Pillar - UFZ Centre for Environmental Research (UFZ)UFZ is a leading institution in environmental toxicology and environmental health research.Expertise in statistical modelling, time-series modelling and epidemiological field studies.

Partner 22 – Swiss Institute of Environmental Science and Technology (EAWAG)EAWAG has multidisciplinary teams of specialists in the fields of Environmental Engineering,Natural and Social Sciences jointly develop solutions to environmental problems. At EAWAGpioneering work on the analysis of organic and inorganic pollutants as well as on their transport andtransformation mechanisms has been done.

Partner 1 – National Environmental Research Institute (NERI)Department of Environmental Chemistry and Microbiology is the leading Danish institution withinthe field of environmental chemistry. Expertise in implementation of new partitioning methodswithin environmental monitoring, with emphasis on the availability of organic contaminants.

Partner 31 – Stockholm University (ITM)ITM has excellence in the following disciplines: atmospheric chemistry and physics, biochemistry,biogeochemistry, environmental organic and inorganic chemistry. Broad experience in studyingenvironmental fate and bioaccumulation phenomena in the laboratory and in the field, and insynthesising the resulting knowledge into algorithms for incorporation into risk assessment modellingtools.

Partner 30 – Lancaster University (ULANC)The Environmental Science Department of the University of Lancaster is internationally recognisedfor its research into the physical and chemical nature of the aquatic, atmospheric and terrestrialenvironments. Expertise in measuring in-situ fluxes of metals in sediments and soils mimickingplant uptake.

Partner 36 – Consejo Superior de Investigaciones Cientificas (CSIC)CSIC is a leading European institution in environmental chemistry. The LC/MS and LC/MS/MSintegrated systems are the main tools of the research group. The environmental chemistry group atCSIC is one of the few groups in Europe with demonstrated analytical expertise for newcompounds.

Partner 35 – Bourgas “Prof. As. Zlatarov” University (LMC)The Department of Physical Chemistry at LMC is a leading institution in Europe within physicalchemistry. The work is devoted to developing and applying quantitative structure-activity methodsaccounting for molecular flexibility and metabolic activation of chemicals.

Partner 26 – ECT Oekotoxikologie GmbH (ECT)ECT has expertise in the development of new ecotoxicological testing methods under laboratoryand field conditions. These include bio-accumulation and toxicity in sediment dwelling organisms,biodegradation of chemicals in surface water and sediment, and determination of fate and effects ofchemicals in terrestrial and aquatic mesocosms.

59

Partner 33 – Universitat Rovira i Virgili (URV)The Department of Chemical Engineering at the University Rovira i Virgili (URV) have expertisewithin the areas of chemical and environmental engineering, chemistry, fluid mechanics-CFD,computer sciences and applied mathematics. The research focuses on the development of patternrecognition and feature extraction algorithms, and the study of structure-activity and structure-property relationships in chemical, biochemical and environmental systems.

Partner 34 – LHASA Ltd (LHASA)LHASA Ltd is the leading supplier of software for the prediction of mammalian toxicity. Itstoxicology program, DEREK for Windows, was originally developed in collaboration with ImperialCancer Research Fund. LHASA staff has also developed the metabolism program, METEOR,which is integrated with DEREK.

Partner 4 – University of Nijmegen (DESUN)The Department of Environmental Studies (DESUN) has excellence within the domain of integratedenvironmental science. The main research topics are ecological restoration of river basins andmodelling of environmental risks posed by chemical contaminants and other stressors.

Partner 23 – National Institute of Public Health and the Environment (RIVM)The National Institute for Public Health and the Environment (RIVM) functions as a center ofscientific expertise for the Dutch Ministry of the Environment on risk assessment of chemicals.RIVM has a long-standing expertise in the field of modelling for risk assessment of chemicalsubstances, as illustrated by e.g. a major role of RIVM in the development of the European UnionSystem for the Evaluation of Substances (EUSES) and the RIVM-contributions to EU-TechnicalGuidance Documents.

NOMIRACLE Effect assessment Pillar (RP 3)

Partner 2 – Head of Effects assessment Pillar - Natural Environment Research Council(NERC)The institute encompasses expertise in ecology, hydrology, virology and environmentalmicrobiology. CEH is part of the Natural Environment Research Council (NERC), which is theUK’s leading body for research, survey, monitoring and training in the environmental sciences.CEH has an international reputation for scientific excellence and high quality research and providesa combination of research capabilities directed towards understanding and managing terrestrial andfreshwater environments within the UK and overseas.

Partner 3 – UFZ Centre for Environmental Research (UFZ)UFZ is a leading institution in environmental toxicology and environmental health research.Research at UFZ departments is aiming at the derivation of mechanistically-based effect markersfor description of multiple exposures relevant to human and ecosystem health. The departmentsinvolved have extensive experiences on the fields of cell toxicology, ecotoxicology, immunologyand human health studies.

Partner 6 – Vrije Universiteit Amsterdam (VU)In the Ecotoxicology section of the Department of Animal Ecology research is focused on theanalysis of the response of soil invertebrates to toxicants in soil. Research involves the investigationof bioavailability and uptake routes, internal distribution and effects of toxicants in these organismsand the relation between internal concentrations and effects. Also mechanisms of adaptation totoxicant stress are studied in soil invertebrates, and the use of molecular biology tools (genomics)has recently been introduced in this research.

60

Partner 1 – National Environmental Research Institute (NERI)Department of Terrestrial Ecology, NERI, has expertise in soil ecotoxicology and ecophysiology ofsoil invertebrates. NERI staff has investigated the effects of simultaneous exposure to climaticstress (extreme temperatures and desiccation) and toxic stress, and was the first to demonstrate thatclimatic stressors can interact synergistically with chemicals.

Partner 5 – University Piemonte Orientale (DSTA)At DSTA biological/ecological studies have been developed to clarify the biological effects ofpollutants at molecular, cellular, organismic and population/community levels both in animals andplants. New research tools in the Ecological Risk Assessment, such as the use of a battery ofbiomarkers have been developed. In addition new tools for the study of biological stress induced bypollutant exposure, such as the use of DNA microarrays and the analysis of the protein content(Proteoma) in the organism, have been developed in the proteomic laboratory

Partner 12 – Wageningen University (WAU)Laboratory of Nematology, Wageningen University, The Netherlands has a longstanding record ofstudying multiple stress effects on the molecular as well as the individual and population level innematodes. Excellent facilities are available to study the impact of stressors on the DNA expressionlevel, the protein level and the partner is an expert in relating these stress-induced changes topopulation level effects.

Partner 13 – University of Aveiro (DBUA)The Ecotoxicology group at DBUA has expertise in the area of aquatic ecotoxicology and ecology.The work is focused on the assessment of the quality of water bodies in Portugal, both structurally(diatoms and macroinvertebrates) and in terms of functioning (testing and development of a suite ofgeneric in situ tools to assess ecological quality of freshwaters).

Partner 11 – University of Tübingen (EKUT)The main expertise of the Animal Physiological Ecology Section of Tübingen University is therelationship of biochemical, molecular, cytological, histological, physiological, and developmentalresponses (biomarkers) to xenobiotics, in invertebrates and fish, and corresponding effects at higherbiological levels. The section is fully equipped to amplify and characterise genes and analyzebiochemical responses to chemicals at the protein or mRNA level, predominantly of the hsp system(hsp70, hsp90), in vertebrates as well as in numerous invertebrate taxa.

Partner 7 – National Institute for Public Health (NIPH)Department of Toxicological Analysis is concentrating on research and development of analyticalmethods for air pollutants, new biological exposure tests, alternative expression methods for acutetoxicity determination, certificated reference materials from human urine and predictive toxicology,involving methods of mathematical statistics for biological processes and techniques such asQSAR.

Partner 17 – Salzburg University (USALZ)A strong focus of USALZ is immune modulation by environmental pollutants. Present researchaims at detecting pollutant-induced modulations in the expression of immunoregulatory cytokines,clarifying underlying molecular mechanisms on the level of promoter regulation, and setting upcell-based assays for large-scale studies. The biochemical section at the department is well equippedfor all relevant methods in cell culture, biochemistry and immunology. An animal care facility is onsite and mouse models of allergic and asthmatic diseases are routinely studied. All animalexperiments are subject to review and approval by national institutions.

61

Partner 25 – Aachen University of Technology (RWTHA)The Department of Ecology, Ecotoxicology and Ecochemistry belongs to the Aachen University(RWTH) in Germany. Research of the department is intensively concerned with development andapplication of biotesting systems. Especially aquatic ecotoxicological modelling and simulation andbiostatistics are in the focus of the department.

Partner 16 – LemnaTec (LEMTEC)LemnaTec is a technology based private company playing a connective role between ecotoxicology,general effect quantification of bioassays and data handling and deriving actions from these data.LemnaTec develops image-processing units specified for bioassays. Based on this, newtechnologies for rapid biomonitoring of biological changes are made available to researchers.LemnaTec develops databases specified for the needs of biological effect monitoring. LemnaTechas a long expertise on bringing bioassays through the whole process from biological research inthe lab to final international standards.

Partner 15 – WRc-NSF Ltd (WRcNSF)WRcNSF has a wide range of technical services and information to help those managing pollution,industrial development and environmental planning. WRcNSF's Environmental Monitoring andAssessment (EMA) team’s contribution to the project will provide expert testing using Microtoxwhich enable clients to understand the availability of chemicals in soils. Specific expertise lie in theareas of Toxicity Identification Evaluation (TIE) approaches and these will be used to identifymajor toxicants in soils with mixed pollution.

Partner 24 – LimCo International (LIMCO)LimCo International GbR is an SME specialised in limnology, aquatic ecotoxicology and (online)biomonitoring. LimCo perform water quality assessments and standard toxicity tests (Daphniamagna, Chironomus riparius, fish) as well as develop new rapid toxicity bioassays and onlinebiomonitors based on fast and sensitive behavioural responses of aquatic and benthic animalspecies. LimCo has developed the Multispecies Freshwater Biomonitor® (MFB), a fully automatedinstrument to continuously record survival and different behaviours quantitatively and in real time,allowing for time-to-death and time-to-response surface modelling as well as time-series analyses.The MFB allows for simultaneous multi-species and multi-exposure in different media, such assediment/soil, water and air.

Partner 10 – Jagiellonian University (UJAG)The project will be performed in the Institute of Environmental Sciences (Department of Ecotoxico-logy), a leading research and educational institution in its field and the European CommunityCenter of Excellence. Department expertise lies within the fields of ecotoxicology of terrestrialinvertebrates, ecosystem ecotoxicology, biogeochemistry and population ecology.

Partner 14 – University of Antwerp (UA)The research of Department of Biology of the University of Antwerp focuses on the molecular andbiochemical aspects and bioavailability of microcontaminants (metals as well as organics) and thephysiological aspects of adaptation to environmental stress.

Partner 8 – Cardiff University (UWC)The University of Wales, Cardiff is among the top 5 percentile of Britain's leading research andteaching universities. UWC staff has established a focus that links environmental science withmolecular genetics in model systems (e.g. nematodes) and sentinel organisms (e.g. earthworms).The know-how and all facilities that are pertinent to the project are available at UWC.

62

Partner 9 – University of Cambridge (UCAM)The project will be performed in the Department of Biochemistry, which is actively involved indeveloping functional genomic techniques including metabolomics, proteomics and structuralfunctional biology. UCAM staff is developing a number of analytical techniques for the large scaleand high throughput analysis of metabolites to be used in conjunction with pattern recognition.

NOMIRACLE Risk assessment Pillar (RP 4)

Partner 4 – University of Nijmegen (DESUN)The Department of Environmental Studies (DESUN) has excellence within the domain of integratedenvironmental science. The main research topics are ecological restoration of river basins andmodelling of environmental risks posed by chemical contaminants and other stressors.

Partner 6 – Vrije Universiteit Amsterdam (VU)The Department of Theoretical Biology is part of the Institute of Health Sciences of the Faculty ofEarth and Life Sciences of the Vrije Universiteit. The main expertise of the Department can belabelled as quantitative bio-energetics. The aim of the research programme is to develop aquantitative and coherent theory for energy and mass transduction that links theories concerning alllevels of organization, from membrane physiology to ecosystem dynamics.

Partner 37 – University of Southampton (USOUTH)The Clinical Pharmacology Group belongs to the School of Medicine of Southampton University.The group has expertise in the field of risk assessment of non-genotoxic carcinogens. He hasdeveloped pathway-related uncertainty factors are developed from databases describing humanvariability and interspecies differences in the major pathways of elimination (phase I, phase II andrenal excretion).

Partner 29 – Ecole Polytechnique Fédérale de Lausanne (EPFL)Having a leading role internationally within the Life Cycle Initiative of the United NationsEnvironmental Program (UNEP) and the Society for Environmental Toxicity and Chemistry(SETAC), EPLF co-ordinates the development of guidance and methods for best available practicein Life Cycle Impact Assessment. EPLF has also initiated international work on Life Cycle Costingwithin the SETAC activities.

Partner 33 – Universitat Rovira i Virgili (URV)The URV staff has expertise in the areas of chemical and environmental engineering, chemistry,fluid mechanics-CFD, computer sciences and applied mathematics.

Partner 3 – UFZ Centre for Environmental Research (UFZ)UFZ is a leading institution in environmental toxicology and environmental health research.Expertise in statistical modelling, time-series modelling and epidemiological field studies.

Partner 21 – Alterra (ALTERRA)Alterra engages in strategic and applied research to support design processes, policymaking andmanagement at the local, national and international level. This includes not only innovative,interdisciplinary research on complex problems relating to rural areas, but also the production ofreadily applicable knowledge and expertise enabling rapid and adequate solutions to practicalproblems.

63

Partner 19 – Finnish Environment Institute (SYKE)SYKE is the leading institute within the environmental administration of Finland. The key tasks ofSYKE include multidisciplinary research in environmental changes and their causes and insolutions to environmental problems, and environmental monitoring and assessment.

Partner 32 – DIALOGIK (DIA)The main expertise of DIALOGIK is investigation of patterns of communication and cooperation inthe areas of high tension within the sectors of politics, economy and civil society. The mainobjective of DIALOGIK is to analyse the conditions and prerequisites for improving purposefulcommunication (such as institutionalised risk communication as an instrument of health protection)and for developing and initiating innovative procedures of participation and cooperation procedures(such as citizen panels or mediations in environmental conflict situations).

NOMIRACLE Management Pillar (Pillar 5)

Partner 1 – National Environmental Research Institute (NERI)NERI is the leading environmental research institution in Denmark. NERI staff is highlyexperienced in the management of large research projects including management of EU R&TDprojects. Since its foundation in 1989, NERI has participated in about 80 EU projects, being co-ordinator of about 15 of these.

Partner 18 – Directorate General Joint Research Centre (JRC)As a Directorate General of the European Commission, the EC-JRC functions as a reference centreof science and technology for the Union. The JRC has develops skills and tools to provide impartialand Europe-wide expertise in the environmental field. The Institute of Environment andSustainability (IES) provides scientific and technical support to EU policies for the protection of theenvironment contributing to sustainable development in Europe. The IES has considerable expertisein the collection, management and analysis of geographically referenced data, spatial modelling,metadata systems, database design and in the development of internet based applications.

Partner 13 – DBUAProf. Amadeu Soares , a current member of CSTEE (EC Scientific Committee on Toxicity,Ecotoxicity and the Environment) and former member of the Scientific Advisory Committee ofECVAM, has wide experience in participating in national and international (bi-lateral or multi-lateral) funded research projects, including with Developing Countries in Asia and South andCentral America, both as partner (ca. 10 projects) and as project co-ordinator (ca. 7 projects). Hasbeen acting as co-ordinator (EV5V-CT91-0009, EV5V-CT94-0422, ERBCIPD-CT94-0119, IC18-CT98-0264, LIFE95/P/A23/P/119/ MLTRG) and partner (TS3*-CT93-0230, ENV4-CT97-0470) ofseveral EC contracts, and has some bilateral R&D projects with Brasil and Mexico. He is also Headof the Department of Biology from Aveiro University, and a former President of SETAC-Europe(Society of Environmental Toxicology and Chemistry - Europe) and of the Institute forEnvironmnetal and Life Sciences, University of Coimbra, Portugal. He has being responsible fororganising a 10-year programme of a series of 2-week intensive courses, aimed for post-graduatestudents, with the participation of postgraduate students from all over the world (Portuguese, north-American, several EU countries, Brasil, Costa Rica, Mexico).

Partner 38 – Institut SYMLOG de France (SYMLOG)Institut Symlog de France is a small private structure listed in the CNRS guide to independentresearch laboratories in the social sciences. Founded in 1982, Symlog has since 1984 conductedresearch and intervention in the area of technological and natural risks, with special emphasis onpublic perception and participation, providing counsel on institutional and organizational decisionmaking. Symlog has performed qualitative studies for large French and international public and

64

private institutions. Interventions include analysis of public communication in risk management,study of public risk perceptions, evaluation of institutional risk regulation systems. Methods appliedinclude interviews (individual or focus groups), content analysis, large surveys, and social valueinventories. Symlog personnel have participated in many ministerial or international expert groups.Claire Mays (Leader of WP 5.4) is Programme Manager at Institut Symlog. Her main interest liesin the empowerment of stakeholder groups through facilitating their expression and analysis ofsocial values. Regarding dissemination activities, Claire Mays has served since 2000 as rapporteurand social sciences consultant to the OECD NEA Radioactive Waste Management CommitteeForum on Stakeholder Confidence. A large component of her work consists of compiling andreporting the lessons learnt in the Forum, collecting member and end-user evaluative feedback, andensuring public internet access to documentary outputs.

Partner 20 – Kaunas University of Technology (APINI)The Institute is involved in research, purpose oriented training and technical assistance toenterprises, and education. Among the research areas of the Institute are the following: chemicalscontrol and management, environmental impact assessment, environmental efficiency, modellingand management of surface water quality, simulation of various pollution load scenarios,development and implementation of the concepts and techniques of pollution prevention/ wasteminimisation/ cleaner production.

B.6 Description of project management

Organisational structure and decision making mechanismsThe management of NOMIRACLE will be headed by the Project Co-ordinator, Hans Løkke,NERI, in close collaboration with a Management Board (MB) and a Project Secretariat (PS)placed at NERI. Other important structural elements are the General Assembly of all partners andthe Advisory Board. At the Research Pillar level four Steering Committees will manage theactivities within the pillars, and at the Work Package level the work package leaders will managethe activities within the work packages. The management structure of the project is shown in FigureB.6-1. This structure will ensure that the decision-making mechanisms are simple and clear, andsuitable to handle unforeseen problems in this relatively large and complex project. Further, thisstructure will consolidate a high degree of integration between partners. It is described below howthe different elements of the project management enable the project to achieve its goals.

65

European Commission

NOMIRACLECo-ordinator (NERI)

Board of Management

WP 5.2 Data Management

WP 5.3 Training

WP 5.4 Dissemination

Four Steering Committees

Research Pillar 1:

Risk scenarios

Research Pillar 2:

Sound exposure

Research Pillar 4:

Risk assesment

Research Pillar 3:

Effect assesment

Advisory Board

General Assembly of Partners

WP

5.1

Pro

ject

Sec

reta

riat

Figure B.6-1: Management structure of NOMIRACLE.

The Project Co-ordinator will hold the contract and have the direct contact with the EuropeanCommission. He is responsible for the scientific content and quality of the project, and for thefinancial, legal and management aspects of the project. The Project Co-ordinator has the direct linkwith the European Commission. He is head of the Project Secretariat and is chairing theManagement Board, the General Assembly, and the Advisory Board. The Project Co-ordinator,Hans Løkke, has a long experience of research management. Since 1990 he held a position asdirector of a research department that counts 37 staff members, including 19 senior scientists. In1992-1996 he co-ordinated the EC project SECOFASE (Development, Improvement andStandardization of Test Systems for Assessing Sublethal Effects of Chemicals on Fauna in the SoilEcosystem, Contract EV5V-CT92-0218 + ERB-CIPD-CT93-0059) with 10 partners. From 1992 –1997 he was the co-ordinator of the Danish Centre for Ecotoxicological Research, within theDanish Environmental Research Programme with participation of 66 scientists from 14 institutions(www.smp.au.dk/smp_uk/programmetscentre/c06/c061/c061.htm). The Deputy Project Co-ordinator, Hanne Bach, assists the Project Co-ordinator. She has long experience with projectmanagement, and since 1999 she has held a position as director of a research department counting33 staff members, including 29 scientists. She has broad experience within environmental impactassessment studies and State of Environment Reporting as editor of the Danish SOE reports.

The Management Board (MB) consists of the Co-ordinators of the 4 Research Pillars (RP), across-cutting co-ordinator with expertise in the integration of environmental and human health, anda cross-cutting co-ordinator for training and dissemination. Each member will have a deputy, whowill be invited to attend the MB meetings. All members of the MB have experience on managementfrom ongoing or former EU projects, and they are highly qualified in research management.

The members of the MB are shown in Table B.6-1. Together, the members and theirdeputies have insight in all aspects of the NOMIRACLE project. The MB will meet at least twice ayear, one time in connection to the General Assembly. The expenses for travelling of the boardmembers are included in the calculation of travelling and subsistence of the project management.

66

Table B.6-1: Members of the management board of NOMIRACLE.

Member of MB Deputy of MBNOMIRACLECo-ordinator

Dr. Hans Løkke(NERI, Denmark)

Prof. Martin Holmstrup(NERI, Denmark)

Research Pillar 1,Risk Scenarios

Dr. Hanne Bach(NERI, Denmark)

Dr. Arwyn Jones(JRC/IES, Italy)

Research Pillar 2,Sound Exposure

Prof. Gerrit Schüürmann(UFZ, Germany)

Prof. Michael McLachan(ITM, Sweden)

Research Pillar 3,Effects Assessment

Dr. Dave Spurgeon(NERC, UK)

Dr. Ryszard Laskowski(UJAG, Poland)

Research Pillar 4,Risk Assessment

Dr. Ad Ragas(DESUN, The Netherlands)

Dr. Jean Lou Dorne(USOUTH, UK)

Human healthcrosscutting co-ordinator

Prof. Olf Herbarth(UFZ, Germany)

Dr. Albert Duschl(USALZ, Austria)

Environmental crosscuttingco-ordinator

Prof. Francesc Giralt(URV, Spain)

Dr. Wim de Coen(UA, Belgium)

Dissemination and training Prof. Amadeu Soares(DBUA, Portugal)

Dr. Timo Assmuth(SYKE, Finland)

The MB will be chaired by the Project Co-ordinator (Hans Løkke). The MB will conductoperational management of the project and support the necessary decisions in co-ordinating andadministrating the project. The MB is responsible for the implementation of the entire project, andfor public relations. In particular, the MB will watch the progress of the project and that themilestones will be met in due time. The MB will assist the Project Co-ordinator in quality assuranceof project deliverables. Whenever required, scientists from the project will attend the meetings ofthe MB to discuss specific matters. Decisions made will depend on consensus, and the Project Co-ordinator will be responsible for the final decisions. In cases of conflicts or conflicting points ofview within the project, the case will be brought for the MB.

The Advisory BoardTo guarantee realism and quality of the research, an Advisory Board will be consulted on matters ofimportance for future changes in the methodologies for risk assessment. The Advisory Board isdescribed in section B.3. The Advisory Board may be extended with new members, which areidentified as key stakeholders during the project period.

67

Table B.6-2: Members of the Advisory Board.

Representing Interests/Expertise1 Open European Commission Primary enduser2 Dr. Anne Fairbrother US-EPA and SETAC Risk assessment/

exchange of European and Americanexperiences/dissemination of results

Dr. Wolf-Rüdiger Bias,BASF Ltd, Germany

Industry Pesticides/end-user

Dr. Ralf Schulz,Syngenta Crop ProtectionLtd, UK

Industry Pesticides/end-user

3

Dr. Ali R. Temara,Procter & Gamble,Belgium

Industry Industrial chemicals andbiocides/enduser

4 Prof. Kees van Leeuwen,Director of IHCP-JRC,Italy

Human Health Integrated Risk Assessment/end-user

Prof. Michael Depledge,Head of Science,Environment Agency, UK

National EPA National authority, end-user

Dr. Jukka Malm,Division ManagerChemicals DivisionFinnish EnvironmentInstitute

National EPA National authority, end-user

5

Open National EPA National authority, end-userOpen Consumer interests NGO6Open Public interests in

environmentNGO

The Advisory Board will be a forum for dialog between the NOMIRACLE Management Board andstakeholders and end users comprising:

1) European CommissionOne or more seats will be kept open to ensure that the interests of the European Commission will berepresented.

2) US EPADr. Anne Fairbrother is chief of the Ecosystems Characterization Branch of US EnvironmentalProtection Agency, National Health and Environmental Effects Research laboratory, WesternEcology Division, Corvallis, Oregon. She will ensure the communication on activities oncumulative risk assessment in the US EPA. As a former president of SETAC North America(Society of Environmental Toxicology and Chemistry) and present member of the SETAC WorldCouncil, she will give advice on the dissemination of the NOMIRACLE results to the globalscientific audience.

3) IndustryDr. Ralf Schulz (BASF Ltd), Dr. Wolf-Rüdiger Bias (Syngenta Crop Protection Ltd.) and

68

Dr. Ali Temara (Procter and Gamble) will contribute to the connection between NOMIRACLE andactual needs from end users in the chemical industry.

4) European Commission, Joint Research CentreProfessor Kees van Leeuwen is Director of the JRC Institute for Health and Consumer Protection(IHCP) where he is responsible for the European Centre for the Validation of Alternative Methods(ECVAM), the European Chemicals Bureau (ECB), Physical and Chemical Exposure,Biotechnology and GMOs, and Medical Technology.

5) National AuthoritiesNational authorities will be invited to cover regional differences in Europe. At present, MichaelDepledge (UK EPA) and Jukka Malm (Finnish Environmental Institute) have accepted toparticipate.

6) NGOsTwo seats will be kept open for consumer interests and public interests in the environment and itsbiodiversity.

The Project Secretariat (PS) will assist the Project Co-ordinator and the Research PillarCo-ordinators in the day-to day management. The PS staff will assist in communication withinNOMIRACLE, with external bodies, in the organisation of seminars and workshops, and thecommon database administrator and the webmaster of a coming NOMIRACLE homepage will bemembers of the PS staff.

Dr. Morten Strandberg from NERI will hold a 50% part time position as project manager.He is a multidisciplinary environmental scientist and adviser with experience from fields includingecology, ecotoxicology, ecological modelling, radioecology, mycology and botany, as well asecological risk evaluation and ecological risk assessment of pesticides and genetically modifiedorganisms. He has been editor of popular scientific journals and books, and he has organised andled national and international workshops on environmental issues. His responsibilities will includemaintaining contact with all partners in the project, ensuring all partners will keep reportingdeadlines. He will be responsible for the daily financial activities in order to ensure an efficientpayment flow in the project. He will be responsible for the internal communication flow through e-mails and for the installation and updating of the NOMIRACLE homepage. He will assist theProject Co-ordinator in preparing the agenda for meetings and writing minutes. He will also beeditor of a quarterly Newsletter for NOMIRACLE to be published on the open communicationwebpage of the project. He will closely follow the progress in Work Package 5 on disseminationand training, and Work Package 5.4 on the internal database of NOMIRACLE. Further, the projectmanager will plan the internal meetings and workshops within the project. A 50 % part time dailytechnical administration assistant, Mrs. Lene Birksø Bødskov, will support the project manager. Shehas experience in the administration of Framework projects and will assist on communication,financial matters and act as webmaster in updating of the open and the internal website. She willalso organise meetings and workshops.

The PS will work closely together with the specialised NERI administrators, who will assistthe PS in financial and legal questions (2 person month per year), and in public relations beside theactivities in WP 5.4.

Research Pillar Steering Committees. The RP level is based on conceptual scientific areas ofspecialisation and ensure efficient management of the project. The 4 RPs will each have a Co-ordination Committee headed by the RP Co-ordinator who will collect audits, reports and otherinformation generated within the module. The RP Co-ordinators will assist the Project Co-ordinatorin ensuring that the partners of the RP fulfil their legal and contractual obligations towards theCommission and the other partners.

69

Table B.6-3: Research Pillar Steering Committees.

RP No Research Pillar Co-ordinator Members

1 Dr. Hanne Bach all participants in WP 1.1 and 1.22 Prof. Gerrit Schüürmann all participants in WP 2.1, 2.2, 2.3 and 2.43 Dr. David Spurgeon all participants in WP 3.1, 3.2, 3.3 and 3.44 Dr. Ad Ragas all participants in WP 4.1, 4.2, 4.3 and 4.4

The WP levelCorresponding to the Research Pillar Committees, the participants in each work package will attendwork meetings on a regular basis. A high degree of integration of the daily work within the workpackages is anticipated. If unforeseen difficulties arise which cannot be solved within the workpackage by the work package leader, he or she will contact the Research Pillar Co-ordinator or theProject Co-ordinator with no delay.

General AssemblyAll partners will meet in a yearly General Assembly to agree on the annual implementation plan,changes in the Consortium or any matter of fundamental importance to the Consortium. TheGeneral Assembly will be held in connection to an important and relevant international scientificconference providing for presentation of the findings of the Consortium.

The General Assembly will guarantee that all partners are informed and agree on theprogress of the project and the need for adjustments and that consensus can be achieved on theannual implementation plan. The General Assembly will also give advise to the Project Co-ordinator and the Management board on expansion of the Consortium or exclusion of partners fromthe Consortium. Any changes of the Consortium Agreement and other matters of fundamentalimportance to the Consortium will be discussed in the General Assemply.

Plan for management of knowledge, of intellectual property and of other innovation-relatedactivities arising in the projectResearch results will belong to the institutions where the research was carried out, and theintellectual property rights will be applied individually according to the rules of the employer underthe European and national legislation. In cases of joint contribution, the ownership of intellectualproperty will be shared in accordance with a notified agreement.

The procedures for dissemination, protection and exploitation of the intellectual propertywill be outlined in a Consortium Agreement to be negotiated by the partners as part of the contractnegotiation process. Special focus will be on SMEs to ensure that relevant project results andmaterials are exploited commercially. The Project Secretariat and the Management Board will assistthe Project Co-ordinator in these management activities.

Before publication of research results, all draft publications shall pass an initial review of itspotential for patenting and intellectual property rights which may lead to commercial exploitation.This review will be conducted by the respective Research Pillar Co-ordinators within 4 weeks afterreceipt of the manuscripts. The partners are responsible that all patentable data will result in anapplication for a patent, with shared intellectual ownership rights negotiated between the relevantpartners on a case-by-case. In cases of patent applications, data may not be published during theperiod of application and evaluation of patents. In cases of commercial exploitation of the results,the involved partners shall sign an agreement with the Project Co-ordinator on property rights.Partners are responsible for all costs in relation to commercialisation or application for patents.

Addition or exclusion of partnersAlthough the structure of the project has been thoroughly elaborated involving all partners into acomplete and well-structured team, the further progress of the project may call for expansion of the

70

Consortium or exclusion of partners. In such cases, the Project Co-ordinator in collaboration withthe Management Board will elaborate a plan for the needed changes. The General Assembly may beinvited to an extraordinary meeting, however, depending on the importance of such changes forNOMIRACLE, the changes may alternatively be sent in hearing at all partners. After consulting theManagement Board and the European Commission, the Project Co-ordinator will take the finaldecision.

A codex for scientific and administrative practise will be set in collaboration with allpartners to secure good scientific practise. Special attention will be given to the gender issue.

B.7 Project resources

Personnel costsResources have been allocated to individual partners according to their workload. Total budgets foreach Partner and their contribution to individual WPs are indicated in Table B.7-1 (Project effortform) and under each Research Pillar description (Section B.4). The core partners UFZ, NERI,DESUN, JRC and NERC are large institutions covering a wide range of disciplines. Therefore, thebudgets of these institutions are comparatively high. Most other partners will contribute within oneor a few specialised fields and accordingly have smaller budgets.

The critical mass for development of novel methodologies for exposure and effects assessments andtheir validation will require relatively more manpower per se than theoretical work, e.g.development or improvement of mathematical models. Therefore the budgets of RP 2 and RP 3 arethan RP 1 and RP 4.

The management pillar involves training. Direct costs for training is 1.7% of the budget, howevertraining of young scientist, e.g. PhD students or master students will allocate considerable amountsof resources within the research pillars.

Institutions with a human health profile receive 15.0% of the budget, 64.2% to institutions withmainly an environmental profile, and 22.3% will be used for partners already involved in integratedenvironmental and human health risk assessment such as SYKE, UFZ and DIA. The activities ofSMEs counts 7.1% of the total budget. The composition of partners in the Consortium and theirskills are carefully selected to create a team, which will be able to meet the objectives of the project.

EquipmentMost of the laboratories of the partners are well equipped with state of the art apparatus availablefor the NOMIRACLE project, and therefore do not have demands for new equipment. In a fewcases, specific new instruments are required. For the developing of models, some partners also needto purchase top-level workstations. However, the share of the budget used for durable equipment islower than 3% of the total budget.

ConsumablesLaboratory studies on exposure and effects, and toxicokinetic work, need a budget for consumables.Especially the molecular techniques used in RP 3 are expensive, and resources have been allocatedin accordance with this.

TravellingTo ensure a proper integration of partners, all partners will allocate sufficient travelling budgets toparticipate in meetings of the Consortium and its workshops and technical seminars.Subcontracting

71

Only one Partner (EKUT) will use subcontracting to fulfil the objectives. This concernshistopathological analysis if tissues in RP 3. The requested funding used for subcontracting is in theorder 100 Keuro.

Audit costsManagement costs will include costs for carrying out anticipated technical, financial, technologicaland ethical audits.

OverheadsThe rates for overheads are the standard 20% for ACF calculation basis. Some research institutes(e.g. NERI, NERC and ALTERRA) have used much higher overhead rates (80-116%). These haveall been negotiated and agreed upon between EU and the respective institutions to cover the actualcosts of the research activities.

AdministrationOf the total budget 5.2% is used for administration. This includes administration of a commondatabase, a webmaster for an interactive webside, workshops, collaboration with the AdvisoryBoard, training, demonstration, dissemination and audit activities. NERI (Project Co-ordinator) andthe Pillar Co-ordinators (NERI, UFZ, NERC and DESUN) will have regular meetings requiringsufficient budgets.

Table B.7-1: NOMIRACLE-IP Project Effort Form 1Full duration of project

(person-months for activities in which partners are involved)Project acronym –NOMIRACLE

Partner 1NERI

Partner 2NERC

Partner 3UFZ

Partner 4DESUN

Partner 5DSTA

Partner 6VU

Partner 7NIPH

Partner 8UWC

Partner 9UCAM

Partner 10UJAG

Partner 11EKUT

Partner 12WU

Partner 13DBUA

TotalForm 1

RTD/Innovation activitiesWP 1.1 23 6.5 0.5 30WP 1.2 24 6 2 32WP 2.1 98 98WP 2.2 43 43 86WP 2.3 95 95WP 2.4 48 48WP 3.1 12 36 86 6 18 36 20 12 18 244WP 3.2 60 48 18 12 30 168WP 3.3 24 43 12 48 36 163WP 3.4 36 42 30 36 36 22 24 226WP 4.1 52 48 100WP 4.2 49 46 95WP 4.3 0WP 4.4 1 7.5 2 3WP 5.2 Data Management 1 1WP 5.4 Dissemination 3 3Total research 167 102 462.5 150.5 48 114 36 36 36 84 60 48 48 1,392

Demonstration activitiesWP 5.3 1 1Total demonstration 0 0 0 0 0 0 0 0 0 0 0 0 1 1

Training activitiesWP 5.3 3 1 0.5 1 5.5Total training 3 0 1 0.5 0 0 0 0 0 0 0 0 1 5.5

Management activitiesWP 5.1 31.5 1.5 1.5 1.5 36Total management 31.5 1.5 1.5 1.5 0 0 0 0 0 0 0 0 0 36

TOTAL ACTIVITIES 201.5 103.5 465 152.5 48 114 36 36 36 84 60 48 50 1,434.5

Table B.7-1: NOMIRACLE-IP Project Effort Form 2Full duration of project

(person-months for activities in which partners are involved)Project acronym –NOMIRACLE

Partner 14 Partner 15 Partner 16 Partner 17 Partner 18 Partner 19 Partner 20 Partner 21 Partner 22 Partner 23 Partner 24 Partner 25 Partner 26 TotalUA WRcNSF LEMTEC USALZ JRC SYKE APINI ALTERRA EAWAG RIVM LIMCO RWTHA ECT Form 2

RTD/Innovation activitiesWP 1.1 16 4 10 30WP 1.2 2 6 8WP 2.1 36 36WP 2.2 16 16WP 2.3 47 47WP 2.4 37 4 41WP 3.1 30 12 36 30 12 120WP 3.2 12 18 30WP 3.3 12 18 30WP 3.4 36 36WP 4.1 4 4WP 4.2 22 22WP 4.3 18 54 72WP 4.4 3 8 11WP 5.2 Data Management 14 14WP 5.4 Dissemination 0Total research 48 48 12 48 106 60 4 40 36 8 48 12 47 517

Demonstration activitiesWP 5.3 0.5 0.5Total demonstration 0 0 0 0 0 0 0.5 0 0 0 0 0 0 0.5

Training activitiesWP 5.3 2 1.5 8 11.5Total training 0 0 0 0 2 1.5 8 0 0 0 0 0 0 11.5

Management activitiesWP 5.1 0Total management 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TOTAL ACTIVITIES 48 48 12 48 108 61.5 12.5 40 36 8 48 12 47 529

Table B.7-1: NOMIRACLE-IP Project Effort Form 3Full duration of project

(person-months for activities in which partners are involved)Project acronym –NOMIRACLE

Partner 27UNIMIB

Partner 28ENVI

Partner 29EPFL

Partner 30ULANC

Partner 31ITM

Partner 32DIA

Partner 33URV

Partner 34LHASA

Partner 35LMC

Partner 36CSIC

Partner 37USOUTH

Partner 38SYMLOG

TotalForm 3

TotalPartners

RTD/Innovation activitiesWP 1.1 9 9 69WP 1.2 5 8 4 17 57WP 2.1 0 134WP 2.2 36 48 65 149 251WP 2.3 16 40 72 128 270WP 2.4 48 40 88 177WP 3.1 0 364WP 3.2 0 198WP 3.3 36 36 229WP 3.4 0 262WP 4.1 18 36 54 158WP 4.2 42 42 159WP 4.3 21 21 93WP 4.4 4 6 10 24WP 5.2 Data Management 0 15WP 5.4 Dissemination 10 10 13Total research 18 8 18 36 96 21 108 40 108 65 36 10 564 2,473

Demonstration activitiesWP 5.3 0.25 0.25 1.75Total demonstration 0 0.25 0 0 0 0 0 0 0 0 0 0 0.25 1.75

Training activitiesWP 5.3 2.5 0.5 3 20Total training 0 2.5 0 0 0 0.5 0 0 0 0 0 0 3 20

Management activitiesWP 5.1 0 36Total management 0 0 0 0 0 0 0 0 0 0 0 0 0 36

TOTAL ACTIVITIES 18 10.75 18 36 96 21.5 108 40 108 65 36 10 567.25 2,530.75

75

B.8 Detailed implementation plan – first 18 months

Structure of the detailed 18-month implementation planTo achieve the objectives of NOMIRACLE outlined in section B.1 we will initiate work in mostResearch Pillars as well as Pillar 5 on management from the starting date of the project. This is dueto the nature of the research planned in each RP. Even though the project is highly integrated it issensible to initiate the work of the different WPs at the same time, thus establishing newmethodologies where necessary before the chosen risk scenarios will be addressed. An exceptionfrom this is RP 3, where the initiation of the work will wait for the first outcome of WP 1.2.

A Gantt chart (Figure B.8-1) shows the planned schedule of the different WPs and the time wheredeliverables are to be made available for other project components. A graphical presentation of theproject is found in Figure B.4-1 showing the overall structure of NOMIRACLE and how thecomponents are linked together. This structure is also valid for the first 18 months of the project.The NOMIRACLE budget for the first 18 months are shown in the 18-month version of the A3form.

In the following section a detailed description of the work content during the first 18 months aredescribed for each RP. Included here is also a description of how WPs are linked together withinand between the 4 RPs. After this follows the descriptions of the first 18 months’ work of theindividual WPs.

76

Month from the start of the project

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60

Research Pillar 1 – Risk Scenarios

WP 1.1 Establishment of data background for scenario selection

WP 1.2 Scenario selection and ranking

D.1.1.1 Data package for pesticides

D.1.1.2 Data package for pharmaceuticals, biocides and first sub set of VO Cs/semi-VOCs

D.1.1.3 Data package for VOCs/semi-VOCs

Data package for landscape classifi cation and land use

Full documentation to the data management section

D.1.2.1 Scenario selection and criteria setting for pesticides

D.1.2.2 Scenario selection and criteria setting for pharmaceuticals, biocides, a first sub set of VOCs/semi-VOCs

D.1.2.3 Scenario selection and criteria setting for VOCs/semi-VOCs

D.1.2.4 Documentation of the first version of selection procedure including initial uncertainty evaluation

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60

Research Pillar 2 – Sound Exposure

WP 2.1 Matrix-compound interaction

WP 2.2 Available exposure

WP 2.3 Metabolic fate

WP 2.4 Region-speci fic environmental fate

D.2.1.1 Compilation of phase parameters, and database of Abraham descriptors

D.2.1.2 Validated procedure for the determination of phase descriptors for soils, and first series of data

D.2.1.3 Validated experimental procedure for the determination of membrane-water partitioning

D.2.1.4 Computerised increment method to estimate Abraham descriptors from molecular structure

D.2.1.5 Comparative analysis and evaluation of molecular descriptors for modelling hydrogen bonding

D.2.2.1 Analytical methods for multifunctional xenobiotics such as biocides, pesticides and pharmaceuticals

D.2.2.2 Temporal and spatial variability of human exposure-related VOC

D.2.2.3 Sampling methods for soils and sediments and relevant biotest systems

D.2.2.4 Lab and field data of compound exposure, and assessment of availability parameters

D.2.2.5 Assessment of the (source-dependent) indoor/outdoor ratio of VOC

D.2.3.1 Validated procedures for the determination of compound turnover in water, sediments and soils

D.2.3.2 Biodegradation kinetics for some model compounds in water, sediment and different soils

D.2.3.3 Evaluation of relevant parameters for deriving a QSPR for degradation rate parameters

D.2.3.4 Preliminary computer approaches to simulate terrestrial biodegradation

D.2.4.1 Multimedia fate and exposure model with various spatial resolutions at the European level

D.2.4.2 Indication of the spatial detail

D.2.4.3 Inventory of improvement options in existing models, based on empirical model evaluation

D.2.4.4 Cognitive neural network-based intelligent system to identify the most important variables

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60

Research Pillar 3 – Effect Assessment

WP 3.1 Interactive toxicological effects on diverse biological systems

WP 3.2 Combined effects of natural stressors and chemicals

WP 3.3 Toxicokinetic modelling

WP 3.4 Molecular mechanisms of mixture toxicity

D.3.1.1 Finalisation of mixture experiment design and analysis framework

D.3.1.2 Development of screening and targeted methods for human health

D.3.1.3 Dose response profiles for the first set of prioritised chemicals and mixtures

D.3.2.1 Baseline assessment of response to natural stressors

D.3.2.2 Methods to assess the impacts of relevant combinations of cumulative stressors

D.3.2.3 Generation of multiple stressor data for combinations of chemicals and environmental stresses

D.3.3.1 Standard operating procedures fo chemical analysis in different species

D.3.3.2 Preliminary report on the experimental toxicokinetics of two selected chemicals and their mixtures

D.3.3.3 Literature report on the toxicokinetics of chemicals in mixtures in diverse species

D.3.4.1 Repository of protocols for specifi c and global profiling of stress response pathways

D.3.4.2 Completion of baseline analysis of stress response proÞ les

D.3.4.3 Analyses on the effects of two compounds and one additive mixture on biochemical responses

D.1.1.4

D.1.1.5

Figure B.8-1: Work planning showing the timing of the various work packages and theirdeliverables.

77

Month from the start of the project

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60

Research Pillar 4 – Risk Assessment

WP 4.1 New concepts and techniques for probabilistic risk assessment

WP 4.2 Explicit modelling of exposure and risk in space and time

WP 4.3 Dealing with risks in a management context

WP 4.4 Risk presentation and visualisation

D.4.1.1 Report on separation of true uncertainty and interindividual variability in predicted risks for a population

D.4.1.2 Paper on the model formulation for effects of a mixture of 2 compounds

D.4.1.3 Report describing metabolism, pharmacokinetic and pharmacodynamic data in humans and species

D.4.1.4 Report describing method for comparative risk assessment of multiple stressors

D.4.2.1 Random-walk model describing pollutant accumulation in a floodplain

D.4.2.2 Fuzzy ARTMAP neural classifier for cluster analysis and demonstration of a variable selection approach

D.4.2.3 Demonstration of up-scaling methods, based upon the small-scale modelling

D.4.2.4 Demonstration of the CAR-model assessing health risk in the city area of Leipzig

D.4.2.5 A temporal model for the indoor air pollution with volatile organic compounds

D.4.3.1 Conceptual models and mental maps of risk information and cognition

D.4.3.2 Frameworks for the characterisation of knowledge of risks

D.4.3.3 Models to develop integrated approaches and strategic guidance to risk assessment partners

D.4.3.4 Review of relevant risk communication research and communication models

D.4.3.5 Interim report and plan for further work, meetings reports and discussion papers.

D.4.4.1 Methods and tools to provide cumulative risk maps

D.4.4.2 Guidelines for the aggregation of spatial data

D.4.4.3 Examples of cumulative risk maps for EU and a limited number of selected EU regions

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 60

Pillar 5 – Management

WP 5.1 General Project Management

WP 5.2 Establishment of data management system

WP 5.3 Training and Demonstration

WP 5.4 Dissemination and Exploitation

D.5.1.1 Financial and technical reports to the European Commission

D.5.1.2 Meetings of the Management Board, the Advisory Board and the General Assembly

D.5.1.3 Dissemination plan including internal and external homepage and a Quarterly Newsletter

D.5.1.4 Organisation of two scientific sessions

D.5.2.1 Questionnaire on the nature aof data sets utilised and generated by NOMIRACLE

D.5.2.2 Draft design document on NOMIRACLE data management system and implementation strategy

D.5.2.3 Draft NOMIRACLE data management system and implementation strategy

D.5.2.4 Operational data management system

D.5.2.5 Review of the operation and functionality of the NOMIRACLE Data Management System

D.5.3.1 Organisation of an internal training session, at the kick-off meeting

D.5.3.2 Organisation of three advanced studies courses

D.5.3.3 Organisation of one technical session

D.5.3.4 Programme for an international PhD programme

D.5.4.1 IP internal and external reporting procedures

D.5.4.2 Templates in the speciÞ c goal of integrating user feedback

D.5.4.3 Procedures to analyse and redirect user feedback reports to the relevant WPs

D.5.4.4 Stakeholder mapping, focusing on key actors and representatives of public

D.5.4.5 Highlights from First Periodic Report of NOMIRACLE

Figure B.8-1 (continued): Work planning showing the timing of the various work packages andtheir deliverables.

78

Detailed implementation plan for research pillar 1 - (first 18 months)

IntroductionDuring the first 18 months a brutto-type or first version of a scenario selection procedure will bedeveloped and be further tested and improved after the 18 months period. The development willtake place by investigating a series of substance categories used as training substances for themethod development. Scenarios based on these substance categories and the data used for thescenario selection will be available as inputs during the first 18 months in several parts ofNOMIRACLE. This will yield the following main output from RP 1 to be the inputs for otherresearch pillars during the first 18 months: (1) Set of test scenarios possessing the highest riskpotential for cumulative assessment of mixture toxicity to be analysed in RP 3. (2) Evaluation ofuncertainty in the scenario selection as input to WP 4.3. (2) Input to the exposure assessment in WP2.4. (3) Supporting the risk mapping and other means of risk visualisation in WP 4.4.

Especially the activities in RP 3 require input in form of the selected testing scenarios earlyin the project. It is therefore very important for RP 1 to enable a rapid scenario selection. Thisdemand is met through intense activity during the first 18 months and by applying a tiered approachin which a sequence of selected chemical categories are analysed subsequently and used as trainingset for development of the scenario selection methodology. In order to facilitate the methoddevelopment the substance categories need to cover a large variety in product types and usesincluding different properties in fate and toxicity. In this way four categories of chemical substanceswill be analysed for method development purposes during the first 18 months resulting in outputs inthe following order: pesticides, pharmaceuticals, biocides and VOCs/semi-VOCs. The VOCs/semi-VOCs will be treated as two sets: A sub set of the most well known substances for which results canbe made relatively quickly and a more comprehensive set including lesser known substances wheremore project time is needed for establishing the data background. In the case of pesticides therelease rate to the environment is relatively well defined and heavily related to agricultural practice.For pharmaceuticals the release will be governed by both wastewater treatment, including sludgedisposal, and the veterinary use and following release from agri- and aqua-cultures. The release rateof biocides and VOCs/semi-VOCs can be more difficult to quantify and closely linked to specifictype of products for e.g. wood conservation, ship painting and consumer products in general.However, biocide are not poorly described in general and they are judged to be easier to handle thanVOCs/semi-VOCs. In principle all chemicals that can be classified as VOCs/semi-VOCs must beincluded in the analysis, which implies that it is essential to have an explicit definition ofVOCs/semi-VOCs. The definition of VOCs/semi-VOCs is, however, not consistent. Someguidelines are suggested by the EMEP-guidelines for calculation and reporting of emissions. Theamount of chemicals that fulfil these criteria is large and during the first 18 months the project willinclude as many substances as possible from the category of VOCs/semi-VOCs.

The data collection during the first 18 months includes three scaling levels: (1) Europeanscale, where the available data may be rather limited with respect to parameters. (2) Regional scaleincluding better defined conditions where more detailed data are available. (3) Local level, having aone to one relationship to the detailed risk assessment in the context of many high individualexposures at selected areas. An ideal data set for method development will consist of detailed andrepresentative data. However, such an ideal data set hardly exists and the following strategy will beapplied in order to solve that problem:• European wide data set will be applied for all possible type of data;• The highly developed regions Denmark, the Netherlands and northern Italy will be used as

primary pilot areas for detailed data sets;• Lithuania and southern Italy will be focus areas for detailed data as much as possible;• An explorative study will be undertaken identifying other regions in southern and eastern

Europe

79

During the first 18 months there will be a close interaction between WP 1.1 and WP 1.2 and after 9month from the project start a 3-day RP 1-organised workshop will be held between theNOMIRALCE partners to discuss the following issues:• The use of standardised formats for spatial data to enhance the exchange of these data within the

NOMIRACLE project (based on input from the data-management section of NOMIRACLE,WP 5.2);

• The selection of regions in southern and eastern Europe based on the results of the explorativestudy in the scenario ranking;

• Identification of data gaps for the scenario selection and options to deal with these gaps;• Data demands for criteria setting necessary for the scenario selection in WP 1.2;• Evaluation of the scenario selection procedure based on the developed praxis for pesticides.

Detailed Implementation Plan for Research Pillar 2 – First 18 Months

Cumulative risk assessment aims at the analysis, characterization and possible quantification of thecombined risks to human health and the environment from aggregate exposures to multiple chemi-cal and non-chemical stressors. For each stressor, aggregate exposure includes consideration of allrelevant routes of contact with the organisms under investigation.

Research Pillar 2 (RP 2) addresses the problems associated with characterizing exposure in arealistic way. It aims at developing methodologies that allow the quantification of exposure tochemical stressors in environmental compartments in the field, and in typical indoor situations suchas homes, kindergartens and public buildings that are of primary relevance for environmentaleffects on human health. A further goal of RP 2 is to enable a sound quantification of exposure inlaboratory biotest systems that are needed to quantify toxic and ecotoxic potencies of chemical viaexposure-effect curves (see RP 3). This activity relates to soil-, sediment- and cell-based systemswhere sorption to abiotic or biotic matrices forms an important confounding factor in the context ofa sound exposure characterization, and to co-solvency effects in multiple exposure situations.

At present, the predictive exposure characterization is hampered by serious shortcomings:Multimedia fate models are based on phase partitioning algorithms that are not suited for morecomplex bioactive agents such as biocides, pesticides and pharmaceuticals. Secondly, such modelsprovide only (generically derived) total compound concentrations, and do not specify the compoundfraction that is actually available under the relevant (realistic) matrix conditions. Thirdly, exposurepredictions suffer from the lack of adequate models for media-specific degradation rates, andfourthly the models currently used to support risk assessment in the regulatory context are genericin nature and do not account for spatial and temporal variation. Fifthly, indoor exposure is nottreated adequately, despite its high relevance for human health, including emissions from consumerproducts.

Major current experimental deficiencies relate to the proper determination of the contents ofmodern bioactive agents in environmental media, and in particular to the quantification of theavailable exposure under matrix-specific conditions. Moreover, experimental techniques are neededto characterize the sorption of compounds to complex environmental matrices in a realistic way,which will allow the subsequent development of mechanistically sound phase partitioning models.

In the first 18 months, four interlinked research areas will be pursued, which will set thestage for developing a suite of methods for the characterization and quantification of available ex-posure. The research will be organized in four work packages (WPs) as outlined below.

Figure B.8-2 shows the major information flow across the four WPs of RP 2 as well as withthe other research pillars.

WP 2.1 Matrix-compound interaction (Leader: Gerrit Schüürmann, UFZ)The WP aims at developing a methodology that allows to parameterize the phase partitioning ofmore polar and in particular multifunctional chemical agents (e.g. biocides, pesticides, pharmaceuti-

80

cals) in a realistic way. The challenge is to consider major interaction forces such as dispersioninteractions, polar interactions and hydrogen bonding explicitly.

In the first 18 months, model systems will be built to characterize the relevant interacttioncapabilities of those soils that represent major European types as identified in WP 1.2 (EAWAG). Acomplementary effort with investigate membrane-water partitioning to obtain a more realistic hy-drophobicity descriptor for the bioconcentration of more complex xenobiotics in organisms (UFZ).

The parameterization of the phase partitioning requires both (measured) phase parametersand so-called Abraham descriptors of the compounds. For the latter, a method will be developed tocalculate the relevant parameters from molecular structure. The initial focus will be on a parameter-isation of the hydrogen bond donor and acceptor capabilities as one of the major drivers of phasepartitioning. To this end, computational chemistry will be employed, covering both a 2D and 3Dlevel of description, e.g. electrotopological parameters as well as quantum chemical parametersbased on perturbational molecular orbital theory (UFZ).

By the end of the first phase, experimental data will be available for some reference soilsand judiciously selected compounds, and the hydrogen bonding descriptors will be evaluated toidentify the most promising route for their final parameterization.

WP 2.2

Organismic dose

WP 2.4

WP 2.1

WP 2.3

Metabolic ratesand pathways

Cumulative effects

Matrix-specific available exposure Indoor exposure

RP 4 Cumulative risk

QSAR screening

RP 1 Compound selection

RP 3 Biotest systems

Phase partitionalgorithms

RP 5 Data management, Training, Dissemination

Management Board

Spatially explicit exposure

Risk scenarios

Reference soils/sediments Reference compounds

RP 2 Analytical competence

Metabolism

Figure B.8-2: Major information flow within RP 2 as well as between RP 2 and the other researchpillars.

81

WP 2.2 Available exposure (Leader: Philipp Mayer, NERI)The goal is to develop a suite of methods that allows to determine available exposure. This creates aunifying assessment strategy across matrix-specific contamination situations and associatedenvironmental processes. Complex exposure situations such as soil contamination are characterizedby a superposition of competing sorption and transfer processes, and matrix-bound compoundfractions may form a reservoir that can be mobilized under changing pH and water regimes. Asregards human health, indoor exposure including emission from consumer products is of primaryconcern, and here the inhalative pathway plays the major role.

With respect to environmental exposure, major activities in the first 18 months include theestablishment of analytical competence for determining contents of modern bioactive agents (se-lected in WP 1.1) in waters, sediments and soils (CSIC, ITM), and the investigation of availableexposure pursuing two innovative approaches for organic contaminants (available quantity vs. equi-librium sampling; NERI, ITM), and of one new method for heavy metals (diffusive flux; ULANC).Besides reference soils and sediments as selected in WP 1.1, soil-, sediment- and cell-based biotestsystems under investigation in RP 3 will be addressed, to enable the derivation of sound exposure-effect curves and associated effect parameters (EC50 etc.). By the end of the first project phase, anassessment strategy will be developed that integrates the different availability parameters in thecontext of cumulative risk assessment.

As regards human exposure, two major features are addressed: Firstly, the concept of avail-able exposure will be transferred to human health-relevant biotest systems (ciliates, immunocom-petent and other human cells) as employed in RP 3. This will allow mechanistically soundexposure-effect investigations of VOC compounds relevant for the inhalative pathway, and willform the basis for developing the respective structure-activity relationships as foreseen in WP 3.3.Secondly, indoor measurement campaigns in homes, kindergartens and public buildings will beundertaken in major European cities in close cooperation with local authorities. Here, a major resultof the first 18 months will be the characterization of indoor vs. outdoor VOC patterns andassociated indicator components, which will allow the development of targeted strategies for futurehuman health surveys.

WP 2.3 Metabolic fate (Leader: Ovanes Mekenyan, LMC)A major determinant for the fate and exposure of environmental contaminants is their susceptibilityto undergo degradation under natural conditions. Existing methods to predict biodegradation ratesare not targeted to soil and sediment compartments, and their parameterization did not consideremerging bioactive compounds. Similarly, the currently available reference scheme to predictphotolysis rates from molecular structure is an increment method that falls short for moremultifunctional compounds with complex intramolecular interactions. The goal of this WP is todevelop a suite of methods that allow the prediction of degradation rates in major environmentalcompartments and of metabolic pathways from molecular structure. To this end, both newexperimental data and new computational methodologies are needed.

In the experimental branch, degradation kinetics and metabolite formation in relevant soiltypes (UFZ), sediment and water (ECT) are investigated for the reference set of bioactive com-pounds (WP 1.1). A particular focus will be on soils that represent most complex exposure situa-tions. Here, regression and configuration frequency analysis will be employed to determine the soilfactors that govern the biodegradability (pH, TOC, CEC, clay, microbial mass content; UFZ).

In the theoretical branch, three complementary research lines will be pursued: Firstly, sta-tistical methods based on artificial intelligence will serve to explore a mapping between existingbiodegradation data according to OECD 301 and degradation in soil, sediment and water. Here, thefocus is on crucial system parameters such as soil type, environmental milieu, microbial populationcharacteristics as well as relevant compound properties. Secondly, predefined moleculartransformations form the starting point for the development of rule-based systems to predictdegradation pathways and prevalent metabolites, employing arithmetic (LMC) as well as non-

82

numerical (LHASA) probabilistic strategies. Thirdly, the prediction of photolysis rate constants willbe addressed through the development of a method based on perturbational molecular orbital theory(UFZ).

WP 2.4 Region-specific environmental fate (Leader: Mark Huijbregts, DESUN)Multimedia models are suitable instruments for exposure assessments on the regional and con-tinental scale. At present, however, predicted environmental concentrations of compounds sufferfrom several shortcomings as outlined above. Besides the lack of algorithms to adequately treatmultifunctional compounds and their metabolic fate (see WPs 2.1 – 2.3), these models do notproperly account for the spatial variation across Europe. This WP aims at developing a suite ofempirically evaluated fate and exposure models of various spatial scales. In particular, the modelsuite will offer a tiered approach to accommodate data-poor and data-rich situations

The theoretical work focuses on four interlinked research lines. One focus is on building aspatially explicit soil module that includes, in a multimedia fate modelling context, pertinent in-formation from the European Soils Database and related sources (JRC). Because the spatial scaleneeded for a given scenario depends on the environmental processes of concern and on the propertyprofile of the compounds under investigation, the model suite will contain a component to supportthe selection of the optimal spatial resolution for the scenario of interest. Here, probabilisticmodelling techniques will be employed to develop respective rules, considering in particular therelative importance of spatial specification as compared to the uncertainty in substance-specificparameters (DESUN, ITM, RIVM).

A further major activity focuses on the evaluation of model results using measured fielddata, which forms a complementary strategy to identify the optimal spatial resolutions for givenscenarios (ITM, DESUN). Finally, targeted sensitivity studies will be undertaken to unravel syste-matic relationships between property profiles of compounds and predicted exposure patterns,employing cognitive neural network-based intelligent systems that will be trained with multimediasystem behaviour (URV).

Detailed implementation plan for research pillar 3 – first 18 months

In the first 18 months, Research Pillar 3 will take forward detailed effect assessments of prioritysingle chemicals, chemical mixtures in WP 3.1 and multiple stressors in WP 3.2, in complement towork in RP 2 that will develop new approaches to exposure prediction. Quantitative informationgenerated will be collated, providing a basis for risk prediction and uncertainty analysis in WP 4.1.This process of data feed will be ongoing throughout the project, but the approaches and methodsfor primary data collection and analysis will be defined at the outset of each workpackage. Theanalysis of single chemical and combined stressor responses in diverse species in WP 3.1 and WP3.2 will be supported by experimental work that will elucidate the toxicokinetic (WP 3.3) andmolecular (WP 3.4) mechanisms of the observed responses. This will allow the potential forconservation of combined effects between species, life-stages, sexes and specific population sub-groups to be identified.

WP 3.1 Interactive toxicological effects on diverse biological systems (Leader: AlmutGerhardt, LIMCO)This work package will establish the toxicological responses of diverse species to prioritised singlechemicals and chemical mixtures. Measurement of the toxicity effects on each cell/tissue type,species, sex and life-stage will feed data directly into the project database to augment any existingdata. Collection of this data as a single resource will increase the power of probabilistic based riskassessments conducted for susceptible species and population sub-groups that are a key componentof WP 4.1 and 4.2.Work in WP 3.1 will start by liasing with RP 1 and RP 2 to derive priority lists of substances to takeforward into effect assessment. As work in NOMIRACLE proceeds this priority list will be revised

83

and updated as part of WP 1.2, meaning that at any time, effects assessment in WP 3.1 can betargeted on compounds and mixtures predicted to have the highest probability of co-occurrence andtoxic effect. Another key aspect of mixture analysis that will be dealt with early in the workpackage is finalisation of a robust framework for the mixture experiments design and analysis. Theapproach that will be used is based on that developed by Jonker (2003) and Jonker et al. (2004, inpress) during the MIXTOX (ENV4-CT97-0507) project. The data analysis method is based on thelong-standing models of concentration addition (for chemicals with similar modes of action) andindependent action (for chemicals with different modes of action), which are applicable to allchemical classes of compounds. These existing response models are used as a reference againstwhich any mixture toxicity dataset can be analysed. Using this modelling approach, fourbiologically relevant deviation patterns from the reference model can be identified by means oflikelihood analysis: no deviation, absolute deviation (synergism/antagonism), dose effect leveldeviation and dose ratio dependent deviation (see B1 and B4 for details on these response). Toensure this design is implemented across all mixed stressor exposures in NOMIRACLE (forchemical mixtures in WP 3.1 and multiple stressors in WP 3.2), a guidance document for the designof scientifically robust combined stressor experiments will be written. To support this, the existingsoftware used for data analysis by Jonker et al. (2004, in press) will be written as a softwareprogram in the early stage of this work package. This will be disseminated to all partners and alsomade available through the NOMIRACLE website to the wider risk analysis community.Initial assessment of the top 25 single contaminant and top 12 mixture scenarios prioritised in RP 1will be undertaken using a set of rapid “screening” bioassays. These will comprise the Vibriofischeri Microtox system, the single cell system Tetrahymena pyriformis, benthic invertebrateexposures (e.g. with Tubifex tubifex or Chironomidae), fish embryo tests and a set of novel andspecifically developed human cell lines. The mixtures analysed will comprise combinations ofchemicals with similar, dissimilar and unknown modes of action. The aim of screening assays willbe to give a rapid quantitative analysis of the nature of dose response relationships, includingresponse surfaces for mixtures. Data for single compounds will be used to generate QSARs foreffect of compounds in different environmental compartments as applicable during NOMIRACLE.Response surfaces for mixtures will be modelled and the data analysis software used to identifycombinations that both adhere to the concentration addition and independent action models and alsocombinations that show significant deviations from these ideals. All experimental data and analysisresults will be returned to WP 1.2 where it can be used to refine the prioritisation of specificcombinations.

The detailed assessment for single compounds will encompass a taxonomic diversityranging from single celled organisms (Dictyostelium), through plants (Lemna, Trifolium pratense,Lolium perenne, Sinapis alba), invertebrates Daphnia magna, Caenorhabditis elegans,Chironomida, Mytilus galloprovincialis, earthworms, Folsomia candida, carabids) to in particularcases (where there is a justified ethical argument), fish and higher vertebrates such as rodents. Asapplicable, different population sub-groups (based on e.g. age, immune status, and reproductivecondition) will be tested. Co-ordination with WP 1.2 will define the species and life-stages that willbe used for this detailed analysis, depending on the most realistic exposure scenario e.g. throughurban/indoor air, soil and water. During exposures, a series of responses (see Section B4 for acomprehensive list) will be measured. This will provide essential information on the inter-specificcomparability of toxicity for the analysis of sensitivity functions in WP 4.1.

As well as testing the effects of single compounds, two simple (binary) mixtures will also beinvestigated in all species in the first phase (with further simple and complex mixtures measuredlater in NOMIRACLE). One will consist of chemicals of similar mode of action whose toxicity inthe screening assays is adequately described by the concentration addition model; the second ofchemicals with different modes of action whose toxicity in the screening assays is correctlydescribed by independent action model. As for the single compounds, the chemical mixture studieswill span a diversity of species. Exposure will be conducted using similar methods as for the singlecompounds, with experiments designed and analysed according to the previously outlined

84

framework. Completion of this detailed assessment of each species will establish the conservationof joint response effects between different biological systems. All data collected will be collated inthe project database, making it available for species sensitivity and deviation function probabilityanalysis in WP 4.2.

WP 3.2 Combined effects of natural stressors and chemicals (Leader: Martin Holmstrup,NERI)This work package is key to the aims of NOMIRACLE as it will place the classical empiricalmixture toxicity analysis in the context of the diverse environmental conditions that exist acrossEurope thereby unifying the fields of ecotoxicology and stress ecology. The environmental andpopulation specific factors considered will be selected in the prioritisation and scooping processundertaken in RP 1. Scenarios selected will be relevant for the systems under study. For instance,the influence of pathogens/ allergens may be most relevant in humans/mammals, the influence ofdrought most relevant for soil organisms, anoxia for aquatic organisms and so on. Biologicalsystems used for these mixed stressor studies will overlap closely with those used in WP 3.1 (listedin Section B4) allowing us to establish response conservation for single and later multiple stressors.The approach to identifying the interactive effects of chemical and environmental stressors willmirror that used for contaminant mixtures in WP 3.1. First the effects of the single prioritisedenvironmental stressor on relevant organisms will be described. These responses will be collatedthereby providing spin-off information concerning the limits of the tolerance of diverse taxa toenvironmental variation and change. Once we have established the effects of single environmentalfactors the effects of these in combination with chemicals taken from the prioritisation exercise inWP 1.2 will be investigated. The same experimental design and data analysis approach will be usedin the multiple stressor studies as used for the contaminant mixtures investigated in WP 3.1.Interactions will be modelled initially assuming independent action, although, as the data modellingapproach can be used without a priori knowledge of the interaction mechanism, this assumption canbe validated for each multiple stressor experiment conducted. For both analyses, likelihood analysiswill be applied in order to identify deviations (absolute synergism/antagonism, effect level andration dependent) from the underlying model. When clear and consistent interactions betweenenvironmental and chemical stress factors are identified these responses will be catalogued forsubsequent analysis of underlying toxicokinetic and molecular mechanisms in WP 3.3 and 3.4. Thisdata will be collated and stored, thereby providing essential information for cumulative riskassessment for vulnerable populations and natural systems.

WP 3.3 Toxicokinetic modelling (Leader: C. Van Gestel, VU)To give a more mechanistic understanding of mode of action and interaction for single compounds,chemical mixtures and multiple stressors WP 3.3 will establish substance toxicokinetics(accumulation, compartmentalisation, metabolism, elimination) in taxa from different systems withvarying physiologies. In the first instance the most suitable existing toxicokinetic models for eachspecies will be identified in collaboration with partners in WP 4.1 and further refinement anddevelopment made to these as required. A further early component will be the generation ofanalytical protocols for substance families and compounds. Colleagues involved in WP 2.2 willhelp identify suitable methods and WRcNSF will provide advice and technical supervision to assistpartners in further refining these test methods to make them applicable for measuring chemicalconcentrations in each test organism. The resulting Standard Operating Procedures provided by WP2.2 or generated in this WP, will then be used in all future analyses. To ensure properimplementation of the analytical methods, a series of research training visits will be initiatedbetween the partners in WP 3.3 (particularly WRcNSF) and WP 2.2. Further, suitable qualitycontrols (e.g. spiked samples) and reference materials (if available) will be identified to providequality control for all analyses.

85

With initial toxicokinetic models and analytical protocols in place, an initial set oftoxicokinetic assessment will be made. These will be made for two compounds selected from theprioritisation list in all species. This initial data will be used to investigate the utility of thetoxicokinetic models refined for each species. If required, further work will take place incollaboration with partners in WP 4.1 to optimise each toxicokinetic model. As will be the casethroughout the project, data concerning toxicokinetic parameters for compounds (both singly and incombination) will be recorded into the project database. This will provide data for input into riskassessment models that consider bioaccumulation.

On completion of the initial experimental work and modelling exercise, the toxicokineticmodels will be ready for use in single compound and, importantly, for multiple compoundassessments. As a prelude to this, a full literature review on the toxicokinetics of chemicals inmixtures will be performed, including details of any descriptive QSARs. This will be an essentialstep towards the analysis of multiple stressor effects on compound toxicokinetics that will takeplace in the later stages of NOMIRACLE.

WP 3.4 Mechanisms of mixture toxicity (Leader: Aldo Viarengo, DSTA)Toxicokinetics data generated in WP 3.3 will only partly explain the toxic effects of singlecompounds and ultimately the interactive effects of chemical mixtures (WP 3.1) and multiplestressors (WP 3.2). To further elucidate the conserved and distinct molecular changes that underpinthe systemic response of species to complex exposure, a set of analysis will be undertaken usingestablished and state of the art histopatological, cellular, biochemical and molecular methods. Thesewill include measurement of metabolic responses, cellular and tissue specific histopathologicalchanges and the activities of proteins in stress response, chemical metabolism and antioxidantpathways. These specific assays will be supplemented by the use of global response analysis usingtranscriptomic, proteomic and metabolomic profiling methods, supported by appropriatebioinformatic analysis and pattern recognition techniques. Biochemical response analysis isconcentrated into a series of animal phyla, including mammalian cells and rodents. This is done toincrease the relevance of the data collected for predicting mode of toxic effect for human subjects.Work package 3.4 will start by identifying the most suitable protocols available for the specificanalysis in each species. Next, as applicable where possible, methods will be transferred betweenpartners so that a particular assay technique (e.g stress protein analysis, metabolomics, microarray,detoxification enzymes) can be utilised within each species. Formal training courses will beorganised and exchange visits organised. As reliable biochemical and molecular biological methodsare developed for each species, these will be posted on the NOMIRACLE web site. Ultimately apublisher will be sought for this material. Following refinement of protocols for each species, anassessment of baseline responses will be conducted. This will focus on measurement of variationbetween individuals held under environmental conditions within the optimum range. Thesebaselines will help to derive performance criteria that can be used as part of an algorithm to discernthe status of a given organism or receptor in subsequent exposures.

With baselines established for each species, focus will move to identification of mode ofaction, initially for single compounds. To optimise the use of resources in NOMIRACLE,information available on mode of action of particular compounds this will be collated. If there is nosuch information or it is not relevant to the biological system (e.g. herbicides in animals),toxicogenomics and metabolic pathway resources (e.g. KEGG metabolism, and pathDB metabolicpathway databases) containing information on chemical and stress response pathways will bescreened in order to derive any information on likely mechanisms. In cases where there is noinformation, experimental work to identify mode of action will be undertaken. As a prelude to this,the utility of our analysis tools will be established for two compounds with known specific mode ofaction (provisional a cholinesterase-inhibiting pesticide and a respiratory toxicant). This work willconfirm the suitability of the approach for mode of action assessment and highlight anymethodological refinements needed for optimum use of each biochemical method. After completion

86

of this validation phase, experimental work will begin for compounds for which mode of action isnot known or not relevant. In this work, profiling techniques (transcriptomics, metabolomics) willbe deployed and pattern recognition methods used to screen responses against existing data in orderto identify matching pathways and probable route of toxic effect. Targeted experiments will then beundertaken to link observed molecular differences with effects on specific biochemical pathways.This work will act as the prelude to our later work investigating modes of interaction betweenchemical mixtures and combinations of chemical and environmental stressors.

Detailed implementation plan for research pillar 4 – first 18 months

Effect datafrom RP3

Environmentalconcentrationdata from RP2

Scenarioselection anddata from RP1

New concepts andtechniques for

probabilistic riskassessment

Dealing with multipleand complex risks in

a managementcontext

Risk presentationand visualization

Explicit modeling ofexposure and risksover space and timeeffect predictions

PRA techniquesWP4.1

WP4.3

WP4.2

WP4.4

Dissemination in WP5.4

Research Pillar 4

Figure B.8-3: Relations within the work packages of RP 4 and with the other RPs.

Research Pillar 4 integrates the results of RP 1, RP 2 and RP 3 within a probabilistic and spatiallyexplicit modelling framework that is tailored to support risk management decisions. The work in RP4 is organised into four work packages. Figure B.8-3 illustrates the relations between these workpackages and with the other RPs.

WP 4.1 concentrates on the development of new concepts and techniques for probabilisticrisk assessment. Although some techniques developed in WP 4.1 will be generally applicable (e.g.,separation of uncertainty and variability), an important part of the work will concentrate on theprediction of probabilistic effect estimates based on an analysis of the effect data gathered in RP 3.Within WP 4.2, the effect predictions are combined with exposure estimates of RP 2 in a spatiallyand temporally explicit modelling framework. The situations and regions modelled in WP 4.2

87

depend on the scenarios identified in RP 1. Dealing with cumulative and complex risks in riskmanagement will be the main topic of WP 4.3. This WP is strongly linked with WP 4.1 becauseuncertainty is an important issue in dealing with risks. WP 4.3 will also take into account riskperceptions among key actor groups, risk communication strategies and options to implement theprecautionary principle, thus addressing and providing input to dissemination and exploitation inWP 5.4. Finally, the resulting cumulative risk estimates and associated uncertainties will bepresented and visualised for a selected number of EU regions in WP 4.4, which will provide anothermain output for dissemination in WP 5.4.

RP 4 will thus have a highly integrative character and strongly depends on the resultsproduced by the other research pillars, which obviously will take some time. However, instead ofawaiting results from the other RPs, several activities in RP 4 will be initiated at the start of theNOMIRACLE project, i.e., the theoretical studies and the development of general methods andalgorithms that in a later stage will be applied to data and results from other RPs. Development ofsuch methods and algorithms will initially concentrate on situations and locations for whichextensive data are available among the RP 4 partners. Examples are the development of anecological random walk model for a flood plain area along the river Waal in the Netherlands (WP4.2), the modelling of spatially aggregated risks of airborne chemicals for children in the region ofLeipzig in Germany (WP 4.2), the establishment of relationships between exposure and healthimpacts for the Ebro delta in Catalonia (WP 4.2), the separation of uncertainty and variability incontaminant exposure predictions for the Dutch population (WP 4.1), and assessment of risks andmanagement strategies for dioxin-like compounds and related decision and policy analyses (WP4.3). Other research activities that will be initiated from the start of the NOMIRACLE projectinclude the derivation of probabilistic uncertainty factors based on existing databases (WP 4.1), theanalysis of regulatory frameworks and management performance (WP 4.3), and the elicitation ofrisk perceptions and views of experts and stakeholders (WP 4.3).

WP 4.1 New concepts and techniques for probabilistic risk assessment (Leader: Ad Ragas,DESUN)During the first eighteen months of the NOMIRACLE project, the work in WP 4.1 will concentrateon the development of general probabilistic techniques and on research activities that make use ofexisting databases. WP 4.1 starts with a short desk top study in order to identify options forharmonization of the analytical frameworks used in the meta-analysis of human and ecologicaltoxicity data (DESUN). The results of the desk top study will be discussed by both partnersinvolved in the derivation of probabilistic UFs (DESUN and USOUTH) and harmonization willconcentrate on using similar mechanistic descriptors for analyzing human and ecological toxicitydata, e.g., substance parameters, the genetic predisposition of receptors and the toxicological modeof action of substances. Subsequently, probabilistic uncertainty factors (UFs) for human riskassessment will be derived for inter-individual and interspecies differences based onpharmacokinetic data for exposure to single compounds and chemical mixtures that are handled bymajor phase I and phase II metabolic pathways (USOUTH). This activity largely uses existingdatabases that will be supplemented and updated with meta-analyses of published data. Anotheractivity that starts at the beginning of the project period is the extension of a new ecotoxicologicaleffect model that was developed in the FP5 OMNITOX project (EPFL). The main objective is tointegrate toxicity of mixtures and other stressors (e.g., eutrophication, acidification, etc) in thismethod, realizing a truly integrative method for comparative risk assessment. After month 6, asubproject on derivation of a probabilistic NEC will start in WP 4.1 (VU). This research will befocused on modelling the effects on mixtures of 2 compounds, with survival and reproduction asend points. The model will be tested against experimental data from literature, previously gathereddata and to the first experimental data from RP 3. Also after month 6, a subproject will start thatconcentrates on the separation of true uncertainty and interindividual variability in risk predictionsof an integrated human exposure model (DESUN). The outcome of this subproject specifies thepopulation fraction at risk due to interindividual variability in consumption and activity patterns and

88

details the probability of this risk. The results will allow policy-makers to take a better-informeddecision on risks of human exposure to chemical contaminants through multiple environmentalmedia.

The following activities have been scheduled for later stages of WP 4.1 (after month 18):• Derivation of a probabilistic NEC for complex mixtures• Analysis of inter-individual and interspecies differences in pharmacodynamics for single probe

subsrates and chemical mixtures (drug interaction data) to provide a further basis for thederivation of probabilistic UFs for human risk assessment

• Analysis of acute and chronic ecotoxicity data to derive probabilistic UFs for ecological riskassessment

• Development and application of a probabilistic technique to quantify the value of newinformation in human and ecological risk assessment

• Demonstration of the newly developed probabilistic techniques in WP 4.1 in several casestudies that will be selected on the basis of the scenario ranking in WP 1.2

WP 4.2 Explicit modelling of exposure and risk in space and time (Leader: Uwe Schlink,UFZ)During the first eighteen months, the work in WP 4.2 concentrates on the development of novelmodelling techniques to integrate exposure and risk over space and/or time. WP 4.2 starts with twostudies that develop and apply techniques to analyse relations between exposure, receptors andstressors. The first study aims to identify causal parameters that determine human exposure and riskby integration of neural network cluster- and variable-analysis within a GIS environment atdifferent spatial scales (URV). These techniques will be applied to a specific region for whichextensive data are available, e.g., the Ebro delta or another Catalan region.The second studyanalyses relations between species traits and vulnerability in order to identify critical ecologicalpathways and parameters for ecological risk assessment (ALTERRA). Both studies will provideuseful input for the development of human and ecological random walk models. This activity startsin month 7 with the development of a model that describes the accumulation of pollutants inselected ecological receptors (i.e., Little Owl, Badger, shrews and mice) in a floodplain area alongthe river Waal in the Netherlands (DESUN). This area has been chosen as a pilot area becauseextensive pollution and ecological data are available, which can be used to develop, test andvalidate algorithms that describe random walk patterns, ecological interactions andbioaccumulation. During the second stage of the project, these models will be extended to humanreceptors. The results of the random walk models pertain to relatively small-scale areas, whichraises the question of how to interpret risks at different spatial scale levels. This scale level issuewill be addressed in several studies. For ecological risk assessment, extrapolation routines will bederived to up-scale the results of the small-scale ecological modelling to the larger scale(ALTERRA). For human risk assessment, the use of different spatial scales in analysing humanrisks will demonstrate the relative limitations of using large-area scales to identify impact clusters(URV). Problems related to spatial aggregation of risk data are addressed in a study that appliesmultivariate Bayesian techniques with conditional autoregressive terms (CAR) to model the spatialpatterns of risks of airborne chemicals to airway diseases and allergies in children in the city area ofLeipzig (UFZ). The established CAR-model will serve as a reference for further NOMIRACLEwork and will be refined in the second stage of the project for other levels of aggregation, for otherhealth effects and other regions. A final issue addressed during the first stage of WP 4.2 is theestimation of long-term exposure to chemicals utilizing only short-term measurements (UFZ). Dataof the contamination of indoor air with hydrocarbons will be combined with data on behaviour(ventilation practice) and activities (e.g. renovation) of the inhabitants using ANOVA and neuralnetwork models in order to predict the long-term exposure to these hydrocarbons.

The following activities have been scheduled for later stages of WP 4.2 (after month 18):

89

• Extension of the random walk models to human receptors• Refinement of the CAR-model for other levels of aggregation, for other health effects and other

regions• Further exploration of statistical techniques (fuzzy ARTMAP, Kohonen’s self-organizing maps)

to identify causal parameters that determine human exposure and cumulative risk anddemonstration for selected number of EU regions

• Refinement of routines for up-scaling ecological risks and demonstration to selected regionsWP 4.3 Dealing with multiple and complex risks in a management context (Leader: TimoAssmuth, SYKE)During the first eighteen months of WP 4.3, the main objective is to strengthen the knowledge basefor managing multiple and complex risks, with an emphasis on dealing with uncertainty andambiguity. This objective will be realised by studying cognitive, social and contextual aspects ofintegrated risk assessment of chemicals and other stressors. The activities in this work package areorganised in three subprojects, i.e., (1) evaluation of risk perceptions, cognition and knowledge, (2)management strategies and policies in risk assessment, and (3) risk communication. These will beclosely integrated, with more than one partner participating in all of them. All subprojects build ongeneral notions of risk, but the emphasis is on risks caused by chemicals and other stressors.The first subproject will build on previous studies of the various perspectives on chemical risks.Literature and policy documents will be analyzed and expert opinions surveyed in order to describethe relations between cognition, knowledge and perception of risks (SYKE). Furthermore, thedevelopment of knowledge and the modes of reasoning for chemical risks will be analyzed (JRC).The second subproject will address legal and regulatory frameworks and policy principles inintegrated risk assessment. Deliberation process in chemical-related risk management will bestudied (DIA). Strategic and policy issues in relevant risk assessments will be identified andanalyzed, including relevant issues under REACH (SYKE). Finally, guidance will be provided forapproaches to deal with uncertainty and quality assurance in integrated risk assessment of chemicalsand other stressors (JRC).The third subproject will identify options to improve the communicationabout multiple and complex risks, based on desk studies of communication concepts (SYKE) andan initial study of participatory multi-actor risk and uncertainty communication (DIA). Studies willbe made of communication among researchers and between experts and stakeholders, e.g. inconnection with project meetings. From this work, needs, constraints and opportunities for cross-disciplinary and participatory communication will be identified.

The following activities have been scheduled for the later stages of WP 4.3 (after month 18):• In-depth theoretical studies of multi-dimensional risks and associated uncertainties in the EU

chemicals area, including interdisciplinary comparison and evaluation of risks, risk managementpolicy goals and options, and their interrelationships with assessment approaches

• Empirical research in risk perceptions, cognition and opinions among key actor groups, focusingon experts and stakeholders and on case substance classes, regions and assessment themes

• Investigations of needs, qualities and uses of knowledge about complex risks particularly frommixtures of chemicals having specific activities, including evaluations of data and modelsproduced, processed and applied within the project

• Studies of risk and uncertainty communication within the chemicals area particularly in relationto the above cases, as part of wider scientific and social discourses

• In depth analysis, on the basis of documents and data collected through the other tasks, of therelation between assessment and management of multiple risk and its implications forimplementing a precautionary approach. This relation will in particular be studied along threeimportant axes : regional resolution, sector roles and actor roles

• Development and testing of methods and guidance for communicating and addressingmultidimensional risks and uncertainties associated with chemicals in management contexts, forfurther development and for training, dissemination and exploitation

90

WP 4.4 Risk presentation and visualisation (Leader: Joost Lahr, ALTERRA)WP 4.4 integrates the results of the first 18 months of the NOMIRACLE project by producing someexample cumulative risk maps for the EU and selected EU regions. The WP is closely linked to WP1.1 (data gathering) and WP 1.2 (scenario ranking), because these WPs provide the input forproducing the risk maps. Indicators for cumulative risks will be evaluated based on formerexperiences in risk mapping among WP 4.4 partners and literature on risk mapping andcommunication. The most appropriate techniques for presentation and visualization of cumulativerisks will be adopted. Coherence of data formats of the output delivered by the partners from otherwork packages (notably WP 1.1 and WP 1.2) will be guaranteed by adhering to standardised GISformats that will be discussed and adopted by the WP 4.4 partners during a workshop organised inWP 1.1. UFZ will develop guidelines for the aggregation of spatial data mainly in relation tosmoothing and attenuation phenomena, building upon previous experience and work done in WP1.1 and WP 4.2. The final and most important activity of the first stage of WP 4.4 is to identify toolsand techniques to produce risk maps for the EU and selected EU regions. JRC will focus on riskmaps for the EU and Northern Italy; ALTERRA will focus on maps for the Netherlands; NERI forDenmark, UFZ for Leipzig and URV for the Ebro delta. Dissemination of potential cumulative riskmaps will be dealt with in MP5.

The second stage of WP 4.4 (after month 18) will concentrate on:• Updating the ‘potential cumulative risk maps’ with information and new scientific insights from

the other research pillars and work packages• Visualization of the uncertainties in the updated risk estimates• Incorporation of the results of the risk perception and communication studies under WP 4.3 to

finalise the presentation and visualization methods

91

Work package list (18 month plan)

Work-packageNo1

Work package title LeadcontractorNo2

Person-months3

Startmonth4

Endmonth5

Deliv-erableNo6

1.1 Data background for scenarioselection

18 (JRC) 56 0 18 1.1.1.-1.1.5

1.2 Scenario selection andranking

1 (NERI) 48 0 48 1.2.1-1.2.4

2.1 Matrix-compoundinteractions

3 (UFZ) 54 0 48 2.1.1-2.1.6

2.2 Available exposure 1 (NERI) 102 0 48 2.2.1-2.2.9

2.3 Metabolic fate 35 (LMC) 144 0 48 2.3.1-2.3.5

2.4 Region specificenvironmental fate

4 (DESUN) 74 0 54 2.4.1-2.4.4

3.1 Interactive toxicology indiverse biological systems

24 (LIMCO) 130 4 54 3.1.1-3.1.3

3.2 Combined effects of naturalstressors and chemicals

1 (NERI) 68 4 54 3.2.1-3.2.3

3.3 Toxicokinetic – singlechemical uptake andinteractive effects

6 (VU) 50 7 54 3.3.1-3.3.3

3.4 Interactive mechanistictoxicology

5 (DSTA) 62 7 54 3.4.1-3.4.3

4.1 New concepts and techniquesfor probabilistic riskassessment

4 (DESUN) 61 0 48 4.1.1-4.1.4

4.2 Explicit modelling ofexposure and risk in spaceand time

3 (UFZ) 59 0 54 4.2.1-4.2.4

1 Work package number: WP 1 – WP n.2 Number of the contractor leading the work in this work package.3 The total number of person-months allocated to each work package.4 Relative start date for the work in the specific work packages, month 0 marking the start of the project, and all otherstart dates being relative to this start date.5 Relative end date, month 0 marking the start of the project, and all ends dates being relative to this start date.6 Deliverable number: Number for the deliverable(s)/result(s) mentioned in the work package: D1 - Dn.

92

4.3 Dealing with uncertainty andambiguity in riskmanagement

19 (SYKE) 31 0 60 4.3.1-4.3.5

4.4 Risk presentation andvisualisation

21(ALTERRA)

16 13 60 4.4.1-4.4.3

5.1 Administrative management 1 (NERI) 12 0 60 5.1.1-5.1.4

5.2 Data management 18 (JRC) 4 0 60 5.2.1-5.2.5

5.3 Training 13 (DBUA) 6 1 60 5.3.1-5.3.4

5.4 Dissemination 38(SYMLOG)

6 0 60 5.4.1-5.4.5

TOTAL 983

93

Deliverables list (18 month plan)

DeliverableNo7

Deliverable title Deliverydate8

Nature

9

Disseminationlevel10

D.1.1.1 Data package for pesticides. 7 R PP

D.1.1.2 Data package for pharmaceuticals, biocides andfirst sub set of VOCs/semi-VOCs.

12 R PP

D.1.1.3 Data package for VOCs/semi-VOCs 15 R PP

D.1.1.4 Data package for Landscape classification andland use

15 R PP

D.1.1.5 Full documentation to the data managementsection (5.2)

18 R PP

D.1.2.1 Scenario selection and criteria setting forpesticides

9 R PP

D.1.2.2 Scenario selection and criteria setting forpharmaceuticals, biocides, a first sub set ofVOCs/semi-VOCs

12 R PP

D.1.2.3 Scenario selection and criteria setting forVOCs/semi-VOCs

18 R PP

D.1.2.4 Documentation of the first version of selectionprocedure including initial uncertaintyevaluation

18 R PP

D.2.1.1 Compilation of phase parameters, and databaseof Abraham descriptors with chemical structurehandling

12 R PU

D.2.1.2 Validated experimental procedure for thedetermination of phase descriptors for soils, andfirst series of experimental phase partitioningdata of reference compounds

18 R PU

7 Deliverable numbers in order of delivery dates: D1 – Dn8 Month in which the deliverables will be available. Month 0 marking the start of the project, and all delivery datesbeing relative to this start date.9 Please indicate the nature of the deliverable using one of the following codes:

R = ReportP = PrototypeD = DemonstratorO = Other

10 Please indicate the dissemination level using one of the following codes:PU = PublicPP = Restricted to other programme participants (including the Commission Services).RE = Restricted to a group specified by the consortium (including the Commission Services).CO = Confidential, only for members of the consortium (including the Commission Services).

94

D.2.1.3 Validated experimental procedure for thedetermination of membrane-water partitioning,and first series of experimental data at 25°C andpH 7

18 R PU

D.2.1.4 Computerised increment method to estimateAbraham descriptors from molecular structure

18 R PU

D.2.1.5 Comparative analysis and evaluation ofmolecular descriptors for modelling hydrogenbonding

18 R PU

D.2.2.1 Analytical methods for multifunctionalxenobiotics such as biocides, pesticides andpharmaceuticals in water and sediment

12 R PU

D.2.2.2 Temporal and spatial variability of humanexposure-related VOC

12 R PU

D.2.2.3 Sampling methods for soils and sediments andrelevant biotest systems, covering organics(diffusive sampling) and cationic metals(depletive sampling)

18 R PU

D.2.2.4 Lab and field data of compound exposure, andassessment of availability parameters to functionas a mechanistic link between laboratorysystems and the field

18 R PU

D.2.2.5 Assessment of the (source-dependent)indoor/outdoor ratio of VOC

18 R PU

D.2.3.1 Validated experimental procedures for thedetermination of compound turnover in water,sediments and soils

12 R PU

D.2.3.2 Experimental biodegradation kinetics for somemodel compounds in water, sediment anddifferent soils

18 R PU

D.2.3.3 Evaluation of relevant parameters for deriving aQSPR for degradation rate parameters indifferent environmental media

18 R PU

D.2.3.4 Preliminary computer approaches to simulateterrestrial biodegradation

18 R PU

D.2.3.5 Predicted rates of degradation and metaboliteformation for compounds subject to multimediadate analysis in WP 2.4

18 R PU

D.2.4.1 Multimedia fate and exposure model withvarious spatial resolutions at the European level

12 R PU

D.2.4.2 Indication of the spatial detail 18 O PU

95

D.2.4.3 Inventory of improvement options in existingmodels, based on empirical model evaluation

18 R PU

D.2.4.4 Cognitive neural network-based intelligentsystem to identify the most important variablesfor the differences found in partitioningbehaviour, transport pathways and exposureroutes between chemicals

18 P PU

D.3.1.1 Finalisation of the scheme for mixtureexperiment design and data analysis includingdissemination of a guidance document andanalysis software

7 R PP

D.3.1.2 Development of screening and target methodsfor assessment of chemical effects on humanhealth including generation of new stableimmune reporter cell lines

15 R PU

D.3.1.3 Dose response profiles for the screening basedassessment for the first set of prioritisedchemicals and mixtures and a resultingprioritisation of the first set of mixturecombinations (of similar and similar mode ofaction) to be used in the multi-systemassessment

18 R PU

D.3.2.1 Methods to assess and a description of thebaseline responses to the natural stressors underconcern for the various test organisms/cell lines(dose-response relationships).

9 R PU

D.3.2.2 Methods to assess the impacts of relevantcombinations of cumulative stressors for acomprehensive set of taxa covering the entireenvironments and human communities of theEuropean continent.

15 R PU

D.3.2.3 Generation of ecotoxicity data of relevantcombinations of chemical end environmentalstress for the construction of response surfacesof a comprehensive set of taxa

18 R PU

D.3.3.1 Standard Operating Procedures (SOPs) for theanalysis of concentrations of selected chemicalsin the different test organisms

15 R PP

D.3.3.2 Preliminary report on the uptake and eliminationkinetics of two selected chemicals and theirmixture in different test organisms, uponexposure in soil or water

18 R PU

D.3.3.3 Literature report on the toxicokinetics ofchemicals in mixtures as a first step towards thedevelopment of a QSAR for mixture interactions

18 R PU

96

D.3.4.1 Repository of protocols for specific and globalprofiling of stress response pathways in humancells, mammalian models and environmentallyrelevant species

12 R PU

D.3.4.2 Completion of baseline analysis of responseprofiles for single stress response systems andglobal profiling responses in the selectedexperimental systems under ideal conditions

18 R PU

D.3.4.3 Datasets of analyses on the effects of twocompounds with a known specific mode ofaction on global and specific biochemicalresponses

18 R PU

D.4.1.1 Report on separation of true uncertainty andinter-individual variability in predicted risks ofthe Dutch population from exposure topesticides through all relevant environmentalpathways

18 R PU

D.4.1.2 A paper on the model formulation for effects ofa mixture of 2 compounds and a discussion ofthe application of the model to experimental data

18 R PU

D.4.1.3 Consolidated report describing the metabolism,pharmacokinetic and preliminarypharmacodynamic data in human subgroups ofthe population and tests species. Derivation ofuncertainty factors for each subgroup of thepopulation and test species for polymorphicelimination pathways after exposure to a singlechemical or to chemical mixtures

18 R PU

D.4.1.4 Report describing method for comparative riskassessment of multiple stressors

18 R PU

D.4.2.1 A random-walk model that describes theaccumulation of pollutants in selected ecologicalreceptors in a floodplain area along the riverWaal in the Netherlands

18 R PU

D.4.2.2 A fuzzy ARTMAP neural classifier to enablecluster analysis and demonstration of thevariable selection approach for an EU region

18 R PU

D.4.2.3 Demonstration of up-scaling methods, basedupon the small-scale modelling

18 R PU

D.4.2.4 Demonstration of the CAR-model assessinghealth risk in the city area of Leipzig

18 R PU

97

D.4.2.5 A temporal model for the indoor air pollutionwith volatile organic compounds

18 R PU

D.4.3.1 Conceptual models and mental maps of riskinformation and cognition in the IP domain,based on data on risk views and inference fromother RP and IP entities and actors, includingverbal, quantitative and visual perceptionmodels for comparative and cumulative riskassessment

12 R PU

D.4.3.2 Conceptual frameworks for the characterisationof knowledge of risks and its development, itsuse and its quality assurance in the IP domain,including risk information pedigrees.

9 R PU

D.4.3.3 Models of deliberation modes in responses touncertainty and ambiguity, to develop integratedapproaches and strategic guidance to riskassessment partners

15 R PU

D.4.3.4 Review of relevant risk communication researchwith data evaluations for partners, and ofcommunication models, including participatory,interactive comparative and visual approaches

15 R PU

D.4.3.5 Synthetic interim report and plan for furtherwork (including interviews, surveys, casestudies, policy and decision analyses, discourseanalysis; theory and methods development;outreach) and meetings reports and discussionpapers.

18 R PU

D.4.4.1 Methods and tools to provide cumulative riskmaps

15 R PU

D.4.4.2 Guidelines for the aggregation of spatial data 15 R PU

D.4.4.3 Examples of cumulative risk maps for EU and alimited number of selected EU regions

18 R PU

D.5.1.1 Financial and technical reports to the EuropeanCommission

6-18 R CO

D.5.1.2 Meetings of the Management Board (at least 3),the Advisory Board (2) and the GeneralAssembly (2)

3-18 O PP

D.5.1.3 Dissemination plan including internal andexternal homepage and a Quarterly Newsletter

6 O PU

D.5.1.4 Organisation of two scientific sessions (e.g. inSETAC Conferences)

6-18 O PU

98

D.5.2.1 Questionnaire on the nature and type of coredata sets utilised and generated by theNOMIRACLE project

2 O PP

D.5.2.2 Draft design document on NOMIRACLE datamanagement system and implementationstrategy

5 O PP

D.5.2.3 Draft NOMIRACLE data management systemand implementation strategy

8 R PP

D.5.2.4 Operational data management system 10 O PP

D.5.2.5 Review of the operation and functionality of theNOMIRACLE Data Management System

16 O PP

D.5.3.1 Organisation of an internal training session, atthe kick-off meeting

1 O PP

D.5.3.2 Organisation of three advanced studies courses 6-18 O PP

D.5.3.3 Organisation of one technical session 18 O PU

D.5.3.4 Programme for an international PhD programme 18 O PU

D.5.4.1 IP internal (including peer review) and externalreporting procedures

4 O PP

D.5.4.2 Templates for user feedback 4 O PP

D.5.4.3 Procedures to analyse and redirect feedback toWPs

4 O PP

D.5.4.4 Stakeholder mapping, focusing on key actorsand representatives of key civil society groups,including emergent ones

12 O PP

D.5.4.5 Highlights from First Periodic Report ofNOMIRACLE summarised for submission topopular science journals and other relevantmedia

18 R PU

99

Proposal no. Proposal Acronym: NOMIRACLE

RTD and innovation Demonstration Training Consortium Totalrelated activities activities activities activities

Part Cost Costs Requested grant Costs Requested grant Costs Requested grant Costs Requested grant Costs Requested grantno model Euro to the budget (Euro) Euro to the budget (Euro) Euro to the budget (Euro) Euro to the budget (Euro) Euro to the budget (Euro)1 FC 751.361 375.680 0 0 15.000 15.000 100.000 100.000 866.361 490.6802 FC 226.228 113.114 0 0 0 0 11.000 11.000 237.228 124.1143 FC 974.066 487.033 0 0 3.750 3.750 10.700 10.700 988.516 501.4834 AC 271.472 271.472 0 0 1.824 1.824 12.160 12.160 285.456 285.4565 AC 58.764 58.764 0 0 0 0 0 0 58.764 58.7646 AC 171.870 171.870 0 0 0 0 0 0 171.870 171.8707 FC 21.668 10.834 0 0 0 0 0 0 21.668 10.8348 AC 46.250 46.250 0 0 0 0 0 0 46.250 46.2509 AC 54.250 54.250 0 0 0 0 0 0 54.250 54.25010 AC 40.569 40.569 0 0 0 0 0 0 40.569 40.56911 AC 72.600 72.600 0 0 0 0 0 0 72.600 72.60012 AC 111.604 111.604 0 0 0 0 0 0 111.604 111.60413 AC 51.460 51.460 0 0 0 0 0 0 51.460 51.46014 AC 84.250 84.250 0 0 0 0 0 0 84.250 84.25015 FC 110.216 55.108 0 0 0 0 0 0 110.216 55.10816 FC 66.672 33.336 0 0 0 0 0 0 66.672 33.33617 AC 49.688 49.688 0 0 0 0 0 0 49.688 49.68818 FC 561.252 280.626 0 0 16.500 16.500 0 0 577.752 297.12619 FC 186.113 93.056 0 0 5.318 5.318 0 0 191.430 98.37420 AC 26.560 26.560 0 0 15.000 15.000 0 0 41.560 41.56021 FC 175.793 87.896 0 0 0 0 0 0 175.793 87.89622 FC 150.000 75.000 0 0 0 0 0 0 150.000 75.00023 FC 49.850 24.925 0 0 0 0 0 0 49.850 24.92524 FC 67.500 33.750 0 0 0 0 0 0 67.500 33.75025 AC 12.500 12.500 0 0 0 0 0 0 12.500 12.50026 FC 309.034 154.517 0 0 0 0 0 0 309.034 154.51727 AC 113.287 113.287 0 0 0 0 0 0 113.287 113.28728 FCF 32.488 16.244 0 0 5.274 5.274 0 0 37.762 21.51829 AC 97.100 97.100 0 0 0 0 0 0 97.100 97.10030 AC 108.200 108.200 0 0 0 0 0 0 108.200 108.20031 AC 135.000 135.000 0 0 0 0 0 0 135.000 135.00032 AC 148.511 148.511 0 0 3.056 3.056 0 0 151.567 151.56733 AC 196.411 196.411 0 0 0 0 0 0 196.411 196.41134 FC 119.000 59.500 0 0 0 0 0 0 119.000 59.50035 AC 66.500 66.500 0 0 0 0 0 0 66.500 66.50036 FC 110.760 55.380 0 0 0 0 0 0 110.760 55.38037 AC 106.500 106.500 0 0 0 0 0 0 106.500 106.50038 FC 123.864 61.932 0 0 0 0 0 0 123.864 61.932Total 6.059.209 4.041.277 0 0 65.721 65.721 133.860 133.860 6.258.790 4.240.858

NOMIRACLE - Part B 100

Work Package 1.1 description:

Establishment of data background for scenario selection (18 month plan)

Work package number WP 1.1 Start date or starting event: Month 0

Participant id DESUN UNIMIB JRC APINI AlterraPerson-months per participant 0.5 8 9 4 6

Participant id NERI UFZ JRCPerson-months per participant 21 4 3

ObjectivesThe purpose of the first 18 months of work is to secure the core data set for the following activities:(a) Scenario selection and related uncertainty estimates in WP1.2. (b) Input to the exposureassessment in WP 2.4 (in form of emission inventories, environmental characterisation and non-chemical stressors). (d) Supporting the risk mapping and other means of risk visualisation in WP4.4.

Description of work

WP1.1 starts by making an inventory of data required for the scenario ranking in WP1.2. Thisinventory list will be distributed among the NOMIRACLE partners with a request to indicate dataavailability at the participating institutes. Different data sources will be used partly from other partsof NOMIRACLE and partly from external data sources.

This will be done by providing access to data in the area of: (1) Substance related data relevant forintegrating exposure and effects of human and ecological character. (2) Product related data onsubstance composition and release rate including link to human activity. (3) Climatic, physical andecological conditions (e.g. precipitation and eutrophication) relevant for characterizing stressors forcumulative risk assessment. (4) Stressors for cumulative risk assessment linked to human activity(e.g. wastewater, land use and ozone). (5) Landscape classification and land use. (6) Presence andvulnerability of receptors.

The external sources used to supply information on substances will include IUCLID and otherdatabases from ECB, US EPA databases, the QSAR database created by the Danish EPA containingmore than 50 toxicological endpoints and 160.000 substances (Personal agreement with JayNiemela, The Danish EPA). Important information will also be gathered from a series of listingssuch as the CMR (carcinogenic, mutagenic and toxic to reproduction) from Directive 2001/59/EC,OSPARCOM from The Water Frame Work Directive and the PBT (Persistent, Bioaccumulating,Toxic) listing. In addition, UFZ will utilise the ChemProp modeling package, to predict a largevariety of relevant physical-chemical properties as well as toxic modes of action according topublished schemes Russom, Verhaar, Hermens (electrophile alerts). JRC will help to identifyappropriate test substances for the Nomiracle project by providing physico-chemical, toxicological,ecotoxicological and other relevant data by using the ECB databases (e.g. IUCLID).Information about emission will be established in this work package including both the indoor andoutdoor conditions. Production, import and export figures will be extracted from statistic databases,which are available in a number of EU countries. From product databases, such as the Nordic SPIN(Substances in Preparations in Nordic Countries) database, information is available on industrial use

NOMIRACLE - Part B 101

categories and products specified for individual chemicals. The release from products is describedamong other using emission factors obtained from regulators or the different economic sectors.Socio-economic and human population specific (sensitivity of sub-groups) factors need to beregistered as conditions that can change the outcomes of exposure. Information from the productregisters in Denmark and Lithuania will be applied as a detailed case study in order to explore theusefulness of such detailed product related data sets. In the case of pesticides the release rate will beheavily related to agricultural practice For pharmaceuticals the release will be governed by bothwastewater treatment, including sludge disposal, and the veterinary use and following release fromagri- and aqua-cultures. Important product related information about emission will be gathered fromthe EU risk assessment activities (e.g. as outlined at: http://ecb.jrc.it/existing-chemicals). Severalgeo-referenced databases on the emission of pesticides are directly available for the work package,and case studies on emissions to relevant compartments in the landscape will be made available forItaly (UNIMIB), The Netherlands (Alterra), Lithuania (APINI) and Denmark (NERI) and the entireEU (JRC in draft format).Climatic databases and topographic and physio-geographical maps are available from which a seriesof relevant characterizing parameters will be gathered for environmental properties of aquatic andterrestrial ecosystems.Demographic data (e.g. population density) and social structure and housing will be gathered, whichare important both in relation to potential stressors and in relation to input for emissions estimates.Energy consumption in relation (e.g. heating and transport) will be important as an indicator forseveral possible stressors due to air pollution originating from combustion. Transportation in termsof intensity and structure will also be important as input for stressors both for human health (e.g.noise) and ecological health (e.g. fragmentation of habitats).Demographic data (e.g. population density, social structure and vulnerable nature areas) which areimportant in relation to exposure of relevant human and ecological receptors. Socio-economic andhuman population specific (sensitivity of sub-groups) factors need to be registered as conditionsthat can change the outcomes of exposure. Other receptor data include areas for extraction ofdrinking water and agricultural/urban area classifications. This type of data will typically beavailable in GIS.

NERI, JRC and UNIMIB will gather the substance-related data relevant for integrating exposureand effects of human and ecological character. This will be co-ordinated with the other researchpillars.Alterra, APINI, JRC, NERI and UNIMIB and will gather and secure access to product related dataabout substance composition and release rate including a link to human activity.Alterra, APINI, JRC, NERI and UNIMIB will gather data for climatic, physical and ecologicalconditions relevant for characterizing stressors for cumulative risk assessment and also data forstressors due to human activity.Alterra, APINI, JRC, NERI and UNIMIB will gather stressors for cumulative risk assessmentlinked to human activityJRC, NERI and UN will gather data for landscape classification and land use.JRC, NERI, UFZ and UN will gather data on human and ecological receptors

DeliverablesThe schedule for the first 18 months of deliverables will consist of the following data packages,where the numbering of deliverables are shown in parenthesis. A data package for substancecategories contents data or access to data for toxicity and fate parameters; human and ecologicalreceptors; estimated emission and stressors. Evaluation of the uncertainties involved in the data willbe included.7 months: (D1.1.1): Data package for pesticides.12 months (D1.1.2): Data package for pharmaceuticals, biocides and first sub set of VOCs/semi-VOCs.

NOMIRACLE - Part B 102

15 months (D1.1.3): Data package for VOCs/semi-VOCs15 months (D1.1.4): Data package for landscape classification and land use18 months (D1.1.5): Full documentation to the data management section (5.2)

Milestones and expected results for the first 18 months(M1.1.1): The development of criteria setting in WP. 1.2 for the scenario selection procedure willbe governed by limitations in the data background. Therefore a close dialog between the criteriasetting in WP. 1.2 and the work of finding data in WP.1.1 is needed. Also data uncertainty is a keytopic for the criteria setting and has to be communicated from WP. 1.1. to WP. 1.2.

(M1.1.2): The result will be a well designed data system, which in an optimal way supports thework of finding best test scenarios including product oriented emissions estimates and linking them,to existing knowledge and databases.

(M1.1.3): NOMIRACLE will collaborate with the following projects under EU: ALARM (JosefSettele ) by sharing relevant data; HAIR (Robert Luttik) which will enable the use of most recentdata and novel developments in harmonisation of pesticide risk indicators; ERAPHARM (ThomasKnacker) by supporting data about pharmaceuticals.

NOMIRACLE - Part B 103

Work package 1.2 description:

Scenario selection and ranking (18 month plan)

Work package number WP 1.2 Start date or starting event: Month 0

Participant id DESUN UNIMIB ENVI SYKE CEHPerson-months per participant 2 4 5 3 2

Participant id URV NERIPerson-months per participant 4 24

ObjectivesThe objectives in this work package during the first 18 months is, based on data from WorkPackage 1.1, to develop a procedure, which can identify a sub set of scenarios possessing thehighest risk potential for cumulative risk assessment of including mixture toxicity to be analysed inRP 3. Evaluation of uncertainly in the scenario selection will be input to RP 4. Supporting the riskmapping and other means of risk visualisation in WP 4.4 will be possible by using criteria derivedin this work package.

Description of workA first version of a selection procedure will be developed during the first 18 months based on areviewing activity of multi-criteria decision support methods. The criterion setting makes the link tothe data and a large number of different criteria will be needed in order to reflex the complex natureof the risk assessment procedure. The criteria are a highly variable group of data types and can beconsidered as indicators more than input for exact risk assessment procedure. Some criteria will beof the Boolean type (yes/no) others will be a category type (e.g. high, medium, low or indoor,outdoor) and others again will be real numbers (e.g. LC50 or Kow values). Some of the criteria willbe useful for assisting later assessment initiatives in the project in other research pillars while someof the other criteria will only be applied in this work package as a screening indicator. Fourcategories of chemical substances are subsequently analysed and thus used as training set for themethod development in the following order: pesticides, pharmaceuticals, biocides and VOCs/semi-VOCs . During the 18 month period empirical data in form of monitoring data will be gathered inorder to make a platform for evaluation of selection procedure and criteria setting.

During the first 18 months preliminary evaluation of the uncertainty for the first version of theselection procedure will be undertaken. This evaluation will be based on feed back from the otherresearch pillars and by investigating monitoring data.

NERI, UNIMIB, ENVI, URV will lay down specific multi-criteria approaches for ranking andselection.

ENVIMOD, UNIMIB, NERI will formulate the criteria setting for emission patterns.

CEH, UNIMIB, NERI will make the criteria setting for composition of mixtures for mixturetoxicity assessment.

CEH, NERI will make the criteria setting for stress factors for cumulative risk assessment.

NOMIRACLE - Part B 104

UN, NERI will make the criteria setting for characteristics and vulnerability of the biologicalcommunity and ecosystem.NERI, UNIMIB, ENVI, URV, UN, SYKE will make a revision of the selection procedures due tomore detailed input from the other research pillars and collected field scale data.

DeliverablesThe following output will come out for every substance category: Risk visualisation data for GIS atscreening level using criteria for potential risk; Mixtures of highest concern; Mixtures of highestconcern combined with other stressors for cumulative assessment. The schedule for the first 18months of deliverables will be as follows, where the numbering of deliverables are shown inparenthesis:9 months (D1.2.1): Scenario selection and criteria setting for pesticides.12 months (D1.2.2): Scenario selection and criteria setting for pharmaceuticals, biocides, a first subset of VOCs/semi-VOCs.18 months (D1.2.3): Scenario selection and criteria setting for VOCs/semi-VOCs.18 months: (D1.2.4): Documentation of the first version of selection procedure including initialuncertainty evaluation.

Milestones and expected results for the first 18 months(M1.2.1): The development of criteria setting for the scenario selection procedure will be governedby limitations in the data background. Therefore a close dialog between the criteria setting in WP.1.2 and the work of finding data in WP.1.1 is needed. Also data uncertainty is a key topic for thecriteria setting and has to be communicated from WP. 1.1. to WP. 1.2.

(M1.2.2): The development of criteria characterising exposure will be done in close corroborationexposure modelling in WP 2.4. This collaboration is secured by involving URZ from WP 2.4.

(M1.2.3): An important part of the methodological development is the feedback from the otherresearch pillars during the progress of the project in order to make a basis for refinement of theprocedure. This feedback will be secured by personal contact and by participation in a workshopheld by RP 1 (after 9 months).

NOMIRACLE - Part B 105

Work Package 2.1 description:

Matrix-compound interaction (18 month plan)

Work package number WP 2.1 Start date or starting event: Month 0

Participant id UFZ EAWAGPerson-months per participant: 36 18

ObjectivesThe goal is to develop methods to predict the phase partitioning of organic compounds from phase-specific parameters and molecular descriptors. To this end, model systems to measure compoundpartitioning between water or air and different soils (with soil types selected by WP 1.2) will bebuilt (EAWAG), and methods to measure the membrane-water partition coefficient as hydropho-bicity parameter related to the bioconcentration into organisms will be set up (UFZ). Both methodswill be applied to a judiciously selected test set of compounds, making sure that all majorinteraction forces can be sensed experimentally. Computational chemistry will be employed toderive molecular descriptors for characterising hydrogen bond acidity and basicity of organic sub-stances (UFZ), which are needed to predict the phase partitioning from molecular structure and(experimentally determined) phase descriptors.

For the first 18 months of the project, the major objectives are:1. Collect available literature data for partitioning of non-polar and polar organic chemicals in

soils that allow to derive a set of interaction descriptors for the respective soils. Based on theseand other soil specific data (e.g. NMR spectra) a first assessment of the variability of thesorption properties of natural soils in Europe is attempted.

2. Develop experimental model systems for the determination of phase descriptors for soils, andapply the methodology for a first test set of compounds.

3. Develop experimental system to measure membrane-water partitioning that allows also analysisof the impact of pH variation, and perform measurements for first test of compounds.

4. Collect Abraham descriptors from literature and build database with chemical structurehandling, and implement increment method to estimate these descriptors from molecular struc-ture.

5. Collect hydrogen bonding descriptors and calculate their values for a compound set built upfrom literature, and develop a strategy for quantum chemical reference calculations of hydrogenbonding.

Description of workThe work comprises an experimental and a theoretical research line, which are interlinked in orderto develop a methodology for quantifying the phase partitioning of compounds in a realistic way.

Experimental phase partitioning (EAWAG, UFZ)The experimental determination of the phase descriptors requires that the partition constants of adiverse set of organic compounds from water or air to the respective soils are determined. The testset (30-50 compounds) should cover all major functional groups in order to sense all relevant typesof compound-matrix interactions (dispersion forces, Coulomb forces, hydrogen bonding); the phaseparameters do not depend on the specific chemicals used for their analysis. For the European soilsselected in WP 1.1, measurement methodologies based on inverse gas chromatography (IGC) willbe built up and employed (EAWAG). For compounds with longer equilibration times, an alternative

NOMIRACLE - Part B 106

technique based on batch equilibrium experiments with head space analysis may be needed, theanalysis of which is part of the work programme (EAWAG). Finally, the passive-diffusivepartitioning into organisms is accounted for by respective experiments with membrane-watersystems (UFZ). The relevant Abraham descriptors needed for deriving the final model equationswill be compiled or calculated in cooperation with the theoretical branch of this WP.

Prediction of partition-relevant molecular descriptors (UFZ)The prediction of partition coefficients requires both the phase descriptors (experimental branch)and Abraham descriptors. The latter are usually not available for more complex compounds such aspesticides, biocides and pharmaceuticals. In order to develop a method for predicting such parame-ters from chemical structure, two complementary strategies will be pursued: First, an incrementscheme will be optimised and implemented, which is expected to perform reasonably well for moresimple compounds. In case of complex chemical structures, however, the relevant part of the intra-molecular interaction pattern will be nonlinear and context-dependent, which requires a different le-vel of computation. Here, the approach is to develop parameters based on quantum chemical mole-cular orbital theory. In the first 18 months, the respective research will focus on parameterising themolecular hydrogen-bond donor and acceptor capacity, which form a major challenge in the contextof predicting phase partitioning of more complex chemical structures.

Deliverables12 months:D.2.1.1 Compilation of phase parameters, and database of Abraham descriptors with chemicalstructure handling (R, PU).

18 months:D.2.1.2 Validated experimental procedure for the determination of phase descriptors for soils, andfirst series of experimental phase partitioning data of reference compounds (R, PU).D.2.1.3 Validated experimental procedure for the determination of membrane-water partitioning,and first series of experimental data at 25°C and pH 7 (R, PU)D.2.1.4 Computerised increment method to estimate Abraham descriptors from molecular structure(R, PU)D.2.1.5 Comparative analysis and evaluation of molecular descriptors for modelling hydrogenbonding (R, PU).

Milestones11 and expected results for the first 18 months9 months:M.2.1.1 Evaluate performance of IGC method for partitioning experiments with soils. Decisionwhether to switch to batch-headspace technique or further optimisation of IGC technique.

12 months:M.2.1.2 Check database with Abraham descriptors for representation of major chemical classes andphase descriptors as available from the literature.

18 months:M.2.1.3 Theoretical description and computational application of hydrogen bonding descriptors.

11 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 107

Work Package 2.2 description:

Available exposure (18 month plan)

Work package number WP 2.2 Start date or starting event: Month 0

Participant id NERI CSIC ITM JRC UFZ ULANCPerson-months per participant 18 24 18 6 18 18

ObjectivesThe goal is to develop and apply new experimental techniques for the determination of available ex-posure for both laboratory biotest systems and outdoor contamination. To this end, analytical tech-niques will be developed to quantify the compound contents in water, soil and sediment of emer-ging substance classes such as biocides, pesticides and pharmaceuticals (CSIC), and new diffusivesampling strategies for xenobiotics and heavy metals will be further developed and adapted to en-able the quantification of the effective matrix-specific exposure that is ready for interaction withbiota under in situ conditions (ITM, NERI, ULANC). Here, the method optimisation will focus onthe major European soil and sediment types of interest as identified in WP 1.2. As regards indoorexposure, analytical methods will be developed that allow an identification and evaluation of theVOC spectrum that has changed significantly in recent years due to the combined effect of energy-saving measures and the increased use of chemicals in consumer products. Starting from the envi-ronmental fate, the aim is to find out clusters and patterns of combined exposure which characterisespecific and typical health-relevant situations (UFZ, JRC). The developed methods will be used inRP 3 for a proper quantification of sound exposure in the context of deriving exposure-effect rela-tionships, and for the risk assessment of outdoor regions as described in RP 4.

For the first 18 months of the project, the major objectives are:1. Optimise and adapt experimental system to measure the available contents of organic com-

pounds using depletive sampling for the soil and sediment types of interest.2. Optimise and adapt experimental system to measure the matrix-specific compound activity

using non-depletive equilibrium sampling for the soil and sediment types of interest.3. Optimise and adapt experimental system based on diffusive gradient in thin films to measure the

diffusive flux of heavy metals as parameter of their available exposure.4. Assess the importance of the kinetics of chemical transfer from solid phase to solution in

determining available exposure (or availability of exposure).5. Optimise methods to get an overview about the human relevant VOC indoor exposure

Description of workThe experimental work comprises two interlinked research lines:

Environmental exposure (CSIC, ITM, NERI, ULANC)In the initial project phase, the analytical methods to quantify compound contents in water, soil andsediment will be adapted to the NOMIRACLE model substances, focusing on bioactive agents suchas biocides, pesticides and pharmaceuticals. The relevant soil and sediment types are those selectedin WP 1.1, and the programme includes judicious quality assurance measures to arrive at data thataccount for the milieu-specific parameters properly (CSIC). Corresponding activities relate to soil-and sediment-based biotest systems of RP 3, where the development of methods to quantify expo-sure properly will proceed in close cooperation with this WP. With regard to sampling strategies inorder to measure the available exposure of xenobiotics, two complementary approaches (availablequantity vs. activity) will be pursued and comparatively analysed in the context of characterisingsound exposure for both field and laboratory systems. For cationic metals, a depletive sampling ap-

NOMIRACLE - Part B 108

proach will be undertaken, and all three exposure parameters will be tested for their suitability toquantify available exposure through measurements of tissue levels for different types of organismsand soils (cooperation with RP 3). The field component comprises analysis of field samples inclu-ding in situ studies, and a comparison of measured exposure with predictions from the multimediafate model suite of WP 2.4, including the new phase partitioning algorithms developed in WP 2.1.In this way, an iterative improvement of the modelling and experimental techniques will take place,which requires co-ordinated activities across the individual WPs.

Human exposure (JRC, NERI, UFZ)From the human health perspective, the most important exposure towards xenobiotics is the indoorenvironment. This covers homes, public buildings (e.g. schools) and kindergardens, the latter ofwhich are of particular concern for children as particularly sensitive group. Here, the focus is on theinhalative pathway of VOC (volatile organic compounds). Due to energy-saving measures and theincreased use of chemicals in consumer products, the relevant VOC spectrum has changed over thelast years. Accordingly, suitable analytical methods and passive sampling techniques need to bedeveloped (JRC, UFZ), which will be pursued in close cooperation with the analytical work of theenvironmental exposure branch of this WP. Moreover, alternative biotest systems (ciliates, immu-nocompetent and other human cells) will be employed in RP 3 to characterise human health-rele-vant hazardous effects of chemicals, and here the diffusive sampling techniques will be employed tocharacterise available exposure and enable the derivation of sound exposure-effect relationships(NERI, UFZ). Indoor measurement campaigns in major cities across Europe will be undertaken tocharacterise prevalent patterns of the indoor VOC exposure in homes, kindergardens and publicbuildings in close cooperation with local authorities. The measurements will include exposure sam-pling for individuals, the characterisation of indoor/outdoor relationships, and take into accountseasonal variations that have been shown recently to form a further important confounding factor.

Deliverables12 months:D.2.2.1 Analytical methods for multifunctional xenobiotics such as biocides, pesticides andpharmaceuticals in water, soil and sediment, and of indoor VOC (R, PU).D.2.2.2 Temporal and spatial variability of human exposure-related VOC

18 months:D.2.2.3 Sampling methods for soils and sediments and relevant biotest systems, covering organics(diffusive sampling) and cationic metals (depletive sampling) (R, PU)D.2.2.4 Lab and field data of compound exposure, and assessment of availability parameters tofunction as a mechanistic link between laboratory systems and the field (R, PU).D.2.2.5 Assessment of the (source-dependent) indoor/outdoor ratio of VOC (R, PU)

Milestones and expected results for the first 18 months9 months:M.2.2.1 Check performance of analytical methods for European reference soils

12 months:M.2.2.2 Evaluate sampling methods for xenobiotics and heavy metals

18 months:M.2.2.3 Drafting of a generic “available exposure” assessment strategy that integrates the differentavailability parameters (in coop. with RP 3).M.2.2.4 Assessment of the indoor vs. outdoor VOC exposureM.2.2.5 Strategy to derive outdoor- and activity-dependent indicator components and/or VOCpattern.

NOMIRACLE - Part B 109

Work Package 2.3 description:

Metabolic fate (18 month plan)

Work package number WP 2.3 Start date or starting event: Month 0

Participant id LMC ECT LHASA UFZ URVPerson-months per participant: 72 25 18 21 8

ObjectivesThe goal is to provide experimental data that are needed to develop methods for predictingbiodegradation of more complex organic compounds (biocides, pesticides, pharmaceuticals), and todevelop pertinent modules of a modelling suite that allows to predict biodegradation rates andpathways from molecular structure. To this end, experimental biodegradation studies will be per-formed for a set of compounds selected in WP 1.1, covering different soils (UFZ) that are of majorrelevance for the European environment (WP 1.1) as well as sediment and water (ECT). Moreover,artificial intelligence methods will be developed (URV) to unravel relationships between ex-perimental biodegradation rates compiled from published and internal sources (ECT, LHASA,LMC), relevant milieu parameters (e.g. soil type, pH, microbial population characteristics) andbiodegradation rates, as well as relationships across the different environmental media (soil,sediment, water). A complementary modelling strategy focuses on developing rule-based systemsthat allow to simulate terrestrial degradation pathways from molecular structure, employing arith-metical (LMC) as well as non-numerical (LHASA) expressions of probability. For predicting ratesof photolysis in air, a new approach based on pertubational molecular orbital theory will be pursued(UFZ).

For the first 18 months of the project, the major objectives are:1. Build a database biodegradation rates and metabolites of organic compounds in water, soil and

sediment systems.2. Adapt and optimise experimental systems for determining the fate and degradation kinetics of

bioactive compounds (pesticides, biocides, pharmaceuticals) in water, soil and sediment sys-tems.

3. Measure biodegradation rates in some of the relevant systems for a first set of compounds.4. Develop quantitative structure-property relationships for degradation parameters, and explore a

mapping between OECD 301 results and degradation in soil, sediment and water.5. Build transformation descriptions and preliminary example rules for a knowledge-based system

to predict biodegradation and photolysis rates from chemical structure.6. Design links between LHASA and LMC software to maximise the utility of combining these

complementary approaches that are built on different probabilistic strategies.

Description of workThe work comprises three interlinked research lines:

Experimental analysis of biodegradation (ECT, UFZ)Starting point are the OECD guidelines (OECD 301 C, D, OECD 307 and OECD 308), aiming atdata for new bioactive compounds that are – as much as possible – comparable with existing datafor other compounds in water, soil and sediment systems. The use of radio-labelled test compoundswill facilitate the determination of degradation kinetics, metabolites and residues. The rates ofelimination, turnover and mineralisation as well as the formation of non-extractable residues will bedetermined in certain sediments and arable, European reference soils. Rate constants will be calcu-lated and compared with water-based results. The analyses include detailed checks of confoundingfactors such as high volatility, low solubility and high bacteria toxicity of the compounds under

NOMIRACLE - Part B 110

investigation. Using regression and configuration frequency analysis, the soil factors determiningthe biodegradability (pH, TOC, CEC, clay and microbial biomass content) will be assessed.

Correlation analysis of biodegradation rates across chemicals and media (URV)For the exploration of a mapping between OECD 301 results and degradation in soil, sediment andwater, an expert system based on fuzzy methods will be developed. The variables of interest includeparameters of the experimental systems (e.g. soil/sediment chemistry and soil/sediment structure,water and pore water chemistry), microbial population characteristics, temperature, and appropriatephysicochemical parameters and structural features of the chemicals of concern. Quantitative struc-ture-property relationships (QSPRs) will be developed for degradation parameters using descriptorsidentified via self organising maps and dissimilarity measures.

Simulation of degradation pathways (LHASA, LMC, UFZ)Starting point is a set of predefined molecular transformations that are prioritised objectively byusing integral biodegradation data (BOD) and accounting for metabolic pathways. The knowledgebase will include transformations of highly reactive groups and intermediates (oxiranes, ketenes,acyl halides, thiocarboxylic acids, hydroperoxides etc.), chemical equilibrium processes ( carboxy-lic acid hydrolysis, keto-enol tautomerism etc.), and metabolic transformations (oxidation, hydroly-sis, decarboxylation, dehalogenation etc.). Probabilities of rate-determining reactions will beestimated on the basis of experimental metabolic maps, employing both arithmetic and non-numerical reasoning methods. The latter will be designed to support reasoning about diverse infor-mation and problems, using quantitative and qualitative, numerical and non-numerical data. Forairborne xenobiotics, the dominant loss process is their reaction with OH radicals (H abstractionfrom aliphatic carbon, addition to double bonds etc.). To overcome shortcomings of availableincrement methods, a perturbational molecular orbital approach will be developed for the predictionof respective rate constants from molecular structure.

Deliverables12 months:D.2.3.1 Validated experimental procedures for the determination of compound turnover in water,sediments and soils (R, PU).

18 months:D.2.3.2 Experimental biodegradation kinetics for some model compounds in water, sediment anddifferent soils (R, PU).D.2.3.3 Evaluation of relevant parameters for deriving a QSPR for degradation rate parametersdifferent environmental media (R, PU).D.2.3.4 Preliminary computer approaches to simulate terrestrial biodegradation (R, PU).D.2.3.5 Predicted rates of degradation and metabolite formation for compounds subject to mul-timedia date analysis in WP 2.4 (R, PU).

NOMIRACLE - Part B 111

Milestones12 and expected result9 months:M.2.3.1 Evaluate performance of experimental methods to analyse biodegradation.12 months:M.2.3.2 Check biodegradation database for representation of relevant chemical classes,environmental media, experimental conditions and milieu parameters.18 months:M.2.3.3 Comparatively analyse performance of biodegradation simulators with regard to arithmeticprobability calculation vs. non-numerical reasoning.

12 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 112

Work Package 2.4 description:

Region-specific environmental fate (18 month plan)

Work package number WP 2.4 Start date or starting event: Month 0

Participant id DESUN ITM JRC RIVM URVPerson-months per participant 18 18 18 2 18

ObjectivesEnvironmental multimedia fate and exposure models predict environmental concentrations relativeto their emission sources, providing a necessary step in the analysis of the cause-effect chain forboth humans and animals. In this context, a challenging aspect is the optimisation of the spatialresolution in multimedia fate modelling on a European scale. In order to accommodate both data-rich and data-poor situations, a tiered approach is envisaged that provides sound decision support inthe framework of integrated risk assessment.

For the first 18 months of the project, the major objectives are:1. Development of a single multimedia fate and exposure model with various spatial resolutions at

the European level;2. Probabilistic evaluation of spatial model performance;3. Empirical evaluation of spatial model performance;4. Development and application of a cognitive neural network-based intelligent system to identify

crucial input parameters.

Description of workThe components of the work comprise four interlinked research lines:

1. Model development (lead JRC)Building on the existing approaches within DG-JRC, a novel multimedia fate and exposure modelwill be provided with regionalised spatial resolutions at the European level (Deliverable 2.4.1). Thiswill be refined as one input in the study to help identify key spatial characteristics (see Item 2). Amain focus will be on improvements to the soil module by using spatial data from the EuropeanSoils Database and related model prototypes that facilitate various resolutions (see also WP 1.1).Further refinements will be made in collaboration with other partners to account for differenttemperate zones and periods, particularly building on their recommendations in terms of transportand degradation mechanism models.

2. Probabilistic model evaluation (lead DESUN)For a selected number of multimedia fate models with various spatial resolutions on the regionaland European scale the relative importance of spatial specification as compared to the uncertainty insubstance-specific parameters will be explored (Deliverable 2.4.2). Probabilistic modellingtechniques will be employed for this purpose. This will be done for a judiciously selected set ofcompounds, based on input from WP 1.2, including multifunctional substances such as biocides,pesticides and pharmaceuticals that differ in (a) reliability of substance properties, (b) Europeantransport potential, (c) region-specific emission profiles and (d) compartment-specific emissionprofiles (soil, air water). Physicochemical properties and emission profiles of the selectedcompounds, needed in the fate and exposure calculations, will be delivered by WP 1.1.3. Empirical model evaluation (lead ITM)A complementary strategy to identify optimal spatially resolved European multimedia models fordifferent emission scenarios is to evaluate the model results using measured field concentrations

NOMIRACLE - Part B 113

(Deliverable 2.4.3). For this purpose, empirical data on environmental concentrations and emissionprofiles will be collected in close cooperation with WP 1.1 for a subset of the substances applied inthe probabilistic model evaluation (Item 2). The environmental concentrations, calculated with theselected fate models, will be compared with empirical data. The evaluation exercise would aim totest the validity of the models to (i) predict and compare environmental concentrations of parentcompounds and possible metabolites and to (ii) identify key fate processes and wildlife/humanexposure pathways with their proper physical and mathematical description. The evaluation resultswill also be useful to elucidate the effect of varying spatial resolution on the model performance forapplications across the European continent.

4. Sensitivity analysis (lead URV)To aid the identification of dominant relationships between input and output of multimedia fate mo-dels, a cognitive neural network-based intelligent system will be developed that is trained on know-ledge about multimedia system behaviour (Deliverable 2.4.4). The purpose is to unravel systematicrelationships between compound-specific parameters (physicochemical properties) and predictedexposure for given sets of European environment-specific input data (climate, soil type etc.) andemission patterns, considering model-specific intermedia transport algorithms as confounding fac-tors. The preliminary scenario ranking procedure developed in WP 1.2 will provide guidance in thedevelopment of the above neural network system for quantitative identification of intermedia trans-port and exposure pathways.

Deliverables12 months:D.2.4.1 Multimedia fate and exposure model with various spatial resolutions at the European level

18 months:D.2.4.2 Indication of the spatial detail (18 months)D.2.4.3 Inventory of improvement options in existing models, based on empirical model evaluationD.2.4.4 Cognitive neural network-based intelligent system to identify the most important variablesfor the differences found in partitioning behaviour, transport pathways and exposure routes betweenchemicals.

Milestones and expected results for the first 18 months12 months:M.2.4.1 Inventory of existing multimedia fate models with various spatial resolutionM.2.4.2 European-wide field data about selected reference compounds

18 months:M.2.4.3 Strategy to decide, for a given risk assessment context, which degree of spatial resolutionis needed for the evaluation of risks to human health and ecosystems

NOMIRACLE - Part B 114

Work package 3.1 description:

Interactive toxicological effects in diverse biological systems (18 month plan)

Work package number WP 3.1 Start date or starting event: Month 4

Participant id LIMCO NERC VU NERI DSTA WU DBUAPerson-months per participant 12 12 6 6 2 4 6

Participant id UFZ NIPH USALZ RWTHA LemTec EKUT WRcNSFPerson-months per participant 17 15 15 6 5 6 18

ObjectivesThis work package will establish toxic response profiles in diverse species for the single chemicalsand mixtures prioritised in RP 1. Major objectives during the first 18 months of will be:1. Finalising the experimental design and data analysis approach. This will include 1) writing aguidance document on the design and analysis of mixture experiments, and 2) developing themixture experiment framework and analysis model as a workable software tool.

2. Screening single and combined chemical toxicity using rapid bioassays. Assessments will bemade for mixtures of similar, dissimilar or unknown modes of action to evaluate fit to addition andindependent action models. Combinations showing both ideal fit to these existing models and alsosynergism/antagonism, ratio dependent and effect level dependent deviations will then be selectedfor detailed analysis using scenario relevant multiple bioassays.3. Establish methods for assessment of single chemical and mixture toxicity in taxonomicallydiverse systems. Toxicity will be measured first for a set of single contaminants and then for twomixtures, one of compounds with similar modes of action that showed no deviations from thetoxicity predicted due to concentration addition in the screening bioassays, and one of compoundswith dissimilar modes of action that showed no deviation from toxicity predicted by independentaction in the screening bioassays. This will establish the bioassay for each species and also judgethe ability of the screening tools to predict comparative chronic toxicity.

Description of the workExperimental design and data analysis approach

Effect assessment in NOMIRACLE will be unified by use of the single experimental design anddata analysis framework outlined in detail in Section B4. This approach allows the significance ofabsolute synergism/antagonism, ratio and effect level deviations from the established models ofaddition (for chemical of similar modes of action) and independent action (for chemicals withdissimilar modes of action) to be identified. To allow the design and data analysis approach to beused by all partners and the wider risk analysis community, a guidance document will be writtenand software developed and distributed (WU, NERC, VU).

Screening toxicity of chemicals both singularly and in combinationThe top 25 single contaminant and top 12 mixture scenarios prioritised in RP 1 will be screenedusing rapid bioassays such as the Vibrio fischeri Microtox system (WRcNSF), the single celleukaryote Tetrahymena pyriformis (UFZ), benthic invertebrates (e.g. Tubifex or Chironomidae)(NIPH, LIMCO), a fish embryo test (LIMCO, EKUT) and a set of available and specificallydeveloped human cell lines (USALZ, UFZ). Data for single compounds will be used to generateQSARs. For mixtures in these screening assays, and also the latter detailed assessments, it is

NOMIRACLE - Part B 115

anticipated that the toxicity of certain combinations will be correctly described by concentrationaddition and independent action, but because studies will be with specifically acting compoundsthere will also be many cases in which significant deviation (synergism/antagonism, dose effectlevel, dose ratio) from these models will be observed. All such deviations will be catalogued fordetailed investigation of inter-specific conservation and later of mechanism in WP 3.3 and 3.4. Thisinformation will also be returned to RP 1, thereby, allowing the prioritisation of mixtures to beadjusted dependent on whether toxicity can be correctly described using the additive andindependent action models.

Detailed assessment of single chemicals and mixture in diverse biological systemsTo evaluate if the screening tools correctly describe comparative single compound and mixturetoxicity across a diversity of species and systems, a set of detailed investigations will be undertakenin the species listed in Section B4. These detailed assessments will consider the comparative effectsof mixtures that both adhere to and deviate from the concentration addition and independent actionmodels. Initially, exposure protocols will be agreed and harmonised between partner laboratories.This will be particularly important for those species in which it is intended to conduct detailedanalysis of the mechanisms underpinning the toxicity in WP 3.3 and 3.4 (e.g. Danio rerio,Caenorhabditis elegans, Folsomia candida, Daphnia magna and human cell lines). Two prioritisedmixtures will then be investigated in all species and assays in the first 18 months. One ofcompounds of similar mode of action whose toxicity in the screening assays is fully described byconcentration addition and the second of chemicals with different modes of action whose toxicity inthe screening assay is fully described by independent action. During the exposures, a relevant set ofresponses (e.g. metabolic disturbance, immune dysfunction, cellular changes, developmentalparameters, weight change, proliferation, reproduction, behaviour, growth and viability) will bemeasured. Based on response surface modelling, any synergistic/antagonistic, dose level or doseratio dependent deviations from the additive and independent action models will be identified andthis information transferred to WP 4.1 and 4.2 where it can be used for probabilistic analysis.

Deliverables D.3.1.1 Finalisation of the scheme for mixture experiment design and data analysis includingdissemination of a guidance document and analysis software (7 months).D.3.1.2 Development of screening and target methods for assessment of chemical effects on humanhealth including generation of new stable immune reporter cell lines (15 months)D.3.1.3 Dose response profiles for the screening based assessment for the first set of prioritisedchemicals and mixtures and a resulting prioritisation of the first set of mixture combinations (ofsimilar and similar mode of action) to be used in the multi-system assessment (15 months).

Milestones13 and expected result for first 18 monthsM.3.1.1 Compile documentation to support the experimental design and data analysis framework (9months).M.3.1.2 Agreement on the design of the data analysis software (7 months).M.3.1.3 Select screening assays and agree on exposure protocols (9 months).M.3.1.4 Confirmimprovements to existing human health assays including assessment of robustness and suitability ofreporter gene systems (18 months).M.3.1.5 Review prioritisation list generated in WP 1.2 and finalised a set of 25 single compoundsand 12 mixtures for screening assessment (9 months).M.3.1.6 Review screening data for mixtures and return information to RP 1 and revise prioritisation

13 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 116

accordingly (15 months).M.3.1.7 Identify two suitable mixtures for detailed assessment, one of compounds of similar modeof action whose toxicity in the screening assays is fully described by the concentration additionmodel and the second of chemicals with different modes of action whose toxicity in the screeningassays is fully described by the independent action models (15 months).3.1.8 Preliminary comparison of single compound toxicity between single celled organism, animals,plants and human cells and assessment of uncertainty in effect prediction (18 months).

NOMIRACLE - Part B 117

Work package 3.2 description:

Combined effects of natural stressors and chemicals (18 month plan)

Work package number WP 3.2 Start date or starting event: Month 4

Participant id NERI UJAG EKUT DBUA WU UFZ LIMCOPerson-months per participant 17 15 6 9 4 11 6

ObjectivesThe objectives of WP 3.2 are to assess the effects of chemicals and chemical mixtures whencombined with natural environmental stressors. These “natural stressors” will include importantenvironmental parameters that are known to be of significance for the performance of organisms.

1. Working with partners in RP 1, a set of significant (regionally and temporally) naturalenvironmental stressors will be identified for use in the detailed analysis of cumulative stress.Although the choice of exact stressors will be refined during the outset of WP 3.2, at this stage it isexpected that these environmental stressors will include: thermal stress (high and low temperature;assessed in ectothermic organisms); desiccation (assessed in soil organisms); anoxia (assessed inaquatic species); pathogens (toxins; assessed in human cell lines and mice); allergens (bioaerosolsand other routes of delivery; assessed in human cell lines and mice).

2. In the first 18 months, the effects of single stressors will be assessed for the organism systemmost likely to be exposed to the particular stressor. This will be done to maximise the relevance ofthe complete assessment of single and later the cumulative stress. Thus, the influence of pathogensmay be most relevant in humans/mammals, the influence of drought most relevant for soilorganisms and anoxia for aquatic species. As outlined in WP 3.1 the assessment will use a range ofmodel biota covering bacteria, soil invertebrates, freshwater invertebrates, freshwater vertebrates,mammals to human cell lines.

3. After cataloguing the effects of single stressor, an initial trial of combined effects of chemical andnon-chemical stressors will be made. This first set of multiple stressor analyses will be made for thehighest prioritised combination taken from WP 1.2 and will be used to test the ability of the mixedstressor analysis framework, including analysis of which of the reference models (additive,independent action) can best describe different multiple stressor effects. This will lay foundationsfor latter work in WP 3.2 that will contribute to understanding of the effects of relevant cumulativestressors combinations (chemicals combined with environmental stress) on a wide range oforganisms and human immuno-responses to cover most environments and human communities ofthe European continent.

Description of workThe approach to identifying the interactive effects of the environmental stressor will mirror thatused for contaminant mixtures. First the effects of the single prioritised environmental stressor onexposed organisms will be described. These responses will be input into the project database toprovide information concerning the limits of the tolerance of diverse taxa to environmentalvariation and change.Once we have established the effects of specific environmental factors for particular species theimpact of these on the responses of species to chemicals taken through from the prioritisationexercise will be investigated. We will use the same approach to assessing the interactions of specific

NOMIRACLE - Part B 118

stress factors and contaminants as will be used for investigating contaminant mixtures. Usingsimilar experimental designs, exposures will be undertaken in order to generate data suitable for theconstruction of response surfaces. These will then be analysed the same statistical tools as used forthe assessment of chemical mixtures. This will allow identification of the most applicable referencemodel for each multiple stressor effects and also allow identification of and synergistic/antagonistic,stress level or stress ration level deviations from this reference (including nil effect). When clearand consistent interactions between environmental and chemical stress factors are identified theseresponses will be subject to detailed toxicokinetic and molecular mechanism assessment identifiedfor WP 3.3 and WP 3.4 to identify key components underlying interactions effects. All datacollected concerning both single stressor and multiple stressor effects will be fed direct to theproject database and thereby provide essential information for development of a unified approach tocumulative risk assessment for vulnerable populations and natural systems.

Since the particular environmental stressors will be relevant only to particular systems, an initial setof scenario relevant studies are envisaged at the outset of this work package. For human healtheffects, UFZ will carry out laboratory experiments using human T-cell lines, human lung epitheliacell lines, and fibroblasts, for studies of immuno-reactivity responses under the influence ofpathogens and allergens in combination with chemicals. USALZ will validate the results at thewhole-organism level using mice. In terrestrial UJAG will study the effects of interaction betweentoxic chemicals and other environmental factors, such as pathogens (e.g. Bacillus thuringiensis),desiccation and thermal stress in carabids and enchytraeids. NERI are already experience in analysisof the interactions of temperature and soil moisture extremes and will extend this work, combiningthese analysis with the effects of prioritised chemicals. WU will carry out life history studies usingnematodes exposed to combinations of chemicals and thermal stress. DBUA will study effects ofanoxia and thermal stress in combination with chemicals using Daphnia. Finally, EKUT andLIMCO will study effects of thermal stress and anoxia in combination with chemicals in fish

Deliverables

D.3.2.1 Methods to assess and a description of the baseline responses to the natural stressors underconcern for the various test organisms/cell lines (dose-response relationships) (12 month).

D.3.2.2 Methods to assess the impacts of relevant combinations of cumulative stressors for acomprehensive set of taxa covering the entire environments and human communities of theEuropean continent (18 months).

D.3.2.3 Generation of multiple stressor data for relevant combinations of chemical endenvironmental stress for the construction of response surfaces of a comprehensive set of taxa (18month).

NOMIRACLE - Part B 119

Milestones14 and expected result for first 18 monthsM.3.2.1 Identification of the most realistic single environmental stressors for detailed analysis ineach exposure system in conjunction with partners in WP 1.2. (Month 7).M.3.2.2. Adaptation of existing test methods to test realistic environmental exposure scenariosallowing the effects of these to be tested under laboratory conditions (Month 9).M. 3.2.3 Completion of analysis of response profiles for single environmental stressors in the mostrelevant species (Month 15).M.3.2.4 Identification of the first set of environmentally relevant combinations of cumulativestressors for detailed analysis (Month 12)M.3.2.5 Analysis of first set of combinations of environmental stressors and chemical compoundsthat could pose a risk to humans and the environment (Month 18).

14 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 120

Work package 3.3 description:

Toxicokinetic modelling (first 18 month plan)

Work package number WP 3.3 Start date or starting event: Month 7

Participant id VU NERC DSTA UJAG UA WRcNSF LMCPerson-months per participant 12 6 3 9 3 5 12

Participant id UFZPerson-months per participant 12

ObjectivesThe objective of WP 3.3 is to define the role of toxicokinetics in governing the responses of diversetaxa with different physiologies to exposure to single compounds, mixtures and other environmentalstressors. To realise this objective, the following aims have been defined for the first phase of thework package:

1.Further refinement of analytical methods for the determination of selected chemicals in the testorganisms used to determine uptake and elimination kinetics.

2. Determining uptake and elimination kinetics of selected chemicals in aquatic and soil organismsupon exposure to single chemicals and mixtures.

3.Determining the impact of chemicals in a mixture on each others uptake and elimination kineticsin aquatic and terrestrial organisms.

Description of workDuring the first 18 months, the work will focus on the determination of uptake and eliminationkinetics of selected single chemicals and mixtures in a series of aquatic and soil organisms,representing different taxonomic groups and different physiologies.

The decision of which single chemical, chemical mixture and multiple stressor scenarios will betested will be taken on the basis of the output of WP 1.2 and through detailed assessment of theresults of the work planned in WP 3.1. In the initial phase, this WP will mainly focus onestablishing the methods of determining the kinetics in the selected test organisms, by focusing on afew selected chemicals and their mixture(s). At a later stage, emphasis will be placed on mixtures ofchemicals or of chemical and non-chemical stressors that showed typical interactions in toxicityexperiments performed within WP 3.1 and WP 3.2.

Test organisms used in this WP will include Oligochaetes (NERC), Collembola (VU), Carabidae(UJAG), daphnids (UA) and molluscs (DSTA). The first three test organisms will be exposed insoil, while the latter two will be exposed in water. The experiments will include an uptake phase,during which the organisms are exposed to non-toxic concentrations of single chemicals ormixtures. During this phase, at regular time intervals test animals will be sampled and analysed foraccumulated chemical concentrations. After a certain uptake period, typically 2-4 weeks dependingon the organism and test substrate, animals will be transferred to non-treated test substrates andelimination will be determined by analysing test animals at regular time intervals. Colleaguesworking in WP 4.1 will provide methods for the modelling of uptake and elimination kinetics.These models will be applied to the single chemical exposures, and to determine modification of

NOMIRACLE - Part B 121

uptake and/or elimination kinetics by exposure to mixtures of chemicals

Colleagues involved in WP 2.2 will provide methods for the analysis of test chemicals in the testsubstrates and in the test organisms. In collaboration with the other partners involved in this WP,WRcNSF will take the lead in further refining these test methods to make them applicable formeasuring chemical concentrations in the test organisms. The resulting Standard OperatingProcedures provided by WP 2.2 or generated in this WP, will be used by all partners to performchemical analysis for their experiments. To ensure proper implementation of the methods ofchemical analysis, training of partners involved in this WP will take place in the laboratories ofcolleagues of WP 2.2 and WRcNSF will provide advice and technical supervision. Where possible,partners involved in this WP will co-operate in the chemical analysis, with partners mostexperienced in the analysis of a certain type of chemical taking care of the analysis of thosechemicals also in samples of the other partners. A plan for the mutual exchange of protocols,reference materials and samples will be prepared to ensure optimal use of available expertise andfacilities.

Uptake and elimination will not only be related to total (measured) chemical concentrations in thetest substrates, but also to (bio)available concentrations. Available concentrations will be measuredin close co-operation with and using methods provided by colleagues working in WP 2.2.Data of this WP will be fed into the NOMIRACLE database and delivered to WPs 4.1 and 4.2 toform a basis for risk assessment analysis and the development of QSARs for mixture interactions atthe organism and molecular levels under complex exposure conditions (in mixtures and uponexposure to non-chemical environmental stressors). From this basis we will then be able to move onto later work in NOMIRACLE. This will include collecting data on the uptake and eliminationkinetics of a series of selected chemicals in the selected test organisms, upon single and mixtureexposure and under the influence of other environmental stressors (to be defined by RP 1 and in WP3.2). Additionally we will assess the degree of modification of chemical uptake and eliminationkinetics in mixtures and by non-chemical environmental stressors. To develop predictive methodsfor the hazardous effects of indoor exposure on human health, targeted QSAR investigations will beperformed to derive a mapping between VOC compound effect profiles on biotest systems andepidemiological findings about the exposure-related human health status.

DeliverablesD.3.3.1 Standard Operating Procedures for the analysis of concentrations of selected chemicals inthe different test organisms (15 months).D.3.3.2 Preliminary report on the uptake and elimination kinetics of two selected chemicals andtheir mixtures in different test organisms, upon exposure in soil or water (18 months).D.3.3.3 Literature report on the toxicokinetics of chemicals in mixtures as a first step towards thedevelopment of a QSAR for mixture interactions (18 months).

NOMIRACLE - Part B 122

Milestones15 and expected result for first 18 monthsM.3.3.1 Agreement on the first set of chemicals for toxicokinetic analysis (9 months).M.3.3.2 Methods for the analysis of selected chemicals in biological tissues, to be described in theform of Standard Operating Procedures that can be used by all partners in this WP (15 months).M.3.3.3 Identification of the suitable compartment models for each of the species in collaborationwith partners in WP 4.1 (12 months).M.3.3.4 First data on the uptake and elimination kinetics of two selected chemicals in the selectedtest organisms, upon single and mixture exposure (18 months).M.3.3.5 Review of literature data on mixture interactions affecting the uptake and eliminationkinetics of chemicals in aquatic and terrestrial organisms (18 months).

15 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 123

Work package 3.4 description:

Molecular mechanisms of mixture toxicity (18 month plan)

Work package number WP 3.4 Start date or starting event: Month 7

Participants id DSTA NERC UFZ WU UWC UCAM EKUTPerson-months per participant 8 8 7 6 9 9 6

Participants id UAPerson-months per participant 9

ObjectivesBecause of the potential importance of mode of action for understanding combined stressors effectsthis work package will elucidate the conserved and distinct molecular changes that underpin thesystemic response of species to chemical mixtures (WP 3.1) and multiple stressors (WP 3.2).Ultimately we will derive biomarkers of combined exposure and effect that will be validated in fieldstudies in the later stages of WP 3.2. Specific objectives for the first 18 months of the work packageare set out below.

1. Develop a co-ordinated approach that exploits methods for measuring gene and proteinexpression and function for mode of action assessment. We will start by optimising protocols andconducting exchange/training visits to transfer methods between human model systems andenvironmentally relevant species and if applicable vice versa. Standard operating procedures will bemade available to the Consortium and ultimately the wider community as a handbook. Baselinemeasurement will then be made in each species-system under optimal conditions.

2. Establish mode of action for prioritised single compounds from literature and past work. If norelevant information is available, mechanisms will be determined experimentally. Initially,comparative analysis of molecular response patterns will be undertaken to elucidate theconservation of stress responses for compounds with well known modes of action. On this basis wewill define an optimum strategy for the use of transcriptomics, proteomics, metabolomics, targetedbiochemical and pattern recognition methods for mode of action assessment.

3. Use biochemical/molecular tools to assess (confirm) that chemicals with similar modes of actionwhose toxicity can be described by concentration addition act through a single response cascade.

Description of the work

To optimize the use of resources within NOMIRACLE, information on modes of action will first becollected from literature. If there is no information available relevant to the biological system inquestion, public toxicogenomics and metabolic pathways resources (e.g. KEGG metabolism, andpathDB metabolic pathway databases) will be screened and any available data used to elucidatemechanism(s). If this screening yields no data, mode of action will be investigated experimentallyin the most relevant biological system. These include human cell lines (UFZ), rodent (NERC,UCAM), fish (EKUT, UFZ), annelid worm (NERC, UWC), mollusk (DSTA), nematode (UWC,WU), Daphnia (UA) and a single cell organism (DSTA). The work concentrates on animals inorder to increase the relevance to human metabolic systems and ultimately human health.

The biochemical/molecular toolkit for mode of action assessmentThe experimental approach used to establish mode of action (and ultimately interaction) comprise

NOMIRACLE - Part B 124

both comprehensive and targeted methods. DNA microarrays for zebra fish (UFZ), nematode (WU,UWC), earthworm (UWC), Daphnia magna (UA), mussel (DSTA) and Dictyostelium (DSTA) willbe used to identify changes in gene expression. Other gene discovery techniques will also beemployed (UWC, UA). Proteomics will use electrophoresis and chromatography separation withmass spectroscopy (DSTA, UA, UCAM, NERC). For metabolomics, 1H NMR spectroscopy, GC-MS and LC-MS based approaches (UCAM, NERC) will be used. To supplement screeningtechniques, detailed histopathological, biochemical, physiological, and molecular genetic analyseswill be made (see Section B4 for full details) (DSTA, NERC, EKUT, UA). After data acquisition,pattern recognition will be used to interrogate the datasets, allowing us to identify metabolic andgene responses characteristic of specific exposures. Comparative analysis (co-ordinated by DSTA)will establish whether mechanisms are unique to species or common between taxa.

Developing a co-ordinated approach for mode of action and interaction assessmentPrior to the experimental phase, a comprehensive harmonisation process, including a program ofexchange training visits, will be undertaken to optimise procedures for the measurement of stressresponses. These adapted protocols will be compiled on the NOMIRACLE web site.

Establishing compound mode of actionIn the initial phase of WP 3.4, the biochemical/molecular toolkit will be used to confirmmechanisms for compounds with a well known mode of action. These could include for example,cholinesterase inhibiting pesticides, respiratory inhibitors, and blockers of metabolic biosynthesis.Completion of this work will establish the ability to identify mode of action and will suggestrefinements for optimal use of the toolkit for chemicals where mode of action or mixture interactionis not known or not relevant to a particular species.

Mode of interaction assessment for similarly acting compoundsAfter testing for single compounds, the biochemical toolkit will next be used to identify modes ofinteraction for chemical mixtures and later multiple stressors. In the first 18 months theseinvestigations will seek to confirm that chemicals with a known similar mode of action whosetoxicity can be describe using the addition model act through a similar response pathway.Completion of this study with similarly acting chemicals will refine our approach for later analysesof mixtures that i) have dissimilar modes of action (including combinations of chemical and non-chemical stressors), ii) show consistent deviations from the addition and independent action models.

DeliverablesD.3.4.1 Repository of protocols for specific and global profiling of stress response pathways inhuman cells, mammalian models and environmentally relevant species (12 months).D.3.4.2 Completion of baseline analysis of response profiles for single stress response systems andglobal profiling responses in the selected experimental systems under ideal conditions (18 months).D.3.4.3 Datasets of analyses on the effects of two compounds with a known specific mode of actionand one additive mixture on global and specific biochemical responses (18 months).

Milestones16 and expected result for first 18 monthsM.3.4.1 Agree on format for collation of technical guidance and standard operating procedures ontoNOMIRACLE web site (9 months).M.3.4.2 Decide initial series of research training visits (12 months).M.3.4.3 Complete cataloguing of available information on mode of toxic action for prioritisedchemicals in all species (18 months).

16 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 125

M.3.4.4 Identify two compounds with known specific mode of action to be used to confirm theintra-species comparability of biochemical assays (12 months).M.3.4.5 Harmonise testing procedure and agree protocols for exposure and sample collection andhandling with partners involved mixed stressor studies in WP 3.1 and 3.2. (12 months).M.3.4.6 Complete assessment for two compounds with a single specific mode of action (allpartners) (18 months).M.3.4.7 Review and ultimately refine the biochemical toolkit for mode of action and interactionevaluation (18 Months).

NOMIRACLE - Part B 126

Work package 4.1 description:

New concepts and techniques for probabilistic risk assessment (18 month plan)

Work package number WP 4.1 Start date or starting event: Month 0

Participant id VU DESUN USOUTH EPFLPerson-months per participant 12 13 18 18

ObjectivesWP4.1 aims to develop new concepts and techniques for probabilistic risk assessment that arescientifically sound and practicable for management purposes. The specific objectives during thefirst eighteen months of the NOMIRACLE project are:• Separation of true uncertainty and interindividual variability in risk predictions of an integrated

human exposure model.• Extension of the DEB theory for single compounds in order to derive a probabilistic NEC for a

mixture of two known compounds.• Harmonisation of the analytical frameworks for meta-analyses of human and ecological

toxicity data.• Derivation of probabilistic uncertainty factors (UFs) for inter-individual and interspecies

differences based on pharmacokinetic and pharmacodynamic data for exposure to singlecompounds and chemical mixtures that are handled by major phase I and phase II metabolicpathways.

• Development of new methods for comparative risk assessment by integration of mixturetoxicity and multiple stressors (i.e., comparison of toxic stress, eutrophication andacidification).

Description of work

Separation of uncertainty and variability (DESUN)The integrated human exposure model NORMTOX (Ragas & Huijbregts 1999) predicts the risk ofhuman exposure to chemical contaminants from all relevant environmental exposure routes (air,drinking water, surface water, food, soil and dust). During the first stage of WP4.1 (starting aftermonth 6), the influence of true uncertainty in the risk predictions of NORMTOX will be separatedfrom the influence of interindividual variability by means of nested Monte Carlo simulation. Toperform these calculations, detailed data will be gathered on contaminant concentrations in relevantenvironmental media and on human consumption and activity patterns. These data will be analysedto derive the relevant statistics of the input parameters for the nested Monte Carlo simulation. Atfirst, detailed risk predictions will be produced for exposure of the Dutch population to a selectedset of pesticides. If data availability and time allows, these predictions will be extended to otherchemical contaminants and stressors (i.e., pathogens) and other European regions. The outcomespecifies the population fraction at risk due to interindividual variability in consumption andactivity patterns and details the probability of this risk. The results allow policy-makers to take abetter-informed decision on risks of human exposure to chemical contaminants through multipleenvironmental media and can have important implications for data gathering schemes on humanconsumption and activity patterns.

Derivation of a probabilistic NEC for mixtures (VU)The subproject on derivation of a probabilistic NEC will start after month 6. Research will befocussed on modelling the effects on mixtures of 2 compounds, with survival and reproduction as

NOMIRACLE - Part B 127

end points. This model activity includes a toxicokinetic module, a module for effects on (compoundspecific) physiological target parameters and a module for the translation of these effects onendpoints. Statistical methods will be developed to test the model against experimental data; thesemethods will be applied to data from the literature and the archives of the department, and to thefirst experimental data from RP3. Special attention will be given to a particular type of mixture: thatof a molecular and the ionic form of the compound. The relative abundance of these forms dependson the pH, and so on the total concentration of the compound.

Harmonisation of analytical frameworks for meta-analysis of human and ecological toxicity data(DESUN)A desk top study will be performed to identify options for harmonisation of the analyticalframeworks used in the meta-analysis of human and ecological toxicity data. The results of the desktop study will be discussed by both partners involved in the derivation of probabilistic UFs(DESUN and USOUTH) and harmonisation will concentrate on using similar mechanisticdescriptors for analysing the toxicity data, e.g., substance parameters, the genetic predisposition ofreceptors and the toxicological mode of action of substances.

Human and Interspecies uncertainty factors (USOUTH)Most human polymorphic pathways for xenobiotic metabolism (CYP2C9, CYP2C19, CYP2D6,NAT) have been shown to be highly variable so that the current default kinetic uncertainty factorwould not cover healthy adults for contaminants handled via these routes nor the most sensitivesubgroups of the population (neonates and children; CYP2C19 and CYP2D6). During the firsteighteen months of WP4.1, a meta-analysis of human pharmacokinetic data using data for markersof acute and chronic exposure in subgroups of the population (healthy adults, inter-ethnicdifferences, children and the elderly) will be performed for single compounds and chemicalmixtures (drug interaction data) handled partially and totally via these polymorphic enzymes. Thiswill provide new uncertainty factors (to cover the 95th, 97.5th and 99th percentiles) taking intoaccount the full extent of the variability for each polymorphic pathway and for each subgroup of thepopulation. The human kinetic data will then be compared with the available animal data (mouse,rat, rabbit and dog including neonatal animals) to quantify interspecies differences and generateuncertainty factors for this aspect of uncertainty. Finally, a preliminary study of inter-individual andinterspecies differences in pharmacodynamics will be performed to provide a basis for thederivation of uncertainty factors based on mechanistic data. These uncertainty factors can then beused by risk assessors and risk managers to replace the traditional uncertainty factors.

Comparative risk assessment (EPFL)The Assessment of Mean Impacts method (AMI) adapts the traditional concept of ecotoxicologicalrisk that is based on the most sensitive species, to a more discriminating and comparativeassessment using the geometric mean response of species and its associated confidence interval.The method has proved to be a useful tool for the comparison of the potential hazard ofsubstitutable pesticides on freshwater ecosystems (Humbert et al. 2003) with a validation of itsenvironmental relevance (Fawer 2003). Based on the AMI method, a new ecotoxicological effectmodel was developed in the FP5 OMNIITOX project. The current aim is to integrate toxicity ofmixtures and other stressors (e.g., eutrophication, acidification, etc) in this method. Firstly, the mainstressors will be identified and selected for integration in the comparative assessment method.Secondly, the models quantifying the stressors will be analysed to reveal their compatibility withthe comparative risk assessment approach. Finally, a methodological framework for comparativerisk assessment will be developed including chemical mixtures, eutrophication, acidification andother stressors.

Deliverables

D.4.1.1 (month 18): Report on separation of true uncertainty and interindividual variability inpredicted risks of the Dutch population from exposure to pesticides through all relevantenvironmental pathways.

NOMIRACLE - Part B 128

D.4.1.2 (month 18): A paper on the model formulation for effects of a mixture of 2 compounds anda discussion of the application of the model to experimental data.D.4.1.3 (month 18): Consolidated report describing the metabolism, pharmacokinetic andpreliminary pharmacodynamic data in human subgroups of the population and tests species.Derivation of uncertainty factors for each subgroup of the population and test species forpolymorphic elimination pathways after exposure to a single chemical or to chemical mixtures.D.4.1.4 (month 18): Report describing method for comparative risk assessment of multiplestressors.

Milestones17 and expected resultM.4.1.1 (month 10): Selection of relevant stressors in the environment for inclusion in comparativerisk assessmentM.4.1.2 (month 12): Availability of data for separation of true uncertainty and interindividualvariability in human exposure model and decision on the extent of the modelling effortM.4.1.3 (month 14): Analyses of models quantifying the selected stressors for comparative riskassessmentM.4.1.4 (month 18): Uncertainty factors allowing for human variability and interspecies differencesfor polymorphic pathways after single and multiple chemical exposure

17 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 129

Work package 4.2 description:

Explicit modelling of exposure and risk in space and time (18 month plan)

Work package number WP 4.2 Start date or starting event: Month 0

Participant idDESUN URV UFZ ALTERRA

Person-months per participant 12 16 15 16

ObjectivesWP4.2 aims to develop new methods and models that explicitly address temporal and spatialdimensions of cumulative risks, both for human and ecological receptors. The specific objectives ofthe first 18 months are:• Development of a model that facilitates estimation of long-term exposure to airborne

chemicals.• Probabilistic use of species data in ecological assessment based upon a vulnerability

analysis, provision of specific risk limits, and implementation of the results in spatialmodelling.

• Identification of causal parameters that determine human exposure and risk by integrationof neural network cluster/variable analysis within a GIS environment at different spatial scales.

• Development of a model that describes the accumulation of pollutants in selected ecologicalreceptors in a floodplain area along the river Waal in the Netherlands.

• Modelling spatially aggregated risk of airborne chemicals for children in the region ofLeipzig.

• Exploration and development of the possibilities for up-scaling information on exposureand risks.

Description of work

Long-term exposure to airborne chemicals (UFZ)For temporal modelling, a new method will be developed that estimates the long-term exposure tochemicals utilizing only short-term measurements. Here, a particular focus is the contamination ofindoor environment (input from RP2) and the role of behaviour (ventilation practice) and activities(e.g. renovation) of the inhabitants. From previous field studies extensive data of the contaminationof indoor air with hydrocarbons are available (Schlink et al. 2004). ANOVA and neural networkmodels will be used to identify the most important influences and, subsequently, applied to predictthe long-term exposure.

Probabilistic use of species data in ecological assessment and modelling (ALTERRA)Identification of critical ecological pathways and parameters for risk assessment will be performedfor the vulnerable ecological receptors defined in WP1.1. Autecological data will be compiled forrelevant species, including life history, feeding biology, and distribution and mobilitycharacteristics. These data will be categorized and interpreted in terms of vulnerability towardsparticular stresses. Subsequent vulnerability analyses will be done for the species involved usingmulti-criteria analysis. The result is a ranking of species or species assemblages by relativevulnerability (Faber et al. 2003). Relevant critical limits will be derived from literature, and themethod will be further developed towards absolute ranking and combined stresses. The choice ofspecies and end-points will be based upon the outcome of RP1 and the local situation. The resultswill be used as input for random walk modelling.

Identification of causal parameters that determine human exposure and risk (URV)The URV work will focus on establishing relationships between exposure and health impact, using

NOMIRACLE - Part B 130

available data generated in WP1.1. A model will be developed for a specific region (e.g., the Ebrodelta or another Catalan region for which health data are available). This would then serve as a testcase for later implementation for a European model. Analysis of likely correlated predictorvariables, to determine the best set of descriptors for effect clusters (e.g., cancer), will be evaluatedvia Kohonen neural networks and other statistical methods of analysis. Kohonen’s self-organizingmaps and similarity measures will be used to select the best set of predictors for exposure and riskhotspots by classification analysis (Vesanto 1999). Descriptors with similar topology are groupedinto a common class and variables from each class exhibiting the highest correlation or absolutecovariance are pulled out of the class and mixed with the most correlated variables for other classesto define a complete set of cluster descriptors. Changes in map topology caused by progressiveincorporation of the various descriptors will be quantified by dissimilarity measurements so as tofinally identify the most relevant descriptors. The results will be used in a later stage of WP4.2 toselect relevant scenarios and parameters to include in a human random walk model.

Random walk modelling (DESUN, ALTERRA)The development of random walk models will start in month 7 of the NOMIRACLE project.Initially, a model will be developed that describes the accumulation of pollutants in selectedecological receptors (i.e., Little Owl, Badger, shrews and mice) in a floodplain area along the riverWaal in the Netherlands. Extensive pollution and ecological data are available for this site, whichmake it an ideal pilot area to develop, test and validate algorithms that describe random walkpatterns, ecological interactions and bioaccumulation. The project will start by gathering pollution,habitat and auto-ecological data on the study area. Algorithms that describe random walk patterns,ecological interactions and bioaccumulation will be based on previous work and a review ofscientific literature (Van den Brink & Baveco, in press). Finally, predictions will be made andresults analysed. During a later stage of the NOMIRACLE project, the opportunities to extend therandom walk model to other areas, ecological receptors and humans will be explored.

Modelling spatially aggregated risks for children in the region of Leipzig (UFZ)Adjusting for the effects of further stressors, spatial patterns of the risk of airborne chemicals toairway diseases and allergies (based on data of previous studies) will be modelled usingmultivariate Bayesian techniques with conditional autoregressive terms (CAR-model). The CAR-model will be developed for the region of Leipzig, where all required data are available. Initially,the spatial aggregation will refer to the districts of the city area. The model produces spatiallyaggregated risk data with geographical reference (output to GIS in WP4.4) that will be used for riskmapping. For the further NOMIRACLE work, the established CAR-model can serve as a reference,which can be refined for other levels of aggregation, for other health effects and other regions.

Scaling of spatial information on exposure and risks (ALTERRA, URV)Procedures for up-scaling the results of small-scale ecological models to the regional level will bedeveloped. The level of the larger scale will be specified and the relevant data on that scaleidentified, depending on the data availability and on the demands of the end-user. Due to the currentstandard of knowledge, the scale of province is preferable. Furthermore, data on habitat or humanenvironment characteristics will be used in order to produce maps for likelihood of occurrence ofspecies (based upon the SARCH-methods). Finally, the results of the small-scale modelling will beextrapolated to the larger scale, resulting in an assessment of the risks at this larger scale. For thispurpose extrapolation routines will be derived. The results will be demonstrated, depending on theavailability of data, for a region in The Netherlands.The analysis of human exposure and risk clusters will be conducted at different spatial scales todetermine the relationship between cluster size and its potential identification relative togeographical scale and the identifying factors. The use of different scales will demonstrate therelative limitations of using large-area scales to identify impact clusters. In this way, thehomogeneity of associations across various geographic scales will be analysed.

DeliverablesD.4.2.1 (month 18): A random-walk model that describes the accumulation of pollutants in selected

NOMIRACLE - Part B 131

ecological receptors in a floodplain area along the river Waal in the Netherlands.D.4.2.2 (month 18): A fuzzy ARTMAP neural classifier to enable cluster analysis anddemonstration of the variable selection approach for an EU region.D.4.2.3 (month 18): Demonstration of up-scaling methods, based upon the small-scale modelling.D.4.2.4 (month 18): Demonstration of the CAR-model assessing health risk in the city area ofLeipzig.D.4.2.5 (month 18): A temporal model for the indoor air pollution with volatile organic compounds.

Milestones18 and expected resultM.4.2.1 (month 3): Choice of ecological targets for vulnerability analysisM.4.2.3 (month 9): Data for ecological random walk model have been gathered and data at largerscale have been selected for the up-scaling procedure, application and demonstrationM.4.2.4 (month 12): Algorithms for random walk patterns, ecological interactions andbioaccumulation reviewed; decision on inclusion in random walk model and up-scaling routinesM.4.2.5 (month 15): First implementation of random-walk in computer and first demonstration ofup-scaling methods (depending on the results of small scale modelling)M.4.2.6 (month 18): Ecological vulnerability analyses completedM.4.2.7 (month 18): A fuzzy ARTMAP neural classifier established to enable cluster analysisM.4.2.8 (month 18): Random walk model finalized and analysis of the results

18 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 132

Work package 4.3 description:

Dealing with multiple and complex risks

in a management context (18 month plan)

Work package number 4.3 Start date or starting event: Month 0

Participant id SYKE DIA JRCPerson-months per participant 18 7 6

ObjectivesThe overall aim of WP4.3 is to improve the knowledge base for dealing with multiple and complexrisks, uncertainties and ambiguities by studying cognitive and knowledge-related, social andcontextual aspects of integrated risk assessment, and by providing new interdisciplinary, reflexiveand pluralistic approaches to addressing these aspects. The specific objectives during the firsteighteen months of the NOMIRACLE project are:• Review the theoretical and applied frameworks and methods in the areas 'risk, knowledge and

uncertainty', 'perception and cognition', 'policy aspects in risk assessment', and 'riskcommunication', and to make the results of the reviews operational for the NOMIRACLEproject.

• Identify and evaluate data on risk perception and communication to assist other WPs.• Provide inputs for training by interactive research in risk cognition and communication.• Devise a detailed work plan for the next stages of the project.

Description of work

Evaluation of risk perceptions, cognition and knowledgeRisk perception and cognition and knowledge evaluation among key experts and actors withinchemicals control at the European level will be studied by:• Literature and document analyses of views regarding multi-dimensional risks, particularly

chemicals-related, and their integrated treatment (SYKE)• Surveys of expert opinions on risks among project staff including members of advisory bodies

and meeting participants (SYKE)• Analysis of the development of knowledge and modes of reasoning for chemical risks (JRC)We will build on studies of the various perspectives on risks by chemical and other stressors. Theresearch will initially address issues in the production of knowledge in a risk assessment context:multiple sources, qualities and uses of knowledge, plurality of views and reflexive thinking. Riskcomparisons and expressed risk views in case areas of integrative assessment, e.g. under REACH,will be studied in detail. Plans for further research and development will be made includingempirical and in-depth theoretical studies and methodological development (e.g. within multiple riskcognition as a culture and organization dependent process in both science and in science-policy-society interfaces) aimed at improving the framework for multiple risk cognition and assessmentand at subsequent method development.

Management strategies and policies in risk assessmentLinks between risk assessment and management strategies and policies will be analyzed, involvingthe following main tasks:• Conceptual study of deliberation processes in chemicals-related risk management (DIA)• Identification and initial analysis of strategic and policy issues in relevant risk assessments,

including those under REACH (SYKE)• Analysis of and guidance for approaches to deal with uncertainty and quality assurance in

integrated chemicals-related strategic risk management (JRC)The work will initially address legal and regulatory frameworks and policy principles in integrated

NOMIRACLE - Part B 133

assessment, stressing dimensions of integration. Modes of deliberation about risks will be studied atdifferent levels depending on decision context and on actors. Management goals and options,relationships between precautionary and in-depth assessment of chemicals, links betweenassessments, management and information production (with WP4.1) and region-specific approaches(with WP4.2) will be stressed. Preferences in models of and responses to uncertainty will beexamined at subsequent stages among the Consortium mainly, linking the work to the research onrisk perception in WP4.3 and to other WPs. Plans will then be made for further work including in-depth studies of policy aspects of risk assessment and development of methodologies, focusing onthe relations between uncertainty and the framing of risk issues and on the links between riskassessment and precautionary approaches including options assessment.Risk communicationStudies on risk communication will include the following activities:• Desk study of communication concepts and methodological approaches, including results

relevant for integrated risk assessment especially of specifically acting chemicals (SYKE)• Exploratory analysis of risk communication in assessments cases, in interaction with WP5.4

(SYKE)• Initial study of participatory multi-actor risk and uncertainty communication (DIA)• Linking communication studies with training, especially at meetings (SYKE, DIA)The work will focus on discourses about principles and approaches in integrated risk assessment, onways to improve communication about related uncertainties and ambiguities, and on interactions incommunication. Studies will be made of communication among researchers and between expertsand stakeholders, e.g. in connection with project meetings. From this work, needs, constraints andopportunities for cross-disciplinary and participatory communication will be identified. Based onevaluations of relevant experience of communication, new approaches will be explored moreclosely, including dialogue and visual techniques. Plans for further work will stress communicationin dissemination to a wider audience including stakeholders, in researcher and expert training and inexploitation of results by experts e.g. in integrated assessments within and in relation to REACH.

Deliverables

D.4.3.1 (month 12): Conceptual models and mental maps of risk information and cognition in the IPdomain, based on data on risk views and inference from other RP and IP entities and actors,including verbal, quantitative and visual perception models for comparative and cumulative riskassessmentD.4.3.2 (month 9): Conceptual frameworks for the characterization of knowledge of risks and itsdevelopment, its use and its quality assurance in the IP domain, including risk informationpedigrees.D.4.3.3 (month 15): Models of deliberation modes in responses to uncertainty and ambiguity, todevelop integrated approaches and strategic guidance to risk assessment partners.D.4.3.4 (month15): Review of relevant risk communication research with data evaluations forpartners, and of communication models, including participatory, interactive, comparative and visualapproaches.D.4.3.5 (month 18): Synthetic interim report and plan for further work (including interviews,surveys, case studies, policy and decision analyses, discourse analysis; theory and methodsdevelopment; outreach) and meetings reports and discussion papers.

NOMIRACLE - Part B 134

Milestones19 and expected results

M.4.3.1 (month 9): Draft studies (surveys/reviews) in the sub-areas

M.4.3.2 (months 6-16): Dedicated studies in and outputs from meetings within the WP area andwhole RP/IP

M.4.3.3 (month 18): Synthetic report; first meeting papers (reviews and methods); project plan fornext phase finalized

19 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 135

Work package 4.4 description:

Risk presentation and visualisation (18 month plan)

Work package number WP 4.4 Start date or starting event: Month 4

Participant id ALTERRA UFZ NERI URV UNIMIB JRCPerson-months per participant 4 5 1 2 2 1.5

ObjectivesWP4.4 aims to develop and demonstrate the most appropriate and/or novel techniques forpresentation and visualization of cumulative risks. The specific objectives during the first eighteenmonths of the NOMIRACLE project are:• To evaluate the use of (semi-) quantitative indicators for cumulative risks provided by the

integration of data from WP1.1 and the ranking algorithms developed in WP1.2.• To establish ways to make cumulative risk maps for the EU and selected EU regions and

produce some examples.

Description of work

Evaluation of indicators for cumulative risk (ALTERRA)WP4.4 will start in month 4 of the NOMIRACLE project with the evaluation of (semi-) quantitativeindicators for mapping potential cumulative risks. WP4.4 will integrate with activities across thewhole project. The mapping will be based on the data and output from WP1.1, the rankingalgorithms developed in WP1.2, former experiences in risk mapping among WP4.4 partners andliterature on risk mapping and communication. The principal challenges will be to integrateenvironmental and human health risks, to identify the appropriate indicators for cumulative risksand to highlight the importance of spatial analysis of stressors and risks. This will lead to improvedscientific understanding that will enhance the identification of emerging issues, facilitate highquality decision making by regulators and be more resource-effective. An important criterion for theselection of the appropriate methods is their communicability to policy makers, stakeholders and thegeneral public and ease of interpretation. Elements of WP4.4 will therefore be tuned to the workunder WP4.3 (management context) and WP5.4 (dissemination). Close collaboration with thespecific target research project HAIR (to be started in 2004; including partners from ALTERRAand NERI) will enable the use of recent data and novel developments in harmonization of(pesticide) risk indicators. A similar collaboration will be established with the ALARM project(including partners from NERI and UFZ), enabling the use of data on chemical risks forbiodiversity.

Data formats and integration (UFZ and other partners)Important issues in risk mapping are the coherence of Geographic Information System (GIS) dataformats and the aggregation of spatial data. Coherence of data formats of the output delivered bythe partners from other work packages (notably WP 1.1 and WP1.2) will be guaranteed by adheringto standardised exchange formats for spatial and non-spatial data that will be discussed and adoptedby the WP4.4 partners during a workshop organized in WP1.1. This integration will be supportedby the NOMIRACLE Data Management Work Package (5.2) that will undertake a survey of thedata needs and deliverables of the NOMIRACLE project. Building upon previous experience andwork done in WP1.1 and WP4.2, UFZ will develop guidelines for the aggregation of spatial datamainly in relation to smoothing and attenuation phenomena. This work comprises a literaturereview and a case study for the risk to human health.

NOMIRACLE - Part B 136

Production of risk maps (all partners)The final activity of the first stage of WP4.4 is to identify tools and techniques to produce risk mapsfor the EU and selected EU regions. In practise, this activity will focus on the visualisation of therisk calculations in a spatial context. One major technique will be the ability to present results thatare meaningful to policy makers, stakeholders and the general public. One important mechanism inthis respect is the availability of regional and pan-European risk maps and the integration of thisinformation with other environmental/socio-economoic parameters.

WP4.4 will investigate techniques to present risk values :• on a grid basis (e.g. cells of 1 km, 10 km or even 50 km resolution) to show variability on a

continuous basis across an area;• on administrative basis (e.g. Eurostat NUTS nomenclature) to highlight political and socio-

economic context;• or as river basin scale to show the environmental impact (Water Framework Directive etc.).

By using spatial analysis tools embedded within GIS software, the implication of calculated risk forspecific population sectors or environmental compartments could be highlighted (e.g. cumulativerisk and crop yield, land use change, climate change scenarios, etc.).

Selection of the EU regions will take place in WP1.1 based on data availability and the presence ofcumulative stressors. Furthermore, the selection of regions will aim at representing different bio-geographical and socio-economic regions of Europe. Data are already available on Denmark, theNetherlands, and Italy making these interesting areas for examples of the techniques. Anexplorative study on data availability in WP1.1 will identify other EU regions in southern, centraland eastern Europe that are suitable for making risk maps. Furthermore, the modelling activities inWP4.2 will result in detailed spatial data on cumulative risks for specific locations or regions, e.g.the region of Leipzig in Germany and the Ebro delta in Catalonia.

Of particular interest to EU policy makers will be the up-scaling of local and regional scalemodelling to generate a pan-European view. In this context, the JRC will focus on the possibility todevelop methodologies for producing risk maps at EU scale. ALTERRA will focus on maps for theNetherlands; NERI for Denmark, UNIMIB for Italy, UFZ for Leipzig and URV for the Ebro delta.Depending on the outcome of the explorative study in WP1.1, more regions from southern, centraland eastern Europe could be added after these first 18 months. Dissemination of potentialcumulative risk maps will be dealt with in MP5.

Deliverables

D.4.4.1 (month 15): Methods and tools to provide cumulative risk maps

D.4.4.2 (month 15): Guidelines for the aggregation of spatial data

D.4.4.3 (month 18): Examples of cumulative risk maps for EU and a limited number of selected EUregions

Milestones20 and expected result

M.4.4.1 (month 9): Work shop in WP1.1 on data formats and decision on which EU regions toinclude in the mapping effort based on results exploratory study

20 Milestones are control points at which decisions are needed; for example concerning which of several technologieswill be adopted as the basis for the next phase of the project.

NOMIRACLE - Part B 137

M.4.4.2 (month 12): Establishment of harmonized methods for the aggregation of spatial data anddistribution among WP4.4 partners

M.4.4.3 (month 18): Finalization of risk map and other visualization examples (and forwarding toWP5.4).

NOMIRACLE - Part B 138

Work Package 5.1 description:

General Project Management (18 month plan)

Work package number WP 5.1 Start date or starting event: Month 0

Participant id NERI UFZ NERC DESUNPerson-months per participant 10 0.5 0.5 0.5

ObjectivesThe objectives of WP 5.1 are• to manage the resources of the project in a professional and efficient way• to implement decisions by the Management Board• to keep deadlines for reporting• to co-ordinate the necessary interactions between partners, work packages and research pillars• to prepare the overall evaluation of results and deliverables• to organise meetings of the General Assembly, the Management Board, the Advisory Board and

workshops• to ensure efficient communication and dataflow between participants

Description of work

Project Co-ordinator

The Project Co-ordinator will be responsible for the scientific, financial and administrative co-ordination of the project. He will be the intermediary between the partners and the Commission. Hewill be responsible for

• collecting, integrating and submitting project deliverables,

• the submission of technical and financial reports and other required documentation for theEuropean Commission,

• the distribution to the Consortium members of the funds received from the Commission,

• communication of half-annual reports via the password-protected homepage

• informing the affiliated EC scientific officer about meetings and agenda.

The co-ordination of the project will be achieved through the daily work within the ProjectSecretariat, semi-annual meeting of the Management Board and annual meetings of the GeneralAssembly of Partners.

Management Board

The Management Board consists of Research Pillar Leaders and members ensuring the cross-cuttingof environment and human health, and of training and dissemination activities, will be responsiblefor the decision making and communication between project participants and end-users. TheManagement Board will have ordinary semi-annual meetings, and will be in frequentcommunication with the Project Co-ordinator via e-mail. The first meeting of the ManagementBoard will take place in the beginning of the project to adjust the project and set up the frameworkfor the project in a general plan. Further 3-4 meetings will be planned during the first 18 months.The responsibilities of the Management in the first 18 month will be the following:

• to focus the attention and efforts towards the main aims of the project,• to approve technical and financial reports to the Commission

NOMIRACLE - Part B 139

• to review any suggested minor or major changes of the working programme and give theirexplained acceptance or objections,

• to prepare and plan all scientific, logistic and practical arrangements concerning commonactivities across the Research Pillars and WPs,

• to ensure the implementation of the dissemination plan• to ensure the implementation of the training programme• to give advice to the Project Co-ordinator on adjustments of the activities• to approve

Project Secretariat

The Secretariat is anticipated to begin its work during contract negotiations, and at the first day ofthe project, assist the Project Co-ordinator with administrative and financial matters, and beresponsible for the daily contact with all partners. The Project Secretariat will prepare all draftdocuments for internal management and for submission to the Commission, and establish andmaintain a homepage, and it will support the WP leaders to set up a project database (WP 5.2), atraining programme (WP 5.3) and a dissemination plan (WP 5.4). During the first 18 month, theSecretariat will prepare 3 ordinary meetings of the Management Board, two meetings of the GeneralAssembly and two meetings of the Advisory Board. Further, the Secretariat will plan an internaltraining workshop on integration of environment and human health in collaboration with WP 5.4,and two sessions on international conferences, e.g. held by SETAC.

Advisory Board

During the first 18 months, the Advisory Board will have two meetings with the following agenda:

• to comment on the project plans• to comment on the scientific results, publications and reports• to contribute to the training programme• to contribute to the dissemination plan with focus on the end-user

Deliverables

6-18 months:

D.5.1.1 Financial and technical reports to the European Commission

3-18 months

D.5.1.2 Meetings of the Management Board (at least 3), the Advisory Board (2) and the GeneralAssembly (2)

6 months:

D.5.1.3 Dissemination plan including internal and external homepage and a Quarterly Newsletter

6-18 months:

D.5.1.4 Organisation of two scientific sessions (e.g. in SETAC Conferences)

Milestones and expected resultM.5.1.1 First Meeting Management Board (1st or 2nd month of project)M.5.1.2 Meeting of the General Assembly (2nd month)M.5.1.3 Meeting of the Advisory Board (2nd month)M.5.1.4 Internal homepage in work (6th month)M.5.1.5 External homepage in work (7th month)M.5.1.6 First issue of Newsletter (7th month)M.5.1.7 Scientific sessions (at scheduled SETAC Conferences, spring 2005, spring 2006)

NOMIRACLE - Part B 140

Work Package 5.2 description:

Data management (18 month plan)

Work package number WP 5.2 Start date or starting event: Month 0

Participant id JRC NERIPerson-months per participant 10.5 1

ObjectivesThe objective of the NOMIRACLE Data Management System is to bring together data fromdifferent pillars and to provide a single access point for information and deliverables from all WPs,by means of a directory to locate data and resources, providing direct access to data as well as anindication on data quality.

The scope of the first 18 months of activity in WP 5.2is fivefold:a) to gather the necessary information on the nature and type of core data sets utilised and

generated by the NOMIRACLE project;b) to design, develop, test and implement an internet based, platform independent meta-data /

cataloguing system that will be used to record and allow access to NOMIRACLE data sources;c) to fill the data management system with information;d) to maintain the data management system in line with the progress of the project.e) To provide the means for identifying the quality of the data sets utilised and produced by the

activities of NOMIRACLE.

Description of workWP 5.2 will commence by making an inventory of the data requirements and deliverables of theNOMIRACLE project. This inventory will be carried out initially via a survey of the projectdeliverables and a questionnaire that will be distributed among the NOMIRACLE partners. Thequestionnaire will request partners to indicate data sources, data type, availability, etc. at theparticipating institutes. Partners will be asked to indicate which data are official deliverables of theproject and are used as inputs by other WPs.

Following the information gathering exercise, a draft data management system will be designedbased on clearly defined needs of the users and project managers. It is proposed that the datamanagement system will be operated as an on-line system based on current Internet protocols andmeta-data standards. A draft design plan will be circulated among the project participants forcomments and suggestions. The finalised design concept will then be developed as prototypesystem and requested functionality demonstrated to users..

Within the data management system, a number of key issues can be pre-defined. The system willprovide an Internet based, platform independent interface to the information stored in theNOMMIRCALE Consortium. An appropriate meta-data or cataloguing template will be defined torecord the necessary information on the data.

A meta-data gateway will be developed to search the catalogue for data held at variousNOMIRACLE partner sites. The outcome of the search will be a list of relevant data sets. Clickingon any of the links will take the user to the relevant HTML view of the data set record in thecatalogue. A query of the Data Management System will allow searches on any text in a cataloguerecord, and set bounds on the time and space coverage of the dataset. Using a search form the DataManagement System will allow searches on any text in a catalogue record, and set bounds on the

NOMIRACLE - Part B 141

time and space coverage of the dataset.

Where required, the data management user interface will provide access to NOMIRACLEdeliverables sources.

A crucial aspect of data management is traceability and reliability of the information being stored inthe system. The data management system will provide users with tools for identifying the issuesrelating to the quality of the data sets utilised and produced by the various WP of NOMIRACLE.

A draft design plan will be circulated among the project participants for comments and suggestions.The finalised design concept will then be developed as prototype system and required functionalitydemonstrated to users.

Following a testing phase, a full implementation of the system will be carried out by month 10 ofthe project.

On completion of the implementation phase of the Data Management System, WP 5.2 will co-ordinate the data input of the catalogue records and maintain the data management system in linewith the progress of the project.

To support users, the WP 5.2 will maintain a NOMIRACLE Data Management System Help Deskto assist users in the use of the system.

DeliverablesThe schedule for the first 18 months of deliverables will consist of the following data packages.

2 months:D.5.2.1 questionnaire on the nature and type of core data sets utilised and generated by theNOMIRACLE project;

5 months:D.5.2.2 draft design document on NOMIRACLE data management system and implementationstrategy;

8 months:D.5.2.3 draft NOMIRACLE data management system and implementation strategy;

10 months:D.5.2.4 operational data management system

16 months:D.5.2.5 review of the operation and functionality of the NOMIRACLE Data Management System.

Milestones and expected results for the first 18 monthsM.5.2.1 design document for data management system;M.5.2.2 draft NOMIRACLE data management system;M.5.2.3 operational NOMIRACLE data management system.

NOMIRACLE - Part B 142

Work Package 5.3 description:

Training and Demonstration (18 month plan)

Work package number WP 5.3 Start date or starting event: Month 0

Participant id DBUA NERI UFZ DESUN JRC SYKE APINIPerson-months per participant 0.5 1.5 0.5 0.3 1 0.8 4

Participant id ENVI DIAPerson-months per participant 1.3 0.3

ObjectivesThe objectives of WP 5.3, Training, are to train and exchange scientific personnel and students, andto organise training activities orientated to research managers, industrial executives (via theparticipation of SMEs) and other potential end-users (e.g. national/regional regulators, consultants).In addition to within project scientists training, training activities will target EU nationals, accessionand candidate countries and other third countries, such as those North African countries borderingthe Mediterranean, countries from the Baltic region, and China, Brasil, Costa Rica, etc. Aninternational PhD programme will be established.

Description of work

Scientists and students involved in this project will have the opportunity to be trained in variousmethodologies used by different researchers that would not otherwise be possible in a smallerproject. Staff exchanges will be planned during the kick-off meeting, with the likely focus for theiractivities being the various requirements for data and analytical standardisation, technical support,student guidance, analysis and the dissemination of results.

Deliverables1 months:D.5.3.1 Organisation of an internal training session, at the kick-off meeting6-18 months:D.5.3.2 Organisation of three advanced studies courses18 months:D.5.3.3 Organisation of one technical session18 months:D.5.3.4 Programme for an international PhD programme

Milestones and expected results for the first 18 monthsM.5.3.1 planning of staff exchanges (month 1)M.5.3.2 Internal training session (month 1)M.5.3.3 First advanced studies course (month 6)M.5.3.4 Second advanced studies course (month 12)M.5.3.5 Third advanced studies course (month 18)M.5.3.6 Announcement of the International PhD programme (month 18)

NOMIRACLE - Part B 143

Work package 5.4 description:

Dissemination and Exploitation (18 month plan)

Work package number WP 5.4 Start date or starting event: Month 0

Participant id SYMLOG NERIPerson-months per participant 2.5 1

All other partners participate in this activitywith input for scientific and populardissemination

ObjectivesThe objectives of WP5.4 are to:

• Develop and maintain a calendar of expected dissemination products• Contribute to the internal and external dissemination plans• Co-ordinate or perform their implementation.

Description of workFive major functions are assured by the Dissemination and Exploitation WP:A. Contribute to integration of parallel R&D outputs across the IP through the restricted access

homepage of NOMIRACLE in collaboration with the Project Secretariat (WP 5.1).B. Co-ordinate internal peer review of dissemination products.C. Co-ordinate the dissemination of R&D outputs (data, models, reports, articles, conference

presentations, PhD theses; novel laboratory, field and office methods for science-basedcumulative risk assessment and management).

D. Ensure the accessibility of NOMIRACLE outputs through multiple media (electronic, print,including popular science outlets) and in regard to SCALE data sharing.

E. Compile and input feedback from end-users.

Deliverables4 monthsD.5.4.1 IP internal (including peer review) and external reporting proceduresD.5.4.2 Templates for user feedbackD.5.4.3 Procedures to analyse and redirect feedback to WPs12 months

D.5.4.4 Stakeholder mapping, focusing on key actors and representatives of key civil societygroups, including emergent ones18 months

D.5.4.5 Highlights from First Periodic Report of NOMIRACLE summarised for submission topopular science journals and other relevant media

Milestones6 monthsM.5.4.1 Newsletter publicationM.5.4.2 Homepage postings (in collaboration with WP 5.1 and WP 5.2)12 months

NOMIRACLE - Part B 144

M.5.4.3 Identification of issues and needs in user groups through e.g., consultation with theNOMIRACLE Advisory Board of Stakeholders

M.5.4.4 Identification of relevant dissemination outlets including conferences and meetings;development of a staged strategy for deciding when to invest resources in attending suchmeetings

18 monthsM.5.4.5 Press release of highlights from First Periodic Report of NOMIRACLE

NOMIRACLE - Part B 145

B.9 Ethical issues: using animals for science

The requirement for using animalsNOMIRACLE will make use of many techniques to evaluate the impact of toxic stressors on health,including animal experiments. The task of the project requires to compare effects on numerousecological compartments and life forms. Besides of plants and invertebrates, which are considerednot to raise major ethical concerns when used for science, vertebrates are included as well.Specifically, NOMIRACLE includes research using fish, mice and rats.

The use of vertebrate animals can not be avoided without compromising the overall goal ofthe project to an unacceptable extent. However, every effort will be made to establish knowledgewithout animal experiments by using information resources already available. Numbers of animalswill be minimised through sharing of data and materials and by keeping to accepted minimalnumbers to reach statistically significant results in the respective context. Furthermore, efforts forthe development of methods for the replacement of animal experiments are included within theresearch program.

NOMIRACLE will, thus, address the following requirements that have been outlined in the‘Strategy for a future Chemicals Policy’ in order to keep animal testing to a minimum (EC 2001):

1. Use of existing toxicological information on the substances2. Modification of the general testing requirements in order to incorporate exposure-driven testing

where appropriate3. Development of tailor-made testing programmes for higher tier testing and4. Development of further alternative testing methods using fewer or no animals

All studies will be carried out according to the respective national animal welfare regulations of theproject partners, which implement the requirements of the European Directive 86/609/EEC (EC1986) regarding the protection of animals used for experimental purposes.

Specific animal experiments using vertebratesFish: EKUT will use early life stages (eggs, embryos, larvae) of zebrafish for exposure studies.These animals are accepted to act as model organisms for aquatic vertebrates. Currently, there areno accepted in-vitro alternatives to acute and chronic fish tests (IEH 2001). For ethical reasons,preference was given to early life stage exposure rather than to exposure of adults. Eggs will onlybe obtained from a hatchery established at EKUT. Replicate number will be kept to a statisticallysignificant minimum and, to avoid parallel exposure, experiments will be conducted together withLIMCO, and samples will also be shared with UFZ. Exposure studies will be conducted inaccordance to the German law after permission by the local authority (Tübingen districtmagistracy). The experiments are to be supervised by the animal care authority of the BiologicalFaculty at EKUT.

Mice: USALZ has previously established mouse models for allergy to investigate experimentaltherapies of allergic disease and to study influences of environmental factors on allergic processes.In the context of NOMIRACLE it will be tested whether the simultaneous application of chemicalsor mixtures identified as relevant stressors has an influence on the development of allergies or onongoing allergic sensitivities. The results have direct implication for human health and for theevaluation of environmental toxins. All applications will be simulated beforehand in cell cultureexperiments, to determine effective concentrations as well as possible before using animals. Thenumber of animals will be kept to a minimum. The experiments at USALZ are subject to approvaland supervision by the national authority (Bundesministerium für Bildung, Wissenschaft undKultur) and will be performed at a certified animal care facility at USALZ.

Rats: NERC has established capacities for toxicological investigations in the rat. Manytoxicological data are available on rats, so new experiments will be performed only if no data are in

NOMIRACLE - Part B 146

accessible sources. Since combinations are going to be studied, in particular combinations betweenchemical and non-chemical factors, it is possible that the established knowledge does not cover theaspects analysed here. If it should not be necessary to study toxic/stress combinations becauserelevant informations can be obtained in other ways, rat experiments may not have to be performedat all. If they become necessary, all these experiments are subject to approval and supervision by thenational authority (Home Office), and will only be performed by certified personnel in a licencedfacility at NERC.

3R rule: Replace, Reduce, RefineThe accepted rule guiding animal experiments is to replace animal use, reduce the number ofanimals and refine procedures to cause less suffering (3R rule). NOMIRACLE will reduce thenumber of animals by making as much use as possible of available data sources. The fishexperiments will focus on early life stages (eggs, embryos, larvae), which aims at refiningprocedures. It is in this context important to confirm that early life stages are valid for replacingadult fish in complex situations (like chemical plus non-chemical stressors) where this principle hasnot yet been established. Successful experiments within NOMIRACLE may therefore reduce thenumber of adult fish used in the future.

USALZ will adapt assays with already established stable human cell lines for applicationwith the agents investigated in NOMIRACLE, and will also produce new cell lines of particularrelevance for project goals. These cell lines are stably transfected with reporter genes under thecontrol of promoters controlling expression of key proteins from the human immune system. Thegoal is to establish cell culture based tests reflecting suppression or deviation of immune responses.An advantage is that human cell lines are used, and an obvious outcome is that assays for immunefunctions are developed which do not require animals. The mouse experiments performed atUSALZ are required to test whether cell lines and mice perform in a similar way when treated withenvironmental toxins alone or in combination with non-chemical stressors (in this case allergens).Assays developed here may lead to replacement of mice in further studies.

B.10 Gender issues

This proposal will contribute to the activities initiated by the EU to improve the gender balance.During the establishment of the Consortium it became evident that a large effort is needed toenhance the participation of women in this area of research. Four out of 38 of the involved PIs arewomen, and 2 out of 16 of the Management Board members/deputies are women.

B.10.1 Gender Action Plan

Staff within the projectTo promote equality, appropriate measures will be taken during recruitment of post-graduate/post-doctorals to free positions. In addition, efforts and positive measures will be made to attract andencourage the participation of female students/ scientists/ technicians to the planned trainingactivities.

Training programmesThe criteria for selection of participants for the several training activities will be drawn up by theManagement Board. Positive action will be taken to assure a proper gender balance in all trainingactivities.

DisseminationWP 5.4 (Dissemination and Exploitation) will provide leverage in regard to gender issues. The WPleader Claire Mays acts as the IP guardian of special-needs sensitivities, compiling information and

NOMIRACLE - Part B 147

guidance, and co-ordinating attention to sub-populations distinguished by sector, region, language,gender, vulnerability, etc.

The dissemination plan for R&D outputs by NOMIRACLE will include active efforts toadapt and direct information to special groups. WP 4.3 outputs on public risk perception will helpWP 5.4 to characterise special end-user communication needs linked to gender and/or rolesensitivity to risk, as in the case of women, or mothers of young children. These sub-populations aresalient in the context of SCALE focus on children's health, which is to be addressed byNOMIRACLE RP 4.

WP 5.4 will also develop pathways for feedback from stakeholders including women andgroups or institutions serving women and mothers.

B.10.2 Gender issues

The gender issue is particularly relevant in risk perception and risk communication. Studies in manycultural contexts have found that women in general perceive higher risks associated withtechnology, natural hazards, or lifestyle features than do men. (Similarly, male or female membersof minority or disadvantaged groups perceive higher risks than do male members of the dominantsocial group.) Women, and especially mothers of young children, report themselves as more likelyto take protective actions in case of averred risk or major accident. Psychology and sociology alsoindicate significant gender differences in intuitive cost/benefits analyses performed by parentsregarding their own risk, or the risk experienced by their progeny. All these findings are highlysuggestive in regard to the societal interpretation of cumulative exposure risks.

A specific study on the perception/communication of environmental problems onbiodiversity and ecosystems by females is included in WP 4.3. This study will address an aspectthat is considered particularly relevant for the future of sustainable development and therefore of theEU and national policies. As Sustainable Development is characterised by “the protection of futuregenerations” the perception of sustainability, the balance between current and future generations,the societal concerns and the final risk management decision may be influenced by gender. Thisinfluence may be enacted directly, by female decision makers, or indirectly, through femalebehavioural patterns of consumption, or through pressure by female populations upon other actorsin the system.