Sediment Quality Objectives Indirect Effects Project

1
Sediment Quality ObjectivesIndirect Effects
Project
Ben Greenfield Aroon Melwani John Oram Mike
Connor San Francisco Estuary Institute (SFEI)
2
Presentation Overview

• Project conceptual framework
• Description of Multiple Lines of Evidence
• Use of information in assessment context
• Methodological issues and results
• Empirical and mechanistic approaches
• Problems of scale, target species
• BAF vs. BSAF

3
(No Transcript)
4
(No Transcript)
5
Pollutant Groups

• Non-ionic organics
• PCBs
• DDTs
• Chlordanes
• Dieldrin
• Methylmercury

• Dioxins
• PBDEs

6
Conceptual Model
Exposure Assessment
Effects Thresholds For Humans
Effects Thresholds For Wildlife/Fish
Chemical uptake via diet, respiration
Effects Assessment
7
Multiple Lines of Evidence Approach
Exposure Assessment
Effects Thresholds For Humans
Effects Thresholds For Wildlife/Fish
Chemical uptake via diet, respiration
Effects Assessment
8

• Sources of Variability
• Exposure
• Diet
• Lipids Weight
• Spatial movement
• Chemical Partitioning

• Effects
• Consumption Rate
• Size
• Risk management goals

• Uncertainty will be addressed by
• Using multiple lines of evidence
• Incorporating several thresholds into each line
  of evidence
• Unlikely risk
• Potential risk to high-risk consumers
• Potential risk to average consumers
• High risk to average consumers

9
Indirect Effects Weight of Evidence
Human Lines of Evidence
Fish and Wildlife Lines of Evidence
Fish Concentration
Fish Concentration
Laboratory Bioaccumulation Concentration
Laboratory Bioaccumulation Concentration
Sediment Concentration
Sediment Concentration
10
Indirect Effects Approach Compared to Rest of
SQO Program

• Similarities
• Integrate multiple lines of evidence
• Use ordinal scale ranking based on thresholds
• Both exposure and effects are important
• Changes
• All lines of evidence are measures of exposure
• Effects thresholds are determined from
  literature/expert opinion
• If local effects information are available, they
  would be included on a case-by-case basis
• All effects assessments are specific to
  individual contaminants (mixtures not accounted
  for)
• Addition of laboratory bioaccumulation component

11
F
Multiple Effects ThresholdsFish Targets for
Human Health

• Screening values for human consumption of edible
  fish tissue
• Tissue thresholds developed using USEPA and
  CalEPA reference doses and cancer slope factors
• Separate thresholds will be calculated assuming
  varying levels of risk
• Cancer Risk 1x10-4 - 1x10-6
• Assuming 70 kg adult with 70 yr lifetime
• Consumption rate assumptions will also be varied
• OEHHA consumption rate of 21 g/d.
• USEPA consumption rate of 17.5 g/d.
• Other consumption rates will be considered
• E.g., 6.3 g/d rate for all anglers consuming fish
  in SF Bay
• E.g., 142.4 g/d EPA rate for subsistance fishers

12
Multiple Effects ThresholdsFish Targets for
Human Health
F

• Development of four categories
• Category 1 Unlikely risk
• Below all thresholds
• Category 2 Potential risk to high-end consumers
• Above threshold using higher consumption rate
  assumption and protective allowable risk (10-6)
• Category 3 Potential risk to average consumers
• Above threshold using sport fisher consumption
  rate with intermediate allowable risk (10-5)
• Category 4 High risk to average consumers
• Sport fisher consumption rate with less
  protective allowable risk (10-4)

13
Multiple Effects ThresholdsSediment Targets for
Human Health
S

• Numeric targets - again 4 categories
• Based on field sediment concentrations at which
  fish tissue concentrations would exceed target
  concentrations
• When local data are available, targets developed
  for specific water body
• When local data are not available, general
  targets will be recommended
• These will account for uncertainty and will span
  a range of conditions
• Calculated based on concentration ratio between
  sediment and biota
• Using statistical and mechanistic models (more
  later)

14
Multiple effects thresholdsLaboratory
BioaccumulationTargets for Human Health
L

• Numeric targets - again 4 categories
• Based on concentrations observed in 28 day
  laboratory bioaccumulation tests
• Tests on sediments to be evaluated
• Important link between sediments and indirect
  effects
• Confirm whether specific sediments are likely to
  cause exposure to biota
• Also important for contaminants that do not
  bioaccumulate in finfish (e.g., PAHs)
• Our current thinking evaluate risk due to
  consumption of contaminated shellfish

15
F
Multiple Effects ThresholdsFish and Laboratory
BioaccumulationTargets for Wildlife
L

• Thresholds for bird and wildlife consumption of
  fish or shellfish
• Thresholds will be calculated and presented in
  tabular form for sensitive and endangered
  wildlife species
• Tables can be used by local agencies based on
  local species
• For PCBs and DDT, thresholds will be based on
  work of Biological Technical Assistance Group
  (BTAG)
• Low and high Toxicity Reference Values used to
  establish multiple targets
• Field fish samples and laboratory invertebrate
  samples are to be evaluated as separate lines of
  evidence
• All thresholds will be reviewed by a
  Bioaccumulation Work Group, formed specifically
  for the indirect effects task

16
Sensitive and Endangered Target Species

• Least Tern
• Clapper rail
• Brown pelican
• Western snowy plover
• Bald eagle

• Southern sea otter
• Harbor seal
• Tidewater goby
• Salmonids

17
Multiple Effects ThresholdsSediment Targets for
Wildlife
S
Same approach as with sediment targets for
humans. I.e.,

• Numeric targets
• Based on field sediment concentrations at which
  fish tissue concentrations would exceed target
  concentrations
• Calculated based on Biota Sediment Accumulation
  Factor
• Using statistical and mechanistic models (more
  later)

18
Use in AssessmentIntegration of Lines of
Evidence

• Four categories for each line of evidence
• Category 1 Unlikely risk
• Category 2 Potential risk to high-end consumers
• Category 3 Potential risk to average consumers
• Category 4 High risk to average consumers

4
3
2
1
F
A B C D E
S
L
A B C D E
4
3
2
1
4
3
2
1
19
Five Categories For SQO Evaluation
A Sediment meets SQO with high certainty
(i.e., is protective) B Sediment probably
meets SQO, but some uncertainty is present C
Sediment possibly fails SQO, but data are
inconsistent D Sediment likely fails SQO E
Sediment highly likely to fail SQO
20
Use In Assessment
21
Use In Assessment - e.g., "Clean" Site
22
Use In Assessment - e.g., "Clean" Site
23
Use In Assessment - e.g., "Clean" Site
24
Use In Assessment - e.g., "Clean" Site
25
Use In Assessment - e.g., "Clean" Site
26
Use In Assessment - e.g., "Dirty" Site
27
Use In Assessment - e.g., "Dirty" Site
28
Use In Assessment - e.g., "Dirty" Site
29
Use In Assessment - e.g., "Dirty" Site
30
Use In Assessment - e.g., "Dirty" Site
31
Statewide Assessments Will Be Conducted
F
Human Health US EPA SV x 10 90 meet
criteria
Least Tern High Effects SV (1632 ug/kg) 100
meet criteria
Least Tern Low Effects 70
meet criteria
Human Health US EPA SV 10 meet criteria
(Fish Species included Bay Goby, California
Halibut, English Sole, Longfin Sanddab, Pacific
Sanddab, Pacific Staghorn Sculpin, Shiner
Surfperch, Slender Sole, Speckled Sanddab, Starry
Flounder, White Croaker, and White Surfperch)
32
Methodological issues

• Overall approach for development of
  biota-sediment relationship
• Scale of analysis
• At what scale can data be extrapolated for
  biota-sediment relationship development?
• At what scale should movement range be
  extrapolated over?
• Target fish and laboratory bioaccumulation
  species
• BAF vs. BSAF

33
Overall Approach to DevelopBiota to Sediment
Relationship

• Empirical Models Concentrations in Organisms,
  Concentrations in Sediment, Other Factors
• Mechanistic Models Quantification of
  Bioenergetics and Physicochemical Properties and
  Concentrations.
• Data-intensive (e.g., bioenergetics, life
  history, chemical-specific properties)

34

• Empirical modeling approach
• Linear Regression Models Using SQO database and
  other data.

40
30
High toxicity Threshold
4
20
Biota Concentration
Low toxicity Threshold
2
10
0
0
2
4
6
8
10
Sediment Concentration
35
DDTs in San Francisco Bay
Macoma clams vs. sediment
100
2
R
0.6585
10
Tissue DDT (ug/kg dry)
1
1
10
100
1000
0.1
Sediment DDT (ug/kg dry) Results are from 28 day
laboratory bioaccumulation tests
36

• Bivalve concentrations compared to co-located
  sediments.
• Fish concentrations compared with sediments in a
  disk centered at each fish sampling location.
• Disk size ranged from 0.5 - 15 km (0.5 km
  increments)
• No a priori assumptions about fish home range

37
??
Total PCBs
R2 results of distance relationships of sediment
and shiner surfperch data in San Francisco Bay
Linear regression of Total PCB concentration in
sediment vs. Shiner Surfperch tissue in San
Francisco Bay (plt0.05)
38
Total DDTs
R2 results of distance relationships of sediment
and shiner surfperch data in San Francisco Bay
Linear regression of Total DDT concentration in
sediment vs. Shiner Surfperch tissue in San
Francisco Bay (plt0.05)
39
Mechanistic modeling approach

• Calculate Biota-Sediment Accumulation Factors and
  SQO using mechanistic models at local scales
• Demonstrate use of mechanistic model for multiple
  contaminants in two case studies
• Evaluate confounding factors
• Water contamination
• Home range size
• Diet
• Using Gobas model (e.g., TrophicTrace, Arnot and
  Gobas 2004)
• Validating with available empirical data

40
Basic Mechanistic Model Elements

• Uptake
• Dietary
• Gill

• Loss
• Excretion
• Egestion
• Gill Elimination
• Metabolism

Growth
Chemical properties (e.g., Kow) important
41
Data Needs

• Minimum diet and biology
• Dietary preference
• Weight, lipid content
• Preferrable
• Contaminant concentrations in sediment, water,
  inverts, fish

42
Newport Bay case study Developing conceptual
food web model
Preliminary model kindly provided by M. James
Allen, SCCWRP
43
Newport Bay case study Assembling key parameters
44
Develop BSAFs to Set Up SQOs at Appropriate Scale
45
San Diego
San Pedro
SF
Tomales
Linear (SF)
Bivalve Tissue Concentration (log x1, ug/kg, dry
wt.)
Linear (San Diego)
Linear (Tomales)
Linear (San Pedro)
2
2.5
3
3.5
4
4.5
5
5.5
6
Sediment Concentration (log x1, ug/kg, dry wt.)
Macoma nasuta tissue data indicate different
results for different water bodies. E.g., total
PAHs tissue concentrations lower at given
sediment concentration in San Francisco Bay -
suggest water body specific BSAFs
46
Identify Good Target Species
Prey For Humans and Wildlife
Limited Variation in Diet or Home Range
Sediment Linkage
47

• Macoma nasuta is a good species for Laboratory
• Bioaccumulation test
• -Recommended for bed sediment testing (EPA
  guidance)
• -Deposit feeder with high contaminant tolerance
• -Large California database available
• Species with existing data in SQO database

48
Starry Flounder
Summary of regression analysis of individual fish
species vs. summed contaminant concentrations in
sediment collected within 2 km of fish samples
significant linear relationship (plt0.05)
49
Total PCBs
Spatial patterns in total PCB concentrations and
stable isotope signatures suggest site fidelity
for shiner perch in the San Francisco Estuary
50
Map of San Francisco Bay showing locations of
sediment, Shiner surfperch and Macoma nasuta
collections used for empirical modeling of Biota
Sediment Accumulation Factors
51
BSAF vs. BAF
1. BSAF Lipid-normalized tissue conc./ organic
carbon-normalized sediment conc. 2. BAF Tissue
conc. / sediment conc.
52
Lipid and organic carbon normalization
(BSAF) does not improve relationship compared to
BAF
100
2
R
0.2541
10
BAF
Tissue DDT (ug/kg)
2
R
0.6585
BSAF
1
1
10
100
1000
0.1
Sediment DDT (ug/kg)
DDTs in San Francisco Bay
Macoma clams vs. sediment
53
Results and Recommendations

• Overall approach for development of
  biota-sediment relationship
• Empirical (statistical) and mechanistic models
• Target species
• E.g., Shiner surfperch, Macoma clams
• Scale of analysis
• Develop biota-sediment relationships that are
  water-body specific
• BAF vs. BSAF
• Collect data for BSAF (lipid, sediment OC) but
  consider using BAF only

54
Model Methods Toolkit

• Empirical BSAF and BAF models
• Linear Regression (with varying home range size)
• Calculation of average and distribution of BSAFs
  using summary statistics
• Mechanistic BSAF models
• Using established modeling approach (Frank Gobas)
• Species and spatial issues
• Macoma nasuta, shiner surfperch reasonable
• Sediment range optimization routine

55
F
Fish and LaboratoryTargets for WildlifeExample
of Calculations
L

• Example shows prey tissue targets for least terns
  
• Similar tables for other sensitive and endangered
  species
• Only use species that reside in a given water
  body
• Low and high Toxicity Reference Values from BTAG
• Target fish concentrations based on body weight
  (e.g., 40 g)
• e.g., Least Tern high effect threshold
• TRV high Weight / Consumption rate
• 1.5 mg/(kgd) 40 g / 31.1 g/d 1.928 mg/kg
  1928 ppb

Yellow values observed in CA fish
56
Contact Information

• Ben Greenfield ben_at_sfei.org
• Mike Connor mikec_at_sfei.org
• www.sfei.org
• Acknowledgements
• Steve Bay, Doris Vidal, Jim Allen, Steve
  Weisberg, SCCWRP
• Frank Gobas and Jon Arnot, Simon Frasier
  University
• Ned Black, Michael Anderson, Laurie Sullivan,
  Katie Zeeman,
• Robert Brodberg and other members of
  Bioaccumulation Work Group
• Chris Beegan, SWRCB
• Sarah Lowe, Bruce Thompson, Meg Sedlak, SFEI

57
(No Transcript)
58
Bioaccumulation Work Group