Chapter 10 Water-Sediment Studies

Jeremy Dyson Basel, Switzerland

Outline

• • • •

Defining, Estimating & Using Endpoints Parent Kinetics

– – –

Similarities/differences to other test systems Models and flowcharts Statistics and examples Metabolite Kinetics

–

Similarities to other test systems

– –

When are metabolite kinetics not required?

Models and flowcharts Concluding Remarks

Defining, Estimating & Using Endpoints Application of Parent or Metabolite Volatilisation Water Column Well-mixed Aerobic Water+particulates Water-Sediment Interface Sediment Slow-mixing Oxic to anoxic Transfer Processes Metabolism: Formation & Degradation Metabolism: Formation & Degradation

Defining, Using & Estimating Endpoints

• • •

Persistence Endpoints

–

To determine whether various aquatic ecotoxicolgy studies are triggered, e.g. fish accumulations studies

Modelling Endpoints

–

To use in calculating PEC values as part of an aquatic risk assessment, e.g. FOCUS surface water scenarios

Further Aspects of these Endpoints

–

For Parent or Metabolites

– –

For Degradation or Dissipation For Whole System, Water Column or Sediment

Defining, Using & Estimating Endpoints

• • •

The Water-Sediment System & Definitions

– – –

Behaviour can be more complex than in other systems Straightforward definitions e.g. dissipation from compartments Non-straightforward definitions, e.g. degradation in compartments Study Guidelines and Use

– –

Not always clear if dissipation or degradation required Decisions about endpoints used made on a case-by-case basis Difficulties of Estimation

– – –

Main problem over degradation-transfer correlations No simple, robust & reliable constraints procedures Default worst-case approach if lack of degradation in one compartment, implausible transfer rates (Fsed test), or generally inconsistent with other environmental fate studies

Defining, Using & Estimating Endpoints Kinetic Persistence/Modelling Disappearance Level Endpoints Endpoints Level I System (Parent & Metabs) Degradation (1 comp.) Water column (Both) Dissipation Sediment (Both) Dissipation Level II Water column (Parent) Degradation (2 comp.) Sediment (Parent) Degradation

Parent Kinetics

• •

Similarities to Other Test Systems

– –

Data entry and exclusion Selection of fitting routine

– –

Standard constraints, underlying kinetics etc.

Methods of making kinetic decisions Differences to Other Test Systems

– – – –

Day zero data: put all in water column Data in terms of mass or equivalent, e.g. %AR Do not use concentration data Operation of the worst-case default approach at Level P-II

Models and Flowcharts: Level P-I Kinetic Concept Initial Level Mo Compartment wc + sed

or

wc

or

sed Generic Equation

M = Mo

F

(t)

Data for Disappearance Graphs Data for wc + sed Data for Disappearance Times

DT50/90wc+ sed – calculate directly from the fit DT50/90wc – calculate directly from the fit DT50/90sed – calculate directly from the fit

Models and Flowcharts: Level P-I

• • • •

SFO Kinetics

– –

Default first choice Required for modelling endpoints FOMC Kinetics

–

Evaluate if data depart appreciably from SFO kinetics DFOP Kinetics

–

Offers more flexibility than FOMC with extra parameter Hockey Stick Kinetics

–

Data sometimes appear to have some „breakpoint“ in rate

Models and Flowcharts: Level P-I

• •

System Degradation/Compartment Dissipation Persistence Endpoints

– –

Tier 1: Check if SFO is an appropriate model Tier 2: Identify best-fit model if required Modelling Endpoints

– –

Tier 1: Check if SFO is an acceptable model Tier 2: Correction procedures if SFO not an acceptable model

Models and Flowcharts: Level P-II Kinetic Concept Compartment Application Mo Water Column Mw

k

-

w Sediment

r

s-w Ms

r

w-s

k

s Generic Equations

dMw = -

r

w-s Mw +

r

s-w Ms –

k

w Mw dMs = -

r

s-w Ms +

r

w-s Mw – dt dt

k

s Ms

Disappearance Graph Data for wc Data for water column Data for sediment Disappearance Times

DegT50/90w – calculate directly from the fit DegT50/90s – calculate directly from the fit

Models and Flowcharts: Level P-II

• • •

Empirical Transfer Pattern

–

Able to approximate quite closely Simple Transfer Kinetics

– –

No assumptions about sediment concentration gradients Appropriate if gradients are complex and not measured

–

Appropriate to consider before more complex alternatives First-Order Transfer Kinetics

–

Relatively easy to implement in software packages

Models and Flowcharts: Level P-II Example of Transfer Pattern without Degradation

Models and Flowcharts: Level P-II

• • •

The Fsed Test Definition

–

Fraction in sediment at equilibrium in absence of degradation Modelled Fsed Values

–

Calculated from fitted transfer parameters of Level P-II model

Fsed = r

w-s

/ (r

w-s

+ r

s-w

)

Theoretical Fsed Values

–

Based on system/pesticide properties & diffusion assumptions

Fsed = (Kd



b

+



) / [(Z

w

/Z

D

)+(Kd



b

+



)]

Models and Flowcharts: Level P-II

•

Persistence/Modelling Degradation Endpoints

– – –

SFO Fit (Criteria to be met even if fit acceptable) Consistent with environmental fate data Degradation rates k w and k s >0 as demonstrated by t-test The Fsed test needs to be passed Use 1 of 3 default approaches tested to ensure they lead to worst-case PEC values No Criteria met?

Yes Use estimates as required against triggers/ in modelling

Models and Flowcharts: Level P-II

•

Persistence/Modelling Degradation Endpoints Default approach 1 Passes Fsed test but one degradation rate is zero or fails t-test Set degradation rate to overall system half-life in degrading compartment Set degradation rate to 1 000 day half-life in non-degrading compartment Use default as required in modelling

Models and Flowcharts: Level P-II

•

Persistence/Modelling Degradation Endpoints Default approach 2 Fails Fsed test due to zero transfer rate from sediment to water Set water column degradation rate to overall system half-life Yes Fitted degradation faster in water column than in sediment?

Set sediment degradation rate to 1 000 day half-life Set water column degradation rate to estimated half-life Use default as required in modelling No Set sediment degradation rate to overall system half-life

Models and Flowcharts: Level P-II

•

Persistence/Modelling Degradation Endpoints Default approach 3 Fails Fsed test or inconsistent with E Fate data (degradation) Determine and use default that results in worst-case PEC values: Water column degradation half-life=overall system; Sediment half-life= 1 000 days, or vice versa Use default as required in modelling

•

Models and Flowcharts: Level P-II Persistence/Modelling Degradation Endpoints Default approach 3 Compound 2 Default 3A Compound 2 SFO fit Compound 2 Default 3B

Strongly sorbing compound no degration in water column

•

Models and Flowcharts: Level P-II Persistence/Modelling Degradation Endpoints Default approach 3 Compound 6 Default 3A Compound 6 SFO Fit Compound 6 Default 3B

Weakly sorbing compound no degration in water column

Models and Flowcharts: Level P-II

•

What If the Default Options Need Refining?

– –

Fit a diffusion-based model to water-sediment data A TOXSWA example for such refinement is in Appendix 12 Development needed for a user-friendly implementation of TOXSWA, or a diffusion-based model specific to water sediment systems

Statistics and Examples

• • •

Assessing Goodness of Fit

– – –

Visual Assessment Main tool for assessment

– 

2 Plots of model fits & residuals Test

–

Performed for each compartment, even at Level P-II Supplements visual assessment & model comparison Only a guidance value of 15% error value to pass test

– –

t-Test Reliability of individual dissipation/degradation rates Total df with a significance level of 10% to pass test

Statistics and Examples: Level P-I Compound 6 wc + sed Compound 6 wc Compound 6 sed

Statistics and Examples : Level P-I Compartment Modification DegT50/DT50 in days (



2 ) SFO FOMC HS wc + sed Remove outlier 20.1 (3.6) 20.1 (3.6) 19.8 (3.0) wc Remove outlier 19.1 (2.8) 18.6 (2.7) 18.7 (1.9) sed Remove outlier 21.1 (9.4) 15.2 (6.5) 17.7 (7.7)

Statistics and Examples : Level P-II

Compound 6

Statistics and Examples : Level P-II Compartment Modification DegT1/2 Fsed (%) (



2 value) Modelled Theoretical wc Fix Mo



(3.1) sed 2.16 (9.0) 44 27 - 57

Metabolite Kinetics

• •

Similarities to Other Test Systems

– –

Data entry and exclusion Selection of fitting routine

– –

Standard constraints, data exclusion, underlying kinetics etc.

Methods of making kinetic decisions When Are Metabolite Kinetics Not Required?

– – – –

Sometimes not required for minor metabolites If risks implicitly assessed via higher tier studies Sometimes not if also applied as a „parent substance“ Sometimes not if can add metabolite residues to parent

Models and Flow Charts: Level M-I Defining Persistence/Modelling Endpoints Type of Endpoint Compartment Kinetic Model Dissipation System Decline from peak Water Column „ „ „ Sediment „ „ „ Degradation System Formation & degradation

Models and Flowcharts: Level M-I

• • • •

SFO Kinetics

– –

Default first choice Required for modelling endpoints FOMC Kinetics

–

Evaluate if data depart appreciably from SFO kinetics DFOP Kinetics

–

Offers more flexibility than FOMC with extra parameter Hockey Stick Kinetics

–

Not used

Models and Flowcharts: Level M-I

• •

System/Compartment Dissipation/Degradation Persistence Endpoints

– –

Tier 1: Check if SFO is an appropriate model Tier 2: Identify best-fit model if required Modelling Endpoints

– –

Tier 1: Check if SFO is an acceptable model Tier 2: Correction procedures if SFO not an acceptable model

Models and Flowcharts: Dissipation Level M-I Kinetic Concept Initial Level Mo Generic Equation

M = Mo

F

(t)

Compartment wc + sed

or

wc

or

sed Disappearance Graphs Data for wc Data for wc + sed Data for sed Disappearance Times

DT50/90wc+sed – calculate directly from the fit DT50/90wc – calculate directly from the fit DT50/90sed – calculate directly from the fit

Models and Flowcharts: Degradation Level M-I Kinetic Concept Compartment Parent (wc+sed) Application Mo M p

f

m 1-

f

m

Metabolite (wc+sed) M m Generic Equations

M P = Mo

F

P (t) M m (t) = t

f m

 0 Mo

dF

P (t i ) /

d

t i

F

m (t – t i )

d

t i

Disappearance Graph Parent (wc+sed) Metabolite (wc+sed) Disappearance Times

DegT50/90wc+sed – calculate directly from the fit

Models and Flowcharts: Level M-II

• • •

General Recommendations for Development Data/Parameter Requirements

–

Minimise, e.g. do not use sink data as a first step Kinetics

–

Use first-order kinetics for transfer & degradation processes Formation Fraction

– –

Option to use same fraction for water column & sediment Option to use a default fraction, i.e. that estimated at Level M-I

Concluding Remarks

• • •

General Remarks

–

Complex area of kinetics, but the workgroup has increased understanding of strengths & limitations of approaches, bringing greater transparancy & consistency Parent Kinetics

– –

Resolved endpoint definition, use and estimation In a framework and developed degradation refinement process Metabolite Kinetics

–

Resolving endpoint definition, use and estimation

–

Kinetics still need actively developing for Level M-II