ECOLOGICAL RISK ASSESSMENT (EcoRA) Methodology, role of ICTs and its consequence for data processing Dušek L., Holoubek I., Jarkovský J., Koptíková J., Kohout J., Šmíd R., Pavliš P., Mužík J., Hřebíček J. Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Why to discuss risk assessment and environmental informatics I. Very important and ethically binding field of application of information technologies II. III. IV. Field that relies on gathering and processing of large volume of very heterogeneous data Key role of mathematics and modelling: some computational steps really determine the success of the whole methodology Key role of professional communication and information service: risk without communication is not risk Ecological and human risk assessment is a real challenge for development of information systems, tools for automated data processing and for expert communication.
Presentation plan DATA INFORMATION Project XXXX x Data management IS services x Web page Discussion club Help Back Presentations Data browser Expert services << ASSESSMENT PROESS >> I. Role of informatics in EcoRA II. Problem definition and conceptual model III. Assessment scenario and endpoint selection IV. Biological receptors and assessment endpoints V. Parametrization of experimental data from bio-tests VI. Monitoring of real ecosystems: model approach & strategies
ECOLOGICAL RISK ASSESSMENT I. Role of informatics EcoRA methodology process analysis data model Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Ecological risk assessment (EcoRA): key methodical steps Problem formulation Hazard identification VALIDATION OF RESULTS Multicomponent exposure assessment Biological effects assessment Risk characterization COLLECTION & PROCESSING OF DATA Interpretation & Management COMMUNICATION OF RESULTS
EcoRA and related key information sources PROBLEMS Descriptive phase Problem formulation Libraries, databases, Information systems Modelling Standardized tests Chemical monitoring Exposure assessment Biological effect(s) assessment Biological monitoring Risk characterization INFORMATION Data aggregation & Synthesis COMMUNICATION
There are many demands on information service from EcoRA COMMUNICATION OF RESULTS DATA INFORMATION COMMUNICATION TARGET INFLUENTIAL FACTOR EXPOSURE RECEPTOR CHARACTERIZED RISK Clear conclusions Value communication Public Quantified risk Cost analyses Exact methodology Significant & verified results Effectiveness Managers Minimized uncertainties Experts
Informatics: indispensable part of each methodical step RISK CHARACTERIZATION Uncertainty analysis Sampling Modelling Expert judgement EXPOSURE ASSESSMENT Biota Bio-tests Bioindicators BIOLOGICAL EFFECTS CONCEPTUAL MODEL & SCENARIO Biomonitoring Benchmarking Probability estimation Data processing Modelling Multivariate analyses Model of area of interest Experimental design Problem definition Hazard identification Retrospective problem Actual / urgent situation Prospective evaluation Data gathering Data aggregation Information services Optimization & Processing in information systems & Communication
ECOLOGICAL RISK ASSESSMENT II. Problem definition, conceptual model and experimental plans Data components database processing conceptual model Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Problem definition as obligatory entry to the assessment process Further analyses according to scenario and formal methodology Are we able to prioritize key steps? CONCEPTUAL MODEL & SCENARIO Are we able to define sampling and measurement plan? Problem definition Hazard identification Are we able to stratify the area of interest according to risk? Retrospective problem Actual / urgent situation Prospective evaluation Are we able to recognize risk situation?
Problem definition must deal with all included components Stressor(s) Stress factors Source of stress factors Hazardous compounds SITUATION Area of interest Risk situation Natural factors Biological receptors Assessment endpoints Measurement endpoints
Problem definition in process analysis I. Database A. National or other accepted benchmarks Searching for information Cl x PCB Cl y R13 R12 R11 R10 R14 R9 R1 R8 Database B. R2 R7 R3 R6 R4 R5 Server Regional data sources Pilot results Database C. Geographic information systems & Data processing
Problem definition in process analysis I. Information processing Database A. Server Database B. Sorting Filtering Aggregation - Targeted searching for Information - Summary reports PROBLEM DEFINITION & CONCEPTUAL MODEL Ecological risk assessment SCENARIO Database C. EcoRA methodology Exposure assessment Risk characterization Biological effect(s) assessment
ECOLOGICAL RISK ASSESSMENT III. Assessment scenario and endpoint selection Tiered analyses situation plan assessment scenario Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Problem definition = complex information survey Source of contamination Environmental characteristics Development of exact situation plan integrates also exposure assessment (at least at screening level) Compounds of interest EXPOSURE Pilot (screening) test COMPARTMENTS ACCORDING TO POTENTIAL RISK
Problem definition as a part of TIERED ANALYSIS Problem definition Source Compounds Pilot estimates of exposure Situation Area Situation plan Receptors TIER 1 Chemical analyses Information processing Pilot biological test Assessment scenario Chemical Chemical sampling sampling Exposure assessment Ecosystem Ecosystem monitoring monitoring In situ situ bioindicators bioindicators Biological effects Analyses Analyses of biota biota Biological Biological tests tests Exposure Exposure models models TIER 2... TIER X Chemical analyses Biological test Bioindicators Ecosystem monitoring Estimated actual (or predicted) environmental concentrations AEC Estimated toxicological effect concentrations TEC
Problem definition generates comprehensive SITUATION PLAN??!! Increasing value for assessment EXPOSURE Priority of unit for risk assessment Increasing risk Background sites with no influence of exposure? Uncertain influence and/or uncertain assessment endpoint! Potentially affected sites, still clean or with negligible effect Area with probable and substantial toxic impact Already strongly affected area with remarkable effects
. and related tasks for subsequent analyses reducing uncertainty EXPOSURE What is the optimal and safe sampling plan? How to minimize uncertainty in final risk estimate???! Increasing value for assessment! What is the optimal set of further analyses? Priority of unit for risk assessment We need exact assessment scenario Increasing risk
Assessment scenario and basic principle: Where is the problem Scenario is in direct relation to estimated (predicted) exposure pathways: all further analyses follow from this starting point Stressor(s) Soil Primary exposure Secondary exposure Water Air Sediment Soil Air Sediment Further tests Water Time
Scenario as milestone of the assessment process Source of contamination Environmental characteristics Compounds of interest?!? SITUATION PLAN! Scenario Primary exposure Secondary exposure Soil Soil Pilot tests (screening) Sediment Water Air Biota Sediment Water Air Biota Further tests
ECOLOGICAL RISK ASSESSMENT IV. Biological receptors and assessment endpoints Biological receptors assessment endpoints stochastic processing Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Endpoint selection must follow strict rules Ecological relevance Susceptibility to exposure Reprezentativeness Unambiguous definition Accessibility for measurement Environmental & societal importance
Endpoint selection: there are two basic strategies Stressor Sub-cell systems Population Cell systems Communities Tissuetests Organisms Ecosystems Bio-tests? Bio-indication
Endpoint selection: there are two basic strategies Model systems, biomarkers /BIO-TESTS/ Specificity Risk characterization Assessment on model systems Evaluation of real ecosystems Problem formulation Ecosystem-related interpretation Reprezentativeness Both approaches must be combined in optimal way Biological examination provides always stochastic ( uncertain ) data The analyses are basically multivariate Mechanisms of effects Standardized approach Monitoring of real ecosystems /BIOINDICATION/
Each strategy has its advantages and disadvantages Interpretation?? Time changes? Influence of environmental factors?? Suitable parameters Contribution to final risk estimate and its uncertainty
Stochastic processing of data: key methodical tool In final risk characterization are chemical and biological data evaluated in direct contrast Dichotomous evaluation Stochastic model evaluation Hazard quotient HQ = AEC TEC Actual environmental concentration = Toxicological effect concentration Probability density function X? MATC Concentration of stressor X? MATC Distribution 1: Concentration of stressor in environment Distribution 2: Concentration of stressor in relation to effects Given probability of significant negative influence (events) Maximum accepted toxicant concentration (threshold)
ECOLOGICAL RISK ASSESSMENT V. Problem 1: parametrization of experimental data from bio-tests Dose-response curves parameters - models Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Bio-tests can be employed in any assessment step Problem definition Bio-tests Source Compounds Pilot estimates of exposure Situation Assessment scenario Area Situation plan Receptors 1) 2) Toxicological characteristics of key compounds (NOEC, LC x data from databases) Pilot (screening) tests (search for proved effects, exposure) TIER 1 Chemical Chemical sampling sampling Exposure assessment Ecosystem Ecosystem monitoring monitoring In situ situ bioindicators bioindicators Biological effects Analyses Analyses of biota biota Biological Biological tests tests Exposure Exposure models models 3) Tests in batteries optimized according to assessment scenario (often repeated measurement) TIER 2... TIER X Estimated actual (or predicted) environmental concentrations AEC Estimated toxicological effect concentrations TEC
Standard output of bio-tests: dose-response curve Response of biological systém (%) 90 50 10 Low-effect concentrations NOEC: No Observed Effect Concentration -> the highest concentration of a chemical in a toxicity test that causes effects that are not statistically significant from control LOEC: Lowest Observed Effect Concentration Applied concentration levels of tested compound -> the lowest concentration of a chemical in a toxicity test that causes effects that are statistically significant from control Concentration parameters Probability density function EC TEC X effect concentrations LD 50, LC 50, LD X, LC X -> Median lethal dose (concentration), X- effect lethal dose (concentration) Concentration of stressor
Dose-response problem in its textbook form Standard experimental design Standard dose-response curve Control Concentration X1 Concentration X2 Concentration X3... Growing concentration of tested compounds Concentration Xp Response of biological systém (%) 90 50 10 Applied concentration levels of tested compound Straightforward comparison of experimental variants Model estimates and extrapolation between doses NOEC, LOEC L(E,I)C 50 LC x, EC x, IC x
Dose-response problem in its textbook form Comparative approach? Model approach Control Concentration X1 Concentration X2 Concentration X3... Growing concentration of tested compounds Concentration Xp Response of biological systém (%) 90 NOEC, LOEC 50 10 Asking the Right Questions: Ecotoxicology and Statistics. SETAC, London, 1995. Statistical Analysis of Aquatic Applied concentration levels of tested compound Toxicity Data. OECD Workshop. Braunschweig, 1996 NOEC / LOEC approach criticized and refused: low standard of estimates, dependent on experimental design, low extrapolation potential Preferred way: dose-response models and extracted LC X (EC X ) Calabrese, Baldwin, Hum. Exp. Toxicol. (2000): Chapman et al., SETAC Europe (1995); Chapman et al., Environ. Toxicol. Chem. (1996); Chapman et al. Ecotoxicology (1996); Kooijman, OIKOS (1996)
Model solution for dose-response modelling: regression Model regression standards: example pα Y α Estimate of EC A. Výpočet efektivní koncentrace (EC x ) Y = 1+ α p β (x - μ p) 1 - p e x = ln (koncentrace) μ p = ln ED p pα B. Výpočet prahové koncentrace (EC 0 ) Estimate of threshold concentration Y α Y = α Y = 1+ α p 1 - p eβ [(x - μ o) - (μ p - μ o)] x = ln (koncentrace) μ p = ln ED p μ o = ln ED o x < μ o x > μ o µ p X µ o µ p X Evaluation of hormesis model Y C. Výpočet prahové koncentrace (EC 0 ) - "Hormesis model" OECD, 1996-1999 α α + δx Y = 1+ p β (ln x - g) 1 - p e x = koncentrace μ o = ln ED o µ o ln x Y + resistence estimate 2 ( ) 2 X - X0 = B C + Y0 A
Real dose response problem: bio-tests in environmental matrices Patterns of various dose response data: examples 1.2 1.0 0.8 0.6 0.4 0.2 REZISTENCE 0.0 0.0001 0.01 1 100 10000 1.4 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0001 0.01 1 100 10000 1000000 1.4 1.2 1.0 0.8 0.6 0.4 0.2 1.2 1.0 0.8 0.6 0.4 0.2 LAG & HORMESIS SIGMOIDITY 0.0 0.0001 0.01 1 100 10000 0.0 0.001 0.1 10 1000 100000 SUSCEPTIBILITY
Real dose response problem: bio-tests in environmental matrices Amador et al., 1994: In Function of Somatic Cells in the Testis, Springer Verlag, New York, pp. 293 318. We need robust model tools to estimate the patterns and associated parameters Response Odpověď Response Odpověď I Odpověď B L H S B L H S I II III IV I II Dávka Dávka III IV Dose III Response Odpověď Dose B L H S B L H S I II Dávka III IV I II Dávka III IV Dose Response Dose II IV Response Odpověď Response Odpověď V B L H S B L H S I II III IV Dávka Dose VII B L H S I II III IV Dávka Dose Response Odpověď VI I II III IV Dávka Dose
Median effect equation: theoretical background 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.01 1 100 10000 1000000 F A /F N = (C/EC 50 ) m log (F A /F N ) = = m.log( C i ) - m.log( EC 50 ) Pro F A = F N = 0,5: [m.log(c i )] / m = log(ec 50 ) EC x (LC x ) F A = 1 / [1 + (EC 50 /C)] m C = EC 50 [F A / (1 F A )] 1/m F A, F N fraction of affected subjects, individuals, cells,. C concentration of applied compounds M estimated coefficient (related to sigmoidity of the curve) EC 50 median effect concentration Chou. T.C. 1980-1989 Carcinogenesis; Trends in Pharm. Res.
ECOLOGICAL RISK ASSESSMENT V. Problem 2: monitoring of real ecosystems: model approach and strategies Biondication reference sites analog modelling Developed by RECETOX-TOCOEN & Associates, Masaryk University, Brno
Biondication and ecosystem monitoring in tiered analysis Problem definition Ecosystem monitoring Source Compounds Situation Area Receptors Pilot estimates of exposure Situation plan 1) Key environmental data from the area of interest TIER 1 Assessment scenario Chemical Chemical sampling sampling Exposure assessment Ecosystem Ecosystem monitoring monitoring In situ situ bioindicators bioindicators Biological effects Analyses Analyses of biota biota Biological Biological tests tests Exposure Exposure models models 2) Evidence of toxic impacts in past Benchmarking for the status of valuable ecosystems TIER 2... TIER X Estimated actual (or predicted) environmental concentrations AEC Estimated toxicological effect concentrations TEC
Analog studies in EcoRA : National Research Council Committee on the Applications of Ecological Theory to Environmental Problems (1986) National Academy Press, Washington D.C. Biomonitoring of ecosystems: very complex multivariate data Y Y X1 X2 X3 X4 Xp 1 X 1 X 2 abiotic data n biotic data median value 1Rather až1 (27) than confidence median value 0,974 až 1 (1) 0,799 až 0,974 (15) 1 až1 (38) 0,522 až 0,799 (16) 0,88až1 (11) 0,216 až 0,522 (15) 0,33až0,88 (25) estimates from repeated trials we apply searching for analog situation (sites)
Analog modelling for ecological risk assessment A) Calibration data set B) Environmental gradient Contaminated sites Reference (clean) sites X 1 X 2 X 3... X p X 1 X 2 X 3... X p Increasing disturbance Increasing contamination Gradient calibration New problem Searching for similarities New problem Searching for similarities X 1 X 2 X 3... X p X 1 X 2 X 3... X p RISK ASSESSMENT SCENARIO
Analog modelling for ecological risk assessment Rivers /Surface Water Ecosystems/ Soils /Terrestrial Ecosystems/???? Clustering of sites /aggregation criteria/ Searching for relevant reference sites Definition of representative sites for detailed studies Natural / evolutionary background of the sites in contrast to ecotoxicological and/or chemical measures: is there any added value?
Analog modelling for ecological risk assessment: strategy Surface water ecosystems (1) (2) Representative biomonitoring network Other available data Representative information background abiotic determinants biotic descriptors Basic typology of sites Ecologically-related clustering (3) Centroids of groups Selected set of sites, suitable and inte teresting for mutual comparisons (candidates for reference sites, representative minimized network,,..) x1 x1 x1 x1 x1 Ecotoxicological and chemical measures Additional characteristics leading to agreement or disagreement with general characteristics
Analog modelling for ecological risk assessment: strategy Hot spots /gradient/ Typology of sites Searching for reference sites Investigative typology of sites Representatives of different types of sites Verification/definition of reference points Set of similar sites (variability study) DESIGN + SCENARIO of EcoRA
Risk characterization Biological effects Analysis & assessment of exposure Ecological Risk Assessment: scenario & conceptual plan Problem description Situation plan Hazard identification