United States	Office of Water	EPA-842-D-22-002

Environmental Protection	4304T	April 2022

Agency

DRAFT AQUATIC LIFE AMBIENT WATER QUALITY

CRITERIA FOR

PERFLUOROOCTANE SULFONATE (PFOS)

April 2022

U.S. 1 ji\ ironmental Protection Auencv (M'l'ice of Water, Office of Science and
Technology. I lealtli and Ecological Criteria Division

Washington, D.C.


-------
Acknowledgements

Technical Analysis Leads:

Amanda Jarvis, Office of Water, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC

James R. Justice, Office of Water, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC

Brian Schnitker, Office of Water, Office of Science and Technology. I leal ill and Ecological
Criteria Division, Washington, DC

Mike Elias, Office of Water, Office of Science and Technology, Health and I ฆ cological Criteria
Division, Washington, DC

Reviewers:

Kathryn Gallagher and Elizabeth Behl, Office of Water. Office of Science and Technology,
Health and Ecological Criteria Division, Washington. DC

EPA Scoping Workgroup Reviewers:

Gerald Ankley, Laurence Burkhard, Russ Liickson. Matthew Ltleison, Russ Hockett, Dale Hoff,
Sarah Kadlec, Dave Mount, Carlie LaLone. and Dan Villeneuve, Office of Research and
Development, Center for Computational Toxicology and l-xposure, Great Lakes Toxicology and
Ecology Division, Duluth, MN

Anthony Williams, Office of Research and Development, Center for Computational Toxicology
and Exposure. Chemical Characterization and Exposure Division, Durham, NC (Research
Triangle Park)

Colleen Elonen. Office of Research and Development, Center for Computational Toxicology and
Exposure. Scientific Computing and Data Curation Division, Duluth, MN

Robert Burgess, Office of Research and Development, Center for Environmental Measurement
and Modeling. Atlantic Coastal Environmental Sciences Division, Narragansett, RI

Sandy Raimondo. Office of Research and Development, Center for Environmental Measurement
and Modeling, Gulcl Lcosyslem Measurement and Modeling Division, Gulf Breeze, FL

Susan Cormier, Office of Research and Development, Center for Environmental Measurement
and Modeling, Watershed and Ecosystem Characterization Division, Cincinnati, OH

Mace Barron, Office of Research and Development, Center for Environmental Solutions and
Emergency Response, Homeland Security and Materials Management Division, Gulf Breeze, FL

Cindy Roberts, Office of Research and Development, Office of Science Advisor, Policy, and
Engagement, Science Policy Division, Washington, DC

11


-------
EPA Peer Reviewers:

Jed Costanza, Office of Chemical Safety and Pollution Prevention, Office of Pollution
Prevention and Toxics, Existing Chemical Risk Assessment Division, Washington, DC

Alexis Wade, Office of General Counsel, Water Law Office, Washington, DC

Richard Henry, Office of Land and Emergency Management, Office of Superfund Remediation
and Technology Innovation, Edison, NJ

Kelly O'Neal, Office of Land and Emergency Management, Office of Superfund Remediation
and Technology Innovation, Washington, DC

Russ Hockett, Office of Research and Development, Center for Computational Toxicology and
Exposure, Great Lakes Toxicology and Ecology Division, Duluth, MN

Karen Kesler and Lars Wilcut, Office of Water, Office of Science and Technology. Standards
and Health Protection Division, Washington, DC

Rebecca Christopher and Jan Pickrel, Office of Water, Office of Wastewater Management,
Water Permits Division, Washington, DC

Rosaura Conde and Danielle Grunzke, Office of W tiler. Office of Wetlands, Oceans, and
Watersheds, Watershed Restoration, Assessment, and Protection l)i\ ision, Washington, DC

Dan Arsenault, Region I. W ater Division, Boston, MA

Brent Gaylord, Region 2. Wtiter l)i\ ision, New York, NY

Hunter Pates. Region .v W ater l)i\ision. Philadelphia, PA

Reneti 11 til I. Joel I Itinsel. I.tiuien Petter, andKathryn Snyder, Region 4, Water Division, Atlanta,
GA

Aaron Johnson and Sydney Weiss, Region 5, Water Division, Chicago, IL

Russell Nelson. Region (\ Water Division, Dallas, TX

Ann Lavaty, Region 7. W titer Division, Lenexa, KS

Tonya Fish and Maggie Pierce, Region 8, Water Division, Denver, CO

Terrence Fleming, Region 9, Water Division, San Francisco, CA

Mark Jankowski, Region 10, Lab Services and Applied Sciences Divisions, Seattle, WA

iii


-------
Action Development Process (ADP) Workgroup Members:

Tyler Lloyd, Office of Chemical Safety and Pollution Prevention, Office of Pollution Prevention
and Toxics, New Chemicals Division, Washington, DC

Thomas Glazer, Office of General Counsel, Water Law Office, Washington, DC

Stiven Foster and Kathleen Raffaele, Office of Land and Emergency Management, Office of
Program Management, Washington, DC

Kelly O'Neal, Office of Land and Emergency Management, Office of Superfund Remediation
and Technology Innovation, Washington, DC

Sharon Cooperstein, Office of Policy, Office of Regulatory Policy and Management, Policy and
Regulatory Analysis Division, Washington, DC

Cindy Roberts and Emma Lavoie, Office of Research and Development, Office of Science
Advisor, Policy, and Engagement, Science Policy Division. W ashington, DC

Kay Edly and Sydney Weiss, Region 5, Water Division, Chicago. IL

We would like to thank Russ Erickson, Dave Mount and lluss I lockett, Office of Research and
Development, Center for Compulalional Toxicology and Exposure, Great Lakes Toxicology and
Ecology Division, Duluth, MN. Ibr their technical support and contribution to this document.

We would also like to thank Sandy Raimondo and Crystal Lilavois, Office of Research and
Development. Center for Environmental Measurement and Modeling, Gulf Ecosystem
Measuring and Modeling Division, Gulf Breeze, FL, for their work assisting the Office of Water
in developing the estuarine'inarine benchmarks using Interspecies Correlation Estimates (ICE).

iv


-------
Table of Contents

Acknowledgements	ii

Table of Contents	v

List of Tables	vii

List of Figures	ix

List of Appendices	xii

Acronyms	xiii

Notices	xvi

Foreword	xvii

Executive Summary	xix

1	INTRODUCTION AM) BACKGROUND	 1

1.1	Previously Derived PFOS Toxicity Values and Thresholds			2

1.1.1	Previously Published Acute Water Protective Values for Direct Aqueous
Exposure	3

1.1.2	Previously Published Chronic Water Protecli\ e Values for Direct Aqueous
Exposure	3

1.1.3	Previously Published Chronic I'ish Tissue Criteria . .		4

1.2	Overview of Per- and Polyfluori naled Substances (PI AS) 	10

1.2.1 Physical and Chemical Properties of PFOS		13

2	PROBLEM FORMULATION	16

2.1	Overview of PFOS Sources			16

2.1.1	Manufacturing of PFOS	16

2.1.2	Sources of PFOS to Aquatic Environments	19

2.2	Environmental Fate and Transport of PFOS in the Aquatic Environment	21

2.2.1	Environmental Fate of PFOS in the Aquatic Environment	21

2.2.2	Environmental Transport of PFOS in the Aquatic Environment	22

2.3	Transformation and Degradation of PFOS Precursors in the Aquatic Environment.... 24

2.3.1	I)egradation of perfluoroalkane sulfonamido derivatives	25

2.3.2	Peilluorooctane sulfonamide-based side-chained polymers	28

2.3.3	Fluoroalkyl surfactants used in AFFFs	29

2.4	Environmental Monitoring of PFOS in Abiotic Media	30

2.4.1 PFOS Occurrence and Detection in Ambient Surface Waters	30

2.5	Bioaccumulation and Biomagnification of PFOS in Aquatic Ecosystems	35

2.5.1	PFOS Bioaccumulation in Aquatic Life	36

2.5.2	Factors Influencing PFOS Bioaccumulation and Biomagnification in Aquatic
Ecosystems	37

2.5.3	Environmental Monitoring of PFOS in Biotic Media	39

2.6	Exposure Pathways of PFOS in Aquatic Environments	44

2.7	Effects of PFOS on Biota	45

2.7.1 Mode of Action and Toxicity of PFOS	46

v


-------
2.7.2 Potential for Interactions with Other PFAS	48

2.8	Conceptual Model of PFOS in the Aquatic Environment and Effects	50

2.9	Assessment Endpoints	53

2.10	Measurement Endpoints	54

2.10.1	Overview of Toxicity Data Requirements	54

2.10.2	Measure of PFOS Exposure Concentrations	55

2.10.3	Measures of Effect	60

2.11	Analysis Plan	64

2.11.1	Derivation of Water Column Criteria	64

2.11.2	Derivation of Tissue-Based Criteria following Chronic PI- OS Exposures	64

2.11.3	Translation of Chronic Water Column Criterion to Tissue Criteria 	65

3	EFFECTS ANALYSIS FOR AQUATIC LTFE	68

3.1	Toxicity to Aquatic Life			68

3.1.1 Summary of PFOS Toxicity Studies Used to Derive the Aquatic I .ile Criteria	68

3.2	Derivation of the PFOS Aquatic Life Criteria	97

3.2.1	Derivation of Water Criteria for Direct Aqueous INposure	97

3.2.2	Derivation of Freshwater Chronic Tissue criteria for PFOS	106

3.2.3	Translation of Chronic Water Column Criterion to Tissue Criteria	106

3.3	Summary of the PFOS Aquatic l.ilc Criteria			112

4	EFFECTS CHARACTERIZATION FOR AQl A I K I .M L 	114

4.1	Comparison of Quantitative Data used to Deri \ e I 'l esliuater Criteria	114

4.1.1 Aquatic Insects	114

4.2	Additional Analyses Supporting the Freshwater Criteria	116

4.2.1	Additional Analyses Supporting the Derivation of Acute Water Column
Criterion for Freshwater	116

4.2.2	Additional Analyses Supporting the Derivation of Chronic Water Column
Criterion for Freshwater	121

4.3	Influence of Using Non-North American Resident Species on PFOS Criteria	128

4.3.1	I 'reshwater Acute Water Criterion with Native and Established Organisms
(Species Not Resident to North America removed from dataset)	128

4.3.2	Freshu ater Chronic Water Criterion with Native and Established Organisms
(Species Not Resident to North America removed from dataset)	130

4.4	Qualitatively Acceptable Water Column-Based Toxicity Data	133

4.4.1	Consideration of Qualitatively Acceptable Acute Data	134

4.4.2	Consideration of Qualitatively Acceptable Chronic Data	137

4.5	Evaluation of the Acute Insect Minimum Data Requirement through Interspecies
Correlation Estimates (ICE)	144

4.6	Acute-to-Chronic Ratios	149

4.7	Comparison of Empirical Tissue Concentrations to Translated Tissue Criteria	150

4.7.1 Comparison of Quantitative Studies and Tissue-Based Criteria	154

vi


-------
4.7.2 Comparison of Qualitative Studies and Tissue-Based Criteria	158

4.8	Effects on Aquatic Plants	161

4.9	Summary of the PFOS Aquatic Life Criterion and the Supporting Information	162

5 REFERENCES	163

List of Tables

Table Ex-1. Draft Recommended Perfluorooctane Sulfonate (PFOS) Criteria for the

Protection of Aquatic Life in Freshwaters	xxi

Table Ex-2. Draft Recommended Perfluorooctane Sulfonate (PFOS) benchmark for the

Protection of Aquatic Life in Estuarine/Marine Waters	xxi

Table 1-1. Previously Derived PFOS Toxicity Values and Thresholds	5

Table 1-2. Two Primary Categories ofPFAS	11

Table 1-3. Classification and Chemical Structure of Perfluoroalkyl Acids (PFAAs).1	12

Table 1-4. Chemical and Physical Properties of PFOS	14

Table 2-1. Summary of Assessment Endpoints and Measures of Effect Used in the Criteria

Derivation for PFOS			63

Table 2-2. Evaluation Criteria for Screening Bioaccumulalion I-'actors (BAFs) in the Public

Literature	67

Table 3-1. Summary Table of Minimum Data Requirements per the I ^85 Guidelines

Reflecting the Number of Acute and Chronic (ienus and Species Level Mean

Values in the Freshwater and Saltwater Toxicity Dalasets for PFOS	69

Table 3-2. The Five Most Sensitive Genera Used in Calculating the Acute Freshwater

Criterion (Sensitivity Rank 1-5)	72

Table 3-3. Ranked Freshwater Genus Mean Acute Values	78

Table 3-4. The Four Most Sensitive Acute Estuarine/Marine Genera	81

Table 3-5. Ranked Estuarine/Marine Water Genus Mean Acute Values	84

Table 3-6. The I 'our Most Sensitive Genera Used in Calculating the Chronic Freshwater

Criterion		85

Table 3-7 Ranked l-'ieshwater Genus Mean Chronic Values	92

Table 3-S The Three Ranked I Estuarine/Marine Genus Mean Chronic Values	95

Table 3-1) l-'ieshwater Final Acute Value and Criterion Maximum Concentration	99

Table 3-10. l-'i eshwater Final Chronic Value and Criterion Continuous Concentration	102

Table 3-11. Summary Statistics for PFOS BAFs in Fish and Invertebrates1	107

Table 3-12. Draft Recommended Perfluorooctane Sulfonate (PFOS) Criteria for the

Protection of Aquatic Life in Freshwaters	113

Table 4-1. Additional Analyses Supporting the Derivation of the Acute Water Column

Criterion for Freshwater	117

Table 4-2. GMAVs Used in Derivation of Acute Criterion and Additional Analyses

Supporting the Acute Criterion for Freshwater	119

Table 4-3. Additional Analyses Supporting the Derivation of the Chronic Water Column

Criterion for Freshwater	122

Table 4-4. GMCVs Used in Derivation of Chronic Criterion and Additional Analyses

Supporting the Chronic Criterion for Freshwater	126

vii


-------
Table 4-5. Ranked Freshwater Genus Mean Acute Values with Native and Established

Organisms, excluding Species Not Resident to North America	129

Table 4-6. Calculation of Freshwater Acute Water Column Concentration with Native and
Established Organisms (Species Not Resident to North America Removed from

Dataset)	130

Table 4-7. Ranked Freshwater Genus Mean Chronic Values with Native and Established

Organisms	132

Table 4-8. Calculation of Freshwater Chronic Water Column Concentration with Native and

Established Organisms	133

Table 4-9. All ICE models available in web-ICE v3.3 for predicted insect species based on

surrogates with measured PFOS	146

Table 4-10. ICE-estimated Insect Species Sensitivity to PFOS. Values in bold and

underlined are used for estimated insect SMAVs	148

Table 4-11. Comparison of Empirical Tissue Concentrations to Chronic Tissue Criteria and

Additional Tissue Values	152

Table L-l. Surrogate Species Measured Values for PFOS and Corresponding Number of

ICE Models for Each Surrogate	L-7

Table L-2. Comparison of ICE-predicted and measured \ allies of PFOS for species using
both scaled values (entered as mg/L) and values potentially beyond the model

domain (entered as (J,g/L) (Raimondo et al. in prep)	L-10

Table L-3. All ICE Models Available in web-lCF. \ 3 3 for Saltwater Predicted Species

Based on Surrogates with Measured PI 'OS . 	 	L-15

Table L-4. ICE-Estimated Species Sensitivity to PFOS 	L-17

Table L-5. Ranked Estuarine/Marine Genus Mean Acute Values	L-20

Table L-6. Estuarine/Marine Final Acute Value and Protective Aquatic Acute Benchmark. ...L-21

Table N-l. Global Sediment Concentration of PFOS	N-18

Table O-1. Comparison of Pairs Nominal and Measured PFOS Concentrations across

Experimental Conditions		0-9

Table 0-2. Proportion of Paired Nominal and Measured PFOS Concentrations Outside

the 20% Threshold	0-17

Table ()-3. Freshwater Nominal and Measured Concentrations for PFOS	0-29

Table ()-4 Saltwater Nominal and Measured Concentrations for PFOS	0-50

Table ()-5 I'reshwater PFOS Toxicity Studies with Systematic Discrepancies between

Nominal and Measured Concentrations that were > 20%	0-56

Table 0-6. Saltwater PFOS Toxicity Studies with Systematic Discrepancies between

Nominal and Measured Concentrations that were > 20%	0-59

Table Q-l. Characteristics of adult fish sampled for the calculation of PFOS reproductive

tissue BAFs	Q-2

Table Q-2. Summary Statistics for PFOS BAFs in Additional Fish Tissues1	Q-3

Table Q-3. PFOS Concentrations for Additional Fish Tissue.1'2	Q-4


-------
List of Figures

Figure 1-1. Chemical Structure of Linear Perfluorooctane Sulfonate (PFOS)	13

Figure 2-1. Synthesis of PFOS by electrochemical fluorination (ECF)	17

Figure 2-2. Aerobic Biodegradation of EtFOSE in Activated Sludge	28

Figure 2-3. Map Indicating Sampling Locations for Perfluorooctane Sulfonate (PFOS)

Measured in Surface Waters across the United States (U.S.)	31

Figure 2-4. Distribution of the Minimum and Maximum Concentrations (ng/L) of
Perfluorooctane Sulfonate Measured in Surface Waters for Each State or
Waterbody (excluding the Great Lakes) with Reported Data in the Publicly

Available Literature	33

Figure 2-5. Conceptual Model Diagram of Sources, Compailmental I'arlilioning, and

Trophic Transfer Pathways of Perfluorooctane Sulfonate (PI OS) in the Aquatic

Environment and its Bioaccumulation and Effects in Aquatic I .ile 	52

Figure 3-1. Freshwater Acute PFOS GMAVs Fulfilling the Acute MDRs	79

Figure 3-2. Acceptable Estuarine/Marine GMAVs			84

Figure 3-3. Ranked Freshwater Chronic PFOS Used Quantitatively to Derive the Criterion	94

Figure 3-4. Acceptable Estuarine/Marine GMCVs	97

Figure 3-5. Ranked Freshwater Acute PFOS Used Quanlilati\ ely to Derive the Criterion	100

Figure 3-6. Ranked Freshwater Chronic PFOS I sed Quantilali\ ely to Derive the Criterion.... 102

Figure L-l. Example ICE Model for Rainbow Trout (surrogate) and Atlantic Salmon

(predicted)	L-4

Figure L-2. Ranked Estuarine/Marine Acute PFOS GMAVs used for the Aquatic Life

Acute Benchmark Calculation	L-21

Figure L-3. Americamysis bahia (X-axis) and Daphnia magna (^Y-axis) regression model

used for ICE predicted values	L-25

Figure L-4. Americamysis bahia (X-axis) and Oncorhynchus mykiss (Y-axis) regression

model used for ICE predicted values	L-26

Figure L-5 . \mericamysis bahia (X-axis) and Pimephalespromelas (Y-axis) regression

model used lor ICE predicted values	L-26

Figure I .-(ฆ> / hinio rcno -embryo (X-axis) and Daphnia magna (Y-axis) regression model

used for ICE predicted values	L-27

Figure I .-7 / hinio rerio - embryo (X-axis) and Oncorhynchus mykiss (Y-axis) regression

model used for ICI- predicted values	L-27

Figure L-8. Damn rcrio - embryo (X-axis) and Pimephales promelas (Y-axis) regression

model used lor ICE predicted values	L-28

Figure L-9. Daphma magna (X-axis) and Americamysis bahia (Y-axis) regression model

used for ICE predicted values	L-28

Figure L-10. Daphnia magna (X-axis) and Lampsilis siliquoidea (Y-axis) regression model

used for ICE predicted values	L-29

Figure L-l 1. Daphnia magna (X-axis) and Lithobates catesbeianus (Y-axis) regression

model used for ICE predicted values	L-29

Figure L-l2. Daphnia magna (X-axis) and Oncorhynchus mykiss (Y-axis) regression

model used for ICE predicted values	L-30

Figure L-13. Daphnia magna (X-axis) and Pimephales promelas (Y-axis) regression

model used for ICE predicted values	L-30

IX


-------
Figure L-14. Lampsilis siliquoidea (X-axis) and Daphnia magna (Y-axis) regression

model used for ICE predicted values	L-31

Figure L-15. Lampsilis siliquoidea (X-axis) and Ligumia recta (Y-axis) regression model

used for ICE predicted values	L-31

Figure L-16. Lampsilis siliquoidea (X-axis) and Oncorhynchus mykiss (Y-axis) regression

model used for ICE predicted values	L-32

Figure L-17. Lampsilis siliquoidea (X-axis) and Pimephalespromelas (Y-axis) regression

model used for ICE predicted values	L-32

Figure L-18. Ligumia recta (X-axis) and Lampsilis siliquoidea (Y-axis) regression model

used for ICE predicted values			L-33

Figure L-19. Lithobates catesbeianus (X-axis) and Daphnia magna (Y-axis) regression

model used for ICE predicted values	L-33

Figure L-20. Lithobates catesbeianus (X-axis) and Oncorhynchus mytuss (Y-axis)

regression model used for ICE predicted values	L-34

Figure L-21. Lithobates catesbeianus (X-axis) and Pimephales promelas (Y-a\is)

regression model used for ICE predicted values	L-34

Figure L-22. Oncorhynchus mykiss (X-axis) and Americamysis hahia (Y-axis) regression

model used for ICE predicted values	L-3 5

Figure L-23. Oncorhynchus mykiss (X-axis) and Daphnia magna (Y-axis) regression

model used for ICE predicted values	L-3 5

Figure L-24. Oncorhynchus mykiss (X-axis) and / ampsilis siliquonlea (Y-axis) regression

model used for ICE predicted \ allies .		L-36

Figure L-25. Oncorhynchus mykiss (X-axis) and Lilliobaies catesbeianus (Y-axis)

regression model used for ICE predicted values 	L-36

Figure L-26. Oncorhynchus mykiss (X-axis) and Pimephales promelas (Y-axis) regression

model used for ICE predicted values	L-37

Figure L-27. Pimephales promelas (X-axis) and Americamysis bahia (Y-axis) regression

model used for ICE predicted values	L-37

Figure L-2S Pimephales promelas (X-axis) and Daphnia magna (Y-axis) regression model

used lor ICE predicted \ allies	L-38

Figure I .-29. Pimephales promelas (\-axis) and Lampsilis siliquoidea (Y-axis) regression

model used for ICF, predicted values	L-38

Figure I .-3d Pimephales promelas (X-axis) and Lithobates catesbeianus (Y-axis)

regression model used for ICE predicted values	L-39

Figure L-3 I Pimephales promelas (X-axis) and Oncorhynchus mykiss (Y-axis) regression

model used for ICI- predicted values	L-39

Figure L-32. Pimephales promelas (X-axis) and Xenopus laevis (Y-axis) regression model

used for ICE predicted values	L-40

Figure L-33. Xenopus laevis (X-axis) and Pimephales promelas (Y-axis) regression model

used for ICE predicted values	L-40

Figure 0-1. Comparison of PFOS measured and nominal concentrations for freshwater (A)

and saltwater (B) data	0-7

Figure 0-2. Comparison of PFOS measured and nominal concentrations in freshwater (top)
and saltwater (bottom) tests with acute (A and C, respectively) and chronic
durations (B and D, respectively)	0-11

x


-------
Figure 0-3. Comparison of PFOS measured and nominal concentrations in freshwater
(top) and saltwater (bottom) tests conducted in glass (A and C, respectively)

and plastic test vessels (B and D, respectively)	0-14

Figure 0-4. Assessment of measured concentrations as a percent of nominal in relation

to a 20% and 30% threshold for freshwater (A) and saltwater (B) data	0-18

Figure 0-5. Assessment of measured concentrations as a percent of nominal in relation

to a 20% and 30% threshold for saltwater acute (A) and chronic (B) tests	0-20

Figure 0-6. Assessment of measured concentrations as a percent of nominal in relation
to a 20% and 30% threshold for freshwater (top) and saltwater (bottom) tests
conducted in glass (A and C; respectively) and various plaslic (B and D;

respectively) test vessels	 	0-22

Figure 0-7. Assessment of measured concentrations as a percent of nominal in relation

to a 20%) and 30% threshold for freshwater tests with sub si rale (A) and without

substrate (B)	0-24

Figure 0-8. PFOS exposure concentrations measured at the end of the renewal period or
static test (old solutions) vs. concentrations measured in the same solution
immediately after it was introduced into the exposure chamber (new solutions)... 0-25

XI


-------
List of Appendices

Appendix A Acceptable Freshwater Acute PFOS Toxicity Studies	A-l

Appendix B Acceptable Estuarine/Marine Acute PFOS Toxicity Studies	B-l

Appendix C Acceptable Freshwater Chronic PFOS Toxicity Studies	C-l

Appendix D Acceptable Estuarine/Marine Chronic PFOS Toxicity Studies	D-l

Appendix E Acceptable Freshwater Plant PFOS Toxicity Studies	E-l

Appendix F Acceptable Estuarine/Marine Plant PFOS Toxicity Studies	F-l

Appendix G Other Freshwater PFOS Toxicity Studies	 	G-l

Appendix H Other Estuarine/Marine PFOS Toxicity Studies	H-l

Appendix I Acute to Chronic Ratios	I-1

Appendix J Unused PFOS Toxicity Studies			J-l

Appendix K EPA Methodology for Fitting Concentration-Response Data and Calculating

Effect Concentrations	K-l

Appendix L Derivation of Acute Protective PFOS Benchmarks for Estuarine/Marine

Waters through a New Approach Method (NAM): WeblCE	L-l

Appendix M Environmental Fate of PFOS in the Aquatic Environment	M-l

Appendix N Occurrence of PFOS in AMolic Media			N-l

Appendix O Meta-Analysis of Nominal Test Conccnlialions Compared to Corresponding

Measured Test Concentrations	0-1

Appendix P Bioaccumulation Factors (BAFs) Used lo Calculate PFOS Tissue Values	P-l

Appendix Q Translation of Chronic Water Column Criterion into Other Fish Tissue Types

(liver, blood, reproductive tissues)	Q-l

Appendix R Example Data F.vahiation Records (DERs)	R-l

xii


-------
Acronyms

6:2 Cl-PFESA

6:2 chlorinated polyfluorinated ether sulfonate

ACR

Acute-to-Chronic Ratio

AFFF

Aqueous film-forming foams

AIC

Akaike information criteria

AMV

Acute Maximum Value

ASW

artificial sea water

AWQC

National Recommended Ambient Water Quality Criteria

BAF

Bioaccumulation factor

C8-PFPA

Perfluorooctyl phosphonic acid

C8/C8-PFPIA

Bis(perfluorooctyl) phosphinic acid

CAS/CASRN

Chemical Abstracts Service registry numbers

CC

Chronic Criterion

CCC

Criterion Continuous Concentration

C-F

carbon-fluorine

CMC

Criterion Maximum Concentration

C-R

concentration response

C-S

carbon-sulfur

CWA

Clean Water Act

DER

Data Evaluation Record

DMSO

dimethyl sulfoxide

dpf

days post fertilization

drc

dose-response curve

dw

dry weight

ECF

Electrochemical fluorination

ECOTOX

ECOTOXicology database

ELS

Early life-stage

EPA

I S F.n\ ironmental Protection Agency

EtFASAAs

A-ethyl perfluoroalkane sulfonamidoacetic acids

EtFASAs

iV-elhyl periluoroalkane sulfonamides

EtFOSAA

TV-ethyl peril uorooctane sulfonamidoacetic acid

EtFOSE

TV-ethyl peril uorooctane sulfonamidoethanol

FACR

Final Acute to Chronic Ratio

FASAAs

IVrfluoroalkyl sulfonamidoacetic acids

FASAs

Periluoroalkane sulfonamids

FASEs

peril uoroalkyl sulfonamidoethanols

FAV

Final Acute Value

FCV

Final Chronic Value

FFTG

Canadian Federal Fish Tissue Guideline

FIFRA

Federal Insecticide, Fungicide, and Rodenticide Act

FOSA

Perfluorooctane sulfonamide

FWQG

Federal Water Quality Guideline

GLI

U.S. EPA Great Lakes Initiative

GMAV

Genus Mean Acute Value

GMCV

Genus Mean Chronic Value

Xlll


-------
HC1

1% Hazardous Concentration

GSD

Genus sensitivity distribution

hpf

hours post fertilization

ICE

Interspecies Correlation Estimation

Kow

n-octanol-water partition co-efficient

LOD

limit of detection

LOEC

Lowest Observed Effect Concentration

LOQ

limit of quantification

MATC

Maximum Acceptable Toxicant Concentration

MC

Maximum Criterion

MDL

Method detection limit or Minimum Detection Limit

MDRs

Minimum data requirements

NAMS

New Approach Methods

NCCA

National Coastal Condition Assessment

NOEC

No Observed Effect Concentration

NPDES

National Pollutant Discharge Elimination System

NRSA

National Rivers and Streams Assessment

OCSPP

Office of Chemical Safety and Pollution Prevention

OECD

Organization for Economic Co-operation and Development

ORD

Office of Research and Development

OSF

Octane sulfonyl fluoride

OW

Office of Water

PFAAs

Perfluoroalkyl acids

PFAS

Polyfluorinated substances

PFCA

Perfluoroalkyl carboxylic acids or Perlluoroalkvl earboxylates

PFDA

Perfluorodecanoate or Perfluorodecanoic acid

PFdiCAs

Perfluoroalkyl dicarboxylic acids

PFdiSAs

Perfluoroalkane disulfonic acids

PFECAs

Perfluoroalkylether carboxylic acids

PFESAs

Perlluoroalkylelher sulfonic acids

PFDoA

Perlluorododecanoate or Perfluorododecanoic acid

PFOA

Perfluorooctanoie acid or Perfluorooctanoate

PFOS

Perfluorooetane sulfonate or Perfluorooctane sulfonate acid

PFOSI

Perfluorooetane sulfinic acid

PFOS-K

PFOS potassium salt

PFOS-Li

PFOS lithium salt

PFPAs

Perfluoroalkyl phosphonic acids

PFPIAs

Perfluoroalkyl phosphinic acids

PFSAs

Perfluoroalkane (or -alkyl) sulfonic acids or Perfluorokane sulfonates

PFSIAs

FASA A'-glucuronides or Perfluoroalkyl sulfinic acids

pKa

Acid dissociation constant

POSF

Perfluorooctanesulfonyl fluoride

PPAR-a

Nuclear peroxisome proliferator activated receptor-alpha

ppt

parts per thousand

SMACR

Species mean acute-to-chronic ratios

SMAV

Species mean acute value

xiv


-------
SMCV

Species mean chronic value

SNUR

Significant New Use Rules

SOP

Standard Operating Procedure

SSD

Species Sensitivity Distribution

TMDLs

Total Maximum Daily Loads

TSCA

Toxic Substances Control Act

U.S.

United States

UCMR

Unregulated Contaminant Monitoring Rule

weblCE

Web-based Interspecies Correlation Estimation

WQS

Water Quality Standards

WW

wet weight

WWTPs

Wastewater treatment plants

XV


-------
Notices

This draft document provides information to states and tribes authorized to establish
water quality standards under the Clean Water Act, to protect aquatic life from toxic effects of
perfluorooctane sulfonic acid (PFOS). Under the CWA, states and tribes are to establish water
quality criteria to protect designated uses. State and tribal decision makers retain the discretion to
adopt approaches that are scientifically defensible that differ from these criteria to reflect site-
specific conditions. While this document contains the Environmental Protection Agency's (EPA)
draft scientific recommendations regarding ambient concentrations of PFOS thi.il protect aquatic
life, the draft PFOS Criteria Document does not substitute for the Clean Water Act or the EPA's
regulations; nor is it a regulation itself. Thus, the document when final would not impose legally
binding requirements on the EPA, states, irihcs. or the regulated community, and might not apply
to a particular situation based upon the circumstances The l-IW intends to finalize this document
in the future. This draft document lias been approved for puhIication by the Office of Science
and Technology. Office of Water. I S Environmental Protection Agency.

Mention of trade names or commercial products does not constitute endorsement or
recommendation lor use This document can he downloaded from:
https	>v/wqc/ :ic-life-crit.eria-perfluorooctane-sulfonate-pfos.

xvi


-------
Foreword

The Clean Water Act (CWA) Section 304(a)(1) (P.L. 95-217) directs the Administrator of
the EPA to publish water quality criteria that accurately reflect the latest scientific knowledge on
the kind and extent of all identifiable effects on health and welfare that might be expected from
the presence of pollutants in any body of water, including groundwater. This document is a draft
ambient water quality criteria (AWQC) document for the protection of aquatic life based upon
consideration of all available information relating to effects of peiiluorooctane sulfonate (PFOS)
on aquatic organisms.

The term Water Quality Criteria is used in two sections of the CWA, Section 3< '4(a)(1)
and Section 303(c)(2). The term has different meanings in each section. Under CWA section
304, the term represents a non-regulatory, scientific assessment of ecological and human health
effects. Criteria presented in this draft document are such a scientific assessment of ecological
effects. Under CWA section 3<)3. when water quality criteria associated with specific surface
water uses are adopted In a slate or authorized tribe and approved by EPA as water quality
standards, they become the CWA water quality standards applicable in ambient waters within
that stale or authorized tribe \Y'tiler quality criteria adopted in state/tribal water quality standards
could ha\ e the same numerical values as recommended criteria developed under CWA section
304. However, in some situations, states/tribes might want to adjust water quality criteria
developed under CWA section 304 to reflect local water chemistry or ecological conditions.
Alternatively, states and authorized tribes may develop numeric criteria based on other
scientifically defensible methods that are protective of designated uses. Guidelines to assist the
states and authorized tribes in modifying the criteria presented in this draft document are
contained in the Water Quality Standards Handbook (U.S. EPA 2014).

xvii


-------
This document presents draft recommendations only. It does not establish or affect legal
rights or obligations. It does not establish a binding requirement and cannot be finally
determinative of the issues addressed. The EPA will make decisions in any particular situation
by applying the CWA and the EPA regulations on the basis of specific facts presented and
scientific information then available.

Deborah (i \auk-

Director

Office of Science and Technology

xviii


-------
Executive Summary

U.S. Environmental Protection Agency (EPA) developed the draft recommended
perfluorooctane sulfonate (PFOS) aquatic life ambient water quality criteria in accordance with
the provisions of section 304(a) of the Clean Water Act (CWA). This document provides EPA's
basis for and derivation of the draft national PFOS ambient water quality criteria
recommendations for freshwater environments to protect aquatic lilc I .IW has drafted the PFOS
aquatic life criteria to be consistent with methods described in EPA's "(iiin/c/mes for Deriving
Numerical National Water Quality Criteria for the Protection of Aquatic (hgamsms and Their
Uses" (U.S. EPA 1985).

PFOS is an organic, human-made perfluorinated compound, consisting of an eight-carbon
backbone and a sulfonate functional group PI-OS (and other related chemicals that are
perfluoroalkane sulfonic acids) is used in a \ aricly of industrial and commercial products,
including surface treatments of soil, surface treatments of textiles, paper, and metals, and in
specialized applications such as in firefighting foams. This document provides a critical review
of all aquatic toxicity data identified in NWs literature search for PFOS, including the anionic
form (CAS \o 452lM-l)<)-(•>). the acid form (('AS No. 1763-23-1), potassium salt (CAS No.
2795-3l>-3). an ammonium salt (CAS \o 50773-42-3), sodium salt (CAS No. 4021-47-0), and a
lithium salt (CAS \o 29457-72-5). It also quantifies the toxicity of PFOS to aquatic life and
provides draft criteria to protect aquatic life in freshwater from the acute and chronic toxic
effects of PFOS.

The draft Aquatic Life Ambient Water Quality Criteria for PFOS includes water column-
based acute and a water column-based chronic criteria, as well as chronic tissue-based criteria for
freshwaters. Quantitatively-acceptable estuarine/marine toxicity data only fulfilled five of the

xix


-------
eight minimum data requirements (MDRs) for deriving acute estuarine/marine criteria and three
of the eight MDRs for deriving chronic estuarine/marine criteria per the 1985 Guidelines. EPA
did, however, include an acute aquatic life benchmark for estuarine/marine environments in
Appendix L, using available estuarine/marine species toxicity data and the New Approach
Methods (NAMS) application of ORD's peer-reviewed weblCE tool (Raimondo et al. 2010).
Both the freshwater criteria and estuarine/marine benchmarks arc draft recommendations for
states/authorized tribes to consider as protective values in their state/trihal water quality
protection programs. However, the acute estuarine/marine benchmark is less certain than the
freshwater criteria since it was based on both empirical and estimated acute toxicity data
(Appendix L).

The draft freshwater acute water column-bused criterion magnitude is 3.0 mg/L and the
draft chronic water column-based criterion magnitude is <) oi)S4 mg/L. The draft chronic
freshwater criteria also contains tissue-based criteria with magnitudes of 6.75 mg/kg wet weight
(ww) for fish whole-body. 2 ^ I mg kg \\w for fish muscle tissue, and 0.937 mg/kg ww for
invertebrate whole-body tissue All criteria arc intended to be equally protective against adverse
PFOS effects and arc intended to be independently applicable. The three tissue criteria
magnitudes (for lish and in\ crtebratc tissues) are translations of the chronic water column
criterion for freshwater using bioaccumulation factors (BAFs) derived from a robust national
dataset of BAFs (liurkhard 2021). The assessment of the available data for fish, invertebrates,
amphibians, and plants indicates these criteria are expected to be protective of the freshwater
aquatic community.

xx


-------
Table Ex-1. Draft Recommended Perfluorooctane Sulfonate (PFOS) Criteria for the
Protection of Aquatic Life in Freshwaters.				

Type/Media

Acute Water
Col 1111111
(CMC)1-4

Chronic
Water
Coliim 11

(cccy5

Chronic
Invertebrate
Whole-
Hod v1-2

Chronic

lisli
Wliole-
liodv12

Chronic

Fish
Muscle12

Magnitude

3.0 mg/L

0.0084 mg/L

0.937
mg/kg ww

6.75

mg/kg ww

2.91

mg/kg ww

Duration

1-hour average

4-day average

Instantaneous3

Frequency

Not to be
exceeded more
than once in
three years on
average

Not to be
exceeded more
than once in
three years on
average

Not to be exceeded more than once in
ten years 011 average

1	All five of these water column and tissue criteria are intended to be independently applicable and 110 one criterion takes
primacy. All of the above recommended criteria (acute and chronic water column and tissue criteria) are intended to be
protective of aquatic life. These criteria are applicable throughout the year.

2	Tissue criteria derived from the chronic water column concentration (CCC) with the use of bioaccumulation factors and are
expressed as wet weight (ww) concentrations.

3	Tissue data provide instantaneous point measurements that reflect integrative accumulation of PFOS over time and space in
aquatic life population(s) at a given site.

4	Criterion Maximum Concentration; applicable throughout the water column.

5	Criterion Continuous Concentration; applicable throughout the water column.

Table l.\-2. Draft Recommended I'orfluorooclano Sulfonate (I'l'OS) Benchmark for the
Protection of Aquatic l.il'e in Lsluarinc/Marinc W aters.	

Type/Media

Acute W ater Column Benchmark

Magnitude

i) 55 mu 1.

Duration

1 hour 011 a\ eiaue

Frequency

Not lo he exceeded more than once in three years on average

xxi


-------
1 INTRODUCTION AND BACKGROUND

National Recommended Ambient Water Quality Criteria (AWQC) are established by the
EPA under the CWA. Section 304(a)(1) of the CWA states that aquatic life criteria serve as
recommendations to states and tribes and define ambient water concentrations that will protect
against unacceptable adverse ecological effects to aquatic life resulting from exposure to
pollutants found in water. Once EPA publishes final CWA section )4(a) recommended water
quality criteria, states and authorized tribes may consider these criteria and may adopt them or
other scientifically defensible criteria into their w ater quality standards (WQS) lo protect the
designated uses of water bodies. States and authorized tribes may also modify these criteria to
reflect site-specific conditions or use other scientifically defensible methods to develop criteria
before adopting these into standards. Stales and authorized tribes are required to submit new and
revised WQS to EPA for review and appro\ al or clisappix>\ al When approved by EPA, the
state's/tribe's WQS become the applicable WQS for CWA purposes. Such purposes include
derivation of water quality-based effluent limitations in permits issued under the CWA section
402 National Pollutant Discharge Nimination System (NPDES) permit program and
identification of impaired waters and establishment of Total Maximum Daily Loads (TMDLs)
under CWA section 303(d) for PFOS, EPA would recommend the adoption of all criteria,
including the three chronic tissue criteria, to ensure the protection of aquatic life through all
exposure pathways, including direct aqueous exposure and bioaccumulation. The draft
estuarine/marine benchmarks are provided in Appendix L as additional protective values that
states and tribes may consider in their water quality protection programs.

This assessment provides a critical review of all aquatic toxicity data identified in EPA's
literature search for PFOS, including the anionic form (CAS No. 45298-90-6), the acid form

1


-------
(CAS No. 1763-23-1), a potassium salt (CAS No. 2795-39-3), an ammonium salt (CAS No.
56773-42-3), a sodium salt (CAS No. 4021-47-0), and a lithium salt (CAS No. 29457-72-5). It
quantifies the toxicity of PFOS to aquatic life and provides draft criteria to protect aquatic life in
freshwater from the acute and chronic toxic effects of PFOS.

EPA derived the draft recommended criteria using the best available data to reflect the
latest scientific knowledge on the toxicological effects of PFOS io aquatic life. EPA developed
the criteria following the general approach outlined in the EPA's "Gun/dines for Deriving
Numerical Water Quality Criteria for the Protection of Aquatic Organisms am/ I heir Uses"
(U.S.EPA 1985). The draft PFOS freshwater criteria are expected to be protecli\ e of most
aquatic organisms in the community (i.e., approximately 95 percent of tested aquatic organisms
representing the aquatic community) and are clcri \ ed to be pi olccli\ c of aquatic life designated
uses established by states and tribes for freslm tilers The draft esluarine/marine benchmarks are
also intended to be |">rolecli\ e of aquatic lile designated uses, but as they are based on fewer
empirical PFOS dtila hti\ e greater inherent uncertainty. The draft criteria presented herein are
EPA's best esti unite of the maximum concentrations of PFOS, with associated frequency and
duration specifications, that would protect sensitive aquatic life from unacceptable acute and
chronic effects

1.1 Previously Derived PFOS Toxicity Values and Thresholds

Within the IS. no states or tribes have CWA Section 303(c) approved water quality

standards for the protection of aquatic life from the exposure to PFOS. And to date, no state or
tribe has submitted a Water Quality Standard with criteria for PFOS to EPA for approval.
However, two states (Michigan and Minnesota) have acute and chronic protective values that
were developed to be numerical translations of CWA Section 303(c) narrative water quality

2


-------
criteria. And other states have published draft/interim acute and chronic ecological screening
level values/benchmarks for the protection of aquatic life. As such, previously published PFOS
acute and chronic criteria, benchmarks, and thresholds developed by states and international
regulatory authorities were identified, that included values for both freshwater and marine
systems, and are summarized below.

1.1.1	Previously Published Acute Water Protective Values for Pi reel Aqueous Exposure
Previously published freshwater acute values were available lor lour stales (Florida,

Michigan, Minnesota, and Texas) and one geographic region (Europe). These puhlicly available

values for other jurisdictions ranged from 0.021 mu I. in Texas (TCEQ 2021; Giesy et al. 2010)

to 0.78 mg/L in Michigan (EGLE 2010). There were two pre\ iously derived estuarine/marine

acute values, with a benchmark/criterion of n t)t)72 mg/L in l-urope (RIVM 2010) and 0.21 mg/L

in Florida (Stuchal and Roberts 2019; Table 1 -1).

1.1.2	Previously Published Chronic Water Protective Values for Direct Aqueous Exposure
Previously published freshwater chronic values were available for five states (California,

Florida, Michigan, Minnesota, and Texas) and three countries or geographic regions

(Australia New Zealand. Canada, and l-urope). These publicly available values ranged from

0.0005(-> mu L in California (San I'raneiseo Bay RWQCB 2020; SERDP 2019; 99% species

protection) to n 14 mg/L in Michigan (EGLE 2010), 0.000023 mg/L in Europe (RIVM 2010),

0.00013 mg/L in Australia New Zealand (CRC Care 2017; EPAV 2017) 95% species protection

level), and 0.00680 mu I. in Canada (ECCC 2018) (Table 1-1).Previously published freshwater

chronic values were available for five states (California, Florida, Michigan, Minnesota, and

Texas) and three countries or geographic regions (Australia/New Zealand, Canada, and Europe).

These publicly available values ranged from 0.00056 mg/L in California (San Francisco Bay

RWQCB 2020; SERDP 2019; 99% species protection) to 0.14 mg/L in Michigan (EGLE 2010),

3


-------
0.000023 mg/L in Europe (RIVM 2010), 0.00013 mg/L in Australia/New Zealand (CRC CARE
2017; EPAV 2016; HEPA 2020; 95% species protection level), and 0.00680 mg/L in Canada
(ECCC 2018) (Table 1-1).

Previously published estuarine/marine chronic values were available for three states
(California, Florida, and Texas) and two geographic regions (Australia/New Zealand and
Europe). These publicly available values for other jurisdictions mimed from 0.000294 mg/L for
Texas (TCEQ 2021;CRC Care 2017) to 0.013 mg/L in Florida (Stuchal and Roberts 2019) and
were 0.0000046 mg/L in Europe (RIVM 2010) and 0.00013 mg/L in Australia New Zealand
(CRC Care 2017; EPAV 2017); 95% species protection)

1.1,3 Previously Published Chronic Fish Tissue Criteria

Currently there was a single previously derived fish tissue \ alue for other jurisdictions.

This value was a Canadian Federal Fish Tissue (inideline (I'FTG) of 9.4 mg/kg whole-body wet

weight (ww) (ECCC 2<>l S) This \ alue was derived by multiplying Canada's Federal Water

Quality Guideline of (vS uu I. In a Ii.\Iฆ" of 1.378 L/kg.

4


-------
Table 1-1. Previously Derived PFOS Toxicity Values and Thresholds.

State /
Country of
Applicability

Aquatic l.ife Protective Value
(mป/l. unless otherwise iiulicnled)

Criteria or Benchmark and Calculation Approach

Source

Freshwater Acute

Some
European
Countries

0.036

Maximum Acceptable Concentration calculated using the lowest acute
(LC50) value of 3.6 mg/L for mysid (. \mcncamysis bahia) by
assessment factor ol" 100. Dataset includes freshwater and marine aquatic
species, combined

(RIVM
2010)

Texas

0.021

Acute surface water benchmark calculated using U.S. EPA Great Lakes
Initiative (GT.I; U.S I -PA 1 iw5) Tier I Methodology as reported in Giesy
et al. (2o|o) This is an acute surface water benchmark and does not
represent a CAVA Section 303(c) approved water quality standard for
PFOS.

TCEQ
(2021);
Giesy et al.
(2010)

Minnesota

0.085

Final Acute Value (1 AY) calculated as the acute curve-fitted and
extrapolated 10-d EC5<) lor midge (Chironomus tentans) of 170 |ig/L.
And Maximum Criterion (MC) = FAV ^ 2. This protective value is a
translation of narrati ve water quality criteria and does not represent a
CW A Section 303(c) approved water quality standard for PFOS.

(STS/MPCA
2007)

Florida

i) 5.i

Secondary Acute Value (SAV) calculated using U.S. EPA Great Lakes
Initiative (GL1; U.S. EPA 1995) Tier II Methodology. FAV calculated as
the lowest GMAV (unspecified) divided by a safety factor of 6.1. This
\ alnc was released in a White Paper sponsored by Florida Department of
Environmental Protection and is considered a draft eco-based surface
water screening level, it is not a CWA Section 303(c) approved water
quality standard.

Stuchal and

Roberts

(2019)

5


-------
St sill' /
Country of
Applicability

Aquatic l.ife Protective Value
unless otherwise iiuliciilecl)

Criteria or licnchmark and Calculation Approach

Source

Michigan

0.78

Final Acute Value (FAV) calculated from the lowest LC50 by a safety
factor of 6.1 (following US EPA Great Lakes Initiative [GLI; U.S. EPA
1995]). And the Acute Maximum Value (AMV) was the FAV ^ 2. This
protective value is a translation of narrative water quality criteria and
does not represent a CWA Section 303(c) approved water quality
standard for PFOS.

(EGLE 2010)

Marine Acute

Some
European
Countries

0.0072

Maximum Acceptable Concentration calculated using the lowest acute
value (T.C5<~>) of 3.6 mg/L for a mysid (Americamysis bahia) by
assessment factor of 500. Dataset includes freshwater and marine aquatic
species, combined

(RIVM
2010)

Florida

0.21

Secondary Acute Value (SAV) calculated using U.S. EPA Great Lakes
Initiative (Gl.l. 1 S I-PA 11>95) Tier IIMethodology. FAV calculated as
the lowest GMAV (unspecified) divided by a safety factor of 21.9. This
\ nlue was released in a White Paper sponsored by Florida Department of
1 -n\ ironmental Protection and is considered a draft eco-based surface
wiitei' screening level, it is not a CWA Section 303(c) approved water
quality standard.

Stuchal and

Roberts

(2019)

Freshwater Chronic

Some
European
Countries

0 in 10023

Maximum Permissible Concentration calculated using the lowest value
(LOEC) of 0.0023 mg/L for Chironomus tentans (McDonald et al. 2004)
by an assessment factor (100). Dataset includes freshwater and marine
aquatic species, combined.

(RIVM
2010)

6


-------
St sill' /
Country of
Applicsihility

Aqusitic l.ili' Protect ivc Vsiluc
unless otherwise iiuliciilecl)

Crili'i'isi or lii'iichmsirk sind Csilculsition Approsich

Sou ITl'

Canada

0.00680

Federal Water Quality Guideline (1 WQG) calculated as the fifth
percentile \ nine from a Species Sensili\ ily Distribution (SSD) consisting
of 20 species-specific \ allies repiesenlinu fish (5), amphibians (2),
invertebrates (5). and plants and aluae (S)

(ECCC
2018)

Australia,
New Zealand

0.00000023
(99% species protection - high
conservation value systems)

Guidelines calculated from Species Sensitivity Distribution (SSD)
consisting of 18 species-specific values for fish, amphibians, insects,
crustaceans, and algae Ibllowing the guidance of Warne et al. (2017) and
Batley el al (2<)|4)

(CRCCare
2017);

(EPAV
2017); HEP A
(2020)

0.00013
(95% species protection - slightly to
moderately disturbed systems)

0.002

(90% species protection - highly
disturbed systems)

0.031

(80% species protection - highly
disturbed systems)

California

0.00056

(99% species prukvlion)

1IC1 calculated from an acute and chronic NOEC-based SSD as reported
in SI-KDP Project ER18-1614 (SERDP 2019). Acute NOEC values were
coin erted to chronic values using mean acute-to-chronic ratios derived
from (riesy et al. (2010). This value represents an "Interim Final
1 -n\ ironmental Screening Level" and does not represent a CWA Section
303(c) approved water quality Standard for PFOS.

San

Francisco
Bay

RWQCB
(2020);
SERDP
(2019)

Texas

0 005 |

Acute surface water benchmark calculated using U.S. EPA Great Lakes
Initiative (GLI; U.S. EPA 1995) Tier I Methodology as reported in Giesy
et al. (2010). This is a chronic surface water benchmark and does not
represent a CWA Section 303(c) approved water quality standard for
PFOS.

TCEQ
(2021);
Giesy et al.
(2010)

7


-------
St sill' /
Country of
Applicability

Aqusitic l.ili' Protective Vsilue
unless otherwise inilicnlecl)

Crilcrisi or lienchmsirk sind Csilculsition Approsich

Soil ITl'

Minnesota

0.019

Chronic Criterion (CC) calculated as the FAV (170 |ig/L) ^ FACR (9.12)
per Minnesota Rules Chapter 705o Tw o species-specific ACRs and a
default ACR were used to calculate the I-.VCR. This protective value is a
translation of narrative water quality criteria and does not represent a
CWA Section 303(c) approved water quality standard for PFOS.

STS/MPCA
(2007)

Florida

0.037

Secondary Chronic Value (SCV) calculated using U.S. EPA Great Lakes
Initiative (GL1; I S I-PA 1995) Tier II Methodology with acute-to-
chronic (ACR) of 14 5 SCV SAV (530 |ig/L) ACR (14.5) = 37
|ig/L. This value was released in a White Paper sponsored by Florida
Department of Environmental Protection and is considered a draft eco-
based surface water screening level, it is not a CWA Section 303(c)
approved water quality standard.

Stuchal and

Roberts

(2019)

Michigan

0.14

Final Chronic Value (FCV) calculated as the FAV (1,557 |ig/L) FACR
(1 1.35) per U.S. EPA Great Lakes Initiative (GLI; U.S. EPA 1995). Two
species-specific ACRs and a default ACR were used to calculate the
FACR. This protective value is a translation of narrative water quality
criteria and does not represent a CWA Section 303(c) approved water
quality standard for PFOS.

(EGLE 2010)

Marine Chronic

Australia,
New Zealand

0.00000023
(99% species protect ion - liiuh
conservation value s\ stems)

Guidelines calculated from SSD following the guidance of Warne et al.
(2<) 17) and Batley et al. (2014) and consisting of nine species-specific
values representing fish (2), echinoderms (2), crustaceans (2), mollusc
(1), and algae (2).

Note: Per HEPA (2020) freshwater values are to be used on an interim
basis until final marine guideline values can be set using the nationally
agreed process under the Australian and New Zealand Guidelines for
Fresh and Marine Water Quality

(CRCCare
2017);

(EPAV
2017); HEPA
(2020)

n nnoi3

(95% species protection - slightly to
moderaleh disturbed svstems)

0.002

(90% species protection - highly
disturbed svslemsi

0.031

(80% species protection - highly
disturbed systems)

8


-------
St sill' /
Country of
Applicsihility

Aqusitic l.ili' Protect ivc Ysilue
unless otherwise inilicnlecl)

Crili'risi or licnchmsirk sind Csilculsition Approsich

Sou ITl'

Some
European
Countries

0.0000046

Maximum Permissible Concentration calculated using the lowest value
(LOEC) of 0.0023 mg/L for ('/iironoimis teutons divided by an
assessment factor (500). Datasei includes freshwater and marine aquatic
species, combined.

(RIVM
2010)

Texas

0.000294

Default guidelines calculated from SSD following the guidance of Warne
et al. (2017) and Bailey el ill. (2014) and consisting of nine of 16 species-
specific values as reported in CRC CARE (2017). This is a chronic
surface water benchmark and does not represent a CWA Section 303(c)
approved water quality standard for PFOS.

TCEQ
(2021);
(CRCCare
2017)

California

0.0026
(99% species proteclinni

HC1 calculated from an acute and chronic NOEC-based SSD as reported
in SERDP Project ER18-1614 (SERDP 2019). Acute NOEC values were
converted to chronic values using mean acute-to-chronic ratios derived
from (riesy et al. (2010). This value represents an "Interim Final
Ln\ ironmental Screening Level" and does not represent a CWA Section
3<)3(c) approved water quality Standard for PFOS.

San

Francisco
Bay

RWQCB
(2020);
SERDP
(2019)

Florida

i) on

Secondary Chronic Value (SCV) calculated using U.S. EPA Great Lakes
Initiative (GL1; U.S. EPA 1995) Tier II Methodology with acute-to-
cluonic (ACR) of 15.6. SCV = SAV (210 |ig/L) - ACR (15.6) = 13
iiu 1. This value was released in a White Paper sponsored by Florida
Department of Environmental Protection and is considered a draft eco-
based surface water screening level, it is not a CWA Section 303(c)
approved water quality standard.

Stuchal and

Roberts

(2019)

Fish Tissue

Canada

9.4 mg/kg whole body ww fish
tissue

Federal Fish Tissue Guideline (FFTG) where FFTG of 9.4 mg/kg ww =
(FWQG of 6.8 (J,g/L) * (BAFgeomean of 1378 L/kg)

(ECCC
2018)

9


-------
1.2 Overview of Per- and Polyfluorinated Substances (PFAS)

PFOS, and its salts, belong to the per- and polyfluorinated substances (PFAS) group of

chemicals. PFAS are synthetic, organic compounds that consist of a carbon backbone and a
functional group, such as sulfonate or carboxylic acid (CnF2n+i-R). EPA's Office of Pollution
Prevention and Toxics (OPPT) defines a PFAS as: any chemical substance or mixture that
structurally contains the unit R-(CF2)-C(F)(R')R". Both the CF2 and CF moieties are saturated
carbons. And none of the R groups (R, R' or R") can be hydrogen (U.S l-IW 2021b).

Specifically, PFOS consists of an eight-carbon backbone and a sulfonate functional group
(formula is C8F17 SO3; CAS No. 45298-90-6 for anionic form; Buck et al. 2011). The carbon
chain can be fully fluorinated (perfluorinated) or partially fluorinated (polyfluorinated), and
therefore, these chemicals contain the periluoroulkyl moiety (Cnl':,, 1-) The carbon-fluorine (C-
F) bond is strong and stable due to the strong elcclioncuati\ ity and small atomic size of fluorine.
The chemical and thermal suihilily offered by the perfluoroalkv1 moiety, in addition to its
hydrophobic and lipophohic characlciislics. make PFAS water and oil repellent, thermally
stabile, and ha\ e sin lacta 111 properties Due to these properties, PFAS have been used in a wide
range of industrial and consumer products since the 1940s and 1950s, including wetting agents,
lubricants, corrosion inhibitors, firefigluing foams, and stain-resistant treatments to leather,
paper, and clothing The PI AS subgroup of PFOS derivatives have been used in a broad range of
consumer and industrial products since the 1950s, including surface treatments for soil and stain
resistance of textiles, paper, metals, pesticides, and is used in applications such as in firefighting
foams (Ahrens 2011; Ahrens and Bundschuh 2014; Buck et al. 2011; Lindstrom et al. 2011).

There are many families of PFAS and each contains many individual homologues and

isomers (Buck et al. 2011). These PFAS families can be divided into two primary categories:

nonpolymers and polymers. The nonpolymer PFAS include perfluoroalkyl and polyfluoroalkyl

10


-------
substances. Polymer PFAS include fluoropolymers, perfluoropolyethers, and side-chain
fluorinated polymers (Table 1-2).

Table 1-2. Two Primary Categories of PFAS.

VIodified from Buck et al. (2011).

PI'AS Non-polymers:

Structural Klemenls:

r.xamplc PI'AS lamilies:

Perfluoroalkyl Substances

Compounds in which all carbon-
hydrogen bonds, except those on
the functional group, are replaced
with carbon-fluorine bonds

Perfluoroalkyl acids,
perfluoroalkane sulfonamides,
perfluoroalkane sulfonyl
fluorides

Polyfluoroalkyl Substances

Compounds in which all carbon-
hydrogen bonds on at least one
carbon (but not all) are replaced
with carbon-fluorine bonds

IVrll uoroalkane sulfonamido
cleri\ ali\ es. scmilluorinaled n-
alkanes and alkenes

PKAS Polvmers:

Structural Klements:

r.xamplc PI'AS Kamilies:

Fluoropolymers

Carbon only polymer backbone
with fluorines directly attached

Polyetrafluoroethylene,
polyvinylidene fluoride

Perfluoropolyethers

Carbon and oxygen polymer
backbone with fluorines directly
attached

F-(CmF2mO-)nCF3, where the
CmF2mO represents -CF2O, -
CF2CF2O, and/or -
CF(CF3)CF20 distributed
randomly along polymer
backbone

Side-chain fluorinated polymers

\on-lluorinated polymer
backbone with fluorinated side
chains with variable composition

Fluorinated acrylate and
methacrylate polymers,
fluorinated urethane
polymers, and fluorinated
oxetane polymers

PI OS belongs to the perfluoroalkyl acids (PFAAs) of the non-polymer perfluoroalkyl
substances category of PI-AS. PFAAs are among the most researched PFAS (Wang et al. 2017).
The family PFAAs includes perfluoroalkyl carboxylic, sulfonic, sulfinic, phosphonic, and
phosphinic acids (Table 1-3). PFAAs are highly persistent and are frequently found in the
environment (Ahrens 2011; Wang et al. 2017). PFAAs may dissociate to their anions in aqueous
environmental media, soils, or sediments depending on their acid strength (pKa value). The
protonated and anionic forms may have different physiochemical properties.

11


-------
Table 1-3. Classification and Chemical Structure of Perfluoroalkyl Acids (PFAAs).1

Classification

I'linclional Croup

Kxa 111 pies

Perfluoroalkyl carboxylic
acids (PFCAs)

Or

Perfluoroalkyl carboxylates
(PFCAs)

-COOH

Perfluorooctanoic acid (PFOA)

-COO"

Perfluorooctanoate (PFOA)

Perfluoroalkane sulfonic acids
(PFSAs)

Or

Perfluorokane sulfonates
(PFSAs)

-SO3H

Perfluoiooclane sulfonic acid (PFOS)

-SO3-

Peril uorooctanc sulfonate (PFOS)2

Perfluoroalkyl sulfinic acids
(PFSIAs)

-SO2H

IVrlluorooctane sulfinic acid (PFOSI)

Perfluoroalkyl phosphonic
acids (PFPAs)

-P(=0)(OH);

Peril uorooctyl phosphonic acid
(CS-PFPA)

Perfluoroalkyl phosphinic
acids (PFPIAs)

-P(=0)< ()l 1)(("in 1 "2111 1)

Bis(perfluorooctyl) phosphinic acid
(C8/C8-PFPIA)

Perfluoroalkylether carboxylic
acids (PFECAs)

CF^OC F2>nCOOH

Perfluoro (3,5,7-trioxaoctanoic) acid

Perfluoroalkylether sulfonic
acids (PFESAs)

CF3(0CF2)nS03H

(•> 2 chlorinated polyfluorinated ether
sulfonate (6:2 Cl-PFESA)

Perfluoroalkyl dicarboxx Ik-
acids (PFdiCAs)

MOOC-Cnl :„-COOH

9:3 Fluorotelomer betaine

Perfluoroal k a 11 e d i s 111 lb 11 i c
acids (PFdiSAs)

ll();S-C„l :„-S03H

Perfluoro-1,4-disulfonic acid

1 Modified from Buck el ;il (2(>| 11. ()l X I) (2n21).

2 I lie anionic form is nmsi pre\ ;ilenl in llie ;u|iLatic environment

Perfluoi oiilkiine (or -alkyl) sulfonic acids (PFSAs), including PFOS, consist of a general
chemical structure (of CnF2n i SO3H for PFOS; see Figure 1-1). This chemical structure makes
PFOS extremely strong and stable, and resistant to hydrolysis, photolysis, microbial degradation,
and metabolism (see Section 2.3) (Ahrens 2011; Beach et al. 2006; Buck et al. 2011).
Furthermore, PFOS has been classified as persistent, bioaccumulative, and toxic (Ahrens 2011;
Buck et al. 2011; Lindstrom et al. 2011; OECD 2002).

12


-------
o

Figure 1-1. Chemical Structure of Linear Perfluorooctane Sulfonate (PFOS).

Source: United States EPA Chemistry Dashboard; https://com.ptox	' 'ashboard

1.2.1 Physical and Chemical Properties of PFOS

Physical and chemical properties along with other reference information for PFOS are

provided in Table 1-4 These physical and chemical properties help to define the environmental

fate and transport of PI-'OS in the aquatic en\i ronment.

13


-------
Table 1-4. Chemical and Physical Properties of PFOS.

Property

PI'"OS. acidic form1

Source

Chemical Abstracts Service
Registry Number (CAS No.)

1763-23-1



Chemical Abstracts Index
Name

1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-
heptadecafluoro-1 -
octanesulfonic acid



Synonyms

Perfluorooctane sulfonic acid;
heptadecafluoro-1 -octane
sulfonic acid; PFOS acid;
perfluorooctane sulfonate



Chemical Formula

CsHFiyOsS



Molecular Weight (grams
per mole [g/moll)

500.13

Lewis (ed. 2004); HSDB
(2012). SRC (2016)

Color/Physical State

White powder (potassium salt)

OECD (2DD2)

Boiling Point

258-260 ฐC"

SRC (20lo)

Melting Point

No data



Vapor Pressure

2.0 x 10"3 milligrams Mercury
(mm Hg) at 25ฐC (estimate)

HSDB (2012)

Henry's Law Constant

Not measurable: not expected
to volatilize from aqueous
solution ( 2 i) \ |ir")

ATSDR (2015)

Kow

Not measurable

EFSA (2008); ATSDR
(2015)

Organic carbon water
partitioning coefficient (Km )

2.57

Higgins and Luthy (2006)

Estimated pka

3 27 (no empirical
measurements available)

Brooke et al. (2004)

Solubility in W ater

(•>S<) nm 1.

OECD (2002)

Half-Life in Water

Stable

UNEP (2006)

Hal 1-1 .ile in Air

Stable

UNEP (2006)

1 PFOS is a produced ;is ;i potassium sail (CAS No. 2795-39-3). Properties specific to the salt are not
included

PFOS is moderately water soluble, nonvolatile, and stable (Beach et al. 2006; Young and
Mabury 2010). PI'OS is solid at room temperature with a low vapor pressure. No direct
measurement of the acid dissociation constant (pKa) is available. However, PFOS is considered
to have a low pKa, which is based on a calculated pKa of 3.27 provided from Finland in a
comment to Brooke et al. (2004). Therefore, PFOS is deemed to be a strong acid (Brooke et al.
2004). Thus, PFOS introduced as a salt will dissociate into ionic components when in natural

14


-------
water at a neutral pH, and is commonly present as a PFOS anion in solution (Beach et al. 2006;
Giesy et al. 2010; Young and Mabury 2010). The PFOS anion forms strong ion pairs with many
cations, resulting in less solubility in waters that contain great amounts of dissolved solids. Thus,
PFOS solubility in saltwater is approximately 12 mg PFOS/L compared to 589 mg PFOS/L in
pure water (Beach et al. 2006). PFOS is reported to have a mean solubility of 56 mg PFOS/L in
pure octanol (OECD 2002). These solubility data suggest that any form of PFOS discharged into
a water source tends to remain dissolved, unless the PFOS was sorbed lo paniculate matter or
assimilated by organisms (which are both discussed further in Sections 2.2 and 2.5, respectively)
(OECD 2002).

Due to the surfactant properties of PFOS, it forms three layers when added to octanol and
water in a standard test system used to measure an n-oetanol-water partition co-efficient (Kow),
thus preventing direct measurement (Giesy el al 2<) I n. OI X'I) 2002). Although a Kow cannot be
directly measured, a k..ซ lor PI OS has been estimated from its individual water and octanol
solubilities (Giesy el al 2<)|i)). how e\ er. the veracity of such estimates is uncertain (OECD
2002). Lacking a reliable k..ซ lor PI-OS precludes application of Kow-based models commonly
used lo estimate \ arious physiochemical properties for organic compounds, including
bioconcentralion factors and soil adsorption coefficients. Further, the unusual characteristics of
PFOS would bring into question the use of Kow as a predictor of environmental behavior. For
example, bioaccumulation of PFOS is thought to be mediated via binding to proteins rather than
partitioning into lipids (Giesy et al. 2010; OECD 2002), the latter being the theoretical basis for
Kow-based prediction of bioaccumulation.

PFOS is not expected to volatilize from aqueous solution based on its vapor pressure and
predicted Henry's law constant < 2.0 x 10"6 (Beach et al. 2006). In 2002, OECD classified PFOS

15


-------
as a type 2, non-volatile chemical that has a very low or possibly negligible volatility (Beach et
al. 2006; Giesy et al. 2010; OECD 2002).

2 PROBLEM FORMULATION

A problem formulation provides a strategic framework for water quality criterion
development under the CWA by focusing on the most relevant chemical properties and
endpoints. In the problem formulation, the purpose of the assessment is slated, the problem is
defined, and a plan for analyzing and characterizing risk is developed. The structure of this
problem formulation is consistent with EPA's Guidelines for Ecological Risk Assessment
(U.S.EPA 1998).

2.1 Overview of PFOS Sources

2.1.1 Manufacturing of PFOS

PFOS is used in a variety of products including surface treatments for soil and stain

resistance, coating of paper as part of a sizing agent formulation, and in specialized applications

such as firefighting foams. PFOS is produced through I'lectrochemical Fluorination (ECF) in

which an organic raw material, such as octane sulfonyl fluoride (OSF; C8H17SO2F) in the case of

PFOS. undergoes electrolysis in anhydrous hydrogen fluoride solution. This electrolysis leads to

the rc|">laccmenl of all the hydrogen atoms by fluorine atoms and results in

perfhiorooclancsullbin I fluoride (POSF; C8F17SO2F), which is the major raw material used to

manufacture PFOS (Figure 2-1; Buck et al. 2011). The base-catalyzed hydrolysis of POSF

results in PFOS and its salts (Lehmler 2005). ECF results in a mixture of linear and branched

chain perfluorinated isomers and homologues, with ratios of linear to branched perfluorinated

carbon chains of roughly 70 to 80% linear and 20 to 30% branched for PFOS synthesis

depending on how the process is controlled (De Voogt 2010). All compounds produced from

16


-------
POSF and other neutral PFAS with sufficient chain length and a sulfur group have the potential
to degrade or transform into PFOS, and therefore have been considered to be "PFOS
equivalents" and as potential sources of PFOS to the aquatic environment (see Section 2.4;
Ahrens 2011; Lindstrom et al. 2011). PFOS is used in a variety of products including surface
treatments for soil and stain resistance, coating of paper as part of a sizing agent formulation, and
in specialized applications such as firefighting foams.

CH SH

8 17

CH SOF

8 17 2

(OSI )

3 o



cj ฆ c



III. c	Ml. c

C I SO I

X 17 2

(I'OSI )

C F SO 11

8 17 ?

(PFOS)

C F SO X

8 17 2

(POSF Derivatives)

Figure 2-1. Synthesis of PFOS by electrochemical fluorination (ECF).

Modified from Buck et al. (2011).

The manufacture of PFOS started in 1949 with Minnesota Mining and Manufacturing
(later name changed to the 3M Company) (3MCompany 1999). Prior to 2000, the 3M Company

17


-------
was the major producer of POSF, the raw material used to make PFOS (Figure 2-1), with smaller
producers in Europe and Asia (Paul et al. 2009; U.S.EPA 2000a). In 2000, the 3M company
manufactured approximately 78% of the estimated global POSF production (approximately
3,665 tons of the 4,650 tons produced globally; OECD 2002). The estimated total cumulative
production of POSF is between 44,000 and 96,000 tons (Paul et al 2<~)f)0. Prevedouros et al.
2006; Smithwick et al. 2006). Information on previous and current production of POSF from
Asia and other production sources is limited (Paul et al. 2009; Pre-\ edouros el ill. 2006;
Smithwick et al. 2006).

In May 2000, following negotiations between I-PA and 3\l. the 3M Company agreed to
voluntary phase out and find substitutes for PFOS chemistry used to produce all but a few small
applications (i.e., aqueous film-forming foams ( AI'FF). and hard chrome plating mist
suppression) across their range of products In (I .indstrom et al. 2011; U.S.EPA 2000a).
Starting around the same time, a series of Significant New I se Rules (SNUR) were also put into
place by the EPA to restrict the production and use of materials that contain PFOS and its
precursors in the U S (I .indstrom et al 2<)||) In 2009, PFOS and related compounds were listed
under Annex 1} of the Stockholm Convention on Persistent Organic Pollutants; restricting global
manufacturing and use of PI OS (Alliens 2011; OECD 2002). Homologues, neutral precursor
compounds, and new classes ol'PFAS continue to be produced and therefore, are potential
sources of PFOS (Alliens 2011). Assuming there was no step-up production of PFOS and its
precursors to offset the phase-out by the 3M Company, the production is estimated to be
approximately 1,000 tons from 2002 and onward (Paul et al. 2009). However, while
industrialized countries, like the U.S., phased-out the use of PFOS and its precursors, producers
in other countries, such as China and Brazil, have scaled up their production to fill remaining

18


-------
demand (Wang et al. 2013b). Despite the wide use in an array of industrial and consumer
products globally, information on the sources, volumes, and emission of PFOS and its precursors
are limited (Paul et al. 2009; Zhang et al. 2016).

2.1.2 Sources of PFOS to Aquatic Environments

Aquatic environments and soil are thought to serve as a reservoir of PFOS, with 42,000

tons emitted to aquatic environments compared to 235 tons released into to air between 1980 and

2002 (Paul et al. 2009). Unlike other contaminants commonly found in aquatic ecosystems, such

as metals for example, PFAS are synthetic compounds with no natural source Thus, the

occurrence of any PFAS in the environment is an indication of anthropogenic sources (Ahrens

2011). The occurrence of PFOS in aquatic environment can be attributed to both point and non-

point sources, entering aquatic environments from industrial and consumer products during

manufacturing, along supply chains, and during product use and'or disposal (Ahrens 2011;

Ahrens andBundschuh 2<>14: Kannan 2<)| I. Paul et al. 2<)i)1)) I lowever, quantitative

assessments of PI OS production, point and non-point source discharges, and environmental

measurements are limited compared to other persistent, bioaccumulative pollutants (Ahrens and

Bundschuh 2<)|4. /hang et al 2<)!(ฆ>).

Potential point sources of PFOS to the aquatic environment include both industrial

facilities and municipal wastewater treatment plants (WWTPs). Additional point sources may

include surface water runoff from industrial use sites such has metal plating facilities, areas that

have received AFFF applications, landfills, and contaminated soils. Of these, industrial facilities,

specifically those for fluorochemical manufacturing and other use facilities, are a primary source

of PFOS to aquatic systems (Ahrens et al. 201 la; Houtz et al. 2016; Sedlak et al. 2017).

Estimated total global releases to water arising from discharge of PFOS during manufacturing

from 1970 to 2002 ranged between 230 and 1,450 tons (Paul et al. 2009).

19


-------
Potential non-point PFOS sources to aquatic environments include: dry and wet
atmospheric deposition, runoff from contaminated soils, discharge of contaminated groundwater,
runoff from metal plating facilities, the runoff or discharge of contaminated groundwater
particularly from the use of fire-fighting foams and land application of contaminated biosolids
(Ahrens 2011; Kannan 2011; OECD 2002; Paul et al. 2009). Identification of non-point PFOS
sources and understanding their relative contribution to aquatic ecosystems is difficult due to the
lack of sufficient measured environmental data (Ahrens 201 1; Paul el al 2<)<)l>). Overall, the
presence of non-point PFOS sources and their relative contributions are reported to be dependent
on the aquatic system, air, groundwater, and soil le\ els. and nearby land uses. I-'or example,
concentrations of PFAS, including PFOS, have been influenced by urban land use (Ahrens 2011;
Zhang et al. 2016). Overall, PFOS occurrence in aquatic en\ ironmenls is driven by legacy PFOS
sources since PFOS use in the United States was voluntarily phased out by 2002 and significant
new use rules were put into place In N\\ to restrict the production and use of PFOS and its
precursors (Lindstroni et al 2<)| I) And generally, PFAS concentrations in the environment have
been positively correlated with human population density. PFOS was detected in aquatic systems
at ele\ ated concentrations (ranging between ^7 and 1,371 ng/L) in densely populated areas of the
U.S. and Europe (/hang et al. 2016 and Loos et al. 2009; respectively). Paul et al. (2009)
estimated the total global PI OS emissions to air and water from 1970 to 2009 resulting from
consumer use and disposal to he between 420 and 2,100 tons.

Importantly, PFAS are still produced that can transform or degrade into compounds
belonging to the PFSAs family, including PFOS (Ahrens 2011). The metabolic transformation of
PFAS precursors such as perfluoroalkyl sulfonamidoacetic acids (FASAAs) and the degradation
of volatile PFAS such as perfluoroalkyl sulfonamidoethanols (FASEs), are known to degrade to

20


-------
PFOS (Ahrens and Bundschuh 2014; Benskin et al. 2009; Boulanger et al. 2005; Buck et al.
2011; Lange 2000; Liu and Mejia Avendano 2013; Plumlee et al. 2008; Rhoads et al. 2008;

Wang et al. 2017). However, understanding of these transformation processes is limited, and
additional work is needed to fully understand these processes and their role as sources of PFOS
to aquatic environments (Buck et al. 2011; Lau et al. 2007; Liu and Mejia Avendano 2013; Wang
et al. 2017). Degradation of precursors represents a potential]) significant source of PFOS to the
aquatic environment, particularly since PFOS production within the I S lias not occurred since
2002 (Buck et al. 2011; Liu and Mejia Avendano 2013). Nevertheless, PFOS-lrcalcd articles,
such as fabrics, paper, and other treated materials, are still being imported into the I S. and are
ultimately, at least in part, released into the environment (Allied et al. 2015; Lang et al. 2016;
Liu et al. 2014d). The importation of PFOS treated articles is considered as production under the
Toxic Substances Control Act (TSCA) (U.S.EPA 202<))

2.2 Environmental Fate and Transport of Pl'OS in the Aquatic Environment

2.2.1 Environmental Fate of I'l'OS in the Aquatic Environment

PFOS has low volatility in ionized form but can adsorb to particles in air where it can be

transported globally, including remote locations (Benskin et al. 2012; Butt et al. 2010). PFOS is

water soluble and has been found in surface water, ground water, and drinking water. Because of

the relatively low K.v of Pl'OS. it does not easily adsorb to sediments and tends to stay in the

water column (Alli ens 2d I I. |}cach et al. 2006; Giesy et al. 2010; Higgins and Luthy 2006).

PFOS can be re-emitted to aquatic environments from PFOS contaminated soil,

groundwater, ice, and sediment (see Section 2.3). Sediment may be an important sink of PFOS in

the aquatic environment (Ahrens 2011). The movement of PFOS between groundwater, surface

water, and sediment depends on the chemical properties of PFOS and site-specific

physiochemical characteristics (including pH, temperature, organic carbon content, and salinity)

21


-------
of the aquatic environment. In general, PFOS may sorb to sediments (with a Kd greater than 1
mL/g; Giesy et al. 2010). However, this sorption to sediments is limited and PFOS has a Koc of
2.57 indicating that PFOS is relatively mobile in water and the physicochemical characteristics
of the sediment ultimately influence the sorption of PFOS (Ahrens 2011; Higgins and Luthy
2006). While the release of PFOS from the transformation of other PFAS and historical products
still in use (e.g., consumer goods manufactured, imported and 'or obtained before the PFOS
discontinuation and regulations) will continue into the future, the re-emissions of PFOS from
existing sinks are assumed to be decreasing since the restrictions and regulations of PFOS have
gone into place (Ahrens 2011; Ahrens and Bundscluih 2<)|4. Paul el al. 2009; Washington and
Jenkins 2015; Washington et al. 2015).

In the water column, and other cn\ ironmental compartments. PFOS is stable and resistant
to hydrolysis, photolysis, volatilization, and hiodeuradalion (see Appendix M; Beach et al. 2006;
OECD 2002). The persistence of PFOS has been attributed to the strong carbon-fluorine (C-F)
bond. Additionally, there are limited indications that naturally occurring defluorinating enzymes
exist that can break a C-l bond Consequently: no biodegradation or abiotic degradation
processes for PI-'OS are know n The phvsiochemical properties discussed in Table 1-4 result in
PFOS being highly persistent in the aquatic environment (Ahrens 2011). In aquatic
environments, the only dissipation mechanisms for PFOS are physical mechanisms, such as
environmental dilution, oflsile transport, plant uptake, and sorption.

2.2.2 Environmental Transport of PFOS in the Aquatic Environment

The environmental fate of PFOS, outlined in the previous section (Section 2.2) plays a

role in the environmental transport of PFOS (Ahrens 2011). PFOS is either distributed in biota
(via bioaccumulation discussed in Section 2.5) or abiotic matrices (such as water and sediment).

22


-------
Sediment in particular can act as a sink for PFOS. However, the role of sediment as a sink or
source by resuspension is not well understood (Ahrens 2011).

The distribution of PFOS is widespread, including to remote regions despite the limited
number of manufacturing facilities and/or small population sizes typically found in these areas
(Benskin et al. 2012; Butt et al. 2010). PFOS has been detected in water, sediment, and biota
samples from aquatic environments in remote areas (Butt et al 2<)|o. (riesy and Kannan 2001;
Houde et al. 2006; Yamashita et al. 2008). To date, the dominant transport pathway for PFOS to
remote regions has not been conclusively characterized and much of the locus has been on
marine systems, with few studies in freshwater environments (Ahrens 2011; Butt el al. 2010;
Giesy and Kannan 2002). Additionally, the relative importance of each potential transport
pathway is difficult to accurately determi lie (liiilt et al 2010. Young and Mabury 2010). Many
researchers suggest that the dominant mechanism of PI OS transport occurs through water as the
anionic form of PFOS. which is the most commonly found form in the aquatic environment, is
less volatile (see Section 2 2 I aho\ e) and has a high water solubility. These characteristics make
partitioning to and transport through the air less likely (Butt et al. 2010; Giesy and Kannan
2002) I lo\\e\ er. PI OS transport through water is likely the dominant mechanism on more local
scales (eg. within a waterhody or watershed), and is likely not the prevailing transport pathway
of PFOS to remote regions gi\ en the considerations of the long distances. Instead, atmospheric
transport is likely the main mechanism of PFOS transport to remote regions. Another potential
source to remote regions is the indirect formation of PFOS through transformation of other
PFAS, particularly volatile precursors (see Section 2.4; Butt et al. 2010; Wang et al. 2015;

Young and Mabury 2010).

23


-------
Volatile PFOS precursors, which may reach remote locations via atmospheric deposition
themselves, may subsequently be metabolized to PFOS in aquatic organisms (Giesy and Kannan
2002). In all likelihood, the continued presence of PFOS in remote areas may be due to multiple
exposure pathways, including those caused by direct production and use of PFOS itself as well as
degradation and transformation of precursor compounds (Armitage et al. 2009). To better
comprehend both environmental transport and exposure to PFOS. the following needs to be
better understood: 1) the potential transformation, metabolism, and bioaccumulation of PFOS
and its precursors (particularly partitioning beha\ ior. such as tissue distribution and
lipophilicity); 2) explicit biotransformation pathways and pharmacokinetics; and 3) atmospheric
fate and transport of PFOS and its precursors (Armitage et al. 2009).

2.3 Transformation and Degradation of Pl-'OS Precursors in the Aquatic
Environment

Transformation and degradation processes of various PI-AS are potential sources of
PFOS to the aquatic en\ ironmenl (see Section 2.1.2 above). PFAS are still produced that can
transform or degrade into com pounds belonging to the PFSAs family of PFAS, including PFOS
(Ahrcns 2<)| I) Thus, transformation and degradation of PFAS should be considered as an
ongoing potential source of PI OS to the aquatic environment. Currently, the understanding of
these transformation and degradation processes is limited, particularly for PFOS. There is little
understanding of w hicli PI-AS and how much of each has been or will be released into the
aquatic environment (I .in and Mejia Avendano 2013; Wang et al. 2017). Additional work is
needed to fully understand the details of these processes and the occurrence of the compounds to
better comprehend their role as a source of PFOS to aquatic environments (Lau et al. 2007).

These transformation and degradation pathways are dependent on environmental

conditions, degradation kinetics, and the chemical structures and properties of the individual

24


-------
PFAS precursors and volatile PFAS (Buck et al. 2011; Butt et al. 2014; Liu and Mejia Avendano
2013). Of particular importance is the environmental stability of key chemical linkages (such as
esters and ethers) as the stability of these chemical linkages determines the stability of the overall
PFAS (Liu and Mejia Avendano 2013). The most well studied PFAS precursors are
fluorotelomer-based compounds, which are produced through telomerization technology and are
associated with PFOA as the final product (Buck et al. 2011; T.in and Mejia Avendano 2013). In
contrast, perfluoroalkane sulfonamido derivatives and other PFAS, such as side-chain-
fluorinated polymers, are not as well studied.

It is essential to understand the biodegradation of volatile PFAS, such as peril uroalkane
sulfonamido derivatives as their degradation is directly linked with PFOS generation in the
environment (Liu and Mejia Avendano 2d 13) Most published studies on the degradation of
perfluoroalkane sulfonamido derivatives focus on those with eight tluorinated carbons since
PFOS is a final product (linck el al 2<> I I) TV-ethyl perfluorooctane sulfonamidoethanol
(EtFOSE) in particular is the most commonly studied

2.3.1 Deuradation of nerlluoioalkane sulfonamido derivatives

IVrlluoioalkane sulfonamido deii\atives, including perfluoroalkane sulfonamides,

sulfonamidoelhanols, sulfonamidoetln I aciylates, and sulfonamidoethyl methacrylates, are final

products on their ow n and are important building blocks for further synthesis (Buck et al. 2011).

The various deri\ati\ es ha\ e been found to degrade into PFSAs, such as PFOS when sufficient

chain length is present, and are intermediates along the transformation pathway. These

derivatives include members of the TV-ethyl perfluoroalkane sulfonamidoacetic acids

(EtFASAAs), TV-ethyl perfluoroalkane sulfonamides (EtFASAs), perfluoroalkane

sulfonamidoacetic acids (FASAAs), perfluoroalkane sulfonamids (FASAs), FASA N-

glucuronides, and perfluoroalkane sulfinic acids (PFSIAs; Buck et al. 2011). Additionally, in the

25


-------
environment A'-alkyl perfluoroalkane sulfonamidoethyl acrylates and methacrylates (and
polymers based on them) may undergo hydrolysis of the ester linkages to produce A'-alkyl
perfluoroalkane sulfonamidoethanols (FASEs; Buck et al. 2011).

In particular, several studies have demonstrated that EtFOSE, a member of the A'-alkyl
perfluoroalkane sulfonamidoethanols of the perfluoroalkane sulfonamido substances, degrades
into PFOS (Benskin et al. 2009; Boulanger et al. 2005; Hatfield 2<)i) I. Lange 2000; Plumlee et al.
2009; Rhoads et al. 2008). EtFOSE was a product of electrochemical fluori nation and was a
precursor compound for the synthesis of other products such as phosphate osiers that were used
to manufacture paper protectors (3MCompany 1999). Several studies have investigated the
degradation of EtFOSE and all found that it is prone to degradation (Benskin et al. 2009;
Boulanger et al. 2005; Hatfield 2001; Lanue 2<)()i); Plumlee et al 2<)<~>9; Rhoads et al. 2008).

The overall pathway of EtFOSE degradation was determined to be the major difference
between these studies (l-'igure 2-2) Rhoades et al. (200S) determined that EtFOSA could
undergo direct dealk\ lation to form perilnorooctane sulfonamide (FOSA; as shown by the red
arrow in Figure 2-2) I .ange (2<)<)<)) suggested that PFOA could be formed as a minor end
product through an abiotic one-electron transfer mechanism from perfluorooctane sulfinic acid
(PFOSI. demonstrated by the blue arrow in Figure 2-2). In contrast, the other studies did not find
PFOA to be a degradation product (Benskin et al. 2013; Boulanger et al. 2005; Rhoads et al.
2008). Further, in the aerobic biodegradation studies, the rate limiting step was determined to be
the degradation of A'-ethyl perfluorooctane sulfonamidoacetic acid (EtFOSAA) and consequently
EtFOSAA was the major degradation product rather than PFOS (Liu and Mejia Avendano 2013).
Nevertheless, the degradation of EtFOSE resulted in the formation of PFOS as one of the final
degradation products. In contrast, in the abiotic degradation studies, PFOS and PFOSI were

26


-------
either present at trace concentrations or were not observed (Hatfield 2001; Plumlee et al. 2009).
Instead FOSA was considered to be the stable end product (Plumlee et al. 2009). The differences
in the degradation pathways observed in the literature can likely be attributed to environmental
conditions (Buck et al. 2011; Liu and Mejia Avendano 2013). Nevertheless, these pathways
demonstrated that degradation of EtFOSE resulted in the formation of PFOS and should be
considered a potential source of PFOS to the aquatic environment I lowever, currently the
relative contribution of this potential source to the aquatic environment cannot he quantified
(Buck et al. 2011; Liu and Mejia Avendano 2013)

27


-------
cf F. FRFF. F 9 /

WxMr(

FFFFFFFF
EtFOSE

OH

I Wfi

FFFFFFFF	i FFFFFFq

F F, F F F F F

PFOS

F F F R F Fv F O

PFOA
Minor end product
through abiotic
mechanisms

FF F
EtFOSE

W

P F F.F F. F F,F O

WxW

FFFFFFFF
EtFOSAA

FFFFFFFF

EtFOSE Aldehyde

; c F F FF. FF, F 0

FFFFFFFF
!	EtFOSA

H

P F F F F F FF 9

FFFFFFFF

EtFOSA Alcohol

OH

FOSAA

tt

c F c F F. F F. F 9

'V^ScVoT

FFFFFFFF

EtFOSA Aldehyde

Figure 2-2. Aerobic Biodegradation of EtFOSE in Activated Sludge.

Black arrows show the Aerobic Biodegradation pathway as described by Liu and Mejia Avendano (2013). Blue
pathway was observed by Lange (2000). Red pathway was observed by (Rhoads et al. 2008) Semi-stable
compounds are shown inside boxes.

Modified from: Liu and Mejia Avendano (2013).

2.3.2 Perfluorooctane sulfonamide-based side-chained polymers

In contrast to some other PFAS described in Section 0, fluorinated side-chain polymers

do not have the per- or poly fluorinated backbone. Instead, fluorinated side-chain polymers

consist of a variable composition with per- and polyfluoroalkyl side chains (Buck et al. 2011).

The side chains of each of these polymer types may sever to transform into PFAS. Currently,

28


-------
little is known about these transformation processes (Liu and Mejia Avendano 2013). Given the
high production volume of perfluorooctane-sulfonamide-based side-chain polymers prior to
2002, these fluorinated side-chain polymers may contribute to the levels of PFAS in the
environment. It remains unknown how much these polymers contribute to the PFSAs in the
environment (Liu and Mejia Avendano 2013). However, this transformation process is expected
to occur over a long period of time (e.g., > 1,000 years) and may be a relatively small contributor
of PFAS, including PFOS, in the environment (Buck et al. 201 1). In contrast to some other PFAS
described in Section 1.2, fluorinated side-chain polymers do not have the per- or polyfluorinated
backbone. Instead, fluorinated side-chain polymers consist of a variable composition with per-
and polyfluoroalkyl side chains (Buck et al. 2011). The side chains of each of these polymer
types may sever to transform into PFAS. Currently. little is know n about these transformation
processes (Liu and Mejia Avendano 2013) (ii\ en the high production volume of
perfhaorooctane-sullbnamidc-hased side-chain polymers prior to 2002, these fluorinated side-
chain polymers may contribute to the lc\ els of PFAS in the environment. It remains unknown
how much these polymers contribute to the I'l'SAs in the environment (Liu and Mejia Avendano
2013) I lo\\e\ er. this transformation process is expected to occur over a long period of time (e.g.,
> 1,0<)<) years) and may be a relatively small contributor of PFAS, including PFOS, in the
environment (Ikick et al 2<)| I)

2.3.3 Fluoroalkyl surfactants used in AFFFs

The release of AFFF during firefighting activities has been determined to be a substantial

source of PFOS to the aquatic environment (see Section 2.1.2). Since 2002, fluorinated

alternatives to PFAAs have been used to manufacture AFFF (Buck et al. 2011; Wang et al.

2013b). The ten classes of AFFF chemicals have been identified and show that the new

formulations of AFFF include the eight carbon perfluoroalkyl moiety (Place and Field 2012).

29


-------
Some of these fluorinated alternatives may undergo transformation and degradation processes
and therefore may contribute to the levels of PFOS occurring in the aquatic environment (Liu
and Mejia Avendano 2013). However, additional details about the transformation and
degradation processes, including specific transformation pathways, the time to undergo
transformation to produce a final product, and the influence of the environmental condition, are
lacking at this time (Liu and Mejia Avendano 2013; Wang et al 2' > 13h)

2.4 Environmental Monitoring of PFOS in Abiotic Media

PFOS has been detected in a variety of environmental abiotic matrices in aquatic

environments around the globe. These abiotic media include surface water, soils, sediments,
groundwater, air, and ice caps (Butt et al. 2010; Lau et al. 2007). Water is expected to be the
primary environmental medium in which PTOS is found (Lau el al 2<)()7). Occurrence and
detection of PFOS in surface waters is described below and occurrence in other abiotic media is
described in Appendix \

2.4.1 PFOS Occurrence and Detection in Ambient Surface Waters
2.4.1,1 Summary of PI-'OS occurrence and concentrations across the U.S.

PI-OS is one of the dominant PFAS detected in aquatic ecosystems, along with PFOA

(Alliens 2<)| I. Benskin et al 2d 12. Dinglasan-Panlilio et al. 2014; Nakayama et al. 2007;

Remucal 2< > I 'ฆ>. /areitalabad et al. 2013). Despite its wide use and persistence in the aquatic

environment, current information on the distribution of PFOS in surface waters of the U.S. is

relatively limited (Jarvis et al. 2021). Available data are largely collected from freshwater

systems in eastern states, with most of the current, published PFOS occurrence data focused on a

handful of study areas with known manufacturing or industrial uses of PFAS and among areas of

known AFFF use, such as fire-training areas on military bases (Figure 2-3 and Appendix N).

30


-------
/

Figure 2-3. Map Indicating Sampling Locations for Perfluorooctane Sulfonate (PFOS)
Measured in Surface Waters across the United States (U.S.).

Based on data reported in the current, publicly available literature. Sampling locations for the Colorado data were
not available and these data are represented by the dash marks to indicate measured PFOS surface water
concentrations are available. Detailed information on sampling locations, including references, coordinates, and
sampling site identification numbers and names, provided in Appendix N.

Modified from: Jarvis et al. (2021).

31


-------
Concentrations of PFOS in surface waters across the U.S. appear to vary widely, with
observed concentrations ranging over eight orders of magnitude and are generally detected
between picogram and nanogram per liter with reported concentrations in microgram per liter (or
part-per-trillion) ranges (Ahrens 2011; Zareitalabad et al. 2013). For the purposes of this
overview, all concentrations reported here are in nanogram per liter (ng/L). Measured surface
water concentrations of PFOS in peer-reviewed journal articles and publicly available industry
and government reports range between 0.074 and 8,970,000 ng/L with an arithmetic mean
concentration of 786.77 ng/L and a median concentration of 3.6 ng/L (Jan is el al 2021).
However, it should be noted that the mean and median concentrations reported in .larvis et al.
(2021) were calculated from the reported concentrations for individual samples and therefore, are
not fully representative of all the measured PI OS concentrations in I' S. surface waters.
Additionally, as demonstrated by the median concentration of 3 6 ng/L, a majority (roughly
91%) of measured PTOS concentrations were found to fall below 300 ng/L (Jarvis et al. 2021).

32


-------
1000000



C/3

o

u.
P-,

re

•S


-------
the presence of an anthropogenic source, a transport pathway (air, surface water, or ground
water), and the persistence and mobility of the PFOS in the environment. Therefore, PFOS
concentrations in surface water tend to be dependent on the presence of a nearby source and
generally increase with levels of urbanization.

Further, there are insufficient data to quantitatively e\aluate temporal trends of PFOS in
surface waters across the U.S. (Remucal 2019). However, recent studies ha\ e suggested that
PFOS concentrations in surface waters with limited sampling sites in northeastern states appear
to have decreased since the voluntary phase out of PFOS in 2<)i)2 (Pan et al. 2018, Zhang et al.
2016). While these studies observed lower measured PFOS concentrations in surface waters
compared to those reported in earlier reports (I lansen et al 2002. \akayama et al. 2007), few
studies have measured PFOS concentrations from the same sampling sites over time (Jarvis et al.
2021). Eight studies (six Ibaised on the Great Lakes and two in New York on the Hudson River)
measured PFOS in the same wateihody o\er time (Appendix N). Thus, the observed lower
concentrations reported in recent literature could he due to trends of PFOS concentrations
decreasing since the 20<)2 PI OS phaseout. differences in sampling site locations and/or advances
in analylical methods for detecting PFOS that reduced detection limits (Jarvis et al. 2021).

Despite the wide use and persistence of PFOS in aquatic ecosystems and unlike the
extensive sampling of PI OS in drinking water sources1, groundwater, and fish tissue
monitoring2, current information on the environmental distribution of PFOS in ambient surface
waters across the U.S. remains very limited. More recent sampling efforts indicate that PFOS

1	EPA's database for the Unregulated Contaminant Monitoring Rule (UCMR) that includes data for treated surface
waters, (https://www.epa.gov/dwucmr')

2	EPA's National Rivers and Streams Assessment (NRSA; https://www.epa.gov/national-aquatic-resource-
surveys/ncca) and the Great Lakes Human Health Fish Tissue Study component of the EPA National Coastal
Condition Assessment (NCCA/GL)

34


-------
occurrence may be more widespread. PFOS was detected in almost all collected surface water
samples, which can likely be attributed to improvements in analytical methods that lowered the
PFOS detection limit compared to older analytical methods (Gewurtz et al. 2013).

Thus, from the currently available data, which were largely collected from freshwater
systems in eastern states and in the Upper Midwest with known manufacturing or industrial uses
of PFAS or use of AFFF, PFOS concentrations measured in TT S surface waters appear to vary
widely, across eight orders of magnitude (Jarvis et al. 2021). PFOS concentrations in remote
areas (i.e., areas with little to no PFAS manufacturing and/or industrial uses) range between
0.074 to 23.23 ng/L (Jarvis et al. 2021). This contrasts with PFOS concentrations measured in
areas with known PFAS manufacturing, industrial use, and/or application of AFFF, which vary
widely and reach up to the maximum obser\ ed concentration of N.l)70.000 ng/L at a site
impacted by AFFF (Appendix N). While current PI-AS occurrence data illustrate the prevalence
and quantify concentrations of PI OS in surface waters across the U.S., additional data,
particularly in central, southwestern, and western freshwaters as well as saltwater systems, is
needed to better understand PI OS occurrence in aquatic ecosystems across the U.S. (Jarvis et al.
2021) See Appendix \ lor further discussion of PFOA occurrence in surface waters and other
abiotic media such as aquatic sediments, groundwater, air, and ice.

2.5 Bioaccu m illation and Biomagnification of PFOS in Aquatic Ecosystems

PFAS, including PI'OS, are found in aquatic ecosystems around the globe (e.g., Ankley et

al. 2020; Giesy and Kannan 2001; Houde et al. 2008). Although they were used predominantly in
more populated areas, these compounds are resistant to hydrolysis, photolysis, and
biodegradation (see Section 2.2), facilitating their long-range transport to aquatic ecosystems in
the remote arctic and mid-oceanic islands (see Section 2.3.3; Haukas et al. 2007; Houde et al.

35


-------
2006). Several physical-chemical properties of PFOS contribute to bioaccumulation within
aquatic species once they have entered an ecosystem.

2.5.1 PFOS Bioaccumulation in Aquatic Life

In contrast to many persistent organic pollutants, which tend to partition to fats, PFOS

preferentially binds to proteins (Martin et al. 2003a; Martin et al. 2003b). Within an organism

PFOS tends to bioaccumulate within protein-rich tissues, such as the Wood serum proteins, liver,

kidney, and gall bladder (De Silva et al. 2009; Jones et al. 2<)i)3: Martin el al 2003a; Martin et al.

2003b). PFOS also binds to ovalbumin, and the transfer of PFOS to such albumin in eggs can be

an important mechanism for depuration in female o\ iparous species, as well as a mechanism for

maternal transfer of PFOS to offspring (Jones et al. 2003, kannan et al. 2005).

The stability of PFOS contributes lo its Moaccumulalion potential, as it has not been

found to undergo biotransformation within the organism (Martin et al. 2003a; Martin et al.

2003b). Within an organism. PFOS undergoes enterohepatic recirculation, in which PFOS is

excreted from the li\ er in bile lo the small intestine, then reabsorbed and transported back to the

liver (Goecke-Floi a and Reo I lW) This process becomes increasingly more efficient the longer

the pei lliioi inaled chain length is. resulting in longer biological half-lives for chemicals like

PFOS with a relatively long chain length, as they are less readily excreted. PFAS with sulfonate

head groups, such as PFOS. are more efficiently resorbed by the small intestine than carboxylate

PFAS such as PI-OA. resulting in higher bioaccumulation levels (Hassell et al. 2020; Jeon et al.

2010; Martin et al. 2003a).

Sex differences in the elimination rates of PFOS in addition to the transfer of PFOS to

albumin in eggs (e.g., Jones et al. 2003; Kannan et al. 2005) have not been well studied. Some

research suggests lower PFOS elimination rates in female rats than in male rats (Butenhoff et al.

2012; Chang et al. 2012; Pizzurro et al. 2019), suggesting potentially longer retention of PFOS in

36


-------
females. However, this difference was not observed in mice, rabbits, monkeys, or humans
(Pizzurro et al. 2019). In contrast, PFOA elimination rates are higher in females than in males for
both female fathead minnows (Lee and Schultz 2010) and rats (Pizzurro et al. 2019), suggesting
potential longer retention of PFOA in males. These data indicate further research across species
and genders for PFAS elimination rates may be useful.

The structure of PFOS also affects its bioaccumulation potential, with linear forms being
more bioaccumulative than branched forms (Fang et al. 2014; Hassell el ill 2'>20). The
preferential accumulation of linear PFOS occurs because the elimination rale of branched
isomers of PFOS is higher, particularly across gill surfaces (Hassell et al. 2020) This pattern has
also been observed in the field, as the proportion of branched isomers was higher in water and
sediment compared to fish tissue in Taihu l.ake. China (Faim el al 2<)14) and Lake Ontario
(Houde et al. 2008).

2.5.2 Factors Inlliieneinu PI OS ISioacainuilalion and Biomaunification in Aquatic Ecosystems

Because of their affinity lor binding lo proteins, PFAS can enter the base of the food web

through sorption lo organic mailer in sediments or biofilms (Higgins and Luthy 2006; Jeon et al.
2010. IVnland el al 2<>2< >) or can bind lo blood proteins at gill surfaces of aquatic organisms
through respiration (De Sil\ a et al. 2009; llassell et al. 2020; Martin et al. 2003a; Martin et al.
2003b).

PFAS binding lo the surface of sediment organic matter and biofilms is influenced by
both hydrophobic and electrostatic effects, resulting from the hydrophobicity of the
perfluorinated chain and the hydrophilicity of the sulfonate or carboxylate head groups (Higgins
and Luthy 2006; see Section 2.3 for further details on the sorption of PFOS). Overall, these
results suggest that sorption to sediments should be an important mechanism for PFOS entry into

37


-------
an aquatic ecosystem, but that subsequent dietary uptake from benthic feeding organisms will be
more important for PFOS than PFOA.

The importance of the sediment pathway for PFOS bioaccumulation in aquatic
ecosystems has been demonstrated in laboratory studies with Chironomus riparius (Bertin et al.
2014), C. plumosus (Wen et al. 2016), Gammarus fossarum and G. pa/ex (Bertin et al. 2016),
and Lumbriculus variegatus (Lasier et al. 2011), where PFOS concentrations were positively
correlated between sediments and whole-body tissue samples of benthic feeding organisms. The
sediment pathway has also been demonstrated in several field studies, where PI OS was
measured in sediments and biofilms, and was higher in benthic-feeding invert eh rales relative to
pelagic-feeding invertebrates (Lescord et al. 2014; Loi et al. 201 1; Martin et al. 2004; Penland et
al. 2020). In addition, the distribution of PI AS in sediments was more similar to their
distribution in the tissues of benthic invertebrates (I .escord et al 2<> I 5) and fish (Thompson et al.
2011) than they were to their distribution in pelagic organisms.

PFAS can also enter aquatic organisms directly from the water column through
respiration. Because of its binding aflinity to proteins, PFOS can enter the body of gill-breathing
organisms In binding to proteins in the blood at gill surfaces (Jones et al. 2003; Martin et al.
2003a. Martin et al. 20<)3b) The relative distribution of PFOS in tissues is related to the primary
route of exposure (dietary or respiratory). In rainbow trout, the rank order of PFOS
concentrations following aqueous exposure was blood > kidney > liver (Martin et al. 2003a). In
contrast, their rank order following dietary exposure was liver > blood > kidney (Goeritz et al.
2013). Hong et al. (2015) observed the highest concentrations of PFOS in the intestines of green
eel goby, soft tissues, shell, and legs of shore crabs; and gills and intestines of oysters, suggesting

38


-------
bioaccumulation through both dietary and aqueous uptake in invertebrates, and primarily dietary
uptake in fish.

In addition to being bioaccumulative, PFOS has been shown to biomagnify with
increasing trophic level in a variety of freshwater ecosystems (Kannan et al. 2005; Martin et al.
2004; Penland et al. 2020; Xu et al. 2014) and saltwater ecosystems (de Vos et al. 2008; Houde
et al. 2006; Loi et al. 2011; Powley et al. 2008; Tomy et al. 2004) in North America, Europe, and
Asia. PFOS is often the most abundant PFAS in aquatic organisms, and this high relative
abundance is at least partially explained by the biotransformation of PFOS precursor chemicals
into PFOS (see Section 2.4; Haukas et al. 2007; Kannan et al 2005; Kelly et al. 2<)oi). Martin et
al. 2004; Tomy et al. 2004). Higher trophic level organisms have a greater capacity to metabolize
PFOS precursor chemicals, which have been found in lower concentrations in increasing trophic
level (Fang et al. 2014; Kannan et al. 2005. Martin el al 2<>(ป4) This suggests that in addition to
biomagnification, some of the trophic-level increase in PI - OS can be explained by the
biotransformation of precursor chemicals.

2.5.3 Environmental Monitoring of PI-OS in ISiotic Media

PI-OS is one of the dominant PI AS detected in aquatic ecosystems, along with PFOA

(Ahrens 2" I I. TJcnskin el al 2< > 12, Dinglasan-Panlilio et al. 2014; Remucal 2019; Zareitalabad
et al. 201.i) PI-'AS were first detected in human serum samples in the late 1960s, and subsequent
studies across se\ eral continents demonstrated the global distribution of PFAS in humans (Giesy
and Kannan 2001; Houde et al. 2006). Since then, the global distribution of PFAS in tissues of
aquatic species has been demonstrated in studies conducted in freshwater and marine
environments across every continent, including remote regions far from direct sources, such as
the high arctic, Antarctica, and oceanic islands (Giesy and Kannan 2001; Houde et al. 2006).

39


-------
In lentic surface waters of the U.S., one of the most comprehensive studies of PFOS
concentrations included fish muscle tissue data from 157 near shore sites across the Great Lakes
selected following a probabilistic design as part of the 2010 National Coastal Condition
Assessment (Stahl et al. 2014). In this study, PFOS was measured in fish collected at every site,
with a median concentration of 15.2 ng/g ww (Stahl et al. 2014). Lake trout (31% of sampled
species), smallmouth bass (14%), and walleye (13%) were the most commonly-sampled species
from the Great Lakes samples, and the average PFOS concentrations in lake trout muscle were
more than twice as high as PFOS concentrations in muscle of smallmouth hass and walleye
(Stahl et al. 2014).

Martin et al. (2004) measured PFOS in whole body samples of invertebrates and fish in
Lake Ontario, near the town of Niagara-on-the-l .ake PFOS concentrations were much higher in
the benthic amphipod Diporeia hoya (280 ng g ww ) than in the more pelagic Mysis relicta (13
ng/g ww), suggesting sediments are an important source of PI OS in this area (Martin et al.
2004). Among 1he lour fish species sampled, whole body PFOS concentrations were highest in
the slimy sculpin (45<) nu g w w ). w hose preferred food source is D. hoya (Martin et al. 2004).
Although adult lake trout occupy the highest trophic level at this site, based on nitrogen stable
isotope analysis, their PFOS concentrations were less than half (170 ng/g ww) of those measured
in sculpin. as their food web is largely pelagic, and not affected by the high sediment PFOS
concentrations. Based on stomach content analysis, 90% of the adult lake trout diet consists of
alewife, which feed primarily on the more pelagic M relicta, and have the lowest average PFOS
concentration (46 ng/g) among all fish species (Martin et al. 2004).

Guo et al. (2012) measured PFOS in lake trout muscle tissues in Canadian waters of Lake
Superior, Huron, Erie, and Ontario. Average PFOS concentrations correlated with watershed

40


-------
urbanization, and were 0.85, 8.3, 27, and 46 ng/g ww, respectively (Guo et al. 2012.). Delinsky
et al. (2010) measured PFOS in bluegill, black crappie, and pumpkinseed muscle tissue in 59
lakes in Minnesota, including four lakes in the Minneapolis-St. Paul metropolitan area. PFOS
was detected in muscle tissues of fish collected in 13 of the 59 lakes, and concentrations ranged
from 1.08 to 52.4 ng/g ww in lakes where it was detected. In the four lakes in the Minneapolis-
St. Paul metropolitan area, PFOS concentrations in fish muscle tissues ranged from 4.39 to 47.3
ng/g ww (Delinsky et al. 2010).

In flowing surface waters of the U.S., one of the most comprehensi\ e studies of PFOS
concentrations included fish muscle tissue data from 164 urban river sites (5th order or higher)
across the conterminous U.S. selected following a probabilistic design, as part of the 2008 - 2009
National Rivers and Streams Assessment and the National Coastal Condition Assessment (Stahl
et al. 2014). PFOS was detected in 73% of the urban river sites, with a median concentration of
10.7 ng/g (Stahl et al 2< > 14) l.aruemouth bass (34% of sampled species), smallmouth bass
(25%), and channel catfish (II"..) were the most commonly sampled species from the urban
stream sites, and PI-OS concentrations in the muscles of largemouth bass were approximately
twice as high as concentrations in the muscles of smallmouth bass (Stahl et al. 2014).

Ye et al (2008) reported a\eragePFOS concentrations of 83.1, 84.6, and 147 ng/g from
whole body composite samples of multiple fish species from the Mississippi River, Missouri
River, and Ohio Ri\ er. respectively. Delinsky et al. (2010) sampled PFOS in bluegill, black
crappie, and pumpkinseed muscle tissue at several locations along the upper Mississippi River in
2007, and found concentrations ranging from 3.06 ng/g at unimpacted sites to 2,000 ng/g at Pool
2, a heavily impacted site in the Minneapolis-St. Paul metropolitan area (Delinsky et al. 2010).
Malinsky et al. (2011), as reported in Stahl et al. (2014), measured PFOS concentrations ranging

41


-------
from 41.7 to 180 ng/g in fish muscle samples collected along the Mississippi River, with the
lowest concentration reported for sauger and the highest reported for bluegill.

Kannan et al. (2005) measured PFOS in invertebrates and vertebrates from two rivers in
Southern Michigan (Raisin River, St. Claire River), and one in Northern Indiana (Calumet
River). PFOS concentrations were similar across sites for the different taxa and suggested trophic
biomagnification for PFOS. Among invertebrate taxa, zebra mussel PI-OS soft tissue whole body
concentrations ranged from below detection to 3.1 ng/g ww, amphipod u hole body
concentrations ranged from below detection to 2.9 ng/g ww, and crayfish whole body
concentrations ranged from 2.4 to 4.3 ng/g ww. Among fish. PFOS concentrations in round goby
whole body samples ranged from 6.6 to 21.5 ng/g ww, and small mouth bass muscle samples
ranged from below detection to 41.3 ng/g ww (Kannan et al 2<)<)5)

In a more recent study, Penland el al (2<)2
-------
Florida. Charleston Harbor was the more developed of the two sites and had higher overall PFOS
concentrations. Average PFOS concentrations in Charleston Harbor ranged from 19 ng/g in
pinfish to 92 ng/g in spot. In Sarasota Bay, PFOS concentrations averaged 0.2 ng/g in
zooplankton, and ranged from 3.1 ng/g in pigfish to 8.8 ng/g in spotted seatrout, suggesting
evidence of trophic biomagnification.

Lescord et al. (2015) measured PFOS in chironomids, zooplankton, and juvenile and
adult arctic char in six high arctic lakes in Canada. Two of these lakes had been contaminated by
PFAS from a nearby airport while the other lakes were free from point source contamination.
PFOS in chironomid whole body samples was high at the two contaminated lakes, ranging from
28 to 445 ng/g ww, compared to 5.3 to 14 ng/g ww at the reference lakes, indicating the
importance of sediments as a route of exposure into the base of llie food web (Lescord et al.
2015). Whole body concentrations in pelagic zoopWinkion were relatively lower, ranging from 49
to 60 ng/g ww, compared lo <> 12 lo 2.0 ng/g ww at the reference lakes. PFOS in whole body
samples of juvenile char (ISI lo 224 nu u ww ) and muscle tissue of adult char (24 to 117 ng/g
ww) at the two conlaniinaled Wikes were lower lhan whole body PFOS in chironomids, indicating
alack of trophic Momaunilicalion Additionally, PFOS in whole body samples of juvenile char
(0.001 to 15 nu u ww) and muscle tissues of adult char (below detection to 2 ng/g ww) at the
four reference lakes was also lower than whole body PFOS in chironomids at the four reference
lakes.

Tomy et al. (2004) measured PFOS in whole body samples of zooplankton (Calamus
hyperboreus), shrimp (Pandalus sp.), clams (Nya truncata and Serripes groenlandica), and arctic
cod (Boreogadus saida)\ and liver samples of deepwater redfish (Sebastes mentella) collected
from unimpacted marine locations in the Canadian Arctic. PFOS concentrations were low for all

43


-------
taxa, with the lowest concentrations measured in shrimp (0.3 ng/g ww) and clams (0.04 ng/g
ww). PFOS concentrations were similar in zooplankton (1.8 ng/g ww), arctic cod (1.3 ng/g ww),
and redfish (1.4 ng/g ww), indicating little, if any biomagnification from invertebrates to fish
(Tomy et al. 2004). Haukas et al. (2007) found the average liver PFOS concentration (2.02 ng/g
ww) in arctic cod B. saida collected in the Barents Sea off the coast of Svalbard in 2004 to be
similar to whole body concentrations for this species reported by Tomy et al. (2004). The
average whole body PFOS concentration (3.85 ng/g ww) in ice amphipod ((iainmarus wilkitzkii)
samples was higher than the average liver PFOS concentration in in arctic cod. indicating no
biomagnification from invertebrates to fish in this ecosystem (Haukas et al. 2007)

Current data indicate that PFOS concentrations measured in aquatic biota vary widely,
approximately across four orders of magnitude lor both fish (ranging between 0.85 and 2,000
ng/g ww) and aquatic invertebrates (ranging between n 04 and 445 ng/g ww). Like ambient
surface water concentrations. PI-OS concentrations in aquatic biota inhabiting remote areas (i.e.,
areas with little to no PI AS manufacturing and or industrial uses) appear to be lower than those
in areas with known PI AS manufacturing, industrial use, and/or application of AFFF. While
current PI AS monitoring data illustrate the prevalence and quantify concentrations of PFOS in
aquatic biota across the U.S . additional data are needed to better understand PFOS occurrence
and potential bioaccumulalion in aquatic ecosystems across the U.S.

2.6 Exposure Pathways of PFOS in Aquatic Environments

There are multiple exposure pathways of PFOS in the aquatic environment, including: 1)

direct (dermal and respiratory) aqueous exposure; 2) direct exposure to contaminated sediment
(for benthic organisms); 3) dietary and biomagnification; and 4) maternal-transfer (Ankley et al.
2020). Exposure of PFOS through water and sediment occurs through direct contact with the

44


-------
respective media, such as water passing across the gills, or consumption of suspended and
deposited sediments (Prosser et al. 2016). Upon entering an organism, PFAS such as PFOS tend
to bind to proteins, and concentrate preferentially within the blood and protein rich tissues, such
as liver (Haukas et al. 2007; Xia et al. 2013). The affinity of PFOS to bind to proteins contributes
to the bioaccumulation and biomagnification of PFOS (see Section 2 5 above), resulting in
increasing concentrations of PFOS in the diets of higher trophic le\ el organisms, such as
predatory fish and birds (Custer et al. 2019; Haukas et al. 2007; Xu et al. 2d 14). However, as
noted previously in Section 2.2.1, the lack of a meaningful K.™ for PFOS due lo its binding
primarily to protein, not lipids, precludes application of ICv-based models that are commonly
used to estimate bioconcentration factors and predict bioaccumulation for many other important,
environmental contaminants (e.g., PCBs) l.aslly. elevated PI-OS concentrations in eggs and
young of aquatic life suggests that PFOS may be maternally transferred to offspring. This
exposure pathway may be particularly important among egg-laying species because of the
preferential binding of PI OS to egg albumin (Kannan et al. 2005). In summary, PFOS exposure
has been found to occur through multiple exposure routes, including via water, sediment, diet,
and maternal transfer (Jones et al 2<)i)3. Kannan et al. 2005; Sharpe et al. 2010; Wang et al.
2011)

2.7 Effects of PFOS on Biota

The number of PI OS ecotoxicity studies and data are increasing and study designs are

evolving to expand the understanding of the effects of PFOS. Currently, PFOS ecotoxicity
studies are primarily focused on fish, aquatic invertebrates, plants, and algae. Fewer studies are
being conducted on aquatic-dependent birds, reptiles, and mammals. Sections 3 and 4 provide
study summaries of individual publicly available high quality aquatic life toxicity studies, and

45


-------
Appendices A through H summarize current PFOS aquatic life ecotoxicity data, both studies
used here and unused studies due to quality issues.

2.7.1 Mode of Action and Toxicity of PFOS

The mechanism(s) underpinning the toxicity of PFOS is not well-understood and is an

active area of research. Toxicity literature indicate that PFOS causes a wide range of adverse

effects in aquatic organisms, including reproductive effects, de\ elopmental toxicity, and

estrogen, androgen and thyroid hormone disruption (see Sections 3 and 4 and Appendices A.l

through H.l). However, a great deal of research is still needed from a mechanistic perspective to

better understand how the different modes of action elicit specific biological responses. Some

potential PFOS modes of action in aquatic life appear to include: 1) oxidative stress (Li et al.

2017; Sant et al. 2018; Shi and Zhou 2010); 2) autophagic cell death or apoptosis (Sant et al.

2018; Shi et al. 2008); 3) endocrine modulation of estrogen and thyroid receptors (Benninghoff

et al. 2011; Chen el al 2D 18. Dn el al 2D 13. Kim el al. 2d I I. Shi et al. 2008); 4) interference at

the mitochondrial le\el through the uncoupling of oxidative phosphorylation (ECCC 2018); 5)

interference with the homeostasis of l)Y\ metabolism (Hoff et al. 2003); and 6) activation of the

nuclear peroxisome proliterator actuated receptor-alpha (PPAR-a) pathways (Arukwe and

Mortensen 2d I I. Cheng el al 2010, Fang el al. 2013; Fang et al. 2012; Yang et al. 2014).

Following exposure lo PFOS, molecular level events can perturb estrogen-, androgen-

and thyroid-related endocrine systems, as well as neuronal-, lipid-, and carbohydrate-metabolic

systems and lead to cellular- and organ-level disturbances and ultimately result in effects on

reproduction, growth, and development at the individual organism-level (see Ankley et al. 2020

and Lee et al. 2020 for the latest reviews on the subject). The mechanisms of PFOS toxicity to

fish in particular appear to be related to oxidative stress, apoptosis, thyroid disruption, and

alterations of gene expression during development (Lee et al. 2020). Additionally, published

46


-------
research suggested that many of these molecular pathways interact with each other and could be
linked. For example, oxidative stress following exposure to PFOS was correlated with effects on
egg hatching and larval formation, linking reproductive toxicity, oxidative stress, and
developmental toxicity (Lee et al. 2020). The actual mechanism(s) through which PFOS induced
oxidative stress operates still requires additional study, but increased B-oxidation of fatty acids
and mitochondrial toxicity are proposed triggers (Ankley et al 2<)2<). I ,ee et al. 2020). Thus, the
alteration of multiple biological pathways is a plausible explanation lor the di\ ersity of observed
effects of PFOS reported in the literature (Lee et al. 2020). However, the a\ ailahlc data did not
allow for a defined adverse outcome pathway-based understanding of the ultimate reductions to
survival, growth, and reproduction in the various aquatic taxa in which these effects have been
observed or may be expected to occur. Thus. further mechanistic research is warranted.

Notably, PFOS appeared to be related to the disruption of the sex hormone-related
endocrine system at the molecular, tissue, and organ le\els. resulting in observed adverse
reproductive outcomes in freshwater and saltwater fish and invertebrates alike. Further, these
effects have been reported after exposure \ ia multiple exposure routes (i.e., waterborne, dietary,
maternal. I .ee et al. 2<)2<)) And these reproducti\ e effects also appeared to be trans-generational,
as obser\ ed in a multi-generational /chratish (pernio rerio) study by Wang et al. (2011) - see
study summary in Section .1 I 1.3.4).

PFOS is one of the most studied PFAS in the ecotoxicity literature, with reported adverse
effects on survival, growth, and reproduction. However, a great deal of additional research is
needed to better understand the modes of action of PFOS. Specifically, additional research from
a mechanistic perspective is needed to better understand how the different modes of action elicit
specific biological responses in fish, aquatic invertebrates, and amphibians. Potential effects of

47


-------
PFOS involving multiple biological pathways is a research challenge for PFOS and PFAS in
general.

2.7.2 Potential for Interactions with Other PFAS

PFAS occur as mixtures in the environment. Occurrence studies document the presence

of complex mixtures of PFAS in surface waters in the U.S. and across the globe (see also Section

2.4; Ahrens 2011; Ahrens and Bundschuh 2014; Giesy and Kannan 2<)02; Houde et al. 2006;

Keiter et al. 2012; Wang et al. 2017). Although EPA's PFOS recommended aquatic life water

quality criteria are based solely on single chemical exposures to aquatic lile. il is recognized that

PFAS are often introduced into the aquatic environment as end-use formulations comprised of

mixtures of PFAS and/or PFAS-precursors. However, the ecological effects of these potential

PFAS mixtures are poorly understood (Ankley el al. 2020). Il was useful, therefore, to briefly

summarize the types of interactions that might be expected based on the few PFAS mixture

studies involving PFOS and one or more PI AS to date. It should be noted that for purposes of

this document, the reader is referred to Ankley et al. (2020) and elsewhere for more

comprehensive re\ iews of PI AS mixtures in general, and the challenges they are expected to

present in ecological risk assessment I'indings of the studies are as reported by the study authors

without any additional interpretation or analysis of uncertainty.

At both the organismal and cellular levels, studies on zebrafish (D. rerio; Ding et al.

2013), a water flea (/ > magnet, Yang et al. 2019), mosquito (Aedes aegyptii, Olson 2017), a

bioluminescent cyanobacteria (Anabaena sp.; Rodea-Palomares et al. 2012), or with cultured

hepatocytes of the cyprinid, Gobiocypris rarus (Wei et al. 2009), demonstrated that the effects

observed from in vivo and in vitro tests on PFAS mixtures vary and can have unpredictable

exposure and species-specific effects. For example, in a single in vivo exposure of zebrafish (I).

rerio) embryos, synergism, additivity and antagonism were all reported for different

48


-------
combinations/ratios of PFOS and PFOA and endpoints (Ding et al. 2013), illustrating the
complexity and uncertainty associated with mixture studies. Importantly, neither the
concentration addition model nor the independent-action model could predict the combined
effects when strong interactive effects existed. More recently, Yang et al. (2019) exposed the
water flea, D. magna, to single and binary mixtures of PFOS and PFOA. The authors reported
synergism in acute and chronic toxic effects. Conversely, Rodea-Palomares et al. (2012) showed
binary PFOS and PFOA mixture as having an antagonistic interaction across the whole range of
effect levels tested using the bioluminescent cyanobacterium, Anabaena. Olson (2017) exposed
larvae of the mosquito, A. aegypti, to PFOS and perfluorohexane sulfonate (PI I l\S) separately
and as a mixture and reported increased toxicity in a manner greater than would be predicted by
additivity. At both the organismal and cellular le\ els. studies on zebra fish (I). rerio; Ding et al.
2013), a water flea (D. magna; Yang et al. 2d I 'ฆ)). a mosquito (. \edcs aegyptii\ Olson 2017), a
bioluminescent cyanobacterium (. \nabaena sp.\ Rodea-Palomares et al. 2012), or with cultured
hepatocytes of the c\ prinicl. (iubmnpris ranis (Wei et al. 2009), demonstrated that the effects
observed from in vivo and in viim tests on PI AS mixtures vary and can have unpredictable
exposure and species-specilic effects I-'or example, in a single in vivo exposure of zebrafish (I),
rerio) embryos, synergism, additivity and antagonism were all reported for different
combinations ratios of PFOS and PFOA and endpoints (Ding et al. 2013), illustrating the
complexity and uncertainly associated with mixture studies. Importantly, neither the
concentration addition model nor the independent-action model could predict the combined
effects when strong interactive effects existed. More recently, Yang et al. (2019) exposed the
water flea, D. magna, to single and binary mixtures of PFOS and PFOA. The authors reported
synergism in acute and chronic toxic effects. Conversely, Rodea-Palomares et al. (2012) showed

49


-------
a binary PFOS and PFOA mixture as having an antagonistic interaction across the whole range
of effect levels tested using the bioluminescent cyanobacterium, Anabaena. Olson (2017)
exposed larvae of the mosquito, A. aegypti, to PFOS and perfluorohexane sulfonate (PFHxS)
separately and as a mixture and reported increased toxicity in a manner greater than would be
predicted by additivity.

In tests with cultured hepatocytes of the cyprinid, G. rams, co-exposure of PFOS with a
mixture of five other PFAS [PFOA, Perfluorononanoate or Perfluorononanoic acid (PFNA),
Perfluorodecanoate or Perfluorodecanoic acid (PFDA), Perfluorododecanoale or
Perfluorododecanoic acid (PFDoA), and 8:2 FTOH] resulted in highly complex interactions (Wei
et al. 2009). A number of genes differentially expressed in the mixture were not differentially
expressed in the exposure to the individual chemicals, potentially indicating different modes of
action for the mixture compared to the indi\ idual chemicals Tn (his case, the authors reported no
additive responses lor the mixture Consistent with the possible mechanisms of toxicity of PFOS
(see Section 2 7 1). the genes identified in the study are involved in multiple biological functions
and processes, including fully acid metabolism and transport, xenobiotic metabolism, immune
response, and oxidali\ e stress (Wei el al 2009). Finally, U.S. EPA (2021, unpublished) observed
PFOA and PI-OS interacting in an additive manner to reduce pup body weight, pup liver weight,
and maternal li\ er weight in the Sprague-Dawley rat.

2.8 Conceptual Model of PFOS in the Aquatic Environment and Effects

A conceptual model depicts the relationship between a chemical stressor and ecological

compartments, linking exposure characteristics to ecological endpoints. The conceptual model
provided in Figure 2-5 summarizes sources, potential pathways of PFOS exposure for aquatic
life and aquatic-dependent wildlife, and possible toxicological effects.

50


-------
PFOS initially enters the aquatic environment through point sources, including municipal
and industrial dischargers and landfill leachate and non-point sources, including land application
of contaminated biosolids (see Section 2.1.2). PFOS enters the aquatic environment in dissolved
and particle-bound forms and may sorb to surfaces, such as sediment and particulate matter in
the water column (see Section 2.2 and 2.2.2), which is depicted in the conceptual model (Figure
2-5). The conceptual model depicted in Figure 2-5 shows exposure pathways for the biological
receptors of concern (i.e., aquatic life) and potential effects (e.g., on sui\ i\al. growth, and
reproduction) in those receptors. Both direct (i.e.. exposure from the water column which is
represented by **) and indirect (i.e., dietary exposure \ ia the food web *) pathways are
represented in the conceptual model.

51


-------
aii

a

u
I*-*

U

PFOS Source

Point Sources
(from landfill leachate, municipal &
industrial dischargers, and
applications such as surfactants,
textile stain, & soil
repellents )

PFOS id Water

Dissolved & Particle-Bound

Degradation &
Metabolism of
Other PFASs

PFOS iu Sediment

PFOS Source

Nonpoint Sources
(from Aerial deposition AFFF,
land application of paper residuals
contaminated biosolids)



Aquatic Life

Producers

l:t Trophic Transfer

(from phytoplankton, periphyton, macrophytes; e.g., algae, cyanobacteria, waterweed'c ommon eelgrass)

ir
o

cL

ai

0

it

01

Consumers

2nd Trophic Transfer
(to zooplankton. macroinvertebrates:
e.g., cladocerans'copepods &
mayflies/ribbed mussels)



Consumers

3rd Trophic Transfer
(to predatory fish:
e.g., longnose. dace/
American shad)

	*ฆ

Consumers

4th Trophic Transfer

(to predatory fish;
e.g., largemouth bass'
striped bass)

	~



Figure 2-5. Conceptual Model Diagram of Sources, Compartmental Partitioning, and Trophic
Transfer Pathways of Perfluorooctane Sulfonate (PFOS) in the Aquatic Environment and its
Bioaccumulation and Effects in Aquatic Life.

PFOS sources represented in ovals, compartments within the aquatic ecosystem represented by rectangles, and effects in
pentagons. Examples of organisms in each trophic transfer provided as freshwater/marine. Movement of PFOS from water to
receptors indicated by two separate pathways: bioconcentration by producers (*) and direct exposure to all trophic levels within
box (**). Relative proportion of PFOS transferred between each trophic level is dependent on life history characteristics of each
organism.

52


-------
2.9 Assessment Endpoints

Assessment endpoints are defined as "explicit expressions of the actual environmental

value that is to be protected" and are defined by an ecological entity (species, community, or
other entity) and its attribute or characteristics (U.S.EPA 1998). Assessment endpoints may be
identified at any level of organization (e.g., individual, population, community). In the context of
the CWA, aquatic life criteria for toxic pollutants are typically determined based on the results of
toxicity tests with aquatic organisms in which unacceptable effects on growth. reproduction, or
survival occurred. This information is typically compiled into a sensitivity clisirihulion based on
genera and representing the impact on taxa across the aquatic community. Criteria are based on
the 5th percentile of genera and are thus intended to be prolecli\ e of approximately 95 percent of
aquatic genera.

The use of laboratory toxicity tests to protect aquatic species was based on the concept
that effects occurring to a species in appropriate laboratory tests will generally occur to the same
species in comparable field situations. Since aquatic ecosystems are complex and diversified, the
1985 Guidelines recommended acceptable data be available for at least eight genera with a
specified taxonomic di\ ersity (the standard eight-family minimum data requirements, or MDRs).
The i ntent of the eight-family \1DR was to serve as a surrogate sample community
representati\ e of the larger and generally much more diverse natural aquatic community, not
necessarily the most sensitive species in a given environment. The 1985 Guidelines note that
since aquatic ecosystems can tolerate some stress and occasional adverse effects, protection of all
species at all times and places are not deemed necessary (the intent is to protect 95 percent of a
group of diverse taxa, and any commercially and recreationally important species; U.S.EPA
1985).

53


-------
For more details on aquatic life assessment endpoints for PFOS see Section 3.1 below.
This criteria derivation for aquatic life was developed using a genus sensitivity distribution
(GSD), which represents the potential for impact to the survival, growth, or reproductive effects
on taxa across aquatic communities.

2.10 Measurement Endpoints

2.10,1 Overview of Toxicity Data Requirements

To ensure the protection of various components of an aquatic ecosystem, EPA compiles

acute toxicity test data from a minimum of eight diverse taxonomic groups.

•	Acute freshwater criterion require data from the following taxonomic groups

a.	fish in the family Salmonidae in the class Osteichthyes

b.	a second family of fish in the class Osteichlhyes, preferably a commercially or
recreationally important v.armualer species (e.g . Muegill, channel catfish)

c.	a third family in the phylum Choidala (may be in the class Osteichthyes or may
be an amphibian)

d.	a planktonic crustacean (e.g., cladoceran. copepod)

e.	a benthic crustacean (e.g., ostracod, isopod, amphipod, crayfish)

f.	an insect (eg. mayfly. dragonfly, damselfly, stonefly, caddisfly, mosquito,
midge)

g.	a family in a phylum other than Arthropoda or Chordata (e.g., Rotifera, Annelida,
Mollusca)

h a family in any ol der of insect or any phylum not already represented

•	Acute estuarine/marine criterion require data from the following taxonomic groups:

a two families in the phylum Chordata

b.	a family in a phylum other than Arthropoda or Chordata

c.	a fami ly from either Mysidae or Penaeidae

d.	three other families not in the phylum Chordata (may include Mysidae or
Penaeidae, whichever was not used above)

e.	any other family

Additionally, to ensure the protection of various animal components of the aquatic
ecosystem from long term exposures, chronic toxicity test data are recommended from the same
eight diverse taxonomic groups that are recommended for acute criteria. If the eight diverse

54


-------
taxonomic groups are not available to support the chronic criterion derivation using a genus
distribution approach, the chronic criterion may be derived using an acute-to-chronic ratio
(ACR) approach. To apply an ACR approach to derive a chronic criterion, a minimum of three
taxa are recommended, with at least one chronic test being from an acutely sensitive species. To
calculate ACRs, chronic aquatic life criteria require data from the following taxonomic groups:

a.	At least one fish

b.	At least one invertebrate

c.	At least one acutely sensitive freshwater species, for freshwater chronic criterion
(the other two may be saltwater species)

d.	At least one acutely sensitive saltwater species for estuarine murine chronic
criterion (the other two may be freshwater species)

The 1985 Guidelines also specified at least one quantitative test with a freshwater alga or
vascular plant. If plants are among the mosl sensitive aquatic organisms, toxicity test data from a
plant in another phylum should also be a\ ailaMe Aquatic plant toxicity data were examined to
determine whether aquatic plants are likely to be adversely affected hy the concentration
expected to be proteetiv e lor other aquatic organisms. Available data for aquatic plants and algae
were reviewed to determine if they were more sensitive to PFOS than aquatic animals (see
Appendices A. (' and I- for freshwater species)

2.1".J Measure of PFOS l-xnosure Concentrations

This PI OS aquatic life ambient water quality criteria document provides a critical review

of all data identified in NWs literature search for PFOS, including all forms of PFOS used in

toxicity literature (such as the anionic form and salts) and identified in the ECOTOX database:

•	the anionic form (CAS No. 45298-90-6)

•	the acid form (CAS No. 1763-23-1)

•	potassium salt (CAS No. 2795-39-3)

•	an ammonium salt (CAS No. 56773-42-3)

•	sodium salt (CAS No. 4021-47-0)

•	and a lithium salt (CAS No. 29457-72-5)

55


-------
Based on EPA's data review, PFOS toxicity studies typically used the linear PFOS
isomer for dosing with fewer studies using the branched isomer. Studies that conducted PFOS-
only exposures were considered for possible inclusion. For most EPA aquatic life criteria
documents with non-bioaccumulative substances, organisms are exposed to contaminated water
but fed a diet grown in uncontaminated media (not spiked with the toxicant prior to introduction
into the exposure chambers). Such tests were reviewed, and tesls of sufficient quality are
included in this PFOS criteria. Toxicity tests conducted with PFOS-spiked did were also
reviewed and considered suitable for deriving a criterion for this bioaccumulati\ e pollutant;
however, these toxicity tests were limited in the current PFOS toxicity literature Consequently,
toxicity tests with direct aqueous, dietary, and maternal transfer were included in EPA's
derivation of aquatic life criterion for PFOS (see Section 3). Studies not included in the numeric
criteria derivation, including some studies with oilier PI OS exposures (i.e., in vitro studies),
were considered qualitati\ ely as supporting information if lliey were deemed to be of sufficient
quality, and are described in llie I-fleets Characterization section below (Section 4.4).

This set of published literature was identified using the ECOTOXicology database
(ECOTOX.	') as meeting data quality standards. ECOTOX is a

source of high-quality toxicity data for aquatic life, terrestrial plants, and wildlife. The database
was created and is maintained by the EPA, Office of Research and Development, Center for
Computational Toxicology and Exposure. The ECOTOX search generally begins with a
comprehensive chemical-specific literature search of the open literature conducted according to
ECOTOX Standard Operating Procedures (SOPs; Elonen 2020). The search terms are often
comprised of chemical terms, synonyms, degradates and verified Chemical Abstracts Service
(CAS) numbers. After developing the literature search strategy, ECOTOX curators conduct a

56


-------
series of searches, identify potentially applicable studies based on title and abstract, acquire
potentially applicable studies, and then apply the applicability criteria for inclusion in ECOTOX.
Applicability criteria for inclusion into ECOTOX generally include:

a.	The toxic effects are related to single chemical exposure (unless the study is being
considered as part of a mixture effects assessment)

b.	There is a biological effect on live, whole organisms or hi vitro preparation including
gene chips or omics data on adverse outcome pathways potentially of interest

c.	Chemical test concentrations are reported

d.	There is an explicit duration of exposure

e.	Toxicology information that is relevant to OW is reported lor the chemical of concern

f.	The paper is published in the English language

g.	The paper is available as a full article (not an abstract)

h.	The paper is publicly available

i.	The paper is the primary source of the data

j. A calculated endpoint is reported or can be calculated using reported or available
information

k. Treatment(s) are compared to an acceptable control

1. The location of the study (e.g., laboratory vs. field) is reported

m. The tested species is reported (with recognized nomenclature)

Following inclusion in the l-COTOX database, toxicity studies were subsequently
evaluated by the Office of \Y'titer All studies were evaluated for data quality as described by
U.S.EPA (1985) and in NWs Office of Chemical Safety and Pollution Prevention (OCSPP)'s
Ecological I-fleets Test (inidelines (IS N\\ 2<)|6c), and EPA OW's internal data quality SOP,
which is consistent with OCSPP's data quality review approach (U.S.EPA 2018). Office of
Water completed a Data E\ til nation Record (DER) for each species by chemical combination
from the PFOS studies identified by ECOTOX. This in-depth review ensured the studies used to
derive the criteria resulted in robust scientifically defensible criteria. Example DERs are shown
in Appendix R with the intent to convey the meticulous level of evaluation, review, and
documentation each PFOS study identified by ECOTOX was subject to.

57


-------
The 1985 Guidelines document indicates that tests used in criteria should be for North
American resident species. Due to EPA's interest in using all available quality data, particularly
for data-sparse PFOS (relative to cadmium or ammonia, for example), PFOS toxicity studies
were considered for possible inclusion regardless of the test species residential status in North
America, as with other published aquatic life criteria. This approach was also based on the
relative similarity in sensitivities between resident and non-resident species (see Sections 3 and
4). Moreover, non-North American resident species serve as taxonomicallvฆ-related surrogate test
organisms for the thousands of untested resident species. Supporting analyses lo evaluate the
influence of including non-resident species on the freshwater criteria magnitudes were conducted
by limiting toxicity datasets to North American resident species with established populations in
North America (see Section 4.3). These supporting analyses pn>\ ided an additional line-of-
evidence that further suggested it is appropriate to consider nonresident species in PFOS criteria
derivation because of their minimal influence of the criteria magnitudes.

Additionally, a substantial number of PFOS toxicity tests reported only nominal, or
unmeasured. PI-OS concentrations Therefore. I-PA examined whether nominal and measured
concentrations lor PI OS are typically in close agreement with each other. Among the PFOS
studies that were used quantitatively (Sections 3.1.1.1 and 3.1.1.3 and Appendices A.l andC.l)
and qualitath el v (Section 4 4 and Appendix G) in the freshwater water column-based criteria, 65
freshwater studies had measured concentrations, yielding 477 pairs of measured and nominal
concentrations (excluding controls, where PFOS was rarely detected). Furthermore, there were
20 estuarine/marine studies with measured concentrations, yielding 171 pairs of measured and
nominal concentrations. The data were grouped by classifications including water type
(salt/fresh) and experimental conditions (acute/chronic; solvent/no solvent; fed/unfed, etc.). Data

58


-------
displayed a high degree of linear correlation, and the measured and nominal concentrations were
in close agreement. Details of this meta-analysis can be found below in Appendix O.

Therefore, when available, measured PFOS concentrations were used; however, for
several studies measured PFOS concentrations were not reported, and nominal concentrations
were utilized, especially if a concentration-response relationship was observed in another media
(e.g., blood or eggs). Typically, per the 1985 Guidelines, acule toxicity data from all measured
flow-through tests would be used to calculate species mean acute \ allies (S\l A V), unless data
from a measured flow-through test were unavailable, in which case the acute criterion would be
calculated as the geometric mean of all the available acute values (i.e., results of unmeasured
flow-through tests and results of measured and unmeasured static and renewal tests). Chronic
unmeasured flow-through tests, as well as measured and unmeasured static and renewal tests are
not typically considered to calculate chronic \ allies In the case of PFOS, static, renewal, and
flow-through experiments were considered lor possible inclusion for both species mean acute
and chronic values regardless if PI OS concentrations were measured because PFOS is a highly
stable compound (see Section I 2 I). resistant to hydrolysis, photolysis, volatilization, and
biodeuradation (see Section 2 .V (iiesy et al 2010).

Additionally, chronic \ alues were based on endpoints and durations of exposure that
were appropriate to the species Thus, both life- and partial life-cycle tests were utilized for the
derivation of the chronic criteria. However, it should be noted that typically, per the 1985
Guidelines, life-cycle chronic tests would be preferred for invertebrates. The chronic studies used
in the derivation of the PFOS criteria followed taxa specific exposure duration requirements
from various test guidelines (i.e., EPA's 1985 Guidelines and EPA's OCSPP's Ecological
Effects Test Guidelines, U.S. EPA 2016c) when available. For example, EPA's 1985 Guidelines

59


-------
states that daphnid tests should begin with young < 24 hours old and last for not less than 21
days; and this chronic test duration was applied to the consideration of all chronic daphnid tests.
When taxa-specific exposure duration requirements were not available for a particular test
organism in the PFOS toxicity literature, both life- and partial life-cycle tests were considered in
the derivation of the chronic criteria.

PFOS toxicity in aquatic life can be manifested as effects on survival, growth, and/or
reproduction. Measurements of fish tissue, such as whole-body, muscle, and eggs, were most
closely linked to the chronic adverse effects of PFOS, since PFOS is highly persistent and
bioaccumulative. The following subsection of this problem formulation describes the approaches
used to establish PFOS effect concentrations in aquatic life and to relate the various criteria
derived, including for water and tissue.

2,10,3 Measures of Effect

Each assessment endpoinl requires one or more "measures of ecological effect," which

are defined as changes in the attributes of an assessment endpoint itself or changes in a surrogate

entity or attribute in response to chemical exposure. Ecological effects toxicity test data are used

as measures of direct and indirect effects to growth, reproduction, and survival of aquatic

organisms

2,10.3,1 Acute Measures of I-licet

The acute measures of effect on aquatic organisms are the lethal concentration (LCso),

effect concentration (ECso), or inhibitory concentration (ICso) estimated to produce a specific

effect in 50 percent of the test organisms (Table 2-1). LCso is the concentration of a chemical that

is estimated to kill (or immobilize) 50 percent of the test organisms. EC so is the concentration of

a chemical that is estimated to produce a specific effect in 50 percent of the test organisms. The

IC50 is the concentration of a chemical that is estimated to inhibit some biological process (e.g.,

60


-------
enzyme activity associated with an apical endpoint such as mortality) in 50 percent of the test
organisms.

2.10.3.2 Chronic Measures of Effect

The measure of effect for chronic exposures of PFOS was the effect concentration

estimated to produce a chronic effect on survival, growth, or reproduction in 10 percent of the

test organisms (ECio; Table 2-1). EPA selected an ECio to estimate a low level of effect that

would be both different from controls and not expected to be severe enough io cause severe

effects at the population level for a bioaccumulative contaminant, such as PTOS Tlie use of the

ECio, instead of anEC2o, is also consistent with the use of this metric for the bioaccumulative

pollutant selenium in the recent 2016 Selenium Freshwater Aquatic Life Criteria (U.S.EPA

2016a), and is consistent with the harmonized guidelines from OI-CD and the generally preferred

effect level for other countries such as Canada. Australia and New Zealand (CCME 2007; OECD

2001; WarneMSt .1 2<)|X)

Regression analysis was used preferentially to characterize a concentration-response (C-

R) relationship and to estimate concentrations at which chronic effects are expected to occur.

Author-reported \o ()hsei\ ed I-fleet Concentrations (NOECs) and Lowest Observed Effect

Concentrations (I .OLCs) w ere only used for the derivation of chronic criterion when a robust

ECio could not be calculated lor the genus. A NOEC is the highest test concentration at which

none of the obser\ ed effects are statistically different from the control. A LOEC is the lowest test

concentration at which the observed effects are statistically different from the control. When

LOECs and NOECs are used, a Maximum Acceptable Toxicant Concentration (MATC,

geometric mean of the NOEC and LOEC) is calculated. For the calculation of the chronic

criteria, point estimates were selected for use as the measure of effect in favor of MATCs, as

MATCs are highly dependent on the concentrations tested. Point estimates also provided

61


-------
additional information that is difficult to determine with an MATC, such as a measure of effect
level across a range of tested concentrations.

In conformity with the 2013 Ammonia Freshwater Aquatic Life Criteria (U.S.EPA 2013),
a decision rule was also applied to the PFOS toxicity data when an author-reported NOEC or
LOEC was used. The decision rule was not to use "greater than" values for concentrations of low
magnitude or "less than" values for concentrations of high magnitude because they added little
significant information to the analyses. Conversely, if data from studies with only low
concentrations indicated a significant effect (suggesting the test material was highly toxic) or
studies with high concentrations only found an incomplete response for a chronic endpoint
(indicating low toxicity of the test material), those data did significantly enhance the
understanding of PFOS toxicity. Thus, the decision rule was applied as follows: "greater than"
(>) high toxicity values and "less than" (<) low toxicity \ allies were included (U.S.EPA 2013).
Data that met the quality ohjecli\ es and test requirements were utilized quantitatively in deriving
these criteria for aquatic life and are presented in Table 3-3 and Table 3-7.

62


-------
Table 2-1. Summary of Assessment Endpoints and Measures of Effect Used in the Criteria
Derivation for PFOS.

Assessment Kndpoinls for (lie
Aquatic (011111111 nily

Measures of KITecl

Aquatic Life: Survival, growth,
and reproduction of freshwater
and estuarine/marine aquatic life
(i.e., fish, amphibians, aquatic
invertebrates)

For effects from acute exposure:

1.	LC50, EC50, or IC50 concentrations in water

2.	NOEC and LOEC concentrations in water

For effects from chronic exposure

1.	EC 10concentrations in water

2.	NOEC and LOEC concentrations in water. Only
used when an F.C1" could not he calculatedfor a
genus.

Note: only chronic exposures were considered for
derivation o f the tissue-based criteria since l'l ( )S is a
bioaccumulative chemical. These chronic tissue-based
criteria are expected to be protective of acute effects,
because acute effects were observed at much greater
concentrations than chrome effects.

LC50 = 50% Lethal Concentration
EC50 = 50% Effect Concentration

IC50 = 50% Inhibitory Concentration
NOEC = No-observed-effect-concentration
LOEC = Lowest-observed-effecl-co nee 11I ration
EC10 = 10% Effect Concentration

2.10.3.3 Summary of Independent Calculation of Toxicity Values

Where data were a\ ailahle. toxicity \ allies, including LC50 and EC10 values, were

independently calculated using data from the toxicity studies meeting the inclusion criteria

described abo\ e. via independent statistical analysis conducted by EPA. Occasionally, individual

replicate-le\ el data or treatment-level data needed to be obtained from the study authors to

independently calculate toxicity values. All data were analyzed using the statistical software

program R (version 3.6.2) and the associated dose-response curve (drc) package. The R drc

package has several models available for modeling a concentration-response relationship for

each toxicity study. The specific model used to calculate toxicity values was selected following

the details provided in Appendix K and the models performed well on most or all statistical

63


-------
metrics. The independently-calculated toxicity values used to derive the PFOS aquatic life
criteria are included in each study summary below and were utilized to derive this criteria for
aquatic life, where available (for acute criterion in Table 3-9 and chronic criterion in Table 3-10).

2.11 Analysis Plan

2.11.1	Derivation of Water Column Criteria

During CWA section 304(a) criteria development, EPA ic\ iew s and considers all

relevant toxicity test data. Information available for all relevant species and genera are reviewed
to identify: 1) data from acceptable tests that meet data quality standards; and 2) u liether the
acceptable data meet the minimum data requirements (MDRs) as outlined in l-l\\"s 1985
Guidelines (U.S.EPA 1985). The MDRs described in Section 2 10.1 were met for acute and
chronic freshwater criteria derivation, ^ ilh the exception that the acute MDR for an insect was
not fulfilled with quantitatively acceptable data (Appendix A). Therefore, qualitatively
acceptable acute insect data (Appendix G) were evaluated relative to the sensitivities of other
species/MDRs. EPA will continue to seek and evaluate acute PFOS insect data to further
evaluate the sensith ily of aquatic insects Acute and chronic MDRs for PFOS estuarine/marine
criteria dei i\ ation were not met Consequently, EPA used the available toxicity data and aNew
Approach Method (NAA1) to generate protective estuarine/marine benchmarks. A minimal
number of tests from acceptable studies of aquatic algae and vascular plants were also available.
The relative sensiti\ ilv of freshwater plants to PFOS exposures indicates plants are less sensitive
than aquatic vertebrates and invertebrates so plant criteria were not developed.

2.11.2	Derivation of Tissue-Based Criteria following Chronic PFOS Exposures
Chronic toxicity studies (both laboratory and field studies) were further screened to

ensure that they contained the relevant chronic PFOS exposure routes for aquatic organisms (i.e.,

dietary, maternal, or dietary and waterborne PFOS exposure), measurement of chronic effects,

64


-------
and measurement of PFOS in tissue(s). EPA considered deriving tissue-based criteria using
empirical toxicity tests with studies that exposed test organisms to PFOS via water, diet, and/or
maternal transfer and reported exposure concentrations based on measured tissue concentrations.
This approach generally corresponded with the 2016 Selenium Aquatic Life Freshwater
Criterion, which is the only other EPA 304(a) recommended aquatic life criterion with tissue-
based criteria (U.S.EPA 2016a). However, currently, the freshwater chronic PFOS toxicity
dataset with measured tissue concentrations is somewhat limited. There were 14 total chronic
aquatic life studies considered, six quantitative (three fish, one invertebrate, and two amphibian
studies) and eight qualitative studies (see Section 4 7). The quantitative studies pro\ ided data for
three of the eight MDRs. The qualitative studies provided supporting information for only one
additional MDR. Therefore, it was concluded that there are currently insufficient data to derive a
chronic tissue criterion using a GSD approach from empirical tissue data from toxicity studies.
Thus, EPA examined a Bioaccumulation factor (BAF) approach for chronic tissue criteria
development.

2,11,3 Translation of Chronic W ater Column Criterion to Tissue Criteria

To cnaMe use of fish tissue measurements of PFOS in protecting designated uses, chronic

tissue criteria for PL'OS were clei i\ ed hy translating the chronic freshwater column criterion

(summarized in Section 2.1 I I above) into tissue criteria using bioaccumulation factors (BAFs)

and the following equation

Tissue Criteria = Chronic Water Column Criterion x BAF	(Eq. 1)

The resulting tissue criteria correspond to the tissue type from the BAF used in the
equation.

65


-------
2.11.3.1 Aquatic Life Bioaccumulation Factors (BAFs)

A BAF is determined from field measurements and is calculated using the equation:

BAF = ฃMฃta	^Eq 2)

Cwater

Where:

Cbiota = PFOS concentration in organismal tissue(s)

Cwater = PFOS concentration in water where the organism i \ as collected

Given that a BAF is determined from field measurements (as opposed lo controlled
experiments designed to measure bioconcentration of PI OS using specific test guidelines;
(OECD 2001; U.S.EPA 2016a), a BAF is an expression of all exposure routes, i e . dietary,
water, maternal transfer, and contact wilh sediments \ in skin and ingestion. Depending upon the
tissue residue measurement, BAFs can be based upon residues in the whole organisms, muscle,
liver, or any other tissue.

The literature search lor reporting on PFOS bioaccumulation was implemented by
developing a series of chemical-based search terms. These terms included chemical names and
Chemical Abstracts Ser\ ice registry numbers (CASRN or CAS3), synonyms, tradenames, and
other rele\ant chemical forms (i e . related compounds). Databases searched were Current
Contents. lYoQuesl CSA, Dissertation Abstracts, Science Direct, Agricola, TOXNET, and
UNIFY (database internal to I S. EPA's ECOTOX database). The literature search yielded
numerous citations and the citation list was further refined by excluding citations on analytical
methods, human health, terrestrial organisms, bacteria, and where PFOS was not a chemical of
study. The citations meeting the search criteria were reviewed for reported BAFs and/or reported
concentrations in which BAFs could be calculated. Data from papers with appropriate

3 Chemical Abstracts Service registry number (CASRN or CAS) for PFOS is 1763-23-1.

66


-------
information were extracted into a PFOS dataset. The studies meeting these inclusion criteria
were also screened for data quality.

Four factors were evaluated in the screening of the BAF literature: 1) number of water
samples; 2) number of organism samples; 3) water and organism temporal coordination in
sample collection; and 4) water and organism spatial coordination in sample collection.
Additionally, the general experimental design was evaluated. For fun her details on BAFs
compilation and ranking, see Burkhard (2021).

Table 2-2 below outlines the screening criteria for study evaluation and ranking. Only
BAFs of high and medium quality were used to derive the tissue criteria (Appendix P). For
further details on BAFs compilation and ranking, see Burkhard (2021).

Table 2-2. Evaluation Crilcria for Screening liioacciimulalion l-'adors (BAFs) in the Public
Literature.

lahlc modified from Hiirlihanl 12021) / h ah \!aiiHscnpi.		

Screening Kaclor

High Quality

Medium Quality

Low Quality

Number of Water Samples



~ J

1

Number of Organism
Samples'



2-3

1

Temporal Coordination

Concurrent
collection

Within one year

Collection period > 1
year

Spatial Coordination

Collocated
collection

Within 1-2 km

Significantly
different locations
(> 2 km)

General Experimental Design





Mixed species tissues
samples

1 Organismal samples l ioni ihe same species and tissue type.

67


-------
3 EFFECTS ANALYSIS FOR AQUATIC LIFE

3.1 Toxicity to Aquatic Life

All available, reliable studies relating to the acute and chronic toxicological effects of

PFOS on aquatic life were considered in the derivation of the national recommended PFOS
criteria. Data for possible inclusion in the PFOS criteria were obtained from published literature
reporting acute and chronic exposures of PFOS that were associated with mortality, survival,
growth, and reproduction. This set of published literature was identified In the EPA's public
ECOTOX database (ECOTOX: https://cfpub.epa.gov/ecotox/) as meeting data quality standards.
ECOTOX is a source of high-quality toxicity data lor aquatic life, terrestrial plants, and wildlife.
The database was created and is maintained by the EPA. Office of Research and Development,
Center for Computational Toxicology and l-.\posuie. Studies were then further reviewed by EPA,
Office of Water to determine test acceptability for use in the criteria derivation. Additional
literature searches were also conducted to ensure all available toxicity data were captured. The
latest search was conducted through September 2021.

3.1.1 Summary of PI OS Toxicity Studies I scd to Derive the Aquatic Life Criteria

()iuiniilali\ e data lor acute PI-OS toxicity were available for 26 freshwater species,

representing 18 genera and I 5 families in five phyla, and six estuarine/marine species,

representing six genera and li\ e families in four phyla. Chronic PFOS toxicity data were

available for 16 freshwater species, representing 14 genera and 13 families in four phyla, and

three estuarine/marine species, representing three genera and three families in two phyla (Table

3-1).

68


-------
Table 3-1. Summary Table of Minimum Data Requirements per the 1985 Guidelines Reflecting
the Number of Acute and Chronic Genus and Species Level Mean Values in the Freshwater and
Saltwater Toxicity Datasets for PFOS. 	



I'lvshwalor

MDR

(;ma\

S\1 AY

(IMC V

SMC Y

Family Salmonidae in the class

1

1

1

1

Osteichthyes

Second family in the class Osteichthyes,
preferably a commercially or
recreationally important warmwater

2

2

2

2

species









Third family in the phylum Chordata (may
be in the class Osteichthyes or may be an
amphibian, etc.)

5

In



4

Planktonic Crustacean

2

4



3

Benthic Crustacean

2

2

1

1

Insect

oa

na

:

2

Family in a phylum other than Arthropoda
or Chordata (e.g., Rotifera, Annelida, or

5

6

2

2

Mollusca)









Family in any order of insect or any
phylum not already represented

1

1

1

1

Total

IS

2(>

14

16



Saltwater1'

MDR

(;ma\

S\1 AY

c;mc y

SMC Y

Family in the phylum Chordata

i

1

i)

i)

Family in the phylum Chordata

0

0

0

0

Either the Mysidae or Penaeidae family

2

2

l

1

Family in a phylum oilier than Arthropoda
or Chordata

1

1

0

0

Family in a phvlum other than Choidala

1

1

1

1

Family in a phylum other than Chordata

1

1

1

1

Family in a phylum other than Chordata

0

0

0

0

Any other family

0

0

0

0

Total

6

6

3

3

a One MDR, for aquatic insect, u as noi fulfilled with acute quantitative data. However, EPA considered acute qualitative
data for insects, as discussed below. and concluded that there were sufficient data to conclude that not having met the MDR
would not substantively affect the resulting FAV to develop an acute freshwater criterion.

bThe 1985 Guidelines require that data from a minimum of eight families are needed to calculate an estuarine/marine
criterion. Insufficient data exist to fulfill all eight of the taxonomic MDR groups. Consequently, EPA cannot derive an
estuarine/marine acute criterion, based on the 1985 Guidelines. However, EPA has developed draft estuarine/marine
benchmarks through use of surrogate data to fill in missing MDRs using EPA's Web-based inter-species correlation
estimation tool These benchmarks are provided in Appendix L.

69


-------
Below are the summarized studies that provided key acute and chronic freshwater
toxicity data with effect values that were used quantitatively in deriving the acute and chronic
freshwater criteria to protect aquatic life from harmful exposure to PFOS. Study summaries are
also provided for the estuarine/marine toxicity data that could be used quantitatively to derive
acute and chronic estuarine/marine criteria if the MDRs were met.

Study summaries for the most sensitive taxa are grouped In acute or chronic exposure
and sorted by sensitivity to PFOS. Study data were summarized in tabular form in Appendix A
(freshwater acute studies), Appendix B (estuarine/marine acute studies), Appendix C (freshwater
chronic studies), and Appendix D (estuarine/marine chronic studies). Key acute and chronic
toxicity studies used qualitatively as supporting informal ion are described in the Effects
Characterization (Section 4) below and corresponding data are summarized in Appendices E, F,
G and H while the remaining, unused studies are summarized in Appendix J.

Acute and chronic \ allies were presented as reported by the study authors for each
indrvidual study I-PA independently calculated toxicity values if sufficient raw data were
available to conduct statistical analyses All toxicity values, such as LCs, ECs, NOECs, LOECs,
and species- and genus-mean \ allies, were gi\ en to four significant figures to prevent round-off
error in subsequent calculations, not to reflect the precision of the value. The author-reported
toxicity values and I-PA's independently-calculated values (where available) were included for
each study throughout the document (in the study summaries and appendices as applicable), and
the specific value utilized to derive the criteria were identified along with a justification. EPA's
independently calculated toxicity values were used preferentially, where available.

70


-------
3.1.1.1 Summary of Acute PFOS Toxicity Studies Used to Derive the Freshwater Aquatic Life
Criterion

Acceptable data on the acute effects of PFOS in freshwater were available for a total of
26 species representing 18 genera and 15 families in five phyla (Appendix A: Acceptable
Freshwater Acute PFOS Toxicity Studies). More specifically, quantitative data for acute PFOS
toxicity were available for three freshwater fish species (two of the eight MDRs), 13 freshwater
invertebrate species (four of the eight MDRs), and 10 freshwater amphibian species (one of the
eight MDRs). Ranked genus mean acute values (GMAVs) for PFOS in freshwater based on
acute toxicity were identified in Table 3-2 (5 most sensitive genera) and Table 3-3 (all genera)
and plotted in Figure 3-1. The aquatic insect MDR was not fulfilled with quantitati\ ely
acceptable acute data; however, qualitatively acceptable acute insect data (as discussed below)
were evaluated and these data demonstrate the \ aiiahililv in the sensiti\ ity of aquatic insects.
Therefore, EPA will continue to seek additional acute PI-OS insect data to further understand the
sensitivity of this ta\on Additionally. I.IW evaluated the effect of the current qualitative studies
on the Final Acute Value (I AV) in Section 4 2 I Thus, the current development of an acute
freshwater criterion was based on se\en of the eight MDRs.

71


-------
Table 3-2. The Five Most Sensitive Genera Used in Calculating the Acute Freshwater
Criterion (Sensitivity Rank 1-5).

Ranked below from most to least sensitive. Note: five genera are shown with the intent to discuss
sensitive genera to the point where the four most sensitive North American resident species are
represented. The fourth most sensitive genus, Neocaridina, is based on toxicity data from a non-North
American resident.

Rank

Genus

Species

GMAV

(ill"/!.)1

Comment

1

Pimephales

Fathead minnow,
P. promelas

6.950

North American resident
species

2

Oncorhynchus

Rainbow trout,
O. mykiss

7.515

North American resident
species

3

Ligumia

Black sandshell,
L. recta

13.5

North American resident
species

4

Neocaridina

Japanese swamp shrimp,
N. denticulata

15.61

Not a resident species in
North America2

5

Xenopus

African clawed frou.
X laevis

1 5 99

North American resident
species3

1	Values used in additional analyses supporting the criterion calculation In eviminc the effects of less certain toxicity
studies and non-resident species on acute freshwater criterion. See Section 4.1 below for more details.

2	Species not resident to North America were included since the relative sensitivities between native and non-native

species were similar. Therefore, for the PFOS criteria derivation, it was determined that species not native to North
American can serve as surrogates for other sensitive resident North American organisms.

3Not native to North America; however, is considered a resident to North America in the 1985 Guidelines (U.S.EPA
1985).

3.1.1.1.1 Most Sensitive ireshwater (ieims /or Acute Toxicity: Pimephales (fathead minnow)
Drottar and krucger (2000d) e\ aluated the acute effects of PFOS-K (PFOS potassium

salt, CAS 27l)5-3l>-3. I ,ot 21 7 (T-(->2^5) obtained from the 3M Company, 90.49% purity) on

juvenile fathead minnows (1'imephalespmmelas) during a 96-hour measured, static study.

Researchers followed protocols from U.S. EPA Series 850, OPPTS 850.1075 and OECD

Guideline 203 All fish used in the test were from the same source and year class, and the total

length of the longest fish was no more than twice the length of the shortest. The authors reported

a LCso of 9.5 mg/L PFOS. EPA's independently-calculated 96-hour LC50 was 9.020 mg/L and

was used quantitatively to derive the draft acute water column criterion for freshwater.

(3MCompany 2000) provides the results of a 96-hour static, unmeasured acute toxicity

test with the fathead minnow and PFOS-Li (PFOS lithium salt, CAS # 29457-72-5). Fish were

72


-------
79 days old at test initiation with an average length of 2.1 cm and weight of 0.069 g. No
mortality occurred in the control treatment and 100% was observed in the highest treatment (56
mg/L). The study authors reported that the test sample containing 24.5% PFOS-Li exhibited a
96-hour LCso of 19 mg/L, which equates to 4.655 mg/L as PFOS. The independently-calculated
96-hr LCso value was 21.86 mg/L, which equates to 5.356 mg/L as PFOS, and is deemed
acceptable for quantitative use in the derivation of the acute fresh water criterion for PFOS.

The geometric mean of the two acute toxicity values provided abo\ e fori5, promelas
(9.020 and 5.356 mg/L) were used to calculate an SM.W and (i\l.\V of 6.^50 mg/L, which
represents the most sensitive GMAV in EPA's acute dataset used to derive the freshwater aquatic
life criterion.

3.1.1.1.2 Second Most Sensitive Freshwater (ieims for. \cute / oxicity: ()ncorhynchus (trout)

Sharpe et al. (2010) evaluated the acute effects of I'FOS-K (potassium salt, CAS # 2795-

39-3, 98% purity) to (hicorliynchiis my kiss, rainbow trout. \ ia a 96-hour renewal exposure with
measured concentrations (renewal was not stated in paper, but assumed based on other
information pio\ ided. including the test Guideline protocol that the authors cited as the protocol
that was used) There were limited details in the publication about the test protocol; however, it
was noted that the Organization for Economic Co-operation and Development (OECD)

Guideline 203 was followed, and the study authors did not identify any deviations from these test
guidelines. EPA obtained clarification from the study authors on the experimental design
regarding the biomass loading rate, which was 1 to 1.5 g/L (based on four fish weighing a total
of 2 to 3 g per 2 L tank; personal communication with Greg Goss and Rainie Sharpe, March
2021). This biomass loading rate was slightly higher than that stated in OECD Guidelines of 0.8
g/L (OECD 1992). The author-reported 96-hour LCso for the study was 2.5 mg/L. The authors do

73


-------
not specify if this concentration was nominal or measured. Given the clarifications regarding the
biomass loading, the LCso from this study was used quantitatively to calculate the SMAV and
GMAV for derivation of the draft acute water column criterion.

Palmer et al. (2002a) evaluated the acute effects of PFOS-K (potassium salt, identified
as FC-95 obtained from 3M Company) to rainbow trout via a 96-hour static exposure with
measured concentrations. The study author-reported 96-hour TO" Ibr the study was 22 mg/L.
The independently-calculated 96-hour LCso value was 22.59 mg/L. The independently-calculated
LCso was used quantitatively to calculate the SMAV and GMAV for deri\ alion of the draft acute
water column criterion.

The geometric mean of the two toxicity values provided above (2.5 and 22.59 mg/L), was
used to calculate the SMAV and GMAV of 7 5 I 5 mg'I. for rainhow trout, 0. mykiss, which was
used to derive the freshwater aquatic life criterion This GM AV of 7.5 15 mg/L is consistent with
the acute rainbow trout studies cited in OIX'Ds 2<)(P |>| OS I lazard Assessment, from which the
LCso values for rainlxm trout range IV0111 7 S to 22 mu I. (OECD 2002).

3.1.1.1.3 third \ lost Sensitive livslm nicr (icims for. \cute Toxicity: Ligumia (mussel)

lla/cllon (2013): lla/cllon el al. (2012) e\ aluated the acute effects of PFOS (acid form,

> 98" o purity) 011 two freshwater mussels Ligumia recta and Iximpsitis sitiquoidea. The tests

yielded the 3"' and o'1' most sensitive genus values (respectively) in the PFOS freshwater acute

toxicity database ( The /. sitiquoidea results are reported in Appendix A). Acute toxicity was

observed under static conditions over a 24-hour period (< 24-hour old glochidia) or a 96-hour

period (4 to 6-week-old juveniles). The tests followed the ASTM (2006) test method. The 24-

hour EC50 reported by the study authors for glochidia of L. recta was 13.5 mg/L. The 96-hour

LCso reported by the study authors for juvenile L. recta was 141.7 mg/L. Both acute values are

74


-------
acceptable for quantitative use but because the juvenile life stage was less sensitive, only the
glochidia LCso was used to calculate the SMAV. The independently calculated toxicity values
could not be calculated at this time given the lack of data presented in the paper. No other
quantitative toxicity values were available for this species or genus. Therefore, the author-
reported 24-hour ECso of 13.5 mg/L for the glochidia life stage of L. recta served directly as the
SMAV and GMAV which are utilized to derive the freshwater aquatic life acute criterion.

3.1.1.1.4	Fourth Most Sensitive Freshwater Genus for Acute Toxicity: Xeocaridina (shrimp)
Li (2009) conducted three independent repeals of a 96-hour static test 011 PFOS-K

(potassium salt, >98% purity) with the freshwater shrimp species, Neocaridina ilcmiculata (a

non-North American species). The author-reported 96-hour I ,Cso was 10 mg/L based on the

average of three repeat tests. The independently-calculated LCso values for the three independent

experimental repeats were 12.91, 28.55, in 32 mu I., respectively. These independently-

calculated LC50 values were used to calculate the geometric mean of 15.61 mg/L. No other

quantitative toxicity \ allies were a\ ailahle for this species or genus; therefore, the geometric

mean of the three independently-calculated lH->-hour LCsos of 15.61 mg/L served directly as the

SMAV and (i\l.\Y utilized in the acute water column criterion.

3.1.1.1.5	lifi/i \ losi Sensitive Ireshwater Genus for Acute Toxicity: Xenopus (frog)

Palmer niul Krueซor (2001) conducted three independent, renewal assays with Xenopus

laevis. The author-reported 96-hour LC50 values were 13.8, 17.6 and 15.3 mg/L PFOS, the
teratogenesis ECsos were 12.1, 17.6 and 16.8 mg/L PFOS, and the minimum concentrations to
inhibit growth values (effectively LOECs) were > 14.7, 7.97 and 8.26 mg/L for the same three
tests, respectively. LC50 values for teratogenesis and mortality were similar, suggesting there was
no apparent difference in endpoint sensitivity. Mortality was used to derive the X. laevis SMAV

75


-------
since it is the more established endpoint for deriving acute criteria (U.S. EPA 1985). The
independently-calculated 96-hour mortality-based LCso values were 15.53, 18.04, and 14.60
mg/L, which were taken together as a geometric mean to calculate the X. laevis SMAV of 15.99
mg/L. No additional quantitative, acute toxicity data were available for other members of this
genus. Therefore, theX. laevis SMAV of 15.99 mg/L served directly as the Xenopus GMAV.

3.1.1.1.6 Missing Insect MDR

The PFOS acute dataset based on direct aqueous exposures contains IS genera (Table

3-3) representing seven of the eight MDRs. The missing MDR is a represenlali\ e Irom an insect

family. As the derivation of a PFOS acute freshwater criterion is important for the protection of

aquatic life exposed to PFOS, EPA considered qualitative data (see Appendix G) to determine if

the relative sensitivity of aquatic insects could he ascertained There were qualitative data from

two acute studies focused on aquatic insects I 'irst. Yang et al (2<> 14) conducted a 96-hour

renewal, acute test with measured concentrations on the midge, Chironomusplumosus. Second,

Olson (2017) conducted a 4<~>-da\ static test with unmeasured concentrations on the yellow fever

mosquito (AciL-s acgypu) and reported a 4S-hour LCso value. Yang et al. (2014) was classified as

acceptable lor qualitati\ e use because the test organisms were considered to be from a

problematic source since the test organisms were obtained from the Beijing City Big Forest

Flower Market and no further quantification of previous exposure to contaminants or husbandry

was provided (Yang et al 2014). The reported PFOS LCso was 182 mg/L (see Appendix Section

G.2.1.5), indicating that insects, as represented by the midge in this test, is one of the least

sensitive taxonomic groups to acute exposures of PFOS (Figure 3-1). However, as previously

mentioned the source of the test organisms was problematic and it is difficult to ascertain the role

76


-------
any potential previous exposure or husbandry issues might have played in the results of this
toxicity study and the relative sensitivity of chironomids to acute exposures of PFOS.

In contrast, qualitative data from Olson (2017) on the yellow fever mosquito, an invasive
species to the U.S., indicated that this species is relatively sensitive to acute exposures of PFOS.
A LCso of 1.18 mg/L was reported following 48 hours of exposure in this 40-day, static test with
unmeasured test concentrations (Olson 2017; see Appendix G 2 I 5) This study was not
acceptable for quantitative use primarily because the test duration was loo short for the species
and secondarily because the test organism is an invasive, pest species. FIo\\e\ er. the reported
PFOS 48-hour LCso was 1.18 mg/L, which indicated that this species of insect may he one of the
most sensitive species to acute exposures of PFOS (Figure 3-1).

These two qualitatively acceptable studies demonstrated the variability in the sensitivity
and indicated contrasting sensitivity of aquatic insects to acute exposures of PFOS. Therefore,
EPA will continue lo seek additional acute PI OS insect data lo further understand the sensitivity
of this taxon Additionally. N\\ e\ alualed the effect of the current qualitative studies on the
Final Acute Value (I AV) in Section 4 2 1 As such additional insect toxicity data for PFOS are
needed to further examine the ielali\ e sensitivity of insects to PFOS exposures. And the current
development of an acute freshwater criterion was based on seven of the eight MDRs.

77


-------
Table 3-3. Ranked Freshwater Genus Mean Acute Values.

Rank"

GMAV
(ing/l. PIOS)

MI)U

Group'

Genus

Species

S.M.W h
(ing/L PI-OS)

1

6.950

B

Pimephales

Fathead minnow,
Pimephales promelas

6.950

2

7.515

A

Oncorhynchus

Rainbow trout,
Oncorhynchus mykiss

7.515

3

13.50

G

Ligumia

Black sandshell.
Ligumia recta

13.50

4

15.61

E

Neocaridina

Japanese swamp shrimp,
Neocaridina deiuiciilata

15.61

5

15.99

C

Xenopus

African clawed liou.
Xenopus laevis

15.99

6

16.50

G

Lampsilis

Fatmucket,

Lampsilis siliquoidea

16.50

7

17.20

D

Moina

Cladoceran.

Moina niacrocopa

17.20

8

19.88

C

Hyla

Gray lavlYog,
Hyla versicolor

19.88

9

22.48

G

Dugesia

Planaria.

/ Algesia japomca

22.48

10

24.44

B

Danio

Zehalish,
Danio rerio

24.44

11

42.30

1)

I kiphnia

Cladoceran,
Daphnia carinata

11.56

Cladoceran,
Daphnia magna

48.87

Cladoceran,
Daphnia pulicaria

134.0

12

47 4i)

C

Ambystoma

Jefferson salamander,
Ambystoma jeffersonianum

51.71

Small-mouthed salamander,
Ambystoma texanum

30.00

Eastern tiger salamander,
Ambystoma tigrinum

68.63

13

56.49

C

Anaxyrus

American toad,
Anaxyrus americanus

56.49

14

59.87

E

Procambarus

Crayfish,

Procambarus fallax f virginalis

59.87

15

61.80

H

Brachionus

Rotifer,

Brachionus calyciflorus

61.80

16

64.35

G

Elliptio

Eastern elliptio,
Elliptio complanata

64.35

78


-------
Rank3

GMAV
(mg/L PFOS)

MDR
Group0

Genus

Species

SMAVb
(mg/L PFOS)

17

109.2

C

Lithobates

American bullfrog,
Lithobates catesbeiana

133.3

Green frog,
Lithobates clamitans

113.0

Northern leopard frog,
Lithobates pipiens

72.72

Wood frog,
Lithobates sylvatica

130.0

18

172.1

G

Physella

Bladder snail,
Physella acuta

183.0

Snail,

Physella heterostropha pomilia

161.8

a Ranked from the most sensitive to the most tolerant based on Genus Mean Acute Value.
b From Appendix A: Acceptable Freshwater Acute PFOS Toxicity Studies.
0 MDR Groups identified by list provided in Section 2.10.1 above.

~ Chironomus
# Physella
~ Lithobates
• Elliptio

ฆ	Brachionus

ฆ	Procambarus
~ Anaxyras

A Amby stoma
ฆ Daphnia
~ Danio
ฆ Dugesia
A Hyla
ฆ Moina
• Lampsilis
A Xenopus
ฆ Neocaridina
• Ligiunia
~ Oncorhynchus
~ Pimephales

0.1	1	10	100	1000

Genus Mean Acute Value (mg/L PFOS)

Figure 3-1. Freshwater Acute PFOS GMAVs Fulfilling the Acute MDRs.

Qualitative data for insect species taken into consideration to understand the aquatic insect MDR are denoted by the
open white boxes. The GMAVs for these qualitative data were not used to derive the freshwater acute criterion for
PFOS.

l.U T

0.9 -

u
s

th

41
>

0.8 -

0.7 -

0.6 -

0.5 -

ง 0.4
&

41

3 0.3
a


-------
3.1.1.2 Summary of Acute PFOS Toxicity Studies Used to Derive the Estuarine/Marine
Aquatic Life Criterion

Quantitative empirical data for acute PFOS toxicity were available for six saltwater
species, representing only six genera and five families. The data available for saltwater species
fulfilled only five of the eight MDRs. In the interest of providing recommendations to
states/authorized tribes on protective values, EPA developed an estuarine/marine acute
benchmark using the available empirical data supplemented ^iih toxicity \ alues generated
through the use of new approach methods, specifically through the use of the I-PA Office of
Research and Development's peer-reviewed publicly-available Web-based Interspecies
Correlation Estimation (WeblCE) tool (Raimondo et al 2<> I <>) These benchmarks are provided
in Appendix L. Table 3-4 below shows the four most sensitive acute estuarine/marine genera that
could be used quantitatively to derive acute criteria if the MDRs were met. Ranked GMAVs for
saltwater organisms based on acceptable acute toxicity \ alues were identified in Table 3-5 and
plotted in Figure 3-2

80


-------
Table 3-4. The Four Most Sensitive Acute Estuarine/Marine Genera.

Ranked Below from Most to Least Sensitive.			







(/MAN









(mg/l.



Uank

Genus

Species

PI OS)

Comments

1

Mytilus

Mediterranean
mussel,

M. galloprovincialis

1.1

Not a resident species in North
America, but other species in this
genus are resident, commercially,
or ecologically important species

2

Strongylocentrotus

Purple sea urchin,
S. purpuratus

1.7

North American resident species









Not a resident species in North
America, but other species in this

3

Paracentrotus

Sea urchin.

P. lividus

1 7l)5

family (Echinidae) are common
ecotoxicity test species that
serves as a surrogate for untested
urchin species residing in North
America.

4

Americamysis

Mysid,
A. bahia

4 1)I4

North American resident species

3.1.1.2.1 Most Sensitive Estuarine/Marine (iciuis for. Ic/i/c toxicity: Mytilus (mussel)

The acute toxicity of perfluorooctane sulfonate (PI OS. purity not provided) on the

Mediterranean mussel. \ lyii/ns ga/toprovincia/is was evaluated by Fabbri et al. (2014). This

species is not resident to North America, hut is a surrogate for North American mussel species,

including the widespread, commercially, and ecologically important blue mussel, Mytilus edulis.

At test termination (48 hours), the endpoint was the percentage of normal D-larvae in each well,

including malformed larvae and pre-D stages. The acceptability of test results was based on

controls for a percentage of normal D-shell stage larvae, > 75% (ASTM 2004a). The percentage

of normal D-larva decreased with increasing test concentrations. The NOEC and LOEC reported

for the study were 0.00001 and 0.0001 mg/L, respectively. However, the test concentrations

failed to elicit a 50% reduction in malformations in the highest test concentration, and an EC so

was not determined. Therefore, the EC so for the study was greater than the highest test

81


-------
concentration (1 mg/L). The 48-hour ECso based on malformation of > 1 mg/L was acceptable
for quantitative use.

Hayman et al. (2021) report the results of a 48-hour static, measured test on the effects
of PFOS-K (PFOS potassium salt, CAS # 2795-39-3, 98% purity) onMytilusgalloprovincialis.
Authors noted that tests followed (U.S.EPA 1995) and (ASTM 2004b) protocols. Larvae were
enumerated for total number of larvae that were alive at the end of i lie lest as well as number of
normally-developed D-shaped larvae. There were no significant differences between solvent
control and filtered seawater, suggesting no adverse effects of methanol. The author reported 48-
hr ECso, based on normal development, is 1.1 mg/L. EPA was not able to independently
calculate a 48-hour EC50 value as the curve fitted model did not result in a good fit. Therefore,
the author-reported ECso of 1.1 mg/L was considered for quantilali\ e use.

The two ECso values from the two studies both indicated sensitivity of the Mediterranean
mussel to acute exposure of PTOS is above 1 mu I. Ho\\e\ ei\ the ECso forM. galloprovincialis
from Fabbri et al. (2014) was unbounded while the ECso from Hayman et al. (2021) was
definitive, and therefore that latter (I I mu I.) serves as the basis for the SMAV and GMAV to
derive the acute estuarine marine benchmark for PFOS.

3.1.1.2.2 Second Most Sensitive l.siiutrine Marine Genus for Acute Toxicity: Strongylocentrotus
(sea urchin)

The Hayman et al. (2021) study also included the results of a 96-hour static, measured

test on the effects of PI 'OS-K (PFOS potassium salt, CAS # 2795-39-3, 98% purity) on the

purple sea urchin, Strongylocentrotuspurpuratus. Authors noted that tests followed USEPA

(1995) and ASTM (2004b) protocols. At test termination (96 hours), the first 100 larvae were

enumerated and observed for normal development (4-arm pluteus stage). As with the other tests

in the study with different species, there were no significant differences between solvent control

82


-------
and filtered seawater, suggesting no adverse effects of methanol. The author reported 96-hour
ECso, based on normal development, is 1.7 mg/L. EPA was not able to independently calculate a
96-hour ECso value as the curve fitted model did not result in a good fit. Therefore, the author-
reported ECso of 1.7 mg/L mg/L was thus applied for quantitative use and was utilized as the
SMAV and GMAV to derive the acute estuarine/marine benchmark for PFOS.

3.1.1.2.3	Third Most Sensitive Estuarine/Marine Genus for. \ciiie toxicity: Paracentrotus (sea
urchin)

A 72-hour static, unmeasured PFOS (purity not provided) toxicity icsl with the sea
urchin, Paracentrotus lividus (a non-North American species) was conducted In (iiimluz et al.
(2013) The 72-hour ECso based on normal development to the pluteus stage was 1.795 mg/L
PFOS and was acceptable for quantitati\ e use. however, additional consideration needs to be
taken, given to the short test duration.

3.1.1.2.4	Fourthhlosi Sensitive l.siuarine \ /arine (ienns for Acute Toxicity: Americamysis
(mysid)

Along with the Mediterranean mussel and purple sea urchin, Hayman et al. (2021)
conducted a lH->-hour static, measured test to determine the effects of PFOS-K on the mysid,
Americamysis bahia. Authors noted that tests followed USEPA (1995, 2002) and ASTM (2004b)
protocols Only two of the sixty organisms (3.3%) were found dead in the controls at test
termination. The author reported 96-hour LCso is 5.1 mg/L PFOS-K. The independently-
calculated 96-hr LO" \ alue was 4.914 mg/L and is acceptable for quantitative use in the
derivation of the acute estuarine/marine benchmark for PFOS.

83


-------
Table 3-5. Ranked Esi

tuarine/Marine Water Genus

Mean Acute Values.

Rank1

GMAV
(mg/L PFOS)

MDR
Group3

Genus

Species

SMAV2
(mg/L PFOS)

1

1.1

D

Mytilus

Mussel,

Mytilus galloprovincialis

1.1

2

1.7

F

Strongylocentrotus

Purple sea urchins,
Strongylocentrotus
purpuratus

1.7

3

1.795

E

Paracentrotus

Sea urchin,
Paracentrotus lividus

1.795

4

4.914

C

Americamysis

Mysid,

Americamysis bahia

4.914

5

6.9

C

Siriella

Mysid,

Siriella armata

6.9

6

>15

A

Cyprinodon

Sheepshead minnow,
Cyprinodon variegatus

>15

1	Ranked from the most sensitive to the most tolerant based on Genus Mean Acute Value.

2	From Appendix B: Acceptable Estuarine/Marine Acute PFOS Toxicity Studies.

3	MDR Groups identified by list provided in Section 2.10.1 above.

1.0 T

0.9 --

0.8 --

s
V-

V
C3

ฃ

QJ
>

—

13

E

3 0.5

0.7 --

0.6 --

0.4 --

s

cs

04

Ol

ฃ 0.3

a

n

&- 0.2

0.1

0.0

Cyprinodon (non-definitive, greater than value) ~

Siriella

Americamysis

Paracentrotus

Strongylocentrotus

• Mytilus

ฆ Invertebrate (Other)

•	Invertebrate (Mollusk)

~	Fish

0.1

1	10

Genus Mean Acute Value (mg/L PFOS)

100

Figure 3-2. Acceptable Estuarine/Marine GMAVs.

84


-------
3,1,1.3 Summary of Chronic PFOS Toxicity Studies Used to Derive the Freshwater Aquatic
Life Criterion

Chronic toxicity data were available for all of the freshwater MDRs. Chronic PFOS
toxicity data were available for 16 freshwater species, representing 14 genera and 13 families in
four phyla. These data included four freshwater fish species, representing four genera and three
families. The data available for freshwater fish were relatively di\ erse and fulfilled three of the
eight MDRs. Chronic data were also available for nine freshwater invertebrate species,
representing eight genera and eight families. The data available for freshwater in\ crtebrate were
also relatively diverse and fulfilled five of the eight MDRs. The chronic dataset also includes
data for three amphibian species, representing two genera in two families. Ranked GMCVs for
PFOS in freshwater based on chronic toxicity were identified in TaMe 3-6 (four most sensitive
genera) and Table 3-7 (all genera) and plotted in I'iuine 3-3

Table 3-6. The Four Most Sensitive Genera Used in Calculating the Chronic Freshwater
Criterion.

Ranked Below from Most to /.cast Sensitive.

Uank

(ienus

Species

(;\l( V

(ing/l, pros)1

('omincnls

1

('hironomns

Midue.

('luronomus dilutus

0.009676

North American resident species

2

Lampsths

Fatmucket,

Lam/ >st lis siliquoidea

0.01768

North American resident species

3

Enallagma

lilne damselfly,
/ j hi It'agma cyathigerum

0.03162

North American resident species

4

Danio

Zebrafish,
Danio rerio

0.03217

Uncertain as resident species of
North America.2/), rerio is a
common ecotoxicity test species
that serves as a surrogate for
untested fish species residing in
North America.

Values used in additional analyses supporting the criterion calculation to examine the effects of less certain toxicity
studies and non-resident species on chronic freshwater criterion. See Section 4.1 below for more details.
2Not native to North America; however, is considered a resident to North America in the 1985 Guidelines.

85


-------
3.1.1.3.1 Most Sensitive Freshwater Genus for Chronic Toxicity: Chironomus (midge)

(MacDonald et al. 2004) conducted chronic larval and life-cycle tests to determine the

effects of PFOS-K (PFOS potassium salt, 95% purity) on the midge, Chironomus dilutus
(formally known as Chironomus tentans). The test was performed under renewal conditions over
10 days for the larval test and greater than 50 days for the life-cycle test The tests followed the
general guidance given by EPA-600-R99-064 (U.S.EPA 2000b) and ASTM i: 1706-00 (ASTM
2002). (MacDonald et al. 2004) conducted chronic larval and life-cycle tests lo determine the
effects of PFOS-K (PFOS potassium salt, 95% purity) on the midge, Chironomus dilutus
(formally known as Chironomus tentans). The test was performed under renewal conditions over
10 days for the larval test and greater than 5<) days lor the life-cycle test. The tests followed the
general guidance given by EPA-600-R99-<)M (I S l-IW 2<><ป(>h) and ASTME 1706-00 (ASTM
2002).

The author reported I "-day growth and sm \ i\ al ECios for the study were 0.0492 and
0.1079 mg/T.. rcspccti\ cl\ The study authors also reported NOECs of 0.0491 mg/L, LOECs of
0.09<->2 mu I., and MATCs of <) i)(->S7 mu I. lor both endpoints. The author-reported 20-day ECios
for growth. snr\ ival, and total emergence were 0.0882, 0.0864, and 0.0893 mg/L, respectively.
The study authors also reported NOECs of 0.0217 mg/L for growth and survival and < 0.0023
mg/L for emergence. I .OI X's of 0.0949 mg/L for growth and survival and 0.0271 mg/L for
emergence, and MATCs of 0.0454 mg/L for growth and survival and 0.0071 mg/L for
emergence. It is noted here that the paper reported contrasting NOECs for 20-day survival. The
text in the paper stated that the NOEC was 0.0271 mg/L and Table 2 of the paper provided a
value of 0.0949 mg/L. EPA assumed the NOEC in Table 2 of the paper was not correct and that

86


-------
0.0217 mg/L was the correct NOEC based on the data presented in Figure 3A of the paper. This
assumption was applied to the summary of the study results presented in this PFOS draft criteria.
EPA was able to independently calculate an ECio for 10-day growth of 0.05896 mg/L for the
study. The independently-calculated 10-day ECio value for growth of the midge was acceptable
for quantitative use in the derivation of the freshwater chronic water column criterion for PFOS.

McCarthy et al. (2021) also conducted two chronic toxicity tests with PFOS (98%
purity) on the midge, Chironomus dilutus, a 10-day and a 20-day exposure, following standard
protocols (ASTM 2005; U.S. EPA 2000b) with slight modifications. The 10-day exposure was
considered a range finding test, with concentrations spaced by ~100x and only mortality
measured, whereas the 20-day exposure measured both survival and growth. The 20-day
exposure is less than the recommended 5<) - (o day full-life cycle method outlined in U.S. EPA
(2000b) and used in MacDonald et al. (20<)4). and since exposures of midges started on day two
or four, the actual exposure duration is only 16 or 19 days long. The most sensitive endpoint was
survival with an author-reported 10-day ECio of 0.00136 mg/L PFOS. Additionally, the study
authors reported ECi..s of <> oo|(i2 and 0.<)()323 mg/L PFOS for growth as mean biomass and
mean weight. respecti\ cly I-PA was unable to independently calculate ECios for survival and
mean weight. I lowever. I -PA was able to independently calculate an ECio value for mean
biomass of <> <><> I 5SS mg/L. PI OS. The independently-calculated 16-day ECio for mean biomass
was acceptable for <.|uantilali\ e use in the derivation of the freshwater chronic water column
criterion for PFOS.

The most sensitive endpoints from the two toxicity studies with C. dilutus that could be
independently-calculated (see details in Appendix C.2.1) were for 10-day growth with an ECio of
0.05896 mg/L (MacDonald et al. 2004) and 16-day mean biomass with an ECio of 0.001588

87


-------
mg/L (McCarthy et al. 2021). Although over an order of magnitude different, both the ECio of
0.05896 mg/L for 10-day growth and ECio of 0.0015879 mg/L for 16-day mean biomass were
used quantitatively to derive the chronic aquatic life criterion with a SMCV and GMCV equal to
the geometric mean of the two values or 0.009676 mg/L. As mentioned in the Bots et al. (2010)
summary and in Section 4.1.1, the observed effects of PFOS on aquatic insects appeared to be
consistent across the available data for chironomids and odonales I lowever, Bots et al. (2010)
did not measure the effects of PFOS on nymph growth and therefore, the observed effects in that
study cannot be compared with the results of MacDonald et al. (2004) and McCarthy et al.
(2021). The remainder of the toxicity values available for aquatic insects were used as supporting
information to corroborate the toxicity value used to derive the freshwater chronic criterion and
to better understand the effects of PFOS oil aquatic insects in general No other quantitative
toxicity values were available for this species or genus

3.1.1.3.2 Secondlost Sciisiiiw Iresliwaier (ieims for Chronic Toxicity: Lampsilis (mussel)
Hazcllon (2013): Ihizcllon el ;il. (2012) conducted a test of the long-term effects of

PFOS (acid Ibmi. W() puriiy) on glochidia and jmenile life stages from the mussel Lampsilis

siliquoiclea using a unique experimental design for which standard methods have not been

established The lest exposed hooding glochidia (in marsupia) for 36 days followed by a 24-

hour exposure of free glochidia The in marsupia exposure was followed by a 24-hour free

glochidia exposure consisting of a factorial design. As such the free glochidia from the control

group of the marsupia exposure were divided between a control and the two PFOS treatments

and the PFOS treatments were split into control and the same PFOS treatment group as the

marsupia exposure. This factorial design allowed for the comparison of PFOS effects in two

88


-------
different life-stages. See Appendix C.2.2 for additional details on the experimental design and
considerations for the utilization of this study in the criterion derivation.

The data presented in the paper for metamorphosis success were considered for
quantitative use in the derivation of the chronic criterion for PFOS (see Appendix C.2.2). The
author-reported NOEC was 0.0045 mg/L and LOEC was 0.0695 mu T. The reduction in
metamorphosis success at the LOEC was estimated to be 35.4".. I lowe\ er, this was not a
definitive test in that both the study design (which only included two treatment groups) and
level of data presented (which are only presented graphically in Figure 2 of the paper) in the
publication lack the details needed to fully understand the effects of chronic PI OS exposures to
the glochidia and juvenile life stages of Lampsilis siliquoiJea. Additionally, as there were only
two PFOS treatment groups and the gap in these exposure concentrations is large (about 15-
fold), EPA was not able to fit a curve to estimate an I X'i.. in a manner similar to the other
toxicity studies used to deri\ e this criterion Instead, both the use of an MATC and an estimated
ECio were considered lor the chronic \ alue An ECio was estimated by assuming the 0.0695
mg/L treatment represents an I X'i and estimating the ECio using the exposure response slope
from another PI 'OS toxicity study focused on another mussel species (Perna viridis).
SpeciIleally. the chronic exposure of l)ema viridis reported by Liu et al. (2013), which is
summarized in Section 3.1 14 1, was used to derive a ratio of EC10/EC35.4 levels from that
study, which was l-Ci>. IX'35.4 = 0.0033/0.0186 = 0.1770. Applying this ratio to Hazelton et al.
(2012) yields an estimated ECio of 0.0123 mg/L. Given the similarity between this ECio and the
author-reported MATC for Hazelton et al. (2012), the MATC of 0.01768 mg/L was used to
derive the chronic criterion for PFOS. This MATC is currently used quantitatively to derive the
draft chronic water column criterion, and EPA hopes to further refine this estimated ECio by

89


-------
obtaining the treatment level data from the study authors and exploring additional exposure
response slopes from the PFOS dataset. No other quantitative, chronic toxicity values were
available for this species or genus; therefore, the MATC of 0.01768 mg/L served directly as the
SMCV and GMCV used for deriving the chronic aquatic life criterion for PFOS.

3.1.1.3.3 Third Most Sensitive Freshwater Genus for Chronic Toxicity: I .nallagma (damselfly)
Bots et al. (2010) conducted a 320-day partial life-cycle study under renewal test

conditions to examine the effects of PFOS (tetraethylammonium salt, lW\. purity) on the

damselfly Enallagma cyathigerum. Approximately 40% of the nymphs in the control treatment

died during the first 60 days and similar mortality levels were observed in the other treatments.

However, it appeared that control survival plateaued between 60 and 200 days, with 82.57% of

the remaining nymphs in the control treatment sur\ i\ inu during this time, indicating that survival

settled out during this phase of the experiment The initial drop in nymph survival could likely be

attributed to the handling of the test organisms between the \ arious phases of the experiment.

This would explain the ohser\ ed plateau between (ฆ><) and 200 days, as the nymphs were not

handled during this time The obser\ ed control sur\ i val in this test was consistent with other

odonate tests and excessi\ e mortality of nymphs is typically expected within the first 200 days

given the difficulty in maintaining odonates in a lab setting (Abbott and Svensson 2007; Rice

2008). Therefore, the obser\ ed control survival for this study was considered within the

acceptable range lor this species up to the 200-day exposure duration. Further, the control

survival observed in this study was largely consistent with the toxicity testing guidelines for

chironomids (requiring 70% control survival; ASTM 2002; U.S.EPA 2000b), which are currently

the only test guidelines for an emergent aquatic insect as there currently is no test guideline for

odonates. Therefore, considerations regarding the use of these data for chronic criterion

90


-------
derivation were based on best scientific judgement and were restricted to the first 200 days of the
experiment.

The observed effects of PFOS on E. cyathigerum reported in the paper by the study
authors include decreased survival over the exposure duration and decreased metamorphosis
success. The MATC based on metamorphic success was less sensitive than for survival. As such,
the MATC author-reported value of 0.03162 mg/L for nymph sm \ i\ al was considered
quantitatively in the derivation of the aquatic life criteria. The remainder of I lie toxicity values
were used as supporting information to corroborate the toxicity value used lo dei i\ e the
freshwater chronic criterion and to better understand the effects of PFOS on aquatic insects. As
no other quantitative toxicity values were available for this species or genus, the author-reported
MATC of 0.03162 mg/L served directly as (lie SMCY'GMCV Additionally, EPA ran additional
analyses with some of the other toxicity \ allies lor /.. cvaihi^criiin to understand the influence of
this study on the o\eiall chronic criterion (see Section 4 2 2 Mow).

3.1.1.3.4 FourthMosi Sensitive I rcshunicr (icnus for Chronic Toxicity: Danio (zebrafish)

Wang ct nl. (2(111) e\ alualed the full lile-cycle effects of PFOS (> 96% purity) on Danio

rerio \ ia a sialic renewal study thai reported nominal exposure concentrations. This test
evaluated the effects of PFOS on a parental (F0) generation and included breeding trials to assess
the effects of PI-OS on an offspring (Fl) generation exposed via maternal transfer. Following the
receipt of treatment le\ el data from the study authors, EPA independently-calculated an ECio
value of 0.01650 mg/L for Fl survival. While this ECio has some uncertainty given the wide
spacing (lOx) of the treatment concentrations, this toxicity value was supported by others in the
PFOS toxicity literature (see Section 4.4.2.1.4 and Appendix G). Thus, this study and the ECio

91


-------
value for F1 survival was used quantitatively in the derivation of the aquatic life chronic
criterion.

Guo et al. (2019) evaluated the chronic effects of PFOS of unknown form and purity to
AB strain zebrafish (Danio rerio) males in a 21-day static-renewal, unmeasured study. Growth
as weight was the most sensitive endpoint at 21 days, with an author-reported NOEC and LOEC
of 0.02 and 0.04 mg/L PFOS, respectively. Although fish weight was the most sensitive endpoint
identified by the study authors, EPA was only able to independently calculate and ECio based on
mean body length. Therefore, EPA's independently-calculated ECio for the test hased on mean
body length (in cm) at 21 days is 0.06274 mg/L PFOS and was used quantitati\ ely to derive the
draft chronic water column criterion for freshwater.

Given the wide (lOx) spacing of 1 lie treatment concentrations in the full life-cycle test by
Wang et al. (2011) which creates some uncertainty with regards to deriving a more definitive
chronic point estimate lor the species. N\\ is using the two independently-calculated ECio
values from holli studies aho\ e (<> <> I (on and 0.06274 mg/L) to calculate the SMCV and GMCV
for D. rerio The geometric mean of the two I X' values is 0.03217 mg/L, which was used to
derive the freshwater chronic aquatic life criterion.

Table 3-7. Ranked l-"rcsh\valcr Genus Mean Chronic Values.

Rank"

GM( V
(mg/l. PI OS)

MI)U

Group'

Genus

Species

SMCV1'
(ing/l. PI-OS)

1

0.009676

F

Chironomus

Midge,

Chironomus dilutus

0.009676

2

0.01768

G

Lampsilis

Fatmucket,

Lampsilis siliquoidea

0.01768

3

0.03162

F

Enallagma

Blue damselfly,
Enallagma cyathigerum

0.03162

4

0.03217

B

Danio

Zebrafish,
Danio rerio

0.03217

92


-------
Ksink"

(i.MCV
(mg/l. PI OS)

MI)U

Croup'

(it'll IIS

Species

S.MCV1'
(ing/l. PI-OS)

5

0.06329

D

Daphnia

Cladoceran,
Daphnia carinata

0.003162

Cladoceran,
Daphnia magna

1.267

6

>0.1

A

Salmo

Atlantic salmon,
Salmo salar

>0.1

7

0.1555

B

Pimephales

Fathead minnow.
Pimephales promelas

0.1555

8

0.167

E

Procambarus

Crayfish,

Procambarus jallaxJ. vaginalis

0.167

9

0.1789

D

Moina

Cladoceran,
Moina macrocopa

0.1789

10

0.25

H

Brachionus

Rotifer,

Brachionus c nlyciflorus

0.25

11

0.5997

C

Xiphophorus

Swordlail lisli,
Xiphophorus helleri

0.5997

12

0.8872

C

Xenopus

African clawed iVog,
Xenopus laevis

> 1

Clawed frog.

Xeno/ >//.\ tropical is

0.7871

13

1.316

C

Lilhobates

Northern leopard frog,
Lilhobates pipiens

1.316

14

8.831

G

1 'hysella

Snail,

Physella heterostropha pomilia

8.831

a Ranked from the mosl sciimiin c in I lie mnsi inleiaiil basal on Genus Mean Chronic Value.
b From Appendi\ C \ccepiahle l ieshwalei C limine I'I 'OS Toxicity Studies
ฐMDR CuMiips idonlilied In lisi pan ided in Soclioii2.1u.l above.

93


-------
3

s

3

o

ฆ—-

S

S3

04

s

4/

ii

CL

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

0.0

ฆ Invertebrate (Other)

•	Invertebrate (Mollusk)
— Insect

~	Fish

A Amphibian

Pliysella
~ Lithobates
~ Xenopus
~ Xiphophorus
ฆ Brachionus

ฆ	Moina

ฆ	Procambaras
~ Pimephales

~ Salnio (non-definitive, greater than value)
Daphnia

~ Danio
- Enallagma
• Lampsilis
Chiionomus

0.001

0.01	0.1	1

Genus Mean Chronic Value (mg/L PFOS)

10

Figure 3-3. Ranked Freshwater Chronic PFOS Used Quantitatively to Derive the Criterion.

94


-------
3.1,1.4 Summary of Chronic PFOS Toxicity Studies Used to Derive the Saltwater Aquatic Life
Criterion

Data for chronic PFOS toxicity were available for three saltwater invertebrate species,
representing three genera and three families. The data available for saltwater fish fulfilled only
three of the eight MDRs.

Table 3-8. The Three Ranked Estuarine/Marine Genus Mean Chronic Values.

Ranked Below from Most to Least Sensitive. 		







GM( V









(mซ/l.



Uank

Genus

Species

PI-OS)

Comments

1

Perna

Asian green mussel.
i'erna viridis

I) 1)0.1.1

\ol a resident species in North
America

2

Americamysis

Mysid,

Americamysis
bahia

0.3708

North American resident species









Not a resident species in North
America, but other species in this

3

Tigriopus

Copepod,

Tigriopus japonic i is

o 7071

genus (Tigriopus) are common
ecotoxicity test species that serves as
a surrogate for untested copepod
species residing in North America.

3.1.1.4.1 Most Sciisinw r.smarmc \ larinc (icims: Perna (mussel)

I.in et al. (2013) e\ alualed llie chronic effects of PFOS-K (PFOS potassium salt, CAS#

2795->>-.v l>8% purity) on given mussels, Perna viridis, via a 7-day measured, static-renewal

study. Mussels were exposed ill a salinity of 25 ppt (artificial seawater) and a temperature of

25ฐC. PFOS concentrations were verified through water and muscle tissue samples via liquid

chromatography-tandem mass spectrometry. Weights and lengths were determined on days zero

and seven. An author-reported NOEC of 0.0096 mg/L and a LOEC of 0.106 mg/L was

determined for the growth condition index. EPA's independently calculated ECio for growth

condition index is 0.0033 mg/L. This ECio is used quantitatively to represent the chronic

sensitivity of this species to PFOS exposure in a marine/estuarine aquatic life dataset.

95


-------
3.1.1.4.2	Second Most Sensitive Estuarine/Marine Genus: Americamysis (mysid)

Drottar and Krueger (2000h) reported the results of a life-cycle, 35-day flow-through,

measured test of PFOS-K (potassium salt, 90.49% purity) with Americamysis bahia (formerly
Mysidopsis bahia). The 35-day NOEC (reproduction and growth) was 0.25 mg/L, and the
corresponding 35-day LOEC was 0.55 mg/L. An independently-calculated ECio could not be
defined at this time given the level of data that was presented in the paper (Appendix D). The
calculated MATC for the test was 0.3708 mg/L. This chronic \ nine was considered acceptable
for quantitative use despite the control survival of 78% because it was only slightly below the
80%) survival threshold, and because there were no other deficiencies in the study design.

3.1.1.4.3	Third Most Sensitive Estuarine/Marine Genus: / / i*r i opus (copepod)

A 20-day renewal, unmeasured full lile-cycle test with PI OS (analytical grade) was

conducted on the copepod, Tigriopus japonic us (non-Noilh American species) by Han et al.

(2015). The development of the copepod's growth from nauplii to copepodite and from nauplii to

adults was determined daily based on morphological characteristics. Results were presented as

the number of days needed to reach the normal development stages. The highest test

concentration (I mu I. PrOS) significantly increased the amount of time it took the copepods to

reach the de\ elopment stage Additionally, the authors assessed the reproduction of the copepods

by counting the nauplii produced by eight ovigerous females for 10 days in each well exposed to

PFOS. However, it was unclear if this was a subsampling of the organisms used in the 20-day

developmental test or if an independent assay with adult females. Results are presented

graphically as daily nauplii production/individual. There was a statistically significant decrease

in production (daily nauplii production/individual) in the 0.25, 0.5 and 1.0 mg/L PFOS

concentrations compared to the control. It was decreased by approximately 50%> in the highest

concentration (1 mg/L). The 20-day MATC based on time to reach development stage was

96


-------
0.7071 mg/L and was acceptable for quantitative use in the marine/estuarine chronic aquatic
toxicity dataset.

1.0

0.9 -

ฆ Invertebrate (Other)
• Invertebrate (Mollusk)

J 0.8

O

ฃ 0.7
>

C3

3

s

3 0.5

0.6 --

0.4 --

•X
a

cs
0ฃ

o

3 0.3 -
s


-------
the variability in the sensitivity of aquatic insects. Therefore, EPA will continue to seek
additional acute PFOS insect data to further understand the sensitivity of this taxon.

Additionally, EPA evaluated the effect of the current qualitative studies on the Final Acute Value
(FAV) in Section 4.2.1. Thus, the current development of an acute freshwater criterion was based
on seven of the eight MDRs.

GMAVs for 18 freshwater genera are provided in Table 3-.v and the four most sensitive
genera were within a factor of 2.2 of each other. The freshwater FAV. the 5'1' percentile of the
genus sensitivity distribution, for PFOS was 6.01 1 mg/L, and was calculated using the
procedures described in the 1985 Guidelines (U.S.EPA 1985) The FAV was lower than all of
the GMAVs for the tested species. The FAV was then divided by two to obtain a concentration
yielding minimal effects (see Section 2.9) The I AY'2. which is the acute freshwater criterion
(or criterion maximum concentration, CMC), was 3 <) nig I. PFOS (rounded to two significant
figures) and is expected to be protective of approximately l>5"0 of freshwater genera potentially
exposed to PFOS \ia direct aqueous exposure, under short-term duration conditions of one-hour,
when the criterion magnitude is not exceeded more than once in three years on average (Table
3-9)

98


-------
Table 3-9. Freshwater Final Acute Value and Criterion Maximum Concentration.	

Calculated l-'rcshwaler FAY based 011 4 lowest values: T0U1I Numberof(i\l \\ s in Daliisel IS





CMAV









Uank

(iCIlllS

(ing/l.)

ln((;.\l.\Y)

In(CMAV)2

P=U/(\+l)

s(i r l( P)

1

J'ii)ic/>/ia/cs

6.95U

1.94

3.76

U.U53

U.229

2

Oncorhynchus

7.515

2.02

4.07

0.105

0.324

3

Ligumia

13.5

2.60

6.77

0.158

0.397

4

Neocaridina

15.61

2.75

7.55

0.211

0.459



L (Sum):

9.31

22.15

0.53

1.41

S2 =

17.093



S = slope







L =

0.869



L = X-axis intercept





A =

1.794



A = In FAV







FAV =

6.011



P = cumulative probability





CMC =

3.0 mg/L PFOS

(rounded to two significant figures)





99


-------
1.0 T

0.9 --

s

0
+-*

u

1

K
>

0.7 --

0.6 --

s

a

es

Pi

a

4i

a

4J
&H

0.5 --

0.4 --

0.3 --

0.2

0.1 --

0.0

Invertebrate (Other)
Invertebrate (Mollusk)
Fish

Amphibian

Insect (Qualitative Data)
ฆCMC

0.1

~ Chironomus
# Physella
~ Lithobates
# Elliptio

ฆ	Brachionus

ฆ	Procanibams
~ Anaxyrus

~ Ambystoma
ฆ Daphnia
~ Danio
ฆ Dugesia
~ Hyla
ฆ Moina
• Lampsilis
~ Xenopus
ฆ Neocaridina
# Ligimiia
ฆ Oncorhynchus
~ Pimephales

1	10	100

Genus Mean Acute Value (mg/L PFOS)

1000

Figure 3-5. Ranked Freshwater Acute PFOS Used Quantitatively to Derive the Criterion.

Qualitative data for an insect species was taken into consideration to understand the relative sensitivity of aquatic
insects and is denoted by the hollow black box. The GMAV for this qualitative data was not used to derive the
freshwater acute criterion for PFOS.

3.2.1.2 Derivation of Acute Water Criterion for Estuarine/Marine Water

The estuarine/marine acute dataset for PFOS contained six genera (Table 3-5 and

Appendix B) representing only five of the eight taxonomic MDR groups. The missing MDR

groups included one family in the phylum Chordata, a family in a phylum other than Chordata,

and another family not already represented. The GMAVs of the four most sensitive definitive

estuarine/marine genera were within a factor of 4.5 of each other (Table 3-5).

Because data were available for only five of eight MDRs, EPA developed an

estuarine/marine acute benchmark using the available empirical data supplemented with toxicity

values generated through the use of New Approach Methods, specifically through the use of the

100


-------
EPA Office of Research and Development's peer-reviewed publicly-available weblCE tool
(Raimondo et al. 2010). This benchmark is provided in Appendix L.

3.2,1.3 Derivation of Chronic Water Criterion for Freshwater

The PFOS chronic dataset based on direct aqueous exposures contained data for all eight

MDRs, thus the Final Chronic Value (FCV) can be calculated directly without the use of an

ACR. There were GMCVs for 14 freshwater genera (Table 3-7) The four most sensitive genera

were within a factor of 3.3 of each other. The freshwater FCV for PFOS was <") 008398 mg/L,

calculated using the procedures described in the 1985 Guidelines (U.S.EFW llM5) The FCV is

the 5th percentile of the genus sensitivity distribution and is intended to be protecli\ e of 95

percent of the genera. The FCV was lower than all of the (iMCVs of the tested species. Unlike

the FAV, the FCV was not divided by two, as it already represents a low effect level, and was

equal to the water column chronic criterion (or criterion continuous concentration, CCC; Table

3-10). The freshwater ('('(' had a magnitude <> oon4 mg/L PFOS (rounded to two significant

figures), or 8.4 uu I.. and is expected to he protective of 95% of freshwater genera potentially

exposed to PFOS through direct aqueous exposure under long term conditions of four days, if not

exceeded more than once e\er\ three years on average (Table 3-10).

101


-------
Table 3-10. Freshwater Final Chronic Value and Criterion Continuous Concentration.

Calculated Freshwater FCV based on 4 lowest values: Total Number of GMCVs in Dataset = 14





GMCV









Rank

Genus

(mg/L)

ln(GMCV)

ln(GMCV)2

P=R/(N+1)

sqrt(P)

1

Chironomus

0.009676

-4.64

21.51

0.067

0.258

2

Lampsilis

0.01768

-4.04

16.28

0.133

0.365

3

Enallagma

0.03162

-3.45

11.93

0.200

0.447

4

Danio

0.03217

-3.44

11.81

0.267

0.516



ฃ (Sum):

-15.56

61.54

0.67

1.59

S2 =
L =
A =
FCV =

ccc =

26.35
-5.927
-4.780
0.008398

S = slope

L = X-axis intercept
A = InFCV

P = cumulative probability

0.0084 mg/L PFOS (rounded to two significant figures)

l.o T

0.9 -

ง 0.8

io,

ฃ

]ง 0.6 4
s
ฃ

?. 0.5

0.4 -

e

cs

06

o>

3 0.3 4
s

Ol

a

ซ 0.2

0.1 -

0.0
0.001

ฆ

Invertebrate (Other)

•

Invertebrate (Mollask)

-

Insect

~

Fish

~

Amphibian



-ccc

Phy sella #
~ Lithobates
~ Xenopus
~ Xiphophorus
ฆ Brachionus

ฆ	Moina

ฆ	Procambaras
~ Pimephales

~ Salmo (non-definitive, greater than value)
ฆ Daplinia
~ Danio
— Enallagma
# Lampsilis
Chiionomus

0.01	0.1	1

Genus Mean Chronic Value (mg/L PFOS)

10

Figure 3-6. Ranked Freshwater Chronic PFOS Used Quantitatively to Derive the Criterion.

102


-------
3.2,1.4 Deriving A Protective Duration Component of the Chronic Water Column-Based
Criterion

Effects to sensitive life stages was a primary reason why the 1985 Guidelines (U.S.EPA
1985) recommended a 4-day duration for most water column-based criteria. U.S.EPA (1985)
states, "An averaging period of four days seems appropriate for use with the CCC for two
reasons. With one of the two reasons specify being, "for some species it appears that the results
of chronic tests are due to the existence of a sensitive life stage al some lime during the test."

The SMCV for Chironomus dilutus and the Chinmomus GMC V (most sensitive genus)
are based on EPA's independently-calculated ECio of 0.05896 mg/L for a 1 0-da\ larval growth
test by MacDonald et al. (2004) and ECio of 0.001 5SS mg I. for a 16-day larval mean biomass
test by McCarthy et al. (2021). The ECio for a 10-day larval growth by MacDonald et al. (2004)
is slightly higher than the author-reported I X'i" for this effect in the study. The author-reported
ECios for the 20-day test by MacDonald et al were higher than those for the 10-day test, which
is an atypical outcome, and were not used for criteria deri\ alion. Consequently, there was no
clear influence of exposure time on the effects of H OS on this species.

The SMCY lor / ampsilis si/n/noulca and the Lampsilis GMCV (second most sensitive
genus) are hased on a 3(->-da\ study In (I la/elton 2d I 3; Hazelton et al. 2012) using glochidia and
juvenile life stages. The test exposed brooding glochidia (in marsupia) for 36 days followed by a
24-hour exposure of free glochidia. The 24-hour free glochidia exposure consisted of a factorial
design, such that free glochidia from the control group of the marsupia exposure were divided
between a control and the two PFOS treatments and the PFOS treatments were split into control
and the same PFOS treatment group as the marsupia exposure. This factorial design allowed for
the comparison of PFOS effects in two different life-stages.

103


-------
Given the limitations of time points that could be discerned by the test, it appeared that
for reduced viability and or metamorphosis success of free glochidia to occur at concentrations
near the chronic value for the test (0.01768 mg/L), the test's 36-day exposure period would also
be needed. For example, the study authors determined that the in-marsupia (36-day) exposure
held the greatest weight of evidence and explained 78% of the variability in the glochidia
viability (AIC = 22843, Wi = 0.78) and 83% of the metamorphosis success (AIC = 21955, Wi =
0.83). As a result, this species appears to be protected by the chronic 4-day duration component
of the water column criterion. It should also be noted, brief PFOS exposures al elevated
concentrations consistent with the magnitude and I -hour duration of the chronic criterion are not
expected to cause effects to free swimming glochidia based on the 24-hour acute toxicity data for
glochidia.

The SMCV for Enallagma cyathi^cnini and the t.nalla^ma GMCV (third most sensitive
genus) are based on a ป-cla\ partial life-cycle test by liols el al. (2010). Only a single
treatment, 0.1 mg I., showed purl in I effects. The treatment 10X higher (i.e., 1 mg/L) yielded
100%) mortality within 2d days The treatment in times lower (0.01 mg/L) showed no effects
over the entire test The authors pro\ ided the time course of mortality throughout the entire test.
At 0.1 mu I. a marked reduction in survival began at 130 days, and reached zero survival at 250
days, suggesting a relatively long time-to-effect. Because 0.1 mg/L is more than 3-fold higher
than the estimated chronic \ alue for the test, 0.03162 mg/L, it is postulated that the time course
of mortality observed at 0.1 mg/L would be substantially faster than what would be expected to
occur at 0.03162 mg/L. Given the relatively slow manifestation of chronic effects observed in
this study, this species appears to be protected by the chronic 4-day duration component of the
water column criterion.

104


-------
PFOS effects observed for chronically sensitive species indicate that a 4-day chronic
duration is appropriate. For example, the ECio for Danio rerio (fourth most sensitive genus) is
based on survival of the F1 offspring at eight days post fertilization (dpf), suggesting chronic
PFOS effects may occur in relatively brief periods of time following fertilization. Pronounced
effects occurring 8 dpf, with relatively lesser effects later in the course of the 5-month study
suggest the potential presence of a uniquely sensitive life stage occurring from 0-8 dpf.

No chronic PFOS toxicity tests specifically evaluated time-to-cfled. reported effect data
at time intervals at a high enough resolution to model the speed of toxic action, assessed time
variable PFOS exposures, or provided insight into the potential for latent toxicity. However,
chronic tests, including life cycle tests with relatively sensitive species suggested chronic effects
may occur at durations shorter than those of standard chronic toxicity tests (e.g., 28 days ELS)
and a chronic 4-day duration component of the water column criterion was considered protective
for these species genera There lore. N\\ has set the duration component of the PFOS chronic
water column criterion at four days to relied the chronic criterion duration recommended in the
1985 Guidelines This 4-day duration component of the chronic water column is also consistent
with (IS I-PA I1)1) I). which considered the default 4-day chronic averaging period as "the
shortest duration in which chronic effects are sometimes observed for certain species and
toxicants", and concludes that 4-day averaging "should be fully protective even for the fastest
acting toxicants."

3,2,1.5 Derivation of Chronic Water Criterion for Estuarine/Marine Water

The estuarine/marine chronic dataset for PFOS contained GMCVs for three genera.

GMCVs for three estuarine/marine genera are summarized in Section 3.1.1.4 and shown in

Figure 3-4. The eight-family taxonomic (MDR) requirement was not met by the chronic dataset,

as acceptable chronic studies for species representing five MDR groups are not available (two

105


-------
families in the phylum Chordata, a family in a phylum other than Arthropoda or Chordata, a
family in a phylum other than Chordata, and another family not already represented). The 1985
Guidelines allow the use of a Final Acute-Chronic Ratio (FACR) to convert a FAV to an FCV
(i.e., FAV/FACR = FCV), which is equivalent to a CCC. However, since an FAV could not be
calculated with the available data, an FCV also could not be calculated. Consequently, the EPA
could not derive estuarine/marine chronic criteria.

3.2.2	Derivation of Freshwater Chronic Tissue criteria for PFOS

Currently, the freshwater chronic PFOS toxicity data with measured tissue concentrations

were somewhat limited. There are 14 total freshwater aquatic life studies considered for either
quantitative (six studies - three fish, one invertebrate, and two amphibian studies) or qualitative
(eight studies) use in this aquatic life criterion The quantitati\ e studies only comprised data for
three of the eight MDRs. The qualitative studies pro\ ided supporting information for only one
additional MDR. Therefore, it was concluded that there is currently insufficient data to derive a
chronic tissue criterion using a (iSI) approach from empirical tissue data from toxicity studies.
However, these studies pro\ ided context to the translation of tissue criteria as described in
Section .1 2 .1 below This comparison is provided in the Effects Characterization (Section 4).

3.2.3	Translation of Chronic Water Column Criterion to Tissue Criteria

As described in Section 3.2.2 above, there are currently insufficient freshwater chronic

toxicity data with measured tissue concentrations to derive a chronic PFOS tissue criterion using
a GSD approach. Therefore, the chronic tissue criteria for PFOS were derived by translating the
chronic freshwater column criterion (see Section 3.2.1.3) into tissue criteria using
bioaccumulation factors (summarized in Section 3.2.3.1 below) and the following equation:
Tissue Criteria = Chronic Water Column Criterion x BAF (Eq. 1)

106


-------
The resulting tissue criteria corresponded to the tissue type from the BAF used in the equation.

3.2.3,1 PFOS Bioaccumulation Factors (BAFs)

Section 2.11.3.1 above summarizes the literature search, calculation, and evaluation of

the PFOS BAFs for aquatic life. These BAFs were compiled by and can be found in Burkhard

(2021). BAFs used in the derivation of the PFOS tissue criteria consisted of two or more water

and organism samples each and were collected within one year and 2 km distance of one another.

In order to derive more protective tissue criteria and to limit the effects of silc-specific

differences in BAFs, the distributions of BAFs used to derive tissue criteria were based on the

lowest species-level BAF reported at a site. When more than one BAF was available for the

same species within the same waterbody, the species-level BAF was calculated as the geometric

mean of all BAFs for that species at that site Summary statistics for the PFOS BAFs used in the

criteria derivation are presented in Table 3-1 1 and individual BAI-'s are provided in Appendix P.

In hlc 3-11. Sum nisi rv Sl:il isl ics lor PI-'OS liAls in l-'isli ami Invertebrates1.









20"'









(ieonielric

Median

('entile









Mean liA 1

liAl

liAl

Minimum

Maximum

Category

n

(l./kซ-\v\v)

(l./kซ-\v\v)

(l./kซ-\v\v)

(l./kซ-\v\v)

(l./kซ-\v\v)

ln\ ei'lebrales

2X

771.6

924

111.5

2.69

100,000

Fish (\\ hole-Body)

28

3,739

5,905

803.9

4.79

46,098

Fish (Muscle)

21

1,069

1,048

346.4

8.72

50,234

1 Based on tlic lovvesi species-le\ el 13 \F measured at a site (i.e., when two or more BAFs were available for the
same species at Ihc same silc. I lie species-level geometric mean BAF was calculated, and the lowest species-level
BAF was used).

The fish tissue criteria were developed for muscle and whole-body to accommodate the
most commonly sampled tissue types in monitoring programs. Additional tissue values for
various other tissue types (e.g., liver and blood) were also calculated and can be found in
Appendix Q.

107


-------
3.2.3.2 Deriving Protective Tissue Concentrations from the Chronic Water Column Criterion
Invertebrate whole-body and fish muscle and whole-body tissue criteria were derived

separately by multiplying the freshwater chronic water column criterion (see Section 3.2.1.3) by

the respective 20th centile of the distribution of BAFs using Equation (Eq. 1) from Section 3.2.3.

The 20th centile BAF was used to derive tissue-based criteria as a relatively conservative BAF

estimate in order to protect species across taxa and across water bodies with variable

bioaccumulation conditions. That is, use of the 20th centile B.\l protects species and conditions

where the bioaccumulation of PFOS and resultant tissue-based exposures is relali\ ely low as

well as those conditions with the bioaccumulation potential of PFOS is relati\ ely high

The invertebrate whole-body tissue criterion was ca leu kited by multiplying the 20th

centile BAF of 111.5 L/kg ww by the PI-OS freshwater chronic water criterion of 0.0084 mg/L,

resulting in an invertebrate whole-body tissue criterion of <) 937 mg kg ww. The fish whole-body

tissue criterion was calculated by multiplying the 20th centile liAF of 803.9 L/kg ww by the

PFOS freshwater chronic water criterion of 0.0084 mg/L, resulting in a fish whole-body tissue

criterion of 6.75 mg/kg w w The lisli muscle tissue criterion was calculated by multiplying the

20th centile 15.\Iฆ" of 34(-> 4 I. kg ww In the PFOS freshwater chronic water criterion of 0.0084

mg/L.. resulting in a fish muscle tissue criterion of 2.91 mg/kg ww. The chronic tissue-based

criteria are expected to be protective of 95% of freshwater genera potentially exposed to PFOS

under long-term exposures if the tissue-based criteria are not exceeded more than once in ten

years. The duration component of the tissue-based criteria is expressed as an instantaneous

duration because the tissue-based criteria are protective long-term conditions and represent an

integrated measure of bioaccumulated PFOS concentrations over time.

108


-------
3.2,3,3 Deriving A Protective Duration and Exceedance Frequency for the Tissue-based Chronic
Criteria

3.2.3.3.1	Duration: Chronic Criterion Tissue-Based Criteria

PFOS concentrations in tissues are generally expected to change only gradually over time

in response to environmental fluctuations. The chronic tissue-based criteria averaging period, or
duration, was therefore specified as instantaneous, because tissue data provide point, or
instantaneous, measurements that reflect integrative accumulation of PFOS over time and space
in population(s) at a given site.

3.2.3.3.2	Frequency: Chronic Criterion Tissue-In tsci / ('riteria

Ecological recovery times following chemical disturbances are situational-specific, being

largely dependent on: (1) biological variables such as the presence of nearby source populations
or generational time of taxa affected; (2) physical \ ariaMes such as lentic and lotic habitat
considerations where recovery rates in lenlic systems may be slower than lotic systems where the
pollutant may be quickly flushed dou n si ream, and; (3) chemical variables such as the persistence
of a chemical and potential lor residual effects (ii\ en the large variation in possible biological
and physical \ ariaMes inlluencinu ecological reco\ cry, EPA focused on the known chemical
attributes of PFOS to inform a recommended ten-year exceedance frequency for the chronic
tissue-based criteria

Metals and other chemical pollutants may be retained in the sediment and biota, where
they can result in residual effects over time that further delay recovery. Few studies are available
concerning PFOS elimination or depuration half-life in aquatic animals, however the data that
exist indicate a short half-life. For example, the elimination half-life for PFOS in adult rainbow
trout exposed to PFOS for 28 days via the diet followed by 28 days depuration was estimated to
be 8.4 days in muscle tissue (Falk et al. 2015), while the terminal half-life in rainbow trout

109


-------
receiving a one-time intra-arterial injection of PFOS was 86.8 days (Consoer et al. 2016).
Additionally, the depuration half-life in northern leopard frog tadpoles via a 40-day aqueous
exposure to 0.01 mg/L PFOS was estimated to be 2.2 days (Hoover et al. 2017). It is unclear
whether PFOS half-life in aquatic organism tissues is the mechanistic result of rapid depuration
or an artifact of these measurements taken during relatively short testing times (e.g., 28 days)
where a lack of steady state between PFOS and water and tissues has not occurred. Long-term
uptake and subsequent excretion rates of PFOS has been extensively studied in humans relative
to aquatic life. (Li et al. 2018) reported a median PFOS half-life of 3.4 years in human serum
following exposure to PFOS in drinking water, which authors stated was in the range of
previously published estimates. Due to chemical retention in tissues, ecosystems impacted by
discharges of bioaccumulative pollutants (such as selenium) leaner from chemical disturbances
at relatively slow rates. For example, Lemly (1W7) concluded that although water quality in
Belews Lake in North Carolina (a freshwater reservoir) had recovered significantly in the decade
since selenium discharges were halted in 1985, the threat to fish had not been eliminated. The
selenium dischargers that led to se\ ere re|">roductive failure and deformities in fish, was still
measurable (fish deformities) in I wi (se\ en years later) and in 1996 (ten years later). Lemly
(1997. pu 2X<)) estimated based on these data that "the timeframe necessary for complete
recovery from selenium contamination from freshwater reservoirs can be on the order of
decades

Beyond bioaccumulation, chemical-specific considerations such as degradation vs.
persistence may also provide a mechanism influencing ecological recovery rates. The persistence
of PFOS has been attributed to the strong C-F bond, with no known biodegradation or abiotic
degradation processes for PFOS. Somewhat similarly, as elements, metals do not degrade and

110


-------
may persist in aquatic systems following elevated discharge. The persistence of metals may
explain why metals had the second longest median recovery time of any disturbance described in
a systematic review of aquatic ecosystem recovery (Gergs et al. 2016). Gergs et al. (2016)
showed recovery times following metal disturbances ranged from roughly six months to eight
years (median recovery time = 1 year; 75th centile ~ 3 years; n = 20).

The bioaccumulative nature and persistence of PFOS in aquatic systems, including
sediments (Ahrens 2011), in combination with the documented recovery limes of pollutants with
similar chemical attributes (Gergs et al. 2016; Lemly 1997), suggested 10 years was a protective
exceedance frequency for the tissue-based criteria for PFOS The tissue-based criteria are
protective if they are not exceeded more than once in ten years lo allow sufficient time for PFOS
concentrations built up in tissues and source reser\ oirs in the freshwater system to diminish
while simultaneously providing freshwater organisms adequate lime lo recover following
elevated PFOS exposures in tissues

EPA acknowledges thai there is uncertainty in deriving protective tissue criteria
magnitudes In transforming the chronic water column criterion (which was based on tests that
only added PI OS to the water column) into tissue concentrations through field-measured
bioaccumulation data of paired water and tissue concentrations in waterbodies. Nevertheless, the
chronic water column criterion is based on chronic toxicity tests that fed test organisms. In these
tests, PFOS can directly affect species based on direct water column exposure and/or sorb to
added food that is consumed by test organisms before eliciting chronic effects from dietary
exposure. Therefore, the chronic water column criterion magnitude accounts for water column-
based and, to a possible lesser extent, dietary-based effects, while the field-based BAFs account
for water column- and dietary-based PFOS exposure in tissues. The tissue criteria will provide

111


-------
information to states, tribes, and stakeholders on potential effects to aquatic organisms based on
aquatic tissue monitoring data. Quantitatively acceptable data on the effects of dietary exposures
to aquatic species were relatively limited, thus EPA elected to develop protective values for
aquatic organism tissues based on the observed relationship between water column
concentrations and tissue concentrations and observed PFOS toxicity in chronic tests where
PFOS was only added directly to the water column.

3.3 Summary of the PFOS Aquatic Life Criteria

The PFOS aquatic life criteria were developed to protect freshwater aquatic life against

adverse effects, such as mortality, altered growth, and reproductive impairments, associated with
acute and chronic exposure to PFOS. The nationally recommended criteria include water column
based acute and chronic criteria for freshwatci s The freshwater acute water column-based
criterion magnitude is 3.0 mg/L, and the chronic water column-based criterion magnitude is
0.0084 mg/L (Tabic 3-12) The chronic freshwater tissue-based criteria magnitudes are 6.75
mg/kg wet weight (\\\\) lor fish u hole-body. 2.91 mg/kg ww for fish muscle tissue and 0.937
mg/kg ww for in\ ertebrate u hole-body tissue These PFOS aquatic life criteria are expected to
be protect i\ e of aquatic life on a national basis (Table 3-12). All of these water column and tissue
criteria are intended to be independently applicable and no one criterion takes primacy. All of the
above recommended criteria (acute and chronic water column and tissue criteria) are intended to
be protective of aquatic life Acute and chronic water column criteria for estuarine/marine waters
could not be derived at this time due to data limitations; however, an estuarine/marine acute
benchmark protective of aquatic life is provided in Appendix L.

The freshwater chronic water column criterion is more strongly supported than the
chronic tissue-based criteria because the water column-based chronic criterion was derived

112


-------
directly from the results of empirical toxicity tests. The chronic tissue-based criteria are
relatively less certain because they were derived by transforming the chronic water column
criterion into tissue concentrations through BAFs, with any uncertainty and variability in the
underlying BAFs then propagating into the resultant tissue-based criteria magnitudes.

Table 3-12. Draft Recommended Perfluorooctane Sulfonate (I'l-'OS) Criteria for the
Protection of Aquatic Life in Freshwaters.			i	





Chronic

Chronic

Chronic





Acute \\aid-

\Yater

Inverlehrale

lisli





Co! ii m n

Column

Wliole-

\\ hole-

Chronic Fish

Type/Media

(CMC)14

(CCC)15

liodv12

Bod v1-2

Muscle12

Magnitude

3.0 mg/L

0.0084 mg/L

ii l)37 nig kg

WAV

(•> 75 nig kg

WW

2 ^ 1 nig kg

WW

Duration

1 hour average

4 dav averaee

Instantaneous3



Not to be

Not Id he

Not to be exceeded more than once in ten



exceeded more

exceeded more

\ ears on a\ erau

e



Frequency

than once in
three years on
average

than once in
three years on
average







1	All five of these water column and tissue criteria are intended to be independently applicable and no one criterion takes
primacy. All of the above recommended criteria (acute and chronic water column and tissue criteria) are intended to be
protective of aquatic life. These criteria are applicable throughout the year.

2	Tissue criteria derived from the chronic water column concentration (CCC) with the use of bioaccumulation factors and are
expressed as wet weight (ww) concentrations.

3	Tissue data provide instantaneous point measurements that reflect integrative accumulation of PFOS over time and space in
aquatic life population(s) at a given site.

4	Criterion Maximum Concentration; applicable throughout the water column.

5	Criterion Continuous Concentration: applicable throughout the water column.

113


-------
4 EFFECTS CHARACTERIZATION FOR AQUATIC LIFE

The purpose of this section was to describe the supporting information for the derivation
of the PFOS aquatic life criteria that contributed to the weight-of-evidence for the derivation.
This section includes: (1) comparison of quantitative insect data used to derive freshwater
criteria (Section 4.1); (2) additional analyses supporting the criteria that were used as part of the
lines-of-evidence discussion to better understand the influence of using less certain toxicity data
(Section 4.2); (3) assesses the influence of including non-North American resident species in
criteria derivation (i.e., species not resident to North America removed from dataset; Section
4.3); (4) provides summaries of the toxicity studies with apical endpoints (e.g., effects on
survival, growth, or reproduction) that were not used directly to derive the criteria, but were used
qualitatively to support the PFOS criteria (Section 4 4). (5) e\ aliiation of the acute insect MDR
through the use of inter-species correlation esti mates (Section 4 5); (6) discussion of acute to
chronic ratios (Section 4 (•>): (7) comparison of empirical tissue concentrations to translated tissue
criteria (Section 4.7): and (S) discussion of the effects of PFOS on aquatic plants (Section 4.8).
EPA is proposing the national recommended PI OS aquatic life criteria described in the Effects
Analysis Section (see Section 3 abo\e). The additional analyses presented here are solely
intended to support the PFOS criteria through a weight-of-evidence approach that evaluated the
influence of data \ ariation and uncertainties on the PFOS criteria.

4.1 Comparison of Quantitative Data used to Derive Freshwater Criteria

4,1.1 Aquatic Insects

While comparing the effects of PFOS across studies presented several challenges,

especially with differences in test species, methodologies, exposure durations, and observed

endpoints, in general there appeared to be several similarities and few differences between the

three aquatic insect toxicity studies used quantitatively to derive the PFOS chronic freshwater

114


-------
criterion. In all three studies (Bots et al. 2010; MacDonald et al. 2004; McCarthy et al. 2021)
effects of chronic exposures to PFOS on survival and/or emergence were observed in damselfly
(Enallagma cyathigerum) and midge (Chironomus dilutus), respectively. Additionally, all three
studies measured effects of PFOS on growth. However, since the studies focused on very
different life stages for this endpoint (on growth in emerged adult damselfly from Bots et al.
(2010) and on growth in larval midge from MacDonald et al. (2fii)4) and McCarthy et al. (2021),
the toxicity data for growth could not be compared.

During the early phases and exposure durations of the experiments, il appeared that the
effects of PFOS are not similar between the two species and that midge is more sensitive than the
damselfly. This is particularly true for McCarthy et al. (2021) where the author-reported 16-day
ECio was 0.00136 mg/L. However, after this initial phase, the effects of PFOS on damselfly and
midge became more similar. In the later phases of ilie tests, the independently-calculated ECio of
0.0171 mg/L and 1 lie author-reported 20-day M.VI'C of <> 11454 mg/L for chironomid survival in
MacDonald et al (2<)i)4) were similar to the author-reported 150-day MATC of 0.03162 mg/L
for damselfly sui \ i\ al The test organisms at these phases of each respective test likely occurred
in a similar life stage (later de\ elopment and about to undergo metamorphosis). Therefore, they
were more comparable than any of the other survival toxicity values from these studies (i.e., the
10-day values lor damselfly and the 10-day values for midge), which were focused on the effects
of PFOS on much less comparable instars, especially given that odonates have a much longer
development and life span compared to midges.

These results indicated that PFOS exposures to aquatic insects in later life stages are
likely similar. These apparent similarities in the chronic effects of PFOS to aquatic insects
provided support to the toxicity values quantitatively used and to the ranking of these two

115


-------
species (as first and third most sensitive) to derive the chronic freshwater criterion. However,
additional replicate level data would be helpful to fully understand the observed effects of PFOS
individually and to compare across the full PFOS toxicity dataset.

4.2 Additional Analyses Supporting the Freshwater Criteria

4.2.1 Additional Analyses Supporting the Derivation of Acute Water Column Criterion for

Freshwater

In addition to the acute freshwater criterion of 3.0 mg/L PFOS described above, two
additional analyses supporting the derivation of the acute criterion were conducted to consider
the effect of including less certain toxicity data. These additional analyses were conducted as
part of a line-of-evidence discussion to better understand the inlluence of using less certain
toxicity data in the acute dataset. The data considered to be less certain centered on the difficultly
in reliably ascertaining the relative sensili\ ily of aquatic insects to acute exposures of PFOS
based on the current toxicity literature for PI ()S (see Section 3 I 1.1.6).

The two additional analyses presented below either included or excluded data from two
toxicity studies (TaMe 4-1) The additional analyses presented here evaluated the influence of
data variation on the deri\ ation of the PI OS criteria The inclusion of the two qualitatively
acceptable studies demonstrated variability in the sensitivity of aquatic insects to acute exposures
of PFOS Therefore, EPA will continue to seek additional acute PFOS insect data to further
understand 1he sensiti\ ily of this taxon. Thus, the current development of an acute freshwater
criterion was based on se\ en of the eight MDRs. The availability of additional toxicity data for
these particular taxa would reduce the uncertainty in the analysis. The criteria presented in
Section 3.3 are EPA's best estimate of the maximum concentrations of PFOS that will support
protection of aquatic life from acute PFOS exposures.

116


-------
Table 4-1. Additional Analyses Supporting the Derivation of the Acute Water Column
Criterion for Freshwater.

Presented in the order that is summarized in text below. *

Order of
Additional
Analyses

Purpose of
Additional Analysis

Details of Additional
Analysis

Acute Water

Col 1111111
Concentration
for Additional
Analysis

(mg/l.)

Study

1

To explore the
impact of using
qualitative insect
data to fulfill
missing aquatic
insect MDR

Used author-reported
LCso of 1.18 mg/L for
yellow fever mosquito
(Aedes aegypti)

0.72

Olson
(2017)

2

Used author-reported
LCso of 182.12 mg/L
for chironomid
(Chironomus
plumosus)

3.1

Yang et al.
(2014)

*Final derived acute freshwater criterion was 3.0 mg/L PFOS.

In the first additional analysis, the toxicity data for the in\ asive pest species yellow fever

mosquito (Aedes aegypti) from Olson (2017) were used to deri\ c an exploratory acute criterion

magnitude. Including the author-reported I ,('so of 1.18 mu I. lYom Olson (2017) in the derivation

resulted in a freshwater I AY of I 44o mu I. and an acute water column concentration of 0.72

mg/L (Table 4-2. I'.W of I 44o nig I. 2 n 72 mg/L). Including the LCso of 1.18 mg/L for A

aegypti decreases the criterion magnitude In a factor of 4.2 below EPA's recommended estimate

of the maximum concentration of PLOS that will support aquatic life from exposure of 3.0 mg/L.

EPA concluded that the inclusion of a qualitative LCso for A aegypti in the agency's acute

criterion dataset is unwarranted given that: I) A. aegypti is an invasive pest species; 2) the study

was missing important exposure details; and 3) the author-reported LCso and concentration-

response curve could not be assessed by EPA on a statistical basis since model parameters were

not provided, and there were insufficient treatment level data to independently calculate toxicity

values (see Sections 3.1.1.1.6 and G.2.1.5). Additionally, EPA expects that the chronic water

column criterion of 0.0084 mg/L, which is over two orders of magnitude lower than this

117


-------
calculated acute water column value, will likely be the driving magnitude for permits and
assessments. Until such a time when additional toxicity data on aquatic insects are available to
fully understand the potential acute effects of PFOS on aquatic insects, especially considering the
comparison between qualitative data for midge and mosquito, which indicated very different
sensitivities among insects, EPA concluded that the acute criterion derived from the current acute
dataset will adequately protect 95% of the species 99% of the lime.

In the second analysis, the qualitative LCso value for chironomid of I S2.12 mg/L from
Yang et al. (2014) was used (Table 4-2). Including this value increases the "V in the criterion
calculation by one and results in an alternative freshwater l-'.W for PFOS of 6.1 57 mg/L (Section
3.2.1.1; U.S.EPA 1985) and an acute water column concentration of 3.1 mg/L PFOS (rounded to
two significant figures). This second analysis indicated that the qualitative chironomid LCso for
Chironomus plumosus had very little influence on the magnitude of the freshwater CMC for
PFOS. Additionally, this test was considered for qualitati\ e use since the test organisms were
from a problematic source (from the lieijinu City Big Forest Flower Market with no details of
pervious exposures to PI-OS or other chemicals provided) and no further quantification of
previous exposure to contaminants or husbandry was provided (Yang et al. 2014). Therefore,
EPA determined to proceed with the criterion derivation as previously described in Section
3.2.1.1. Again, additional toxicity data on aquatic insects are needed to fully understand the
potential acute effects of PI OS on aquatic insects and modify the acute criterion if necessary.
EPA will continue to seek additional acute PFOS insect data to further understand the sensitivity
of this taxon.

118


-------
Table 4-2. GMAVs Used in Derivation of Acute Criterion and Additional Analyses
Supporting the Acute Criterion for Freshwater.			

MI)U

Croup4

Genus

Species

Acute
Criterion

Additional
Analyses

CMAV
(mg/L PI-OS)1

I'irst2

Second-'

F

Aedes

Yellow lever mosquilo,

Aedes aegypti

-

1.18

-

B

Pimephales

Fathead minnow,
Pimephales promelas



6.950

6.950

A

Oncorhynchus

Rainbow trout,
Oncorhynchus mykiss

7.515

7.515

7.515

G

Ligumia

Black sandshell,
Ligumia recta

13.5

13.5

13.5

E

Neocaridina

Japanese swamp shrimp.
Neocaridina denticulaia

15.61

1 5 M

15.61

C

Xenopus

African clawed frog,
Xenopus laevis

15.99

15.99

15.99

G

Lampsilis

Fatmucket,

Lampsilis silit/iionlea

16 5

16.5

16.5

D

Moina

Cladoceran,
Moina macrocopa

17.20

17.20

17.20

C

Hyla

Gray treefrog.
Hyla versicolor

19.88

19.88

19.88

G

Dugesia

Planaria.

lhigesia /apomca

22.48

22.48

22.48

B

Tkmio

Zebraiish.
/ Vmio rerio

24.44

24.44

24.44

D

/ kip/inia

Cladocuan,
/ ktphma carinata

42.30

42.30

42.30

Cladoceran,
/ ktphnia magna

Cladoceran,
/ ktphnia pulicaria

C

Ambystoma

Jefferson salamander,
Ambystoma jeffersonianum

47.40

47.40

47.40

Small-mouthed salamander,
Ambystoma texanum

Eastern tiger salamander,
Ambystoma tigrinum

C

Anaxyrus

American toad,
Anaxyrus americanus

56.49

56.49

56.49

E

Procambarus

Crayfish,

Procambarus fallax f virginalis

59.87

59.87

59.87

119


-------
MI)U

Croup4

Cenus

Species

Acute
Criterion

Additional
Analyses

CMAV
(mg/l. I'l OS)'

First2

Second-'

H

Brachionus

Rotifer,

Brachionus calyciflorus

61.8

61.8

61.8

G

Elliptio

Eastern elliptio,
Elliptio complanata

64.35

64.35

64.35

C

Lithobates

American bullfrog,
Lithobates catesbeiana

loo 24

109.24

109.24

Green frog,
Lithobates clamitans

Northern leopard frou.
Lithobates pipiens

Wood frog,
Lithobates sylvatica

G

Physella

Bladder snail,
Physella acuta

172.1

172.1

172.1

Snail,

Physella heterosnopha pomilia

F

Chironomus

Midge,

('hironomus plumosus

-

-

182.12

Acute Water Column Concentration (mg/L)

3.0

0.72

3.1

1	GM AVs as presented in lahle ^ in Section 3.1.1.1. Genera presented in order of rank according to the criterion
derivation.

2	Additional analysis Willi I lie inclusion of qualitative yellow fever mosquito data.

3	Additional analysis with I lie inclusion of qualitative chironomid (('. ptumosus) data.

4MDR Groups identified In lisi pro\ ided in Section 2.10.1 above.

120


-------
4.2.2 Additional Analyses Supporting the Derivation of Chronic Water Column Criterion for

Freshwater

In addition to EPA's recommended chronic freshwater criterion of 0.0084 mg/L PFOS
described above in Section 3.2.1.1, six additional analyses supporting the derivation of the
chronic criterion were examined as part of a line-of-evidence evaluation to consider the effect of
including less certain toxicity data on the magnitude of the chronic criterion. The data considered
to be less certain generally centered around two specific areas (I) the difficultly in reliably
estimating a chronic toxicity value given the wide spacing (up to 15-lbkl difference) of the
treatment concentrations in Hazelton et al. (2012) and Bots et al. (2010) (see Section 3 1.1.3.2
and 3.1.1.3.3, respectively) and (2) the uncertainly in the chronic toxicity values ui\ en the level
of data presented in the papers (see Appendices C.2.2 and C.2.3)

The six additional analyses presented below either changed the toxicity value or excluded
data from two toxicity studies (Table 4-3). The additional analyses presented here are solely
intended to support the PTOS chronic criterion through a weiuht-of-evidence approach that
evaluated the influence of data \ arialion on the criterion derivation process. Based on these
additional analyses. I-PA decided to retain the danisellly and fatmucket values as presented in
Section 3 I I 3, to ensure protection of these sensitive taxa as well as the many untested species
for which the damselfly and lalmucket may serve as serve as representative taxonomic surrogate
species. The a\ ailahililv of additional toxicity data for these particular taxa would reduce the
uncertainty in the analysi s The criteria presented in Section 3.3 are EPA's best estimate of the
maximum concentrations of PFOS that will support protection of sensitive aquatic life from
unacceptable chronic exposures.

121


-------
Table 4-3. Additional Analyses Supporting the Derivation of the Chronic Water Column
Criterion for Freshwater.

Presented in the order that is summarized in text below. *

Order of
Additional
Analyses

Purpose of
Additional
Analysis

Details of Additional
Analysis

Chronic Water

Column
Concentration
for Additional
Analysis
(mg/l.)

Study

1

To explore the
impact of using
the various author
reported toxicity
values for
damselfly

Used 10-day MATC of
0.3162 mg/L for damselfly
instead of 150-day MATC of
0.0316 mg/L

() ()069

Bots et al.
(2010)

2

Used 60-day NOEC of 0.1
mg/L damselfly instead of
150-day MATC of 0.0316
mg/L

O.OOW

3

Used 320-day NOEC of <) <>|
mg/L damselfly instead of
150-day MATC of 0.031 o
mg/L

0.0063

4

To explore the
impact of using
the MATC for
fatmucket

Removed MATC ofo <) 1 70S
mg/L for fatmucket

0.0079

Hazelton et
al. (2012)

5

To explore the
impact of using
both the EC io for
fatmucket and the
150-day MATC
for damselflv

Removed both MATC of
0.001768 mg/L for
fatmuekel and 150-day
MATC ofo i)3 16 mg/L for
damselfly

0.0053

Hazelton et
al. (2012) and
Bots et al.

(2010),
respectively

o

To explore the
impact of using
the ECiofor
liilmucket

Use estimated ECio of
0.0123 mg/L for fatmucket
instead of MATC of 0.01768
mg/L

0.0070

Hazelton et
al. (2012)

*Final derived chronic licsliw ;iler criierion was 0.0084 mg/L PFOS.

In the first additional analysis, instead of using the 150-day MATC of 0.0316 mg/L for
Enallagma cyathigerum as described in the final criterion description above in Section 3.2.1.3,
the 10-day MATC of 0.3162 mg/L was used (Table 4-4; Bots et al. 2010), yielding a freshwater
FCV for PFOS of 0.006932 mg/L. This chronic water column concentration of 0.0069 mg/L

122


-------
PFOS (rounded to two significant figures) is slightly lower (more protective) than the final
chronic value of 0.0084 mg/L derived above. This first additional analysis indicated that there is
little difference in the calculated chronic criterion based either on the 150-day or 10-day MATC
for E. cyathigerum. However, as the 150-day MATC was more comparable to the other aquatic
insect data and more representative of life cycle effects than the 10-day MATC, EPA has
concluded that the 150-day MATC should be used quantitati\e1y lo clci i \ e the chronic freshwater
criterion.

In the second analysis, instead of using the 150-day MATC of 0.031 b mu I, for
Enallagma cyathigerum, the 60-day NOEC of 0.1000 mg/I. from the same test was used (Table
4-4;Bots et al. 2010), also yielding an FCV of 0.06932 mg/L (Section 3.2.1.3; U.S.EPA 1985).
Similar to the first analysis, there is little difference in the calculated chronic criterion based
either on the 150-day or 60-day NOEC for /.. cyaihigernm. However, since the 150-day MATC
was more comparable lo the oilier aquatic insect data and representative of life cycle effects than
the 10-day MATC. I-1ป.\ has concluded that the 150-day MATC should be used quantitatively to
derive chronic freshwater criterion

In the third analysis, instead of using the 150-day MATC of 0.0316 mg/L for Enallagma
cyathigerum. the 320-day NOI-C of <> 0100 mg/L from the same test was used (Table 4-4; Bots et
al. 2010), yielding an FCV of <) 006334 mg/L. This analysis indicated that there is about a 1.3-
fold difference (lower) in the calculated chronic criterion if the 320-day NOEC for is.
cyathigerum is used. However, as there were concerns with the control survival of test organisms
(reported as roughly 60% in the first 60 days), EPA has determined that the 150-day MATC
should be used quantitatively to derive chronic freshwater criterion since this toxicity value still

123


-------
represents a life cycle effect and control survival of test organisms was determined to be
acceptable at this time point in the test.

In the fourth analysis, the MATC for fatmucket (Lampsilis siliquoidea) of 0.01768 mg/L
was removed from the chronic dataset to understand the influence of this toxicity value on the
criterion magnitude (Table 4-4). This additional analysis placed the GMCV of 0.06329 mg/L for
Daphnia among the four most sensitive genera, and yielded an FCV of 0.007895 mg/L (Section
3.2.1.3; U.S.EPA 1985). The removal of the chronic toxicity value lor /. siliquoidea has only a
modest influence on the calculated chronic criterion magnitude, but would eliminate mollusks
from the chronic PFOS dataset. EPA decided to retain the fatmucket value to ensure
representation and protection of this sensitive taxon.

In the fifth analysis, the 150-day M ATC of <"> <~>3 162 mu I. for damselfly (Enallagma
cyathigerum) and MATC for fatmucket (/ ampsilis silii/nonk-a) of 0.01768 mg/L were removed
since these values tire less eerltiin compared lo other qutinliltili\ e studies in the chronic criterion
dataset (Table 4-4) As noted tihun e. these toxicity values were considered to be less certain due
to (1) the difficultly in reliably estimating ti chronic toxicity value given the wide spacing (15-
fold difference in I Itizellon et til (2<> 12) and 10-fold difference in Bots et al. (2010) of the
treatment concentrations and (2) the uncertainty in the chronic toxicity values given the level of
data presented in the papers This fifth analysis yielded a freshwater FCV for PFOS of 0.005272
mg/L. Similar to the pre\ ious additional analysis looking at the influence of changing the chronic
value for E. cyathigerum lo the 300-day survival NOEC, the calculated chronic criterion
magnitude was reduced 1.6-fold. EPA decided to retain the damselfly and fatmucket values as
presented in Section 3.1.1.3 above in the current derivation, to ensure representation and
protection of these sensitive taxa.

124


-------
Lastly in the sixth analysis, the estimated ECio for fatmucket of 0.0123 mg/L was used in
the chronic dataset to understand the influence of this estimated toxicity value on the criterion
derivation (Table 4-4), particularly since EPA was not able to fit a curve to estimate an ECio
given that there were only two PFOS treatment groups and the gap in these exposure
concentrations is large (about 15-fold). This additional analysis yielded an FCV of
0.007055mg/L. This additional analysis indicated that the estimated lovicity value from L.
siliquoidea has a modest influence on the calculated chronic criterion. Since the estimated
toxicity value had a modest influence on the recommended ('('(' \ alue, the author-reported
MATC was used instead.

125


-------
Table 4-4. GMCVs Used in Derivation of Chronic Criterion and Additional Analyses Supporting the Chronic Criterion for
Freshwater.

Ml)l\

(ฆ i< >i 11>

(.inn-.



( Inonii
( lilrlioll

Ailililiniiiil An;il\-i-

<;\l< \
mm 1 N <>ni'

1 iiNi-

fremiti"

1 liiiil-

I'Miirlli'

1 Mill1

sixlli'

F

Chironomus

Midge,

Chironomus dilutus

0.009676

0.009676

0.009676

0.009676

0.009676

0.009676

0.009676

G

Lampsilis

Fatmucket,

Lampsilis siliquoidea

0.01768

0.01768

0.01768

0.01768

-

-

0.0123

F

Enallagma

Blue damselfly,

Enallagma cyathigerum

0.03162

0.3162

0.1

0.001

0.03162

-

0.03162

B

Danio

Zebrafish

Danio rerio

0.03217

0.03217

0.03217

0.03217

0.03217

0.03217

0.03217

D

Daphnia

Cladoceran,

Daphnia carinata

0.06329

0.06329

0.06329

0.06329

0.06329

0.06329

0.06329

Cladoceran,

Daphnia magna

A

Salmo

Atlantic salmon,

Salmo salar

>0.1

>0.1

>0.1

>0.1

>0.1

_6

>0.1

B

Pimephales

Fathead minnow,

Pimephales promelas

0.1555

0.1555

0.1555

0.1555

0.1555

0.1555

0.1555

E

Procambarus

Crayfish,

Procambarus fallax f. virginalis

0.167

0.167

0.167

0.167

0.167

0.167

0.167

D

Moina

Cladoceran,

Moina macrocopa

0.17X9

0.1789

0.1789

0.1789

0.1789

0.1789

0.1789

H

Brachionus

Rotifer,

Brachionus calyciflorus

0.25

0.25

0.25

0.25

0.25

0.25

0.25

C

Xiphophorus

Swordtail fish,

Xiphophorus helleri

0.5997

0.5997

0.5997

0.5997

0.5997

0.5997

0.5997

C

Xenopus

African clawed frog,

Xenopus laevis

0.8872

0.8872

0.8872

0.8872

0.8872

0.8872

0.8872

Clawed frog,

Xenopus tropicaus

C

Lithobates

Northern leopard frog,
Lithohi ncs pipiens

1.316

1.316

1.316

1.316

1.316

1.316

1.316

G

Physella

Snail.

Physella heterostropha pomilia

8.831

8.831

8.831

8.831

8.831

8.831

8.831

Chronic Wiitcr Column Concentration

0.0084

0.0069

0.0069

0.0063

0.0079

0.0053

0.0070

' GMCVs as presented in Table 3-7 in Section 3.1.13. Genera presented in order of rank according to the criterion derivation. Order of GMCVs were not
changed for the additional analyses.

2	Additional analysis with changes to toxicity value for /<'. cyathigerum.

3	Additional analysis with the exclusion of L. siliquoidea.

4	Additional analysis with the exclusion of L. siliquoidea and E. cyathigerum.

5	Additional analysis with the changes to toxicity value for L. siliquoidea.

126


-------
6	Not ranked among the four most sensitive taxa so not included in the chronic freshwater criterion calculations per the decision rule for greater than toxicity
values (see Section 2.10.3.2).

7	MDR Groups identified by list provided in Section 2.10.1 above.

127


-------
4.3 Influence of Using Non-North American Resident Species on PFOS
Criteria

EPA conducted two additional analyses of the freshwater criteria by reducing the limited
toxicity datasets to organisms that are resident to, or have been introduced and have established
populations in the conterminous U.S. These analyses were conducted to determine the sensitivity
of the criteria calculations to the inclusion of data for taxa that arc not resident species to North
America but serve as surrogates for other sensitive organisms This analysis was conducted for
both the acute and chronic freshwater datasets only, since the estuarine/marine dalasets are
limited even when all species are included.

4.3.1 Freshwater Acute Water Criterion with Native and I .suihlished Organisms (Species Not

Resident to North America removed from datascl)

Four species were removed in the analy si s of a freshwater acute water criterion with
native, reproducing, or established organism in the conterminous U.S.: Japanese swamp shrimp
(Neocaridina denticiiluiu). planarian (Ihigesiajapotiica). /chralish (Danio rerio) and cladoceran
(Daphnia carinaki). Remo\ al of these species truncated the freshwater acute dataset to 22
species (Table 4-5) The lYeshu ater acute dalasel still retained one missing MDR group (an
insect), which was addressed through inclusion of qualitative data. The Japanese swamp shrimp
ranked fourth (Table 3-3) and all other species mentioned above (planarian, zebrafish, and
cladoceran) were not among the four most acutely sensitive species. The acute water column
concentration was 2 S mg I. PI 'OS (Table 4-6) when using the reduced dataset which was
slightly lower than the recommended CMC of 3.0 mg/L. This value is lower than all of the
GMAVs in Table 3-3. EPA decided to retain the full acute dataset and associated acute criterion
for PFOS of 3.0.mg/L in order to have the largest, high-quality dataset to serve as surrogate

128


-------
species for the broad range of the thousands of untested species present in the freshwater

environment in the U.S.

Table 4-5. Ranked Freshwater Genus Mean Acute Values with Native and Established

Rank"

GMAV
(ing/l. PIOS)

MI)U

Croup'

Genus

Species

S.M.W h
(mg/l. PI OS)

1

6.950

B

Pimephales

Fathead minnow.
Pimephales prom das

6.950

2

7.515

A

Oncorhynchus

Rainbow trout,
Oncorhynchus mykiss

7.515

3

13.5

G

Ligumia

Black sandshell,
Ligumia recta

13.5

4

15.99

C

Xenopus

African clawed frog,
Xenopus laevis

15.99

5

16.5

G

Lampsilis

Fatmuckcl.

Iximpsilis sihquoidea

16.5

6

17.20

D

Moina

Cladoceran,

\ foil hi macrocopa

17.20

7

19.88

C

Hyla

Gray livcfrog,
l/y/a versicolor

19.88

8

47.40

C

Ambysuma

Jefferson salamander,
Ambysioma jeffersonianum

51.71

Small-mouthed salamander,
Am by stoma texanum

30.00

Eastern tiger salamander,
Am by stoma tigrinum

68.63

9

5o 4l>

C

. hiaxyrus

American toad,
Anaxyrus americanus

56.49

10

5l> S7

E

Procamharus

Crayfish,

Procambarus fallax f virginalis

59.87

11

61.S

H

Brachionus

Rotifer,

Brachionus calyciflorus

61.8

12

64.35

G

Elliptio

Eastern elliptio,
Elliptio complanata

64.35

13

80.92

D

Daphnia

Cladoceran,
Daphnia magna

48.87

Cladoceran,
Daphnia pulicaria

134

14

109.24

C

Lithobates

American bullfrog,
Lithobates catesbeiana

133.3

Green frog,
Lithobates clamitans

113

129


-------


CMAV

MI)U





S.MAV1'

Rank"

(mg/1. PIOS)

Croup'

Conns

Species

(mg/L PI OS)









Northern leopard frog,
Lithobates pipiens

72.72









Wood frog,
Lithobates sylvatica

130









Bladder snail,

183.0

15

172.1

G

Physella

Physella acuta

Snail,

161.8









Physella heterosimpha pomilia

a Ranked from the most sensitive to the most tolerant based on Genus Mean Acme Value
b From Appendix A: Acceptable Freshwater Acute PFOS Toxicity Studies
0 MDR Groups identified by list provided in Section 2.10.1 above

Table 4-6. Calculation of Freshwater Acute Water Column Concentration with Native and
Established Organisms (Species Not Resident to North America Removed from Dataset).

Calculated Freshwater FAV based on 4 lowest values: Total Number of GMAVs in Dataset =

= 15





GMAV









Rank

Genus

(mg/L)

ln(GMAV)

ln(GMAV)2

P=R/(N+1)

sqrt(P)

1

Pimephales

6.950

1.94

3.76

0.063

0.250

2

Oncorhynchus

7.515

2.02

4.07

0.125

0.354

3

IJgumia

13.5

2.60

6.77

0.188

0.433

4

Xa tonus

15.99

2.77

7.68

0.250

0.500



— (Sum):

9.33

22.28

0.63

1.54

S2 =

14 wo



S = slope







L =

i) 84 5



L = X-axis intercept





A

1 71 1



A = InFAV







FAV

5 535



P = cumulative probability





Acute Water













Column













Concentration =

2.8 mg/L PI OS

(rounded to two significant figures)





4,3.2 Freshwater Chronic Water Criterion with Native and Established Organisms (Species Not
Resident to \orih America removed from dataset)

Three species were removed from the chronic freshwater dataset that are not native or
established organism in the conterminous U.S.: zebrafish (Danio rerio), cladoceran (Daphnia
carinata) and the clawed frog (Xenopus tropicalis). One species, the zebrafish (Danio rerio) is
not native to the conterminous U.S., although, there have been reports of zebrafish in the wild in

130


-------
several different locations in the U.S. The current status of any such zebrafish populations in the
U.S. is uncertain (U.S.FWS 2018). Additionally, zebrafish are an accepted test species and are
used in analyses in TSCA and FIFRA (U.S.EPA 2016c). Removal of these species truncated the
freshwater chronic dataset to 12 species representing 12 genera (Table 4-7). Additionally, the
non-definitive value for the Atlantic salmon was removed, since the value would rank in the top
four and its use is not indicative of true species sensitivity, as an n niton tided greater than value.
The revised freshwater chronic dataset consisted of seven of the eight MDRs The zebrafish
ranked fourth when all species were included, and the other species mentioned above (cladoceran
and clawed frog) were not among the four most chronically sensiti ve species. Remo\ al of the
species that are not resident to North America reduced the I CV and chronic water column
concentration (Table 4-8). The chronic water column concentration was 0.004001 mg/L PFOS
when using the reduced dataset and was 2.1 times lower than the recommended chronic criterion
of 0.0084 mg/L. The chronic water column concentration from this additional analysis was also
2.1 times lower than the lowest (iMCYs (Table 4-7). Therefore, EPA decided to retain the full
chronic dataset and associated chronic criterion for PFOS of 0.0084 mg/L in order to have the
largest, high quality dataset to ser\ e as surrogate species for the broad range of the thousands of
untested species present in the freshwater environment in the U.S.

131


-------
Table 4-7. Ranked Freshwater Genus Mean Chronic Values with Native and Established
Organisms.					



GM( V

MI)U





S.MCY1'

Uank"

(ing/l. PI OS)

(•roup''

Genus

Species

(ing/l. PIOS)

1

0.009676

F

Chironomus

Midge,

Chironomus dilutus

0.009676

2

0.01768

G

Lampsilis

Fatmucket,

Lampsilis siliquoidea

0.01768

3

0.03162

F

Enallagma

Blue damselflv.
Enallagma crai/ngcrum

0.03162

4

0.1555

B

Pimephales

Fathead minnow.
Pimephales promdas

0.1555

5

0.167

E

Procambarus

Crayfish,

Procambarus fa/lax f virgina/is

0.167

6

0.1789

D

Moina

Cladoceran,
Moina macrocopa

i) 1789

7

0.25

H

Brachionus

Rotifer.

Brachionus calyciflorus

0.25

8

0.5997

C

Xiphophorns

Swordtail fish.
Xiphophorns he/len

0.5997

9

> 1

C

Xenopus

African clawed IVou.
Xenopus laevis

> 1

10

1.267

1)

1 kiplmia

Cladoceran.
Daphnia magna

1.267

11

1.316

(

/ uliohates

Northern leopard frog,
Li! ho hates pipiens

1.316

12

8 831

(i

rhyscllii

Snail,

Physella heterostropha pomilia

8.831

a Ranked from the mosi sciimiin e in I lie mnsi lolerant based on Genus Mean Chronic Value.
bFrom \ppendix C: Acvvpiahle l ieshwalei C hronicPFOS Toxicity Studies
ฐMDK (niuips identifiedb> lisi pm\ ided in Section2.10.1 above.

132


-------
Table 4-8. Calculation of Freshwater Chronic Water Column Concentration with Native
and Established Organisms.	

Calculated Freshwater FCV based on 4 lowest values: Total Number of GMCVs in Dataset = 12





GMCV









Rank

Genus

(mg/L)

In(GMCV)

ln(GMCV)2

P=R/(N+1)

sqrt(P)

1

Chironomus

0.009676

-4.64

21.51

0.077

0.277

2

Lampsilis

0.01768

-4.04

16.28

0.154

0.392

3

Enallagma

0.03162

-3.45

11.93

0.231

0.480

4

Pimephales

0.1555

-1.86

3.46

0.308

0.555



ฃ (Sum):

-13.99

53.19

0.77

1.70

S2 =

99.86



S = slope







L =

-7.756



L = X-axis intercept





A =

-5.521



A = InFCV







FCV =

0.004001



P = cumulative probability





Chronic Water













Column













Concentration =

0.0040 mg/L PFOS (rounded to two significant figures)





4.4 Qualitatively Acceptable Water C olumn-Based Toxicity Data

Several studies were identified as either not meeting EPA's data quality guidelines for

inclusion in the criteria derivation or did not have data available to support the independent

calculation of a toxicity value (e g . I .("*ฆฆ and or ECio). However, these studies were used

qualitatively as supporting information to the PFOS criterion derived to protect aquatic life and

provide additional e\ idence of the observed toxicity and effects of PFOS, including the relative

sensiti\ ities of surrogate, untested species The key studies with apical endpoints (e.g., effects on

survival, growth, or reproduction) that were used qualitatively in the derivation of the PFOS are

summarized below, grouped as either acute or chronic exposures and sorted by relative

sensitivity of genera used to derive the criteria (following the previous study summaries included

in the Effects Analysis). Qualitative study summaries within a factor of two of the final acute and

chronic values were also included below and arranged according to taxonomic relatedness.

NOEC and LOEC values were provided in several of the following study summaries to be

representative toxicity values for comparison to the toxicity values summarized in the Effects

133


-------
Analysis section above (see Section 3) and used in the criteria calculation. The toxicity values
summarized here were not used quantitatively to derive the acute or chronic PFOS freshwater
criteria. Results of each individual study (as well as the rationale why a study was not
quantitatively acceptable) were considered relative to the corresponding freshwater criterion
magnitude to ensure the water column based PFOS criteria were not underproductive and to
provide additional supporting evidence of the potential toxicity of PI OS to aquatic organisms.
The toxicity values summarized as part of this Effects Characterization were not used in any
quantitative analysis or in the derivation of the PFOS aquatic life criterion. Detailed study
summaries and corresponding tabulated data for the studies summarized below, as well as
additional qualitative study summaries of less sensitive taxa, were included in Appendix G.

4.4.1 Consideration of Qualitatively Acceptable Acute Data

4.4.1.1 Qualitatively Acceptable Acute Data for Species Among the 1'our Most Sensitive
Genera Used to Derive the Acute Water Column Criterion

4.4.1.1.1	Most acutely sensin i v genus. Pimephales

There were no qualilali\ ely acceptable acute tests with the genus, Pimephales.

4.4.1.1.2	Second most acutely sensitive genus, Uncorhynchus

There were no qualitati\ ely acceptable acute tests with the genus, Oncorhynchus.

4.4.1.1.3	Ihirtl mosi aciiie/y sensitive genus, Ligumia

There were no qualitatively acceptable acute tests with the genus, Ligumia.

4.4.1.1.4	Fourth most acutely sensitive genus, Neocaridina

There were no qualitatively acceptable acute tests with the genus, Neocaridina.

134


-------
4.4,1.2 Consideration of Relatively Sensitive Tests with Freshwater Species based on
Qualitatively Acceptable Acute Data

4.4.1.2.1 Genus: Danio (zebrafish)

Danio rerio was evaluated by Huang et al. (2010) in a 114-hour static measured exposure

to PFOS (CAS # 1763-23-1, > 96% pure). The 114-hour LCso was 2.2 mg/L PFOS, and the
malformation ECso was 1.12 mg/L PFOS. These data were considered qualitative due to
exposure duration, which was longer (at 114 hours) than fish acute toxicity test guidelines (U.S.
EPA 2016b).

A sub-chronic static unmeasured test was conducted by Ulhaq et al. (2<) 13) to determine
the toxicity of PFOS to D. rerio. The 144-hour LCso was ~ in mg/L PFOS. This toxicity value
was considered qualitative because of test duration, which was longer than the fish acute toxicity
test guidelines (OCSPP 850.1075) recommending exposures of hours (U.S.EPA 1985;
U.S.EPA 2016b).

Martinez el al (2<)|i);i) c\alualcd the acute effects of perfluorooctane sulfonate potassium
salt on zebralish (Ikimo rerio) o\ cr three days, yielding a LOEC value of 10.0 |iM, or 5.382
mg/L PFOS (calculated using the molecular weight of 538.22 g/mol for PFOS-K) for growth as
body length The study is acceptable lor qualilali\ e use because of the short test duration, which
was shorter than the 96-hour exposure period recommended by the fish acute toxicity test
guidelines (US I-PA N85. I S. EPA 2016).

(Ortiz-Villanue\ a et al. 2018) also evaluated the acute effects of perfluorooctane
sulfonate on zebrafish (Danio rerio) following two days of exposure. The author-reported LOEC
was 2.0 |iM (1.0 mg/L) for malformations, and 20 |iM (10 mg/L) for survival (based on a
molecular weight of 500.13 g/mol PFOS). The short test duration made the study acceptable for
qualitative use only.

135


-------
(Vogs et al. 2019) evaluated the acute effects of perfluorooctane sulfonic acid potassium
salt on zebrafish (Danio rerio) embryos in a 118-hour study. An author-reported 118-hr LCso was
3.8 |iM PFOS (or 2.045 mg/L based on molecular weight of 538.22 g/mol PFOS-K) for
mortality. However, this study is acceptable for qualitative use because of the short test duration.

(Truong et al. 2014) evaluated the acute effects of potassium perfluorooctanesulfonate
(PFOS-K, CAS # 2795-39-3) and perfluorooctane sulfonic acid (PI OS. CAS #. 1763-23-1) on
zebrafish (Danio rerio) following 114 hours of exposure. The authors reported 114-hr mortality
LOEC of 61.44 |iM PFOS-K (or 33.07 mg/L based on a molecular weight of 53S 22 g/mol) and
6.4 |iM PFOS (or 3.2 mg/L based on a molecular weight of 5<~>f) 13 mg/L). Both values are
qualitative only because of the atypical test duration, u liich was longer than the 96-hour
exposure period recommended by the fish acute toxicity test guidelines (U.S. EPA 1985; U.S.
EPA 2016). Only the PFOS LOEC of 3.2 nig I. was within a factor of two of the FAV.

The noted 1o\icit\ \ allies provided in each study summary above (2.2, > 10, 5.382, 10,
2.045, 3.2 mg/T,. respecti\ ely). either comprising of author-reported LCso or LOEC values,
indicated that this genus might he more sensiti\ e to acute exposures of PFOS than the
quantitati\e data for the genus (with a (i\l.\V of 24.44 mg/L). These qualitative values were
either below or within a factor of two of the FAV of 6.011 mg/L. EPA, however, concluded
these toxicity studies do not suggest/), rerio is a relatively sensitive species to acute PFOA
exposures because ll\ e of the six quantitatively-acceptable acute tests for this species reported
LCso values (range = 3.502 - 71.12 mg/L; geometric mean = 24.44 mg/L; n = 6) that were more
than two times greater than the FAV.

136


-------
4.4.2 Consideration of Qualitatively Acceptable Chronic Data

4.4.2.1 Qualitatively Acceptable Chronic Data for Species Among the Four Most Sensitive
Genera Used to Derive the Chronic Water Column Criterion

4.4.2.1.1 Most chronically sensitive genus, Chironomus

Stefani et al. (2014) conducted a chronic (10 generation) test of PFOS (form and purity

not reported) with the midge, Chironomus riparius. The NOEC and LOEC were 0.0035 and >

0.0035 mg/L (as time-weighted average) as there were no effects on emergence, reproduction, or

sex ratio at this concentration. The results from this study were not acceptable for quantitative

use because only a single test concentration was used, the chronic value is a greater than low

value and not informative for criterion development, and there was a lack of details pertaining to

the characteristics of the sediment used in the exposure, including details regarding any

differences in measured concentrations over the duration of the exposure. Since this study was

focused on the chronic effects of PFOS to a relatively sensitive species, however, consideration

of the greater than chronic \ alue in the context of other values for the midge was prudent. The

ECios for Chironomus d/lii/ns of1 >5S96 mg/L and 0.001588 mg/L from MacDonald et al.

(2004) and McCarthy et al (2<)21). respecti\ ely. that were used quantitatively in the chronic

criterion cleri\ alion are more robust \ allies than the toxicity value reported in Stefani et al.

(2014). and likely a better estimation of the sensitivity of C. riparius. The chronic value reported

by Stefani et al. (2<) 14) is low er than the chronic criterion but it was only representative of a

NOEC and no effects were observed at 0.0035 mg/L (as time-weighted average). The

Chironomus GMCV is greater than the chronic freshwater criterion of 0.0084 mg/L, and thus,

the species is expected to be protected.

In a companion paper to Stefani et al. (2014), Marziali et al. (2019) similarly conducted a

chronic (10 generation) test of PFOS (form and purity not reported) with C. riparius. The LOEC

based on F1 developmental time and F1 adult weight was < 0.004 mg/L (time-weighted

137


-------
average). The were no effects on F1 exuvia length at this concentration. The results from this
study were not considered for quantitative use because only a single test concentration was used,
there was a lack of consistent observed effects in both the control and the treatment groups
across the generations, and details pertaining to the characteristics of the sediment used in the
exposure were lacking, including details regarding any differences in measured concentrations
over the duration of the exposure. Again, it is prudent to consider the less than chronic value
from the study in the context of the more robust and definitive values lor mi due. That is, the
chronic values established for the related midge, (ihliiiiis. IV0111 MacDonakl el ill (2004) and
McCarthy et al. (2021) are more reliable and definili\ e \ allies representing the sensitivity of the
genus in the chronic criteria dataset.

4.4.2.1.2	Second most chronically sensitive genus. / ani/>sths

There were no qualitatively acceptable chronic tests with the genus, Lampsilis.

4.4.2.1.3	Third most chronically sensitive genus, llnallagma

There were no qualitati\ ely acceptable chronic tests with apical endpoints for this genus.

4.4.2.1.4	l:ourth most chronically sensitive genus, Danio

l)ii et al. (2(H)1)) in\ csliualcd the effect of PFOS (> 99% purity) on the survival, growth

and hepatotoxicity of Danio rerio female fry exposed via renewal, unmeasured conditions for 70
days. The 70-day M.VI'C for increased malformation and decreased survival of F1 fish was
reported as 0.0224 mg/L PFOS. An independently-calculated ECio could not be determined as
the treatment level data required for analysis appears to have been lost (personal communication
with Bingsheng Zhou, corresponding study author). The author reported MATC of 0.0224 mg/L
for increased F1 malformation and decreased survival was similar to the independently-
calculated ECio of 0.01650 mg/L for F1 survival from Wang et al. (2011) and ECio of 0.06274

138


-------
for mean body length from Guo et al. (2019), which were used quantitatively in the freshwater
chronic criterion.

Cui et al. (2017) investigated the toxic effects of PFOS (> 96% purity) to Danio rerio in a
near full life-cycle (unmeasured) static renewal test. Breeding trials were also carried out to
produce F1 offspring (F0 females were paired with F0 males from the same treatment group).
Malformation and survival rate of both generations were evaluated, and the study authors
indicate that F1 offspring derived from the parental fish exposed to 0.25 mu I. were observed to
have severe deformities (including uninflated swim bladders, bent spine, pericardial edema, yolk
sac edema, and necrosis) and low survival rates. A M ATC of 1118 mg/L was calculated from
the author-reported values for effects on altered sex ratio (female dominance) and low F1
offspring survival. However, an independently-calculated toxicity \ alue could not be determined
with the data provided in the paper. The author-reported M.VI'C was used qualitatively since
EPA was unable to independently \ erily the reported toxicity \ alue with the data provided in the
paper. The author reported M.VI'C of <) IMS ing/L for decreased F1 offspring survival was
substantially higher than the independently-calculated ECio of 0.0165 mg/L from Wang et al.
(201 I) and IX"iฆ ฆ of <)  0.40 mg/L). The study author reported
NOEC, LOEC, and M.VI'C for growth as both total body length and weight were 0.20, 0.40, and
0.2828 mg/L, respectively. This 15-day growth MATC of 0.2828 mg/L was roughly one order of
magnitude higher than the FCV of 0.008398 mg/L and higher than the ECios calculated for both
quantitative studies used for criteria derivation. Shi et al. (2009) was for a shorter exposure
duration (15 days in a rapid early-life stage test compared to 150 days in a full life-cycle test and

139


-------
21 day early-life stage test) with a less sensitive endpoint (growth compared to reproduction).
The SMCV of 0.03217 mg/L is expected to be more representative of the sensitivity of this
species.

Danio rerio embryos were also investigated by Keiter et al. (2012) in a long-term flow-
through measured study with PFOS-K (potassium salt, CAS # 2795-39-3, > 98% purity). This
test was considered qualitative as there were complications wilh the lest design (see full
description in detailed study summary in Appendix G.3.2.4) and a poor concent ration-response
relationship with the endpoints evaluated. There was also a wide (> 1 OOx) difference between the
NOEC and LOEC for the study. For example, the NOEC and I OEC for F1 and 12 I SO-day male
lengths and weights, and F2 180-day female weights, was 0.0006 and 0.1 mg/L. The LOEC for
the remaining growth endpoints (lengths and weights across all generations) was < 0.0006 mg/L.
In contrast, the MATC for F2 180-day survival was <> I 732 niu'I. 1'he MATC of 0.1732 mg/L is
substantially higher than ECi-s from the Wang et al. (201 I) and Guo et al. (2019) studies used
quantitatively to dcri\e the chronic criterion

Chen et al (2<) I (ฆ>) e\ aluated the estrogenic effects of PFOS (> 96% purity) to Danio rerio
via renewal, unmeasured tests Body length was measured on individual fish while body weight
was obtained by pooling samples of l<> fish per sample and calculating averages. The 42 dpf
LOEC based on an increase in condition index (beneficial effect) was 0.250 mg/L PFOS. These
data are classified as qualitative because there was only one exposure concentration. The LOEC
of 0.250 mg/L was one order of magnitude higher than the SMCV of 0.03217 mg/L for D. rerio
that is expected to representative of the sensitivity of this species.

Jantzen et al. (2017) evaluated the effects of PFOS on the morphometric, behavioral and
gene expression in Danio rerio exposed via 5-day static, unmeasured exposures (OECD Method

140


-------
212). The five-day (plus nine days for observation) NOEC, LOEC, and MATC for growth as
total body length were 0.02, 0.2, and 0.06325 mg/L, respectively. The LOEC was associated
with only a 3.8% decrease in growth compared to control. This study was considered for
qualitative use due to the short exposure duration and small (negligible) effect. The MATC of
0.06325 mg/L was similar to the SMCV for D. rerio.

Sharpe et al. (2010) examined the bioaccumulation and toxicity of PFOS isomers on
Danio rerio through three different tests, a 96-hour renewal toxicity test on adults, a 48-hour
renewal toxicity test on embryos, and a chronic exposure test that evaluated maternal transfer
and effects on fecundity of PFOS isomers. The 96-hour test was used quantitali\ ely lo derive the
acute water column criterion (see Appendix A). The 48-hour tests were used qualitatively and
are summarized in Section G.2.2.3. A 14-day chronic exposure was conducted to examine PFOS
accumulation and changes in isomer profiles in response lo maternal transfer. A 21-day exposure
was also conducted to test the potential of PFOS to reduce fecundity.

Fecundity was reduced 34".. relative to control in fish exposed to 0.5 mg/L PFOS for 14
days and 47% in fish exposed 21 days. The results of this study were considered qualitative as
the tests consisted of just one experimental concentration of 0.5 mg/L and because one of the two
control replicates was lost, w liich the study authors note was due to unusual aggression among
the test organisms The aulhor-reported LOEC of 0.5 mg/L was one order of magnitude higher
than the SMCV lor / rcno

Chen et al. (2013) examined the behavioral effects of zebrafish resulting from prolonged
chronic exposure to PFOS. Adult Danio rerio (US-AB strain) used for spawning were
maintained following standard protocols. At approximately three months of age, F0 adults from
the same treatment group were bred. Embryos (Fl) hatched from these adults were monitored for

141


-------
developmental progression, and hatched larvae were monitored for 8 dpf for malformation and
mortality. Second generation (Fl) zebrafish were maintained in water free from PFOS. F1 larvae
hatched from FO adults from the 21-120 dpf and 1-120 dpf groups showed much higher rates of
mortality and malformation than the control and the 1-20 dpf groups. The author-reported LOEC
was 0.250 mg/L. The results of this study were considered qualitative as the test consisted of one
experimental concentration (of 0.250 mg/L) prohibiting point estimation. The author-reported
LOEC of 0.250 mg/L was one order of magnitude higher than the SMCV lor /'. rerio.

(Bao et al. 2019) evaluated the chronic effects of perfluorooctane on ju\ enile zebrafish
(Danio rerio), via a 21-day study. An endpoint for fecundity was not reported In the authors, and
there were no significant differences between the exposure concentrations in terms of growth
length or weight (NOEC > 0.2 mg/L PFOS) \o mortality nas ohsei'\ ed Independently
calculated ECios could not be calculated as I'PA was unaMe to fit a model with significant
parameters. Therefore. ui\ en N\\ was unaMe to independently calculate toxicity values based
on the level data pro\ ided in the paper In the study authors, the test duration was a partial-life
cycle test as opposed to the preferred lile-eyele lest for which there were studies on this species
(Wang el al 2<>l I). and the author-reported toxicity values results in a NOEC > 0.2 mg/L this
stud) was used qualitatively to deri\ e the draft chronic water column criterion.

The noted toxicity \ allies provided in each study summary above (0.0224, 0.1118,
0.2828, <0.0006, n 25'). n <)t>325, 0-5, 0.250, and > 0.2 mg/L), comprising of a mix of author-
reported NOEC, LOEC, and MATC values, indicated varying sensitivity to chronic exposure of
PFOS for this genus when compared to the quantitative data for the genus (with an
SMCV/GMCV of 0.03217 mg/L). However, it was difficult to compare these papers since the
present studies either: (1) only included one exposure concentration prohibiting definitive point

142


-------
estimation; (2) were of unacceptably short chronic exposure duration; (3) involved the
assessment of different endpoints; and (4) included complications from test designs. The SMCV
of 0.03217 mg/L that was used quantitatively in the chronic criterion derivation is expected to be
more robust than any of the individual toxicity values reported in these qualitative studies, and
therefore, is expected to be adequately protective of this genus.

4.4,2.2 Consideration of Relatively Sensitive Tests with Freshwater Species based on
Qualitatively Acceptable Chronic Data

4.4.2.2.1 Genus: Daphnia (cladoceran)

Jeong et al. (2016) conducted a 25-day chronic life-cycle renewal, unmeasured lest of

PFOS-K (potassium salt, purity 99%) with Daphnia magna The NOEC based on reproduction in

the F0 generation was 0.010 mg/L. The 25-day LOEC was 0.1 <~>f) mg/L. The calculated MATC

was 0.03162 mg/L, and the independently-calculated IX'io was 0.0<)4I mg/L (Appendix G).

Independent statistical analyses were conducted using data that were estimated (using Web plot

digitizer) from the figures presented in the paper This independently-calculated ECio value was

not considered reliable. howe\ er. as this \ alue was much lower than the author reported NOEC,

because the test concentrations were widely spaced (with each treatment group increasing by one

order of magnitude and ranging between n.0001 and 10 mg/L). Furthermore, the data presented

in the paper were control normalized, and there was no consistent concentration-response

relationship. While the toxicity value from the study was within a factor of two of the FCV of

0.008398 mg/L indicating that this species might be more sensitive to chronic exposures of

PFOS than the quantitative data for the genus indicates, it is still greater than the FCV by an

order of magnitude. Additionally, while the SMCV of 1.267 mg/L is substantially greater than

the chronic value from the study, the preponderance of other acceptable chronic values for D.

magna precludes a major change in species sensitivity.

143


-------
4.4.2.2.2 Genus: Oryzias (medaka)

Ji et al. (2008) evaluated the chronic toxicity of PFOS to the Japanese medaka, Oryzias

latipes, via unmeasured renewal exposures. The author-reported 14-day NOECs for F0 (parental
generation) adult survival, condition factor and adult male GSI and HSI were all > 1 mg/L PFOS,
whereas the 14-day adult female HSI and GSI MATC and LOEC were 0.3162 and < 0.01 mg/L
PFOS, respectively. For the F1 (progeny generation), the MATCs lor percent hatchability, time
to hatch, and swim-up success were all 0.3162 mg/L PFOS, and the lar\ al growth (as organism
weight and length) and non-apical GSI LOECs were < 0.01 mg/L PFOS. In the latter (F1
generation), the reduction in fish weight at 0.01 mu I. was only 12% compared to controls, thus,
the concentration-response curve for weight was shallow I <> mg/L yielded a 29% reduction in
weight. Many of these toxicity values, particularly those for apical endpoints, suggested that this
genus is likely less sensitive than the one of the four most sensitive genera that drive the chronic
criterion, which is 0084 mu T. The magnitude of the chronic criterion is expected to be
protective of the low I .()!ฆ(' for lar\ al growth in the study of < 0.01 mg/L, which appears to be
the most sensitive endpoint in the study The results of the apical endpoints evaluated in this
study were considered to he <.|ualitati\ely acceptable primarily because of a lack of replication
during the egg stage of the I ' I generation

4.5 Evaluation of the Acute Insect Minimum Data Requirement through
Interspecies Correlation Estimates (ICE)

The acute dalaset for PFOS contained 18 genera (Table 3-3) representing seven of the
eight taxonomic MDR groups. The missing MDR was a representative from an insect family.
Evaluation of qualitatively acceptable insect data (i.e., Yang et al. 2014) relative to the acute
criterion magnitude was the primary line of evidence used to inform insect sensitivity to acute
PFOS exposures (see section 3.1.1.1.6). Acute insect LCso data were estimated using web-ICE

144


-------
and compared to the acute criterion as a secondary line of evidence to evaluate insect sensitivity
to acute PFOS exposures.

EPA's web-ICE tool is described in detail in Appendix L.l. Briefly, ICE models are log-
linear regressions of the acute toxicity (EC50/LC50) of two species across a range of chemicals,
thus representing the relationship of inherent sensitivity between those species (Raimondo et al.
2010). ICE models can be used predict the sensitivity of an untested tavon (predicted taxa are
represented by the y-axis) from the known, measured sensitivity of a surrogate species
(represented by the x-axis). This analysis focused on all possible ICE models thai used insects as
a predictor species (i.e., y-axis) and a corresponding surrogate input species (i.e . \-a\is) for
which a SMAV (see Table 3-3) was available. These models are shown in Table 4-9 along with
use classifications for each individual model based on a host of statistical metrics described by
(Willming et al. 2016) see box one of Appendix I. I lor additional discussion on model use
criteria).

145


-------
Table 4-9. All ICE models available in web-ICE v3.3 for predicted insect species based on surrogates with measured PFOS.

Model parameters are used to evaluate prediction robustness. Cross-validation success is the percentage of all model data that were predicted within 5-fold of the
measured value through leave-one-out cross-validation (Willming et al. 2016). Taxonomic distance describes the relationship between surrogate and predicted
species (e.g., 1 = shared genus, 2 = shared family, 3 = shared order, 4 = shared class, 5 = shared phylum. 6 = shared kingdom).

PrcdicU-d Species

Siiitiiป:iIc Species

Sllipc

llllcrccpl

l)cปrccs

<>r

I'rivdiim


0.98

0

0.04

9.91

14500000

100

5

Accepted

Paratanytarsus
parthenogeneticus

Oncorhynchus mvkiss

0.86

1.45

4

0.78

0.0193

1.1

32

9800000

50

6

Rejected

Paratanytarsus
parthenogeneticus

Pimephales promelas

1.05

0.22

4

0.97

0.0002

0.13

92

10600000

83

6

Accepted

Pteronarcella badia

Americamysis bahia

0.72

0.83

4

0.83

0.0112

0.4

0.12

7300

50

5

Accepted

Pteronarcella badia

Oncorhynchus mykiss

0.59

-0.21

15

0.48

0.0018

0.88

0.61

1100000

47

6

Rejected

Pteronarcella badia

Pimephales promelas

U.28

-0.06

8

0.7

0.0023

0.09

1.24

110000

100

6

Rejected

Pteronarcys californica

Daphnia magna

0.63

0.72

24

0.54

0

0.94

0.15

68300

42

5

Rejected

Pteronarcys californica

Oncorhynchus mykiss

0.63

0.05

44

0.25

0.0003

1.7

0.61

70500

35

6

Rejected

146


-------
Table 4-10 shows model outputs from all the rejected, qualitatively acceptable, and
acceptable ICE models listed in Table 4-9. PFOS acute values are typically reported as mg/L and
are, therefore, often greater than the toxicity values used to develop an ICE model, meaning the
input PFOS LCso value of the surrogate was typically outside the model domain. In these cases,
the input toxicity value could be entered as |ig/L and model would be allowed to extrapolate
beyond its range or the input toxicity value could be a "scaled" mu I. \ alue (i.e., estimate the
value as mg/L). Table 4-10 includes a column to denote whether the input toxicity data were
|ig/L or a "scaled" mg/L value for individual models. Please see Appendix L 2 I lor further
discussion on the selection process for identifying whether a iig/L or a "scaled" mu I. value was
used as the input toxicity value for individual models.

Within Table 4-10, bolded and underlined \ nines in the "I-Miniated Toxicity" column
represent the estimated LC50 values from the acceptable models Only estimated toxicity values
from acceptable models were used to de\ clop the estimated insect SMAVs reported in Table
4-10. When more than one acceptable 1(11 model was available for an individual predicted insect
species, all 1he acceptable estimated toxicity \ allies (i.e., LCso values) were taken together as a
geometric mean to represent the estimated SM.W.

147


-------
Table 4-10. ICE-estimated Insect Species Sensitivity to PFOS. Values in bold and
underlined are used for estimated insect SMAVs.

( uiiiniiin Niiiiu-

Pivdkk'd Spicks

SlllTOป;ik- S|K-iil-s

MS"-

in-
mป/l.
input

I'.sliniiili'd

-------
2014; see section 3.1.1.1.6) suggested insects are not among the most sensitive genera to acute
PFOS exposures. Estimated insect SMAVs and qualitatively acceptable insect data were
considered together as multiple lines of evidence to conclude there is variability in the sensitivity
of aquatic insects. EPA will continue to seek additional acute PFOS insect data to further
understand the sensitivity of this taxon. And for the time the current development of an acute
freshwater criterion was based on seven of the eight MDRs.

4.6 Acute-to-Chronic Ratios

The 1985 Guidelines allow the use of a FACR to convert the FAY to the I CV as an

alternative approach to derive the chronic criterion instead of the direct calculation to determine
the FCV (as described in Section 2.10.1) when the eight MDRs are not met (U.S.EPA 1985).
While this alternative approach was not needed lor the deri\ ation of the chronic PFOS criterion,
the calculations of ACRs are as follows. Sixteen ACRs lor se\ en invertebrate species and two
fish species can be calculated from the quantitative acute and chronic toxicity data (Appendix A
and Appendix C). Appendix I includes the ACRs for freshwater aquatic species with quantitative
chronic values for which com parable quanlilali\ e acute values were reported from the same
stud) or same investigator and laboratory combination. For each species where more than a
single ACR was calculated. Species Mean Acute-Chronic Ratios (SMACRs) were also
calculated as the geometric mean value of individual ACRs. In the case of a single ACR within a
species, that ACR was the SMACR.

The ACRs ranged from 4.110 to 13,674 across all species (a factor of 3,327), which
occurs within the Daphnia magna SMACR. There was little explanation for the extreme range in
ACRs among paired tests with D. magna. However, the ACR of 13,674 from paired tests
conducted by Lu et al. (2015) appears to be an outlier. Excluding the 13,674 outlier ACR from

149


-------
the paired tests with D. magna reported by Lu et al. (2015) and from the paired test with
Daphnia carinata (Logeshwaran et al. 2021) produced an SMACR range of 4.11 to 1,030. This
range was greater than a factor of 10 with no relationship between ACR and SMAV apparent.
The Guidelines do not provide for calculation of a FACR under these circumstances. However, if
one were calculated as the geometric mean of the SMACRs excluding the outliers, it would be
122.2, representing the geometric mean of the eight SMACRs highlighted in bold font in
Appendix I.

4.7 Comparison of Empirical Tissue Concentrations to Translated Tissue
Criteria

Measured PFOS tissue data were reported in 14 puhliciiiions focused on freshwater
species, six of which were quantitatively acceptable and eight of u liich were qualitatively
acceptable (Table 4-11). The six quantitati\ elv acceptable studies included data for one
invertebrate, two fish, and one amphibian species, and the eight qualitatively acceptable studies
included data for two in\ eitehiate and four lish species. Results of these studies are summarized
in Section 4.7.1 and Section 4 7 2

Tissue concentration data from these toxicity studies were compared to the translated
tissue \ allies for im e it eh rates and lish to better understand the protectiveness of the aquatic life
tissue criteria While tissue concentrations from the toxicity literature were limited, overall,
translated tissue concentrations for invertebrate whole-body, fish whole-body and fish muscle
were consistent with tissue-based PFOS concentrations from chronic toxicity studies with direct
aqueous exposure. However, tissue concentrations from toxicity studies focused on maternal
transfer indicated that the tissue criteria may be under protective and that a reproductive tissue
criterion may be needed to ensure protection from PFOS through this exposure pathway
(Hazelton et al. 2012; Wang et al. 2011). Hazelton et al. (2012) examined the effects of PFOS on

150


-------
glochidia viability and metamorphic success with the highest test concentration (0.0695 mg/L)
causing an estimated 34.5% reduction in metamorphosis success (see Appendix C.2.2 for more
details). Wang et al. (2011) saw a decrease in the survival of the F1 generation from exposed
adults, with an EPA independently-calculated an ECio value of 0.0165 mg/L (see Appendix C.2.4
for more details). Nevertheless, BAF data for reproductive tissues are currently limited; and
therefore, a reliable reproductive tissue criterion cannot be deri\ ccl ill lliis time (see Appendix Q).

As for other tissue types and taxa with limited data, tissue concentrations from available
toxicity studies suggest that the translated tissue concentrations for fish li\ or and reproductive
tissues may be under protective. While no amphibian tissue criteria are available, tissue
concentrations from two amphibian toxicity tests indicate that the fish tissue criteria may not be
protective of amphibians. However, tissue data lor these tissues and taxa are limited and
additional data are needed.

151


-------
Table 4-11. Comparison o

Empirical Tissue Concentrations to Chronic Tissue Criteria and Additional Tissue Values





Measured Tissue

Chronic Tissue







(onccnlnilion

\ :illies2



Species

LLiidpoint

(nig/kg ww)1

(mg/kg ww)

Tissue Type

Quantitative Studies

Fatmuckct

Probability of successful

LOEC: 0.248

0.937

Invertebrate Whole-body

(Lampsilis siliquoidea)

metamorphosis of glochidia





(adult)

Zebrafish

F1 survival

LOEC: 4.0

6.75

\\ hole-body (adult)

(Danio rerio)









Fathead minnows

Fecundity

NOEC: 8.7-

43 3d

Gonad concentrations

(Pimephales promelas)



LOEC 19.6



(adult male)3





NOEC: 34.8 -

43 3d

Gonad concentrations





LOFC 82 6



(adult female)



Growth (weight in Fl)

LOI ( 37 0

43.30'

Gonad concentrations
(adult F0 male)3





LOi:C 37 4

43 30

Gonad concentrations
(adult F0 female)





I.OEC: 84.5

20.68

Liver (adult F0 male)





I.OEC: 68.2

20.68

Liver (adult F0 female)

Northern leopard frog

Length at metamorphosis al

I.OEC: 18.81

None Available

Whole-body (before

(Lithobates pipiens)

GS 424





metamorphosis)





LOEC: 13.89



Whole-bodv

Qualitative Studies

Red worms

Reduction in superoxide

LOEC: 1.757 (dw)

0.937

Invertebrate Whole-bodv

(L. hoffmeisteri)5

dismutase







Great pond snails

Survival

LOEC: 2,955

0.937

Invertebrate Whole-body

(Lymnaea stagnalis)









European eels

C hanges in protein

NOEC: >5.037

20.68

Liver

(Anguilla anguilla)5

expression







Goldfish

Survn al

NOEC: >39.91

2.91

Muscle

(Carassius auratus)









Common carp

Condition I'aclor

LOEC: 35.97

20.68

Liver

(Cyprinus carpio)









152


-------
Species

Knilpoinl

Measured Tissue
Concciitnilioii
\\\\)'

Chronic Tissue
\ :illies2
(inww)

Tissue Type

Zebrafish

(D. rerio)4

Swimming activity

LOEC: 0.214

6.75

\Vhole-bod>

1 Tissue concentrations are author reported values. EPA did not independently ca

culate to\icil\ \ allies for tissue concentrations.

2	Chronic tissue value concentrations represent chronic tissue criteria (invertebrates, fish muscle, lish u hole body) or additional
tissue values (fish blood, fish liver, fish reproductive tissue) calculated from BAFs for a given tissue l\ |V See Section 3.2.3 and
Appendix Q for details.

3	Fish reproductive tissue value based on female reproductive tissue.

4	Gosner stage associated with this endpoint is not specifically reported by the study authors. However, ihe aulhors define complete
metamorphosis as emergence of the forelimbs which is GS 42 according to Ta\ lor and Kolross (194f\)

5	Toxicity data for non-apical endpoints.

153


-------
4.7.1 Comparison of Quantitative Studies and Tissue-Based Criteria

Tissue concentration data from these toxicity studies were compared to the translated

tissue values for invertebrates and fish to better understand the protectiveness of the aquatic life
tissue criteria. Hazelton et al. (2012) exposed adult fatmucket (Lampsilis siliquoideci) to aqueous
PFOS for 36 days. Measured PFOS concentrations in the control and exposure treatments
averaged 0.0021, 0.0045, and 0.0695 mg/L, respectively. Corresponding tissue concentrations
were 0.009, 0.015 and 0.248 mg/kg wet weight. A statistically significant decrease in the
probability of successful metamorphosis of glochidia produced to the juvenile stage was
observed in the highest PFOS exposure concentration.

(Wang et al. 2011) exposed larval (8 hpf) zebralish (/ kmio rerio) to aqueous PFOS for
five months. Fish were exposed to three nominal PFOS concent rations (0.005, 0.05, and 0.25
mg/L, respectively). Whole-body PFOS tissue concentrations measured after five months in the
two highest exposure concentrations a\ eraued (•> 2 and 1 I I mg kg wet weight, respectively, in
males, and 4 n and 7 7 mu kg wet weight, respectively, in females. PFOS was also measured in
embryos produced from exposed parents and a\eraged 5.75 and 11.0 ng/embryo wet weight.
Weights of embryos were not reported In the study authors, so concentrations could not be
calculated to compare embryo tissue concentrations to the translated tissue criteria. However,
given the study design included tissue measurements in the parental (F0) generation and the
exposure to the offspring generation (Fl) was via maternal transfer, the tissue concentration in
the F0 generation associated with the Fl survival LOEC of 0.05 mg/L was a whole-body tissue
concentration of 6.2 and 4.0 mg/kg wet weight (ww) in male and females, respectively.

Ankley et al. (2005) exposed sexually mature adult fathead minnows (Pimephales
promelas) to aqueous PFOS for 21 days during which time they were allowed to reproduce, and
then held the resulting offspring for an additional 24 days in the same exposure concentrations.

154


-------
Aqueous measured PFOS concentrations in the control and exposure treatments averaged
<0.001, 0.0276, 0.101, 0.281, and 0.818 mg/L, respectively. PFOS was measured in the plasma,
livers, and gonads of adult males and females after 21 days; in embryos; and in whole-body
larval samples after 12 and 24 days. Tissue measurements were not made in organisms from the
highest exposure concentration, where exposed adults were either dead or listless after 14 days.
Plasma PFOS concentrations in the 0.0276, 0.101, and 0.281 mg'l. exposures averaged 28.6,
134, and 355 mg/L in males, and 48.5, 178, and 474 mg/L in females l.i\ er PI OS
concentrations in the 0.0276, 0.101, and 0.281 mg/L exposures averaged 6.9, 19 n. and 109
mg/kg wet weight in males, and 32.8, 82.8, and 262 mg/kg wet weight in females (ionad PFOS
concentrations in the 0.0276, 0.101, and 0.281 mg/L exposures averaged 8.7, 19.6, and 109
mg/kg wet weight in males, and 34.8, 82 (\ and 2M mu'kg ^el weight in females. Embryo PFOS
concentrations in the 0.0276, 0.101, and 0 2SI mg I. exposures were 9.3, 11.5, and 28.6 mg/kg,
respectively. Larval PI-OS concentrations measured after 12 and 24 days of exposure were
similar, with whole-body concentrations corresponding to the 0.0276, 0.101, and 0.281 mg/L
exposures of 19 S. 4S <). and 57 5 mg kg wet weight after 12 days, and 17.8, 49.0, and 83.5
mg/kg wet weight after 24 days The most sensitive apical endpoint was fecundity, with an
aqueous l:X'\-of 0.U51 mg,l. No corresponding tissue-based ECio was calculated, but the
corresponding gonad concentrations would be expected to fall between 8.7 and 19.6 mg/kg in
males and 34.8 and 82 (•> mg/kg in females. No muscle or whole-body measurements in adults are
available to perform a direct comparison to the tissue criteria.

Suski et al. (2021) reported the chronic toxicity of PFOS-K (PFOS potassium salt, CAS#
2795-39-3, > 98%,) on the fathead minnow, Pimephalespromelas. Measured PFOS
concentrations in water were 0.00014 (control), 0.044, 0.088, 0.14, and 0.231 mg/L. The most

155


-------
sensitive endpoint from the study was a significant decrease in the mean mass of individuals in
the larval F1 generation with the author-reported NOEC and LOEC, based on growth in the F1
generation, being 0.044 (6% reduction in growth compared to controls) and 0.088 mg/L PFOS
(associated with an 18% reduction in growth), respectively, with a MATC of 0.06222 mg/L. The
LOEC was associated with measured gonad and liver concentrations in F0 male and females of
37.9, 37.4, 84.5, and 68.2 mg/kg ww, respectively. No corresponding lissue-based ECio was
calculated, but the corresponding gonad and liver concentrations would be expected to fall below
the translated reproductive tissue concentration of 77.52 mg/kg ww and abo\ e the translated liver
tissue concentration of 67.30 mg/kg ww. No muscle or whole-body measurements in adults are
available to perform a direct comparison to the tissue criteria.

Ankley et al. (2004) exposed Northern leopard frogs (/ iiliobuics pipiens) to PFOS from
Gosner stage 8/9 embryos through metamorphosis The time to metamorphosis ranged from 60-
112 days. All frogs in the highest exposure concentration died before metamorphosis. The most
sensitive apical endpoint was length at metamorphosis, which was significantly lower (p < 0.05)
in the second highest exposure relati\ e to the control. The measured aqueous PFOS
concentrations in the NOI-C and I.OI-C exposure concentrations averaged 0.957 and 3.42 mg/L
over the full exposure duration. Corresponding whole-body tissue concentrations measured in
tadpoles exposed for 54 days (before metamorphosis) were 10.1 and 67.42 mg/kg dry weight.
Whole body concentrations were also measured after 35 days and were similar to 54-day
measurements in the 0.957 mg/L exposure (10.2 mg/kg dry weight), but higher in the 3.42 mg/L
exposure (117.4 mg/kg dry weight). Tadpole moisture content was not reported. In order to
convert the reported dry weight concentrations to wet weight concentrations, so that they would
be more directly comparable to the whole-body fish tissue criteria, a whole-body moisture

156


-------
content of 72.1% was applied, calculated as the average for all fish collected as part of the USGS
National Contaminant Biomonitoring Program (NCBP Fish Database (usgs.gov). The resulting
whole-body wet weight concentrations corresponding to the NOEC and LOEC were 2.84 and
32.75 mg/kg wet weight based on day 35 measurements, and 2.82 and 18.81 mg/kg wet weight
based on day 54 measurements.

In a separate study with Lithobatespipiens, Hoover et al (2' > I 7), exposed juvenile
(Gosner stage 26) northern leopard frogs to three PFOS concentrations (7 (o. 78.1, and 884 mg/L
measured PFOS, respectively) for 40 days. Survival, growth (snout-vent length). and
developmental time (days to Gosner stage 40) were measured, and the most scnsiii\e apical
endpoint was time (in days) to reach Gosner stage 40, with a NOEC of 7.65 mg/L and a LOEC
of 78.1 mg/L. Whole body PFOS concentrations in frogs exposed to the NOEC exposure level
averaged 10.53 mg/kg dry weight after 40 days, and concentrations in frogs exposed to the
LOEC exposure le\ el a\ eiaged 4l) 77 mg/kg dry weight after 4<) days. Tadpole moisture content
was not reported in this study, so the 72 I"., moisture content for fish species described above
was applied to coin ei t concentrations to wet weights. Corresponding 40-day NOEC and LOEC
wet weight PI-OS concentrations were 2 and 13.89 mg/kg wet weight, respectively.

Tissue concentration data from the Hazelton et al. (2012) and Wang et al. (2011) suggest
that the tissue criteria may not be protective, as measured whole-body tissue LOECs are lower
than corresponding whole body tissue criteria. However, these studies focused on exposure via
maternal transfer, suggesting that a reproductive tissue criterion may be needed to ensure
protection of PFOS through this exposure pathway. Nevertheless, BAF data for reproductive
tissues is currently limited and therefore, a reliable reproductive tissue criterion cannot be
derived at this time (see Appendix Q). Fathead minnow data reported in Ankley et al. (2005) and

157


-------
Suski et al. (2021) suggest that fish liver and reproductive tissue chronic values may be
protective. Female gonad tissue concentrations associated with decreased fecundity reported by
Ankley et al. (2005) were higher than the reproductive tissue chronic value. Male and female
liver tissue concentrations associated with reduced growth were higher than the liver chronic
value, while female gonad tissue concentrations associated with reduced growth were slightly
lower than the reproductive tissue chronic value. However, results of these fathead minnow
studies were not directly comparable to chronic tissue criteria, because neither whole-body nor
muscle tissue were measured. Finally, although no amphibian tissue criteria are a\ ailable, tissue
concentrations associated with the LOEC in both ( Ankley et al 2004) and (Hoo\ er et al. 2017)
are both higher than the fish whole-body tissue criterion, suggesting that the fish tissue criteria
may be protective of amphibians.

4,7.2 Comparison of Qualitative Studies and Tissue-Based Criteria

Like the com pari son with the quantitative studies, tissue concentration data from these

toxicity studies were compared to the translated tissue values for invertebrates and fish to better
understand 1he |irotecti\ eness of the aquatic life tissue criteria. Liu et al. (2016) exposed 4-5 cm
body length red worms (/ imnoilriliis hoffmcisteri) to two aqueous concentrations of PFOS for 10
days and measured oxidali\ e stress biomarker activity. Measured exposure concentrations were
0.567 and 5 4lM mg L PFOS. and corresponding whole-body tissue PFOS concentrations were
89.5 and 1,757 mu kg dry w eight. Moisture content was not reported. A significant (P < 0.05)
reduction in superoxide dismutase was observed in the highest treatment concentration after 10
days. Apical endpoints were not reported for this exposure. In a separate study with L.
hoffmeisteri, Qu et al. (2016) calculated 48-hour ECsos in response to PFOS at three pH values
(6.2, 7.0, 8.0). PFOS water concentrations used were not reported. However, whole body tissue
concentrations were measured after 48-hours in the control, 0.2 mg/L, and 2.0 mg/L nominal

158


-------
PFOS exposures. Whole-body tissue concentrations in the 2.0 mg/L exposure were 23.41 mg/kg
dry weight at pH 6.2 and 12.61 mg/kg dry weight at pH 8.0. The tissue concentrations are not
representative of what would be measured in water concentrations at the ECso, which ranged
from 23.81 mg/L PFOS at pH 6.2 to 39.80 mg/L PFOS at pH 8.0, nor are they representative of
bioaccumulation in response to a chronic exposure.

Olson (2017) exposed adult great pond snails (Lymnaeu siagimlis) to PFOS for 21 days.
The most sensitive apical endpoint was survival, with a NOEC of 3 mu I. PI ()S nominal, and a
LOEC of 6 mg/L PFOS nominal. Whole-body PFOS tissue concentrations at the NOEC and
LOEC after 21 days were 9,179 mg/kg dry weight and 10 087 mg/kg dry weight, respectively.
Percent moisture was not reported by the study authors, so dry weights were converted to wet
weights using the average whole soft body ".. moisture content of 70 7% for the snail species
Achatina achatina (Achaglinkame et al. 2020) in order to more directly compared, stagnalis
tissue concentrations from this study to the invertebrate tissue criterion. Resulting wet weight
PFOS concentrations at the \OI-C and T.OEC were 2,699 and 2,955 mg/kg, respectively.

Roland et al (Zo|4) exposed ju\ enile F.uropean eels (Anguilla anguilla) to PFOS for 28
days Measured PI-OS water concentrations were 0.00001 mg/L in the control and 0.00081 and
0.011 mu I. in the two exposure concentrations. Corresponding liver tissue PFOS concentrations
after 28 days were o 0338 and 5.037 mg/kg wet weight, respectively. Significant (p < 0.05)
changes in protein expression were reported for both exposure concentrations, but no significant
effects of growth or survival were reported at either exposure concentration.

Feng et al. (2015) conducted a 96-hour study with juvenile goldfish (Carassius auratus)
and measured the effects of PFOS on mortality or antioxidant enzyme activity. Measured PFOS
in the two exposure concentrations were 1.04 |imol/L (0.520 mg/L) and 10.18 |imol/L (5.09

159


-------
mg/L). Liver, gill, and muscle PFOS concentrations were 32.81, 42.13, and 33.08 mg/L wet
weight, respectively, at the lower exposure level, and 58.37, 69.02, and 39.91 mg/L wet weight,
respectively, at the higher exposure level. No effects of mortality were observed during the test.
Among the antioxidant enzyme activity endpoints, glutathione peroxidase was significantly (p <
0.05) lower than the control in the highest exposure concentration.

Hagenaars et al. (2008) exposed juvenile common carp (('\ />riims carpio) to three
exposure concentrations of PFOS plus a control for 14 days and measured relative condition
factor and several non-apical endpoints related to liver function. Nominal PFOS exposure
concentrations were control, 0.1, 0.5, and 1 mg/l. Corresponding liver PFOS concentrations
after the 14-day exposure were 0.97, 35.97, 168.4, and 283.0 mg/kg wet weight. The most
sensitive endpoint was condition factor, which was significantly (p " n 0001) higher than
controls among fish in the 0.1 mg/L exposure after 14 days, although larger condition factors are
associated with increase uell-heing in fish. The lowest exposure level where a significantly (p <
0.0001) lower condition factor than controls was observed was at the 0.5 mg/L exposure
concentration

SpulIxT et al (2<)|4) exposed Ikiwo rerio embryos (2 hpf) to 0.1 mg/L and 1.0 mg/L
nominal PI OS concentrations lor se\ en days. Corresponding whole-body PFOS concentrations
in 7-day-old lar\ ae were 0.<)22 and 0.214 mg/kg wet weight, respectively. Spulber et al. (2014)
reported no effects of PI-OS on viability, time to hatch, or deformities. The most sensitive
endpoint was swimming activity, where fish exposed to 1.0 mg/L PFOS responded more slowly
(p < 0.05) to a startle response, followed by a longer (p < 0.05) hyperactive response to the
stimulus.

160


-------
Relationships between tissue PFOS concentrations associated with effect concentrations
and tissue criteria among qualitative studies were varied. Among invertebrates, Lymnaea
stagnalis whole-body concentrations were well above the criterion at the NOEC for survival. The
absence of wet weight measurements made comparisons less direct for Limnodrilus hoffmeisteri,
but the measured whole-body dry weight NOEC for reduced enzyme activity reported by Liu et
al. (2016) was nearly 100 times higher than the invertebrate tissue criterion, strongly suggesting
that the wet weight NOEC would also be greater than the tissue criterion Among fish,
concentrations in Anguilla anguilla were below the tissue criterion, but no effects were observed.
In Carassius auratus, the highest muscle tissue concentration was greater than the criterion, and
no apical effects were reported. The measured liver LOEC for ( yprinus carpio condition factor
was greater than the chronic liver value. I'inalk. the whole-hody I .OF.C for Danio rerio was
lower than the whole-body criterion, but the I .()!ฆ(' was hascd 011 the non-apical behavioral
endpoint swimming a c t i \ ily ()\ oral I. tissue concentration data from qualitative studies either
suggest the tissue criteria u onkl he protecti\ e, or do not provide any evidence that the PFOS
tissue criteria would not he |">rotecti\e of aquatic species.

4.8 Kf'f'ccts on Aquatic Plants

A\ ailable data for aquatic plants and algae were reviewed to determine if aquatic plants

were likely to he ad\ ersely affected by PFOS and if they were likely to be more sensitive to
PFOS than aquatic animals (see Appendix E). Toxicity values for freshwater plants were well
above the freshwater chronic criterion as effect concentrations for freshwater plants and algae
ranged from 0.19 to 252 mg/L relative to animal chronic values of 0.001588 to 16.35 mg/L
(Appendix C: Acceptable Freshwater Chronic PFOS Toxicity Studies). Therefore, it was not

161


-------
necessary to develop a criterion based on the toxicity of PFOS to aquatic plants. The PFOS
freshwater acute and chronic criteria are expected to be protective of freshwater plants.

4.9 Summary of the PFOS Aquatic Life Criterion and the Supporting
Information

The PFOS aquatic life criteria were developed to protect aquatic life against adverse
effects, such as mortality, altered growth, and reproductive impairments, associated with acute
and chronic exposure to PFOS. The national recommended criteria include water column-based
acute and chronic criteria for freshwaters. The freshwater acute water column-based criterion
magnitude is 3.0 mg/L, and the chronic water column-based criterion magnitude is n 0084 mg/L
(8.4 |ig/L). The chronic freshwater criterion also contains tissue-based criteria expressed as 6.75
mg/kg wet weight (ww) for fish whole-body. 2 ^ I rag/kg u u lor lish muscle tissue and 0.937
mg/kg ww for invertebrate whole-body tissue These PI-OS aquatic life criteria are expected to
be protective of freshwater aquatic life, such as fish and aquatic in\ ertebrates on a national basis.
Although empirical PI-OS toxicity data for esluarine, marine species were not available to fulfill
the eight MDRs directly. I-PA included an acute aquatic life benchmark for estuarine/marine
environments in Appendix I., using a\ ailable estuarine/marine species toxicity data and a NAM
application of ORD's peer-re\ iewed web ICE tool. The estuarine/marine acute water column-
based benchmark magnitude is 0.55 mg/L and is expected to protect estuarine/marine aquatic life
from acute PFOS exposures I-PA conducted additional analyses supporting the derivation of the
water column criteria for PFOS (all summarized above in Sections 4.2 and 4.3) and confirmed
that the criteria and benchmark calculations presented in this document accurately reflect the
latest and best available scientific knowledge.

162


-------
5 REFERENCES

3MCompany. 1999. Fluorochemical Use, Distribution and Release Overview. U.S.
Environmental Protection Agency, Washington, DC. AR226-0550.

https://www.fliioridealert.ore/wp-content/pesticides/pfos.fr.final.docket.0008.pdf

3MCompany. 2000. Information on Perfluorooctane Sulfonates: Post-1975 Studies Pertaining to
Environmental Effects, Fate & Transport, and Health Effects.Cvr Ltr Dtd 050400 [FC-95
DATA], EPA/OTS, St Paul, MN.

3MCompany 2001. Environmental Monitoring - Multi-City Study Water, Sludge, Sediment,
POTW Effluent and Landfill Leachate Samples. Executive Summary 3M Environmental
Laboratory 1-12. https://static.ewe.ore/files/MultiCity execsum.pdf

Abbott, J. K. andE. I. Svensson. 2007. Ontogeny of sexual dimorphism and phenotypic
integration in heritable morphs. Evolutionary Ecology. 22(1): 103-121.

Achaglinkame, M. A., E. Owusu-Mensah, A. A. Boakyc and I Oduro. 2020. Effect of Size and
Drying Time on the Rehydration and Sensory Properties of freeze-Dried Snails (Achatina
achatina). International Journal of Food Science. 2020: 1-5

Ahrens, L. 2011. Polyfluoroalkyl compounds in the aquatic cm iron incut: a review of their
occurrence and fate. J Environ Monit. 13( I). 2<)-3 I

Ahrens, L. and M. Bundschuh. 2014. Fate and effects of poly- and perfluoroalkyl substances in
the aquatic environment: a review. Environ Toxicol Chem. 33(9): 1921-1929.

Ahrens, L., M. Shocik T I larncr. S C l.ee, R. Guo and E. J. Reiner. 2011a. Wastewater
treatment plant and landfills as sources of polyfluoroalkyl compounds to the atmosphere.

Environ Sci Technol 45(N) XiWK-K|i>5

Ahrens. I.., S. Taniyasu. I.. \V Ycunu. \ Yamashita, P. K. Lam andR. Ebinghaus. 2010.
Distribution of polylluoroalkx I compounds in water, suspended particulate matter and sediment
from Tokyo Bay, Japan. Chcmosphere. 79(3): 266-272.

Ahrens, L., I. \V Ycung, S Taniyasu, P. K. Lam andN. Yamashita. 2011b. Partitioning of
perfluorooctanoale (PI OA). perfluorooctane sulfonate (PFOS) and perfluorooctane sulfonamide
(PFOSA) between water and sediment. Chemosphere. 85(5): 731-737.

Allred, B. M., J. R. Lang, M. A. Barlaz and J. A. Field. 2015. Physical and Biological Release of
Poly- and Perfluoroalkyl Substances (PFASs) from Municipal Solid Waste in Anaerobic Model
Landfill Reactors. Environ Sci Technol. 49(13): 7648-7656.

Amraoui, I., N. Khalloufi and S. Touaylia. 2018. Effects to Perfluorooctane Sulfonate (PFOS) on
the Mollusk Unio ravoisieri Under Laboratory Exposure.Chem. Ecol., 34: 324-339.

Anderson, R. H., G. C. Long, R. C. Porter and J. K. Anderson. 2016. Occurrence of select
perfluoroalkyl substances at U.S. Air Force aqueous film-forming foam release sites other than

163


-------
fire-training areas: Field-validation of critical fate and transport properties. Chemosphere. 150:
678-685.

Ankley, G. T., P. Cureton, R. A. Hoke, M. Houde, A. Kumar, J. Kurias, R. Lanno, C. McCarthy,
J. Newsted, C. J. Salice, B. E. Sample, M. S. Sepulveda, J. Steevens and S. Valsecchi. 2020.
Assessing the Ecological Risks of Per- and Polyfluoroalkyl Substances: Current State-of-the
Science and a Proposed Path Forward. Environ Toxicol Chem.

Ankley, G. T., D. W. Kuehl, M. D. Kahl, K. M. Jensen, B. C. Butterworth and J. W. Nichols.
2004. Partial life-cycle toxicity and bioconcentration modeling of periluorooctanesulfonate in the
northern leopard frog (Rana pipiens). Environmental Toxicology and Chemistry: An
International Journal. 23(11): 2745-2755.

Ankley, G. T., D. W. Kuehl, M. D. Kahl, K. M. Jensen, A. Linnum, R I.. I.eino and D. A.
Villeneuve. 2005. Reproductive and Developmental Toxicity and Bioconcentration of
Perfluorooctanesulfonate in a Partial Life-Cycle Test with the Fathead Minnow (Pimephales
promelas). Environ. Toxicol. Chem. 24: 2316-2324.

Annunziato, K. M. 2018. Low Molecular Weight PFAS Alternatives (C-6) Result in Fewer
Cellular and Behavioral Alterations than Long Chain (C-S C-l->) PFASS in Larval Zebrafish.
Ph.D.Thesis, Rutgers, The State University of \ew Jersey, New Brunswick, NJ. 188 p.

Anselmo, H. M., L. Koerting, S. Devito, .1 11 \ an den Berg. M. DuMx-idam, C. Kwadijk and A.
J. Murk. 2011. Early life developmental effects of marine persistent organic pollutants on the sea
urchin Psammechinus mili a lis Ecotoxicol Environ Saf. 74(S) 2182-2192.

Anselmo, H. M., J. 11 \ an den Berg. I VI. Rietjens and A. J. Murk. 2012. Inhibition of cellular
efflux pumps invoh ed in niulli \enoMolic resistance (MXR) in echinoid larvae as a possible
mode of action for increased ecoloxicological risk of mixtures. Ecotoxicology. 21(8): 2276-2287.

Aquilina-Beck, A. A.J I. Reiner, K. W. Chung, M. J. DeLise, P. B. Key and M. E. DeLorenzo.
2020 I 'ptake and Biological Effects of Perfluorooctane Sulfonate Exposure in the Adult Eastern
Oyster Crassostrea virginica Arch I-n\ iron. Contam. Toxicol., 79: 333-342 (Supplemental
journal materials).

Arias, E. Y . M Mallavarapu and R. Naidu. 2015. Identification of the source of PFOS and
PFOA contamination at a military air base site. Environ Monit Assess. 187(1): 4111.

Armitage, J. M., U. Schenker, M. Scheringer, J. W. Martin, M. MacLeod and I. T. Cousins.
2009. Modeling the global fate and transport of perfluorooctane sulfonate (PFOS) and precursor
compounds in relation to temporal trends in wildlife exposure. Environmental science &
technology. 43(24): 9274-9280.

Arukwe, A. and A. S. Mortensen. 2011. Lipid Peroxidation and Oxidative Stress Responses of
Salmon Fed a Diet Containing Perfluorooctane Sulfonic- or Perfluorooctane Carboxylic Acids.
Comp. Biochem. Physiol. C Toxicol. Pharmacol. 154: 288-295.

164


-------
ASTM (American Society for Testing Materials). 1994. Standard Guilde for Conducting Acute
Toxicity Tests with Fishes, Macroinvertebrates, and Amphibians. E729-88a In Annual Book of
ASTM Standards.

ASTM. (American Society for Testing and Materials). 2002. Test method for measuring the
toxicity of sediment-associated contaminants with freshwater invertebrates. E 1706-00. In
Annual Book of ASTM Standards, Vol 11.05. Philadelphia, PA.

ASTM. (American Society for Testing and Materials). 2004a. International Standard Guide for
Conducting Static Acute Toxicity Tests Starting with Embryos of Four Species of Salt Water
Bivalve Mollusks. E 724-98.

ASTM. (American Society for Testing and Materials). 2004b. Standard (inide for Conducting
Toxicity Tests with Bioluminescent Dinoflagellates,. West Conshohocken. PA Method E 1924-
97.

ASTM. (American Society for Testing and Materials). 2006. Standard guide for conducting
laboratory toxicity tests with freshwater mussels. E 2455-iH-. In Annual Book of ASTM
Standards. Philadelphia, PA.

ATSDR (Agency for Toxic Substances and Disease Registry) 2d I 5. Draft Toxicological Profile
for Perfluoroalkyls. United States Department of Health and Human Services, Agency for Toxic
Substances and Disease Registry, Public Health Sen ice. Atlanta, (i.\ Accessed May 2016.

http://www.atsdr.cdc.eov/ToxProfiles/tp200.pdf.

Awkerman, J. A . S Raimondo. (' R. Jackson and M. (i lianon. 2014. Augmenting aquatic
species sensitivity distributions with interspecies toxicity estimation models. Environmental
toxicology and chemisti\ 33(3) (->88-695.

Bao, M., W. Huang. \V \\ Au. S Zheng, C. Liu, Y. Huang and K. Wu. 2019. Exposure to
perfluorooctane sulfonate based on circadian rhythm changes the fecundity and expression of
certain genes on the hypolhalamic-pituilaiy-gonadal-liver axis of female zebrafish. Toxicol Appl
Pharmacol 381: 114715

Barber. .1 I.. I Iterger, C Chaemfa, S. Huber, A. Jahnke, C. Temme andK. C. Jones. 2007.
Analysis of per- and polyfluorinated alkyl substances in air samples from Northwest Europe.
Journal of Environmental Monitoring. 9(6): 530-541.

Beach, S. A., J. L. Newsled, K. Coady and J. P. Giesy. 2006. Ecotoxicological evaluation of
perfluorooctanesulfonate (PFOS). Reviews of environmental contamination and toxicology. 133-
174.

Bejarano, A. C., S. Raimondo and M. G. Barron. 2017. Framework for optimizing selection of
interspecies correlation estimation models to address species diversity and toxicity gaps in an
aquatic database. Environmental science & technology. 51(14): 8158-8165.

165


-------
Bejarano, A. C. and J. R. Wheeler. 2020. Scientific Basis for Expanding the Use of Interspecies
Correlation Estimation Models. Integrated Environmental Assessment and Management. 16(4):
528-530.

Benninghoff, A. D., W. H. Bisson, D. C. Koch, D. J. Ehresman, S. K. Kolluri and D. E. William.
2011. Estrogen-Like Activity of Perfluoroalkyl Acids In Vivo and Interaction with Human and
Rainbow Trout Estrogen Receptors In Vitro. Toxicol. Sci. 120: 42-58 (Supplemental Journal
Materials).

Benninghoff, A. D., G. A. Orner, C. H. Buchner, J. D. Hendricks. A \1. Duffy and D. E.
Williams. 2012. Promotion of Hepatocarcinogenesis by Perfluoroalkyl Acids in Rainbow
Trout.Toxicol. Sci., 125: 69-78 (Supplemental Journal Materials).

Benskin, J. P., A. Holt and J. W. Martin. 2009. Isomer-specific biotransformation rates of a
perfluorooctane sulfonate (PFOS)-precursor by cytochrome P450 isozymes and human liver
microsomes. Environmental science & technology. 43(22): 8566-8572.

Benskin, J. P., M. G. Ikonomou, F. A. Gobas, T. H. liegley. M B. Woudneh and .1 R Cosgrove.

2013.	Biodegradation of N-ethyl perfluorooctane sulfonamido ethanol (EtFOSE) and EtFOSE-
based phosphate diester (SAmPAP diester) in marine sediments Environ Sci Technol. 47(3):
1381-1389.

Benskin, J. P., D. C. Muir, B. F. Scott, C Spencer, A O De Silva, H. Kylin, J. W. Martin, A.
Morris, R. Lohmann, G. Tomy, B. Rosenberg, S Taniyasn and N. Yamashita. 2012.
Perfluoroalkyl acids in the Atlantic and Canadian Arctic Oceans. Environ Sci Technol. 46(11):
5815-5823.

Bertin, D., B. Ferrari. P l.ahadie. A Sapin, J. Garric, H. Budzinski, M. Houde andM. Babut.

2014.	Bioaccuulation of perlluoioalkyl compounds in midge (Chironomus reparius) larvae
exposed to sediment Ijniion Pollul IX1) 27-34

Bertin. D , P. Lalxulie. li .1 I) I'enari. A Sapin, J. Garric, O. Geffard, H. Budzinski and M.
Babut 2<) 16 Potential exposure routes and accumulation kinetics for poly- and perfluorinated
alkyl compounds for a freshwater amphipod: Gammarus spp. (Crustacea). Chemosphere. 155:
380-387.

Bots, J., L. De Biuyn. T Snijkers, B. Van den Branden and H. Van Gossum. 2010. Exposure to
perfluorooctane sulfonic acid (PFOS) adversely affects the life-cycle of the damselfly Enallagma
cyathigerum. Environ Pollut. 158(3): 901-905.

Boudreau, T. M. 2002. Toxicity of Perfluorinated Organic Acids to Selected Freshwater
Organisms Under Laboratory and Field Conditions. M.S. Thesis, University of Guelph, Ontario,
Canada. 145 p.

Boudreau, T. M., P. K. Sibley, S. A. Mabury, D. G. Muir and K. R. Solomon. 2003a. Laboratory
evaluation of the toxicity of Perfluorooctane Sulfonate (PFOS) on Selenastrum capricornutum,
Chlorella vulgaris, Lemna gibba, Daphnia magna, and Daphnia pulicaria. Arch Environ Contam
Toxicol. 44(3): 307-313.

166


-------
Boudreau, T. M., C. J. Wilson, W. J. Cheong, P. K. Sibley, S. A. Mabury, D. C. Muir and K. R.
Solomon. 2003b. Response of the zooplankton community and environmental fate of
perfluorooctane sulfonic acid in aquatic microcosms. Environmental Toxicology and Chemistry:
An International Journal. 22(11): 2739-2745.

Boulanger, B., J. Vargo, J. L. Schnoor and K. C. Hornbuckle. 2004. Detection of perfluorooctane
surfactants in Great Lakes water. Environmental science & technology. 38(15): 4064-4070.

Boulanger, B., J. D. Vargo, J. L. Schnoor and K. C. Hornbuckle. 2005. Evaluation of
perfluorooctane surfactants in a wastewater treatment system and in a commercial surface
protection product. Environmental science & technology. 39(15): 5524-5530.

Brill, J. L., S. E. Belanger, J. G. Chaney, S. D. Dyer, S. Raimondo, M G liarron and C. A.
Pittinger. 2016. Development of algal interspecies correlation estimation models for chemical
hazard assessment. Environmental toxicology and chemistry. 35(9): 2368-2378

Brooke, D., A. Footitt and T. Nwaogu. 2004. Environmental risk evaluation report
Perfluorooctanesulphonate (PFOS). Environmental Agency I

Brown, S. R., R. W. Flynn and J. T. Hoverman. 2021. Periluoroalkyl Substances Increase
Susceptibility of Northern Leopard Frog Tadpoles to Trematode Infection. Environ Toxicol
Chem. 40(3): 689-694.

Buck, R. C., J. Franklin, U. Berger, J. M. Conder, I. T Cousins. P de Voogt, A. A. Jensen, K.
Kannan, S. A. Mabury and S. P. van Leeuwen. 201 1. Perfluoroalkyl and polyfluoroalkyl
substances in the environment terminology, classification, and origins. Integr. Environ. Assess.
Manag. 7(4): 513-541.

Burkhard, L. P. 2021. Evaluation of published bioconcentration factor (BCF) and
bioaccumulation factor (Ii.\Iฆ') data lor pei-and polyfluoroalkyl substances across aquatic
species. En\ ironmental Toxicology and Chemistry.

ButenholV. .1 I. . S C Chang. G \V Olsen and P J. Thomford. 2012. Chronic dietary toxicity
and carcinogenicity study with potassium pertluorooctanesulfonate in Sprague Dawley Rats.
Toxicol. 2l)3 1-15

Butt, C. M., U. IkTger. R liossi and G. T. Tomy. 2010. Levels and trends of poly- and
perfluorinated compounds in the arctic environment. Sci Total Environ. 408(15): 2936-2965.

Butt, C. M., D. C. Muir and S. A. Mabury. 2014. Biotransformation pathways of fluorotelomer-
based polyfluoroalkyl substances: a review. Environ Toxicol Chem. 33(2): 243-267.

CCME. (Canadian Council of Ministers of the Environment). 2007. Protocol for the derivation of
water quality guidelines for the protection of aquatic life. Winnipeg, Canada.

https://www.ccme.ca/files/Resoiirces/siipportine scientific documents/protocol aql 2007e.pdf

167


-------
Chang, S. C., P. E. Noker, G. S. Gorman, S. J. Gibson, J. A. Hart, D. J. Ehresman and J. L.
Butenhoff. 2012. Comparative pharmacokinetics of perfluorooctanesulfonated (PFOS) in rats,
mice, and monkeys. Reprod. Toxicol. 22: 428-440.

Chen, H., C. Zhang, Y. Yu and J. Han. 2012. Sorption of perfluorooctane sulfonate (PFOS) on
marine sediments. Mar Pollut Bull. 64(5): 902-906.

Chen, J., S. R. Das, J. La Du, M. M. Corvi, C. Bai, Y. Chen, X. Liu, G. Zhu, R. L. Tanguay, Q.
Dong and C. Huang. 2013. Chronic PFOS exposures induce life stage-specific behavioral
deficits in adult zebrafish and produce malformation and behavioral deficits in F1 offspring.
Environ Toxicol Chem. 32(1): 201-206.

Chen, J., R. L. Tanguay, T. L. Tal, Z. Gai, X. Ma, C. Bai, S. C. Tilion. I) Jin, D. Yang and C.
Huang. 2014a. Early life perfluorooctanesulphonic acid (PFOS) exposure impairs zebrafish
organogenesis. Aquatic toxicology. 150: 124-132.

Chen, J., R. L. Tanguay, T. L. Tal, Z. Gai, X. Ma, C. Bai, S. C. Tilton, D. Jin, D. Yang. C.

Huang and Q. Dong. 2014b. Early Life Perfluorooctanesulphonic Acid (PFOS) Exposure Impairs
Zebrafish Organogenesis.Aquat. Toxicol., 150: 124-132 (Supplemental Journal Materials).

Chen, J., X. Wang, X. Ge, D. Wang, T. Wang. T. Zhang. R I. Tanguay, M. Simonich, C. Huang
and Q. Dong. 2016. Chronic perfhioroocianesul phonic acid (PI 'OS) exposure produces
estrogenic effects in zebrafish. Environ Pollul 2IS. 7<)2-7<)8.

Chen, J., L. Zheng, L. Tian, N. Wang, L. Lei, Y. Wang. () Dong. C. Huang and D. Yang. 2018.
Chronic PFOS Exposure Disrupts Thyroid Structure and Function in Zebrafish. Environ.

Contam. Toxicol. 1<>I(I) 75-71)

Cheng, J., S. Lv, S. \ie. J I.in. S 'long. \ Kang, Y. Xiao, Q. Dong, C. Huang andD. Yang.
2016. Chronic Perfluoiooctane Sulfonate (PI'OS) Exposure Induces Hepatic Steatosis in
Zebrafish. Aquat Toxicol 17(-> 45-52

Cheng. Y . Y Cui, 11 M Chen and \V P Xie. 2011. Thyroid Disruption Effects of
Environmental Level Pertluorooctane Sulfonates (PFOS) in Xenopus laevis.Ecotoxicology. 20:
2069-2078.

Coats, K., G. Long and J Adelson. 2017. Quality Systems Manual (QSM) for Environmental
Laboratories. Department of Defense, Washington, DC.

Cochran, R. S. 2015. Evaluation of perfluorinated compounds in sediment, water, and passive
samplers collected from the Barksdale Air Force Base. MS Thesis. Texas Tech University,
Lubbock, TX United States. https://ttu-ir.tdl.org/bitstream/handle/2346/63633/CQCHRAN-
THESIS-2015.pdf? sequent sAllowed=v

Cormier, B., A. Batel, J. Cachot, M. Begout, T. Braunbeck, X. Cousin and S. H. Keiter. 2019.
Multi-Laboratory Hazard Assessment of Contaminated Microplastic Particles by Means of
Enhanced Fish Embryo Test with the Zebrafish (Danio rerio).Front. Environ. Sci., 7: 14 p.
(Supplemental Journal Materials).

168


-------
Cousins, I. T., R. Vestergren, Z. Wang, M. Scheringer and M. S. McLachlan. 2016. The
precautionary principle and chemicals management: The example of perfluoroalkyl acids in
groundwater. Environ Int. 94: 331-340.

CRCCare. (CRC for Contamination Assessment and Remediation of the Environment). 2017.
Assessment, management and remediation guidance for perfluorooctanesulfonate (PFOS) and
perfluorooctanoic acid (PFOA). Technical Report no. 38. Part 3: ecological screening levels.
Newcastle, Australia.

https://www.crccare.com/files/dmfile/CRCCARETechReport38Part3 Assessmentmanagementan
dremediationforPFOSa	, s2.pdf

Cui, Y., S. Lv, J. Liu, S. Nie, J. Chen, Q. Dong, C. Huang and D. Yang 2d I 7. Chronic
Perfluorooctanesulfonic Acid Exposure Disrupts Lipid Metabolism in /ebralisli. Hum. Exp.
Toxicol. 36: 207-217.

Custer, C. M., T. W. Custer, R. Delaney, P. M. Dummer, S. Schultz and N. Karouna-Renier.
2019. Perfluoroalkyl Contaminant Exposure and Effects in Tree Swallows Nesting ill Clarks
Marsh, Oscoda, Michigan, USA. Arch Environ Contain Toxicol. 77( 1): 1 -13.

Dang, Y., F. Wang and C. Liu. 2018. Real-time PCR array to study the effects of chemicals on
the growth hormone/insulin-like growth factors ((iH/IGFs) axis of zebrafish embryos/larvae.
Chemosphere. 207: 365-376.

Dasgupta, S., A. Reddam, Z. Liu, J. Liu and I). C Vol/. 2<)2<) I liuh-Content Screening in
Zebrafish Identifies Perfluorooctanesulfonamidc as a Potent l)e\ clopmental Toxicant.Environ.
Pollut., 256: 113550 (Supplemental Journal Materials).

De Silva, A. O., C. Spencer. li I ' Scott. S Backus and D. C. Muir. 2011. Detection of a cyclic
perfluorinated acid, peril uoroethyฆ level ohexane sulfonate, in the Great Lakes of North America.
Environ Sci Technol 45(N) X<)(-<<)-X< )(•<(•<.

De SiKa. A O , P. J Tseng and S A Mabuiy. 2009. Toxicokinetics of Perfluorocarboxylate
Isomers in Rainbow Trout. l-n\iron Toxicol Chem. 28: 330-337.

De Voogl, I' 2<)|0. Reviews of Environmental Contamination and Toxicology Volume 208:
Perfluorinated Alkylated Substances. 208. Springer Science & Business Media pp

de Vos, M. G., M. A. I luijhregts, M. J. van den Heuvel-Greve, A. D. Vethaak, K. I. Van de
Vijver, P. E. Leonards, S. IJ. van Leeuwen, P. de Voogt and A. J. Hendriks. 2008. Accumulation
of perfluorooctane sulfonate (PFOS) in the food chain of the Western Scheldt estuary:

Comparing field measurements with kinetic modeling. Chemosphere. 70(10): 1766-1773.

Delinsky, A. D., M. J. Strynar, P. J. McCann, J. L. Varns, L. McMillan, S. F. Nakayama and A.
B. Lindstrom. 2010. Geographical distribution of perfluorinated compounds in fish from
Minnesota lakes and rivers. Environ. Sci. Technol. 44: 2549-2554.

169


-------
Desjardins, D., C. A. Sutherland, R. L. VanHoven and H. O. Krueger. 2001. PFOS: A 7-Day
Toxicity Test with Duckweed {Lemna gibba G3).Project Number: 454A-111, Wildlife
International, Ltd. < Easton, MD.

Ding, G., L. Wang, J. Zhang, Y. Wei, L. Wei, Y. Li, M. Shao and D. Xiong. 2015. Toxicity and
DNA Methylation Changes Induced by Perfluorooctane Sulfonate (PFOS) in Sea Urchin
Glyptocidaris crenularis.Chemosphere. 128: 225-230 (Supplemental Journal Materials).

Ding, G., J. Zhang, Y. Chen, G. Luo and C. Mao. 2012. Acute Toxicity Effect of PFOS on
Zebrafish Embryo.Adv. Mater. Res., 356: 603-606.

Ding, G., J. Zhang, Y. Chen, L. Wang, M. Wang, D. Xiong and Y. Sun. 2013. Combined effects
of PFOS and PFOA on zebrafish (Danio rerio) embryos. Arch Environ Contain Toxicol. 64(4):
668-675.

Dinglasan-Panlilio, J. M., S. S. Prakash and J. E. Baker. 2014. Perfluorinaled compounds in the
surface waters of Puget Sound, Washington and Clayoquot and Barkley Sounds. British
Columbia. Mar Pollut Bull. 78(1-2): 173-180.

Drottar, K. R. and H. O. Krueger. 2000a. A 96-Hour Static Acute Toxicity Test with the Fathead
Minnow (Pimephalespromelas).Project Number 454A-102, Wildlife International Ltd., Easton,
MD.

Drottar, K. R. and H. O. Krueger. 2000b. PI-OS A 4S-I lour Static Acute Toxicity Test with the
Cladoceran (Daphnia magna).Project 454-A-1 <>4. Wildlife International Ltd., Easton, MD. 60 p.
(OPPTS 850.1010).

Drottar, K. R. and II O kiueuer	PFOS: A 96-Hour Shell Deposition Test with the

Eastern Oyster (Crassoslrea \ iruinica) Project 454-A-106, Wildlife International Ltd., Easton,
MD. 56 p

Drottar. k R and II O kiueuer Zoood PFOS: A 96-Hour Static Acute Toxicity Test with the
Fathead Minnow (Pimephales piomelas) Project 454A-102, Wildlife International Ltd., Easton,
MD. 5S |i

Drottar, K. R and 11 O. Kiueuer. 2000e. PFOS: A 96-Hour Static Acute Toxicity Test with the
Freshwater Mussel (I 'nio complamatus).Project 454-A-105, Wildlife International Ltd., Easton,
MD. 60 p. (OPPTS K5<) H>75).

Drottar, K. R. and H. O. Krueger. 2000f. PFOS: A 96-Hour Static Acute Toxicity Test with the
Saltwater My si d (Mysidopsis bahia).Project 454-A-101, Wildlife International Ltd., Easton, MD.
58 p.

Drottar, K. R. and H. O. Krueger. 2000g. PFOS: A 96-Hour Toxicity Test with the Freshwater
Alga (Selenastrum capricornutum).Project 454-A-103A, Wildlife International Ltd., Easton, MD.
66 p. (OPPTS 850.5400).

170


-------
Drottar, K. R. and H. O. Krueger. 2000h. PFOS: A Flow-Through Life-Cycle Toxicity Test with
the Saltwater Mysid (Mysidopsis bahia).Project 454-A-107, Wildlife International Ltd., Easton,
MD. 76 p. (OPPTS 850.1350).

Drottar, K. R. and H. O. Krueger. 2000i. PFOS: A Semi-Static Life-Cycle Toxicity Test with the
Cladoceran (Daphnia magna).Project 454-A-109, Wildlife International Ltd., Easton, MD. 93 p.
(OPPTS 850.1300).

Drottar, K. R. and H. O. Krueger. 2000j. PFOS: An Early Life-Stage Toxicity Test with the
Fathead Minnow (Pimephales promelas).Project 454-108, Wildlife International Ltd., Easton,
MD. 88 p. (OPPTS 850.1400).

Du, G., J. Hu, H. Huang, Y. Qin, X. Han, D. Wu, L. Song, Y. Xia and X Wang. 2013.
Perfluorooctane sulfonate (PFOS) affects hormone receptor activity, steroidogenesis, and
expression of endocrine-related genes in vitro and in vivo. Environ Toxicol Cliem. 32(2): 353-
360.

Du, J., J. Cai, S. Wang and H. You. 2017. Oxidali\ e stress and apotosis to zebralish (Danio
rerio) embryos exposed to perfluorooctane sulfonate (PI OS) and ZnO nanoparticles. Int J Occup
Med Environ Health. 30(2): 213-229.

Du, J., S. Wang, H. You, R. Jiang, C. Zhuanu and X Zhang	I )e\ elopmental Toxicity and

DNA Damage to Zebrafish Induced by Perfluorooctane Sulfonate in the Presence of ZnO
Nanoparticles. Environ. Toxicol. 31: 360-371 (Supplemental journal materials).

Du, Y., X. Shi, (' I.in. k Yu and li. Zhou. 2009. Chronic effects of water-borne PFOS exposure
on growth, survi\al and hepatotoxicity in zebrafish: a partial life-cycle test. Chemosphere. 74(5):
723-729.

Dyer, S. D., D. .1 Yeisteeg. S I- lielanger, J. G. Chaney and F. L. Mayer. 2006. Interspecies
correlation estimates predict protecti\ e environmental concentrations. Environmental science &
technology 40(9): 3102-31 I I

Dyer, S. I). I) .T Versteeg. S \i. Belanger, J. G. Chaney, S. Raimondo and M. G. Barron. 2008.
Comparison of species sensi ti \ ity distributions derived from interspecies correlation models to
distributions used to deii\ e water quality criteria. Environmental science & technology. 42(8):
3076-3083.

ECCC. (Environment and Climate Change Canada). 2018. ECCC PFOS Guidelines: Canadian
Environmental Protection Act, 1999 Federal Environmental Quality Guidelines Perfluorooctane
Sulfonate (PFOS), pp. 1 -22.

EFSA (European Food Safety Authority). 2008. Opinion of the scientific panel on contaminants
in the food chain on perfluorooctane sulfonate (PFOS), perfluorooctanoic acid (PFOA) and their
salts. The EFSA Journal 653: 1-131.

EGLE (Environment, Great Lakes and Energy). 2010. PFOS Aquatics.

171


-------
EPAV. (Environment Protection Authority Victoria). 2017. Incoming water standards for aquatic
ecosystem protection: PFOS andPFOA. Cariton, VIC. 1633.2.

Fabbri, R., M. Montagna, T. Balbi, E. Raffo, F. Palumbo and L. Canesi. 2014. Adaptation of the
bivalve embryotoxicity assay for the high throughput screening of emerging contaminants in
Mytilus galloprovincialis. Mar Environ Res. 99: 1-8.

Fairbrother, A., 2008. Risk Management Safety Factor. In. Encyclopedia of Ecology, vol. 4. SE
Jorgensen andBD Fath. ed. Elsevier publishing.

Fang, C., Q. Huang, T. Ye, Y. Chen, L. Liu, M. Kang, Y. Lin, 11 Shell and S. Dong. 2013.
Embryonic Exposure to PFOS Induces Immunosuppression in the I'ish I .arvae of Marine
Medaka. Ecotoxicol. Environ. Saf. 92: 104-111.

Fang, C., X. Wu, Q. Huang, Y. Liao, L. Liu, L. Qiu, H. Shen and S. Dong 2<)|2 I'I'OS elicits
transcriptional responses of the ER, AHR and PPAR pathways in Oryzias melastiuma in a stage-
specific manner. Aquat Toxicol. 106-107: 9-19.

Fang, S., X. Chen, S. Zhao, Y. Zhang, W. Jiang, L. Yang and I. Zhu. 2014. Trophic
magnification and isomer fractionation of perfluoroalkyl substances in the food web of Taihu
Lake, China. Environ. Sci. Technol. 48: 21 73-21S2.

Feng, C., F. Wu, Y. Mu, W. Meng, S. D. Dyer. M I'an. S Raimondo and M. G. Barron. 2013.
Interspecies correlation estimation-applications in water quality criteria and ecological risk
assessment. ACS Publications.

Feng, M., Q. He, L. Meng, X. /hang. P. Sun and Z. Wang. 2015. Evaluation of Single and Joint
Toxicity of Perfluorooctane Sulfonate. Perfluorooctanoic Acid, and Copper to Carassius auratus
Using Oxidative Stress Biomai kei s Aquat Toxicol. 161: 108-116 (Supplemental Journal
Materials).

Flynn. R \V . M I' Chislock. M I-. (iannon. S. J. Bauer, B. J. Tornabene, J. T. Hoverman and
M. SepuK eda 2<)|i) Acute and Chronic I-fleets of Perfluoroalkyl Substance Mixtures on Larval
American liullfrogs (Rana catesbeiana). Chemosphere. 236: 7 p.

Flynn, R. W . M lacchetta. C de Perre, L. Lee, M. S. Sepulveda and J. T. Hoverman. 2021.
Chronic Per-/Pol\ lluoroalkx I Substance Exposure Under Environmentally Relevant Conditions
Delays Development in Northern Leopard Frog (Rana pipiens) Larvae. Environ Toxicol Chem.
40(3): 711-716.

Foguth, R. M., T. D. Hoskins, G. C. Clark, M. Nelson, R. W. Flynn, C. De Perre, J. T.

Hoverman, L. S. Lee, M. S. Sepulveda and J. R. Cannon. 2020. Single and Mixture Per- and
Polyfluoroalkyl Substances Accumulate in Developing Northern Leopard Frog Brains and
Produce Complex Neurotransmission Alterations.Neurotoxicol. Teratol., 81: 106907-106907.

Fojut, T. L., A. J. Palumbo and R. S. Tjeerdema. 2012a. Aquatic life water criteria derived via
the UC Davis Method: III. Diuron. In. Tjeerdema, R.S., ed., Rev Environ Contam Toxicol. 216.
ed. Springer, New York, NY.

172


-------
Fojut, T. L., A. J. Palumbo and R. S. Tjeerdema. 2012b. Aquatic life water quality criteria
derived via the UC Davis method: II. Pyrethroid insecticides. In. Tjeerdema, R.S., ed., Rev
Environ Contam Toxicol. 216. ed. Springer, New York, NY.

Fort, D. J., M. B. Mathis, P. D. Guiney and J. A. Weeks. 2019. Evaluation of the developmental
toxicity of perfluorooctanesulfonate in the Anuran, Silurana tropicalis. J Appl Toxicol. 39(2):
365-374.

Funkhouser, M. 2014. The toxicological effects of Perfluorooctane sulfonate (PFOS) on a
freshwater gastropod, Physa pomilia, and a parthenogenetic decapod. Procambarus fallax f.
virginalis.

Furdui, V. I., P. W. Crozier, E. J. Reiner and S. A. Mabury. 2008. Trace le\ cl determination of
perfluorinated compounds in water by direct injection. Chemosphere. 73( I Suppl): S24-30.

Furdui, V. I., N. L. Stock, D. A. Ellis, C. M. Butt, D. M. Whittle, P. W. Crozier, E. J. Reiner, D.
C. Muir and S. A. Mabury. 2007. Spatial distribution of perfluoroalkyl contaminants in lake trout
from the Great Lakes. Environ Sci Technol. 41(5): I 554-1 55l>

Gergs, A., S. Classen, R. Strauss, R. Ottermanns, T. Brock. 11. Ratte, U. Hommen and T. Preuss.
2016. Ecological Recovery Potential of Freshwater Organisms Consequences for Environmental
Risk Assessment of Chemicals. Rev. Environ. Contam. Toxicol. 23o 259-294.

Gewurtz, S. B., S. M. Backus, A. O. De Silva, L. Alliens. A Aimellin, M. Evans, S. Fraser, M.
Gledhill, P. Guerra, T. Harner, P. A. Helm, H. Hung, N. Kheia. M. G. Kim, M. King, S. C. Lee,
R. J. Letcher, P. Martin, C. Marvin, D. J. McGoldrick, A. L. Myers, M. Pelletier, J. Pomeroy, E.
J. Reiner, M. Rondeau, M. C. Sauve, M. Sekela, M. Shoeib, D. W. Smith, S. A. Smyth, J.

Struger, D. Spry, J. Syrgiannis and J. Waltho. 2013. Perfluoroalkyl acids in the Canadian
environment: multi-media assessment of current status and trends. Environ Int. 59: 183-200.

Giesy, J. P. and k. kannan 2001 Global distribution of perfluorooctane sulfonate in wildlife.
Environmental science & technology 35(7): 1339-1342.

Giesy, J. P. and K. Kannan 2<)(i2. Peer reviewed: perfluorochemical surfactants in the
environment I jivironmenlal science & technology. 36(7): 146A-152A.

Giesy, J. P., J. E Naile. .1 S kliim, P. D. Jones and J. L. Newsted. 2010. Aquatic toxicology of
perfluorinated chemicals Rev Environ Contam Toxicol. 202: 1-52.

Goecke-Flora, C. M. and N. V. Reo. 1996. Influence of carbon chain length on the hepatic
effects of perfluorinated fatty acids. A 19F and 31P-NMR investigation. Chem. Res. Toxicol. 9:
689-695.

Goeritz, I., S. Falk, T. Stahl, C. Schafers and C. Schlechtriem. 2013. Biomagnification and
Tissue Distribution of Perfluoroalkyl Substances (PFASs) in Market-Size Rainbow Trout
(Oncorhynchus mykiss). Environ. Toxicol. Chem. 32(9): 2078-2088.

173


-------
Gunduz, G., H. Parlak, O. C. Arslan, M. Boyacioglu and M. A. Karaaslan. 2013. Embryotoxic
Effects of Perfluorooctane Sulfonate Compounds in Sea Urchin Paracentrotus lividus. Fresenius
Environ. Bull. 22(la. 2013): 171-177.

Guo, J., P. Wu, J. Cao, Y. Luo, J. Chen, G. Wang, W. Guo, T. Wang and X. He. 2019. The PFOS
disturbed immunomodulatory functions via nuclear Factor-kappaB signaling in liver of zebrafish
(Danio rerio). Fish Shellfish Immunol. 91: 87-98.

Guo, R., E. J. Reiner, S. P. Bhavsar, P. A. Helm, S. A. Mabury, E. Braekevelt and S. A.
Tittlemier. 2012. Determination of polyfluoroalkyl phosphoric acid diesters, perfluoroalkyl
phosphonic acids, perfluoroalkyl phosphinic acids, perfluoroalkyl carhoxylic acids, and
perfluoroalkane sulfonic acids in lake trout from the Great I .akcs Region. Anal. Bioanal. Chem.
404: 2699-2709.

Hagenaars, A., D. Knapen, I. J. Meyer, K. van der Ven, P. Hoff and W. Do Coon 2008. Toxicity
evaluation of perfluorooctane sulfonate (PFOS) in the liver of common carp (C\ piinus carpio).
Aquat Toxicol. 88(3): 155-163.

Hagenaars, A., E. Stinckens, L. Vergauwen, L. Bervoels and I) Knapen. 2014. PFOS Affects
Posterior Swim Bladder Chamber Inflation and Swimming Performance of Zebrafish
Larvae.Aquat. Toxicol., 157: 225-235.

Hagenaars, A., L. Vergauwen, W. De Coen and D. Knapen 201 I Structure-Activity
Relationship Assessment of Four Perfluorinated Chemicals I 'sing a Prolonged Zebrafish Early
Life Stage Test. Chemosphere. 82: 764-772.

Han, J. and Z. Fang. 2010. Estrogenic Effects, Reproductive Impairment and Developmental
Toxicity in Ovoviparous Swordtail Fish (Xiphophorus helleri) Exposed to Perfluorooctane
Sulfonate (PFOS). Aquat. Toxicol. 99: 281-290.

Han, J., E. J. W on, M ('. I.ee, .1 S Seo, S. J. Lee and J. S. Lee. 2015. Developmental
retardation, reduced fecundity: and modulated expression of the defensome in the intertidal
copepod Tigriopus japoniais exposed to BDE-47 and PFOS. Aquat Toxicol. 165: 136-143.

Hansen, k .1.11 Johnson, .1 l-ldridge, J. Butenhoff andL. Dick. 2002. Quantitative
characterization of trace le\ els of PFOS and PFOA in the Tennessee River. Environmental
Science & Technology 36(8) 1681-1685.

Hanson, M. L., P. K. Sibley, R. A. Brain, S. A. Mabury and K. R. Solomon. 2005. Microcosm
Evaluation of the Toxicity and Risk to Aquatic Macrophytes from Perfluorooctane Sulfonic
Acid.Arch. Environ. Contam. Toxicol., 48: 329-337.

Hassell, K. L., T. L. Coggan, T. Cresswell, A. Kolobaric, K. Berry, N. D. Crosbie, J. Blackbeard,
V. J. Pettigrove and B. O. Clarke. 2020. Dietary Uptake and Depuration Kinetics of
Perfluorooctane Sulfonate, Perfluorooctanoic Acid, and Hexafluoropropylene Oxide Dimer Acid
(GenX) in a Benthic Fish. Environ. Toxicol. Chem. 39(3): 595-603.

174


-------
Hatfield, T. L. (3M Company, Project W2783). 2001. Screening studies on the aqueous
photolytic degradation of 2-(N-ethylperfluorooctanesulfonamido)-ethyl alcohol (N-EtFOSE
alcohol). AR226-1030a80.

Haukas, M., U. Berger, H. Hop, B. Gulliksen and G. W. Gabrielsen. 2007. Bioaccumulation of
per- and polyfluorinated alkyl substances (PFAS) in selected species from the Barents Sea food
web. Environ Pollut. 148(1): 360-371.

Hayman, N. T., G. Rosen, M. A. Colvin, J. Conder and J. A. Arblaster. 2021. Aquatic toxicity
evaluations of PFOS and PFOA for five standard marine endpoints Chemosphere. 273: 129699.

Hazelton, P. D. 2013. Emerging methods for emerging conlaminanls no\ el approaches to
freshwater mussel toxicity testing. University of Georgia,

Hazelton, P. D., W. G. Cope, T. J. Pandolfo, S. Mosher, M. J. Strynar, M (' liainhart and R. B.
Bringolf. 2012. Partial life-cycle and acute toxicity of perfluoroalkyl acids lo freshwater mussels.
Environ Toxicol Chem. 31(7): 1611-1620.

Higgins, C. P. andR. G. Luthy. 2006. Sorption of perfluorinated surfactants on sediments.
Environ Sci Technol. 40(23): 7251-7256.

Hoff, P. T., K. Van de Vijver, W. Van Dongen, E. L. Esmans, R. Blust and W. M. De Coen.
2003. Perfluorooctane sulfonic acid in bib (Trisopterus luscus) and plaice (Pleuronectes platessa)
from the Western Scheldt and the Belgian North Sea: distribution and biochemical effects.
Environmental Toxicology and Chemistry: An International Journal. 22(3): 608-614.

Hong, S., J. S. Khim. T Y Wang. .1 I- Naile, J. Park, B. O. Kwon, S. J. Song, J. Ryu, G.
Codling, P. D. Jones. Y I. Lu and J P Giesy. 2015. Bioaccumulation characteristics of
perfluoroalkyl acids (PI 'A As) in coastal organisms from the west coast of South Korea.
Chemosphere 129: 157-163

Hoover. G. M., M. F. Chislock, 1} J Tornabene, S. C. Guffey, Y. J. Choi, C. De Perre, J. T.
Hoverman. L S. Lee and M S Sepulveda. 2017. Uptake and Depuration of Four
Per/Polylluoioalkyl Substances (PFASS) in Northern Leopard Frog Rana pipiens Tadpoles.
Environmental Science & Technology Letters. 4(10): 399-403.

Houde, M., G. C/.nh. J M Small, S. Backus, X. Wang, M. Alaee and D. C. Muir. 2008.
Fractionation and bioaccumulation of perfluorooctane sulfonate (PFOS) isomers in a Lake
Ontario food web. Environmental science & technology. 42(24): 9397-9403.

Houde, M., J. W. Martin, R. J. Letcher, K. R. Solomon and D. C. Muir. 2006. Biological
monitoring of polyfluoroalkyl substances: A review. Environ Sci Technol. 40(11): 3463-3473.

Houtz, E. F., R. Sutton, J. S. Park and M. Sedlak. 2016. Poly- and perfluoroalkyl substances in
wastewater: Significance of unknown precursors, manufacturing shifts, and likely AFFF impacts.
Water Res. 95: 142-149.

175


-------
HSDB (Hazardous Substances Data Bank). 2012. Hazardous Substances Data Bank. National
Institutes of Health, Health & Human Services, U.S. National Library of Medicine, Bethesda,
MD. Accessed May 2016. http://toxnet.nlm.nih.eov/cei-bin/sis/htmleen7HSDB.

Huang, H., C. Huang, L. Wang, X. Ye, C. Bai, M. T. Simonich, R. L. Tanguay and Q. Dong.
2010. Toxicity, uptake kinetics and behavior assessment in zebrafish embryos following
exposure to perfluorooctanesulphonicacid (PFOS). Aquat Toxicol. 98(2): 139-147.

Huang, J., L. Sun, J. A. Mennigen, Y. Liu, S. Liu, M. Zhang, Q. Wang and W. Tu. 2021.
Developmental toxicity of the novel PFOS alternative OBS in developing zebrafish: An
emphasis on cilia disruption. J Hazard Mater. 409: 124491.

Huang, Q., C. Fang, X. Wu, J. Fan and S. Dong. 2011. Perfluorooclane sulfonate impairs the
cardiac development of a marine medaka (Oryzias melastigma). Aqual Toxicol. 105(1-2): 71-77.

Huang, S. S., J. P. Benskin, B. Chandramouli, H. Butler, C. C. Helbing and J R Cosgrove. 2016.
Xenobiotics Produce Distinct Metabolomic Responses in Zebrafish Larvae (Danio i crio).

Environ Sci Technol. 50(12): 6526-6535.

Jantzen, C. E., K. A. Annunziato, S. M. Bugel and K. R. Cooper. 2017. PFOS, PFNA, and PFOA
sub-lethal exposure to embryonic zebrafish ha\ e different toxicity profiles in terms of
morphometries begavioral and gene expression Aqual Toxicol. 175 168-170.

Jarvis, A. L., J. R. Justice, M. C. Elias, B. Schnilker and k (iallagher. 2021. Perfluorooctane
Sulfonate in US Ambient Surface Waters: A Review of Occurrence in Aquatic Environments
and Comparison to Global Concentrations. Environmental Toxicology and Chemistry. 40(9):
2425-2442.

Jeon, J., H. K. Lim. k kannan and S I) kini 2<)|o Effect of perfluorooctanesulfonate on
osmoregulation in marine fish. Sehasles schleueli. under different salinities. Chemosphere. 81(2):
228-234

Jeong. T Y . M. S. Yuk. .1 Jeon and S I) kini 2<)|e>. Multigenerational Effect of
Perfluorooclane Sulfonate (PI 'OSj on the Individual Fitness and Population Growth of Daphnia
magna. Sci. Total Environ. 5(->l)'570: 1553-1560.

Ji, K., Y. Kim, S Oh. 1} Aim. IT Jo andK. Choi. 2008. Toxicity of Perfluorooctane Sulfonic
Acid and Perfluorooclanoic Acid on Freshwater Macroinvertebrates (Daphnia magna and Moina
macrocopa) and Fish (Oryzias latipes). Environ. Toxicol. Chem. 27: 2159-2168.

Jones, P. D., W. Hu, W. De Coen, J. L. Newsted and J. P. Giesy. 2003. Binding of perfluorinated
fatty acids to serum proteins. Environmental Toxicology and Chemistry: An International
Journal. 22(11): 2639-2649.

Kang, J. S., T. G. Ahn and J. W. Park. 2019. Perfluorooctanoic Acid (PFOA) and Perfluooctane
Sulfonate (PFOS) Induce Different Modes of Action in Reproduction to Japanese Medaka
(Oryzias latipes). J. Hazard. Mater. 368: 97-103 (Supplemental Journal Materials).

176


-------
Kannan, K. 2011. Perfluoroalkyl and polyfluoroalkyl substances: current and future perspectives.
Environmental Chemistry. 8(4).

Kannan, K., L. Tao, E. Sinclair, S. D. Pastva, D. J. Jude and J. P. Giesy. 2005. Perfluorinated
compounds in aquatic organisms at various trophic levels in a Great Lakes food chain. Arch
Environ Contam Toxicol. 48(4): 559-566.

Keiter, S., L. Baumann, H. Farber, H. Holbech, D. Skutlarek, M. Engwall and T. Braunbeck.
2012. Long-Term Effects of a Binary Mixture of Perfluorooctane Sulfonate (PFOS) and
Bisphenol A (BPA) in Zebrafish (Danio rerio). Aquat. Toxicol. 118' I 19: 116-129.

Kelly, B. C., M. G. Ikonomou, J. D. Blair, B. Surridge, D. Hoover. R Grace and F. A. Gobas.
2009. Perfluoroalkyl Contaminants in an Arctic Marine Food Web. Trophic Magnification and
Wildlife Exposure. Environ. Sci. Technol.(43): 4037-4043.

Kim, S.-K. and K. Kannan. 2007. Perfluorinated acids in air, rain, snow, surface runoff, and
lakes: relative importance of pathways to conlami nation of urban lakes. Environmental science
& technology. 41(24): 8328-8334.

Kim, S., K. Ji, S. Lee, J. Lee, J. Kim, S. Kim, Y. Kho and K Choi. 2011. Perfluorooctane
sulfonic acid exposure increases cadmium toxicity in early life stage of zebrafish, Danio rerio.
Environ Toxicol Chem. 30(4): 870-877.

Kim, W. K., S. K. Lee and J. Jung. 2010. Integrated assessment of biomarker responses in
common carp (Cyprinus caipio) exposed to perfluorinated organic compounds. J Hazard Mater.
180(1-3): 395-400.

Kimmel, C. B., W. \V Ballard, S R kimmel, B. Ullmann and T. F. Schilling. 1995. Stages of
embryonic development of the zehalish. Developmental dynamics. 203(3): 253-310.

Kimmel, C li , I) S Sepich and B Ti evarrow. 1988. Development of segmentation in zebrafish.
Development 104(Supplcmcnt) ll)7-2<~>7

KonwicL li 1, G. T. Tomy. \ Ismail, J. T. Peterson, R. J. Fauver, D. Higginbotham and A. T.
Fisk. 2008. Concentrations and patterns of perfluoroalkyl acids in Georgia, USA surface waters
near and distant to a major use source. Environmental Toxicology and Chemistry: An
International Journal 27(1") 2011-2018.

Lammer, E., G. Carr, k. Wendler, J. Rawlings, S. Belanger and T. Braunbeck. 2009. Is the fish
embryo toxicity test (FET) with the zebrafish (Danio rerio) a potential alternative for the fish
acute toxicity test? Comparative Biochemistry and Physiology Part C: Toxicology &
Pharmacology. 149(2): 196-209.

Lang, J. R., B. M. Allred, G. F. Peaslee, J. A. Field and M. A. Barlaz. 2016. Release of Per- and
Polyfluoroalkyl Substances (PFASs) from Carpet and Clothing in Model Anaerobic Landfill
Reactors. Environ Sci Technol. 50(10): 5024-5032.

177


-------
Lange, C. C. (3M Company). 2000. The aerobic biodegradation of N-EtFOSE alcohol by the
microbial activity present in municipal wastewater treatment sludge. 3MProject ID: LIMS E00-
2252 St. Paul, MN.

Lanza, H. A., R. S. Cochran, J. F. Mudge, A. D. Olson, B. R. Blackwell, J. D. Maul, C. J. Salice
and T. A. Anderson. 2017. Temporal monitoring of perfluorooctane sulfonate accumulation in
aquatic biota downstream of historical aqueous film forming foam use areas. Environmental
toxicology and chemistry. 36(8): 2022-2029.

Lasier, P. J., J. W. Washington, S. M. Hassan and T. M. Jenkins. 2<~> I I Perfluorinated chemicals
in surface waters and sediments from northwest Georgia, USA. and their bioaccumulation in
Lumbriculus variegatus. Environ Toxicol Chem. 30(10): 2194-22<)|

Lau, C., K. Anitole, C. Hodes, D. Lai, A. Pfahles-Hutchens and J. Seed 2<)()7 Perfluoroalkyl
acids: a review of monitoring and toxicological findings. Toxicol Sci. 99(2) 306-394.

Lee, J. J. and I. R. Schultz. 2010. Sex Differences in the Uptake and Disposition of
Perfluorooctanoic Acid in Fathead Minnows after Oral Dosing Environ. Sci. Technol. 44(44):
491-496.

Lee, J. W., K. Choi, K. Park, C. Seong, S. I) Yu and P. Kim. 2<)2<) Adverse effects of
perfluoroalkyl acids on fish and other aquatic organisms: A re\ ieu Sci Total Environ. 707:
135334.

Lehmler, H. J. 2005. Synthesis of environmentally rele-^ mil lluorinated surfactants—a review.
Chemosphere. 58(11): 1471-1496.

Lemly, A. D. 1997 Ixosystem icco\ cry following selenium contamination in a freshwater
reservoir. Ecotoxicolouy and I ji\ironnieiilal Safety. 36(3): 275-281.

Lescord, (i I.., k A kidd. A () l)e Sil\a, M. Williamson, C. Spencer, X. W. Wang andD. C.
G. Muir 2014. Peril iioiinaled and poly lluorinated compounds in lake food webs from the
Canadian High Arctic I jniion Sci Technol. 49: 2694-2702.

Lewis. R .1 Sr. cd. 2004. Sax's Dangerous Properties of Industrial Materials, llthed. Wiley-
Interscience. Wiley & Sons Inc., Hoboken, NJ.

Li, M. H. 2008. Effects of nonionic and ionic surfactants on survival, oxidative stress, and
cholinesterase activity of planarian. Chemosphere. 70(10): 1796-1803.

Li, M. H. 2009. Toxicity of perfluorooctane sulfonate and perfluorooctanoic acid to plants and
aquatic invertebrates. Environ Toxicol. 24(1): 95-101.

Li, M. H. 2010. Chronic Effects of Perfluorooctane Sulfonate and Ammonium
Perfluorooctanoate on Biochemical Parameters, Survival and Reproduction of Daphnia magna. J.
Health Sci. 56: 104-111.

178


-------
Li, Y., T. Fletcher, D. Mucs, K. Scott, C. H. Lindh, P. Tallving and K. Jakobsson. 2018. Half-
lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water.
Occupational and environmental medicine. 75(1): 46-51.

Li, Y., Z. Han, X. Zheng, Z. Ma, H. Liu, J. P. Giesy, Y. Xie and H. Yu. 2015. Comparison of
Waterborne and In Ovo Nanoinjection Exposures to Assess Effects of PFOS on Zebrafish
Embryos.Environ. Sci. Pollut. Res. Int., 22: 2303-2310.

Li, Y., B. Men, Y. He, H. Xu, M. Liu and D. Wang. 2017. Effect of single-wall carbon
nanotubes on bioconcentration and toxicity of perfluorooctane sulfonate in zebrafish (Danio
rerio). Sci Total Environ. 607-608: 509-518.

Liang, R., J. He, Y. Shi, Z. Li, S. Sarvajayakesavalu, Y. Baninla. I' (iuo. .1 Chen, X. Xu and Y.
Lu. 2017. Effects on Perfluorooctane Sulfonate on Immobilization, Heaill">eal. Reproductive and
Biochemical Performance of Daphnia magna. Chemosphere.( 168): 1613-1M K

Lindstrom, A. B., M. J. Strynar and E. L. Libelo. 2<> I I Polyfluorinated compounds past,
present, and future. Environ Sci Technol. 45(19) 7^54-7^(->l

Liu, C., V. W. Chang andK. Y. Gin. 2013. Environmental Toxicity ofPFCs: An Enhanced
Integrated Biomarker Assessment and Structure-Activity Analysis linviron. Toxicol. Chem., 32:
2226-2233 (Supplemental Journal Materials)

Liu, C., V. W. C. Chang andK. Y. H. Gin. 2<)|4a Genoloxicily of perfluorinated chemicals
(PFCs) to the green mussel (Penia \ iridis). Science of the Total l-nvironment. 487: 117-122.

Liu, C., V. W. C. Chang and k Y. 11 Gin. 2014b. Oxidative Toxicity of Perfluorinated
Chemicals in Green Mussel and liioaccumulation Factor Dependent Quantitative Structure-
Activity Relationship. I -n\ iron Toxicol Chein. 33: 2323-2332 (Supplemental journal materials).

Liu, C. and k Y Gin 2<)|S Immunoloxicily in Green Mussels Under Perfluoroalkyl Substance
(PFAS) Exposure Re\ eisiMe Response and Response Model Development. Environ. Toxicol.
Chem 37 I 138-1145 (Supplemental journal materials).

Liu, C . k Y Gin and V. W Chang. 2014c. Multi-biomarker responses in green mussels
exposed to PI Cs effects at molecular, cellular, and physiological levels. Environ Sci Pollut Res
Int. 21(4): 2785-27lM

Liu, J. and S. Mejia Avendano. 2013. Microbial degradation of polyfluoroalkyl chemicals in the
environment: a review. Environ Int. 61: 98-114.

Liu, J., R. Qu, L. Yan, L. Wang and Z. Wang. 2016. Evaluation of single and joint toxicity of
perfluorooctane sulfonate and zinc to Limnodrilus hoffmeisteri: Acute toxicity, bioaccumulation
and oxidative stress. J Hazard Mater. 301: 342-349.

Liu, X., Z. Guo, K. A. Krebs, R. H. Pope and N. F. Roache. 2014d. Concentrations and trends of
perfluorinated chemicals in potential indoor sources from 2007 through 2011 in the US.
Chemosphere. 98: 51-57.

179


-------
Logeshwaran, P., A. K. Sivaram, A. Surapaneni, K. Kannan, R. Naidu and M. Megharaj. 2021.
Exposure to perfluorooctanesulfonate (PFOS) but not perflurorooctanoic acid (PFOA) at ppb
concentration induces chronic toxicity in Daphnia carinata. Sci Total Environ. 769: 144577.

Loi, E. I., L. W. Yeung, S. Taniyasu, P. K. Lam, K. Kannan andN. Yamashita. 2011. Trophic
Magnification of Poly- and Perfluorinated Compounds in a Subtropical Food Web. Environ. Sci.
Technol.(45): 5506-5513.

Lou, Q. Q., Y. F. Zhang, Z. Zhou, Y. L. Shi, Y. N. Ge, D. K. Ren, H. M. Xu, Y. X. Zhao, W. J.
Wei and Z. F. Qin. 2013. Effects of perfluorooctanesulfonate and peril uorobutanesulfonate on
the growth and sexual development of Xenopus laevis. Ecotoxicolouy 22(7): 1133-1144.

Lu, G., J. Liu, L. Sun and L. Yuan. 2015. Toxicity of perfluorononanoic acid and
perfluorooctane sulfonate to Daphnia magna. Water Science and Engineering 8(1): 40-48.

MacDonald, M. M., A. L. Warne, N. L. Stock, S. A. Mabury, K. R. Solomon and P. K. Sibley.
2004. Toxicity of Perfluorooctane Sulfonic Acid and Perfluorooctanoic Acid lo Chironomus
tentans.Environ. Toxicol. Chem., 23: 2116-2123.

Martin, J. W., B. J. Asher, S. Beesoon, J. P. Benskin and M. S. Ross. 2010. PFOS orPreFOS?
Are perfluorooctane sulfonate precursors (Prel OS) important determinants of human and
environmental perfluorooctane sulfonate (PI-OS) exposure? J Environ Vlonit. 12(11): 1979-2004.

Martin, J. W., S. A. Mabury, K. R. Solomon and D. (' Muir 2<)()3a. Dietary accumulation of
perfluorinated acids in juvenile rainbow trout (Oncorhyncluis my kiss). Environmental
Toxicology and Chemistry: An International Journal. 22(1). 189-195.

Martin, J. W., S. A. Mabury, K. R. Solomon and D. C. G. Muir. 2003b. Bioconcentration and
tissue distribution of perfluorinated acids in rainbow trout (Oncorhynchus mykiss). Environ.
Toxicol. Chem. 22: 196-204.

Martin. .1 W., M. M. Smilhwick. 1} M Braune, P. F. Hoekstra, D. C. Muir and S. A. Mabury.
2004 Identification of long-chain perilnorinated acids in biota from the Canadian Arctic.
Environmental science & technology. 38(2): 373-380.

Martinez, R.I. I lencro-Nouaieda, M. Van Antro, M. P. Campos, M. Casado, C. Barata, B. Pina
and L. Navarro-Martin 201 Morphometric Signatures of Exposure to Endocrine Disrupting
Chemicals in Zebralish l-leutheroembryos.Aquat. Toxicol., 214: 17 p. (Supplemental Journal
Materials).

Martinez, R., L. Navarro-Martin, C. Luccarelli, A. E. Codina, D. Raldua, C. Barata, R. Tauler
and B. Pina. 2019b. Unravelling the Mechanisms of PFOS Toxicity by Combining
Morphological and Transcriptomic Analyses in Zebrafish Embryos.Sci. Total Environ., 674:
462-471 (Supplemental Journal Materials).

Marziali, L., F. Rosignoli, S. Valsecchi, S. Polesello and F. Stefani. 2019. Effects of
Perfluoralkyl Substances on a Multigenerational Scale: A Case Study with Chironomus riparius
(Diptera, Chironomidae). Environ Toxicol Chem. 38(5): 988-999.

180


-------
McCarthy, C. J., S. A. Roark, D. Wright, K. O'Neal, B. Muckey, M. Stanaway, J. N. Rewerts, J.
A. Field, T. A. Anderson and C. J. Salice. 2021. Toxicological Response of Chironomus dilutus
in Single-Chemical and Binary Mixture Exposure Experiments with 6 Perfluoralkyl Substances.
Environmental toxicology and chemistry. 40(8): 2319-2333.

Meyers, A. L., P. W. Crozier, P. A. Helm, C. Brimacombe, V. I. Furdui, E. J. Reiner, D.
Burniston and C. H. Marvin. 2012. Fate, distribution, and contrasting temporal trends of
perfluoroalkyl substances (PFASs) in Lake Ontario, Canada. Environ Int. 44: 92-99.

Mhadhbi, L., D. Rial, S. Perez and R. Beiras. 2012. Ecological risk assessment of
perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PI-OS) in marine environment
using Isochrysis galbana, Paracentrotus lividus, Siriella armala and Psetta maxima. J Environ
Monit. 14(5): 1375-1382.

Moody, C. A. and J. A. Field. 2000. Perfluorinated surfactants and the en\ iron mental
implications of their use in fire-fighting foams. Environmental science & technology. 34(18):
3864-3870.

Moody, C. A., G. N. Hebert, S. H. Strauss and J. A. Field 2<><>3 Occurrence and persistence of
perfluorooctanesulfonate and other perfluorinated surfactants in groundwater at a fire-training
area at Wurtsmith Air Force Base, Michigan. USA. J Environ Monit. 5(2): 341-345.

Nagel, R. 2002. DarT: The embryo test with the Zebialish Danio rei io—a general model in
ecotoxicology and toxicology. Altex. 19: 38-48.

Nakata, H., K. Kannan, T. Nasu, H..-S. Cho, E. Sinclair and A Takemura. 2006. Perfluorinated
contaminants in sediments and aquatic organisms collected from shallow water and tidal flat
areas of the Ariake Sea, Japan: environmental fate of perfluorooctane sulfonate in aquatic
ecosystems. Environmental science & technology. 40(16): 4916-4921.

Nakayama, S., \1. J Su \ nar, L. I lellant, P. Egeghy, X. Ye and A. B. Lindstrom. 2007.
Perfluorinated compounds in the Cape Fear drainage basin in North Carolina. Environmental
science & technology. 41(15) 5271 -5276.

Nalbantlar, B and O. C. Aislan. 2017. Determination of the Perflorooctane Sulfonate-Induced
Genotoxic Response in Mylilus galloprovincialis Using a Micronucleus Assay.Zool. Ecol., 27:
161-167.

Newsted, J. L., R. Holem, G. Hohenstein, C. Lange, M. Ellefson, W. Reagen and S. Wolf. 2017.
Spatial and temporal trends of poly- and perfluoroalkyl substances in fish fillets and water
collected from pool 2 of the Upper Mississippi River. Environ Toxicol Chem. 36(11): 3138-
3147.

Nguyen, T. V., M. Reinhard, H. Chen and K. Y. Gin. 2016. Fate and transport of perfluoro- and
polyfluoroalkyl substances including perfluorooctane sulfonamides in a managed urban water
body. Environ Sci Pollut Res Int. 23(11): 10382-10392.

181


-------
NJDEP. (New Jersey Department of Environmental Protection). 2019. Investigation of Levels of
Perfluorinated Compounds in New Jersey Fish, Surface Water, and Sediment. New Jersey
Department of Environmental Protection Division of Science, R., and Environmental
Health,SRI 5-010. pp. 1-46.

https://www.ni.gov/dep/dsr/publications/Investigation%20of%20Levels%20of%20Perfluorinated
%20Compounds%20in%20New%20Jersev%20Fish.%20Surface%20Water.%20and%20Sedime

nt.pdf

Nowak, C., A. Czeikowitz, C. Vogt, M. Otetken, B. Streit and K. Schwenk. 2008. Variation in
sensitivity to cadmium among genetically characterized laboratory strains of the midge
Chironomus riparius. Chemosphere. 71: 1950-1956.

Nowak, C., T. Hankeln, E. R. Schmidt and K. Schwenk. 200o. Development and localization of
microsatellite markers for the sibling species Chironomus riparius and (luronomuspiger
(Diptera, Chironomidae). Mol. Ecol. Notes, 6: 915-917.

Nowak, C., D. Jost, C. Vogt, M. Oetken, K. Schwenk and J. Oehlmann. 2007a. Consequences of
inbreeding and reduced genetic variation on tolerance to cadmium stress in the midge
Chironomus riparius. Aquat. Toxicol. 85: 278-284.

Nowak, C., C. Vogt, J. B. Diogo and K. Schwenk 2<)07b. Genetic inpoverishment in laboratory
cultures of the test organism Chironomus ri/Hirius I ji\ iron. Toxicol Chem. 26: 1018-1022.

Nowak, C., C. Vogt, M. Pfenninger, K. Schwenk. .1 Oehlmann, B. Streit andM. Oetken. 2009.
Rapid genetic erosion in pollutant-exposed experimental chironomid populations. Environ.

Pollut. 157: 881-886.

NRC. 2013. Assessing risks to endangered and threatened species from pesticides. National
Academies Press Washington, DC. pp 142

Oakes, K. D., I\ k. Sibley. J. \V. Martin, D. D. MacLean, K. R. Solomon, S. A. Mabury and G.
J. Van der Kraak. 2<)i)5 Short-Term I Exposures of Fish to Perfluorooctane Sulfonate: Acute
Effects on Fatty Acyl-Co.\ Oxidase Activity, Oxidative Stress, and Circulating Sex
Steroids Ijniron. Toxicol. Chem., 24: 1 172-1181.

OECD. (Organisation for Economic Cooperation and Development). 1992. Fish, acute toxicity
test. Guideline 2<)3 In Guidelines for Testing of Chemicals. Paris, France.

OECD. (Organisation for Economic Co-operation and Development). 2001. Guidance Document
on the Use of the Harmonised System for the Classification of Chemicals Which Are Hazardous
for The Aquatic Environment. Joint Meetings of the Chemicals Committee and the Working
Party on Chemicals, Pesticides and Biotechnology. ENV/JM/MONO(2001)8. https://www.oecd-

ilibrarv.org/docserver/9789264078444~

en.pdf?e	ฃ0&id=id&accname=guest&checksum=2AA44B183B2EED270D64D

182


-------
OECD. ( Organization for Economic Co-Operation and Development). 2002. Hazard
Assessment of Perfluorooctane Sulfonate (PFOS) and its Salts. ENV/JM/RD(2002)17/FINAL.
pp.1-363.

OECD. (Organisation for Economic Cooperation and Development). 2004a. Chironomid toxicity
test using spiked sediment; Guideline for Test-ing of Chemicals No. 218. Paris, France.

OECD. (Organisation for Economic Cooperation andDevelopment). 2004b. Chironomid toxicity
test using spiked water; Guideline for Testingof Chemicals No. 219. Paris, France.

OECD. (Organisation for Economic Cooperation and Development) 2010. Sediment-Water
Chironomid Life-Cycle Toxycity Test using spikedwater or spiked sediment; Guideline for the
Testing of Chemicals No. 233. Paris, France.

OECD. (Organisation for Economic Cooperation and Development). 2013 l isli embryo acute
toxicity (FET) test; Guideline for the Testing of Chemicals No. 236. https://	sd-

ilibrarv.ore/docserver/9789264203709-

en.pdf?eM.'U^ I I	A.id=id&accname= ..	)D8ADD33E	I *>

OECD. (Organisation for Economic Co-operation and De\ elopmenl). 2021. Reconciling
Terminology of the Universe of Per- and Pok lluoioalkx I Substances Recommendations and
Practical Guidance. Series on Risk Management \o.M Paris, France.
ENV/CBC/MONO(2021)25. pp.34.

https://www. oecd.org/official documents/	df/?cote=ENV/CB C/MONO(2

021 )25&docLanguage=en

Olson, A. D. 2017. An investigation into the toxicity, bioconcentration, and risk of
perfluoroalkyl substances in aquatic ta\a

Ortiz-Villanue\ a, L . .1 Janmot, R. Martinez, L. Navarro-Martin, B. PinaandR. Tauler. 2018.
Assessment of Endocrine Disi iiptors Effects on Zebrafish (Danio rerio) Embryos by Untargeted
LC-HRMS Metabolomic Analysis.Sci. Total Environ., 635: 156-166 (Supplemental journal
materials)

Padilla, S., I) Coiiim, B. Padnos, D. L. Hunter, A. Beam, K. A. Houck, N. Sipes, N.

Kleinstreuer, T. knudscn. I) .1 Dix and D. M. Reif. 2012. Zebrafish developmental screening of
the ToxCast Phase I chemical library. Reprod Toxicol. 33(2): 174-187.

Palmer, S. J. and H. O. Krueger. 2001. PFOS: A Frog Embryo Teratogenesis Assay - Xenopus
(FETAX).Project 454A-116, Wildlife International Ltd., Easton, MD. 181 p.

Palmer, S. J., R. L. Van Hoven and H. O. Krueger. 2002a. Perfluorooctanesulfonate, Potassium
Salt (PFOS): A 96-Hr. Static Acute Toxicity Test with the Rainbow Trout (Oncorhynchus
mykiss).Project 454A-145, Wildlife International Ltd., Easton, MD. 42 p.

183


-------
Palmer, S. J., R. L. VanHoven and H. O. Krueger. 2002b. Perfluorooctanesulfonate, Potassium
Salt (PFOS): A 96-Hour Static-Renewal Acute Toxicity Test with The Sheepshead Minnow
(Cyprinodon Variegatus).PROJECT NUMBER: 454A-146A, Wildlife International, Ltd.

Palumbo, A. J., P. L. TenBrook, T. L. Fojut, I. R. Faria and R. S. Tjeerdema. 2012. Aquatic life
water criteria derived via the UC Davis Method: I. Organophosphate Insecticides. In.

Tjeerdema, R.S., ed., Rev Environ Contam Toxicol. 216. ed. Springer, New York, NY.

Pan, Y., H. Zhang, Q. Cui, N. Sheng, L. W. Y. Yeung, Y. Sun, Y. Guo and J. Dai. 2018.
Worldwide Distribution of Novel Perfluoroether Carboxylic and Sulfonic Acids in Surface
Water. Environ Sci Technol. 52(14): 7621-7629.

Park, K., C. Nikapitiya, T.-S. Kwak and I.-S. Kwak. 2015. Anlioxidali\e-ielated genes
expression following perfluorooctane sulfonate (PFOS) exposure in the inlei'lidal mud crab,
Macrophthalmus japonicus. Ocean Science Journal. 50(3): 547-556.

Paul, A. G., K. C. Jones and A. J. Sweetman. 2009. A first global production, emission, and
environmental inventory for perfluorooctane sulfonate I ji\ iron mental science & technology.
43(2): 386-392.

Penland, T. N., W. G. Cope, T. J. Kwak, M J Strynar, C. A. Grieshaber, R. J. Heise and F. W.
Sessions. 2020. Trophodynamics of Per- and Pol\ lluoroalkyl Substances in the Food Web of a
Large Atlantic Slope River. Environ. Sci Technol 54( I I) 6800-6SI I

Pizzurro, D. M., M. Seeley, L. E. Kerper and B. D. Beck. 2<)|i) Interspecies differences in
perfluoroalkyl substances (PFAS) toxicokinetics and application to health-based criteria. Reg.
Toxicol. Pharma. 106: 239-250.

Place, B. J. and J. A. Field. 2012. Identification of novel fluorochemicals in aqueous film-
forming foams used In the I S military I-n\ iron Sci Technol. 46(13): 7120-7127.

Plumlcc. \1 11 . J l.aialxv and M Reinhard. 2008. Perfluorochemicals in water reuse.
Chemosphere 72(l<)) l 541 -1 547

Plumlee, Mil. k McNeill and M. Reinhard. 2009. Indirect Photolysis of Perfluorochemicals:
Hydroxyl Radical-Initiated Oxidation of N-Ethyl Perfluorooctane Sulfonamido Acetate (N-
EtFOSAA) and Other Periluoroalkanesulfonamides. Environmental Science & Technology.
43(10): 3662-360S

Powley, C. R., S. W. George, M. H. Russell, R. A. Hoke and R. C. Buck. 2008. Polyfluorinated
chemicals in a spatially and temporally integrated food web in the Western Arctic. Chemosphere.
70: 664-672.

Preus-Olsen, G., M. O. Olufsen, S. A. Pedersen, R. J. Letcher and A. Arukwe. 2014. Effects of
Elevated Dissolved Carbon Dioxide and Perfluorooctane Sulfonic Acid, Given Singly and in
Combination, on Steroidogenic and Biotransformation Pathways of Atlantic Cod. Aquat.

Toxicol., 155: 222-235.

184


-------
Prevedouros, K., I. T. Cousins, R. C. Buck and S. H. Korzeniowski. 2006. Sources, fate and
transport of perfluorocarboxylates. Environ Sci Technol. 40(1): 32-44.

Prosser, R. S., K. Mahon, P. K. Sibley, D. Poirier and T. Watson-Leung. 2016. Bioaccumulation
of perfluorinated carboxylates and sulfonates and polychlorinated biphenyls in laboratory-
cultured Hexagenia spp., Lumbriculus variegatus and Pimephales promelas from field-collected
sediments. Sci Total Environ. 543(Pt A): 715-726.

Qu, R., J. Liu, L. Wang and Z. Wang. 2016. The toxic effect and bioaccumulation in aquatic
oligochaete Limnodrilus hoffmeisteri after combined exposure to cadmium and perfluorooctane
sulfonate at different pH values. Chemosphere. 152: 496-502.

Raimondo, S. and M. G. Barron. 2020. Application of Interspecies Correlation Estimation (ICE)
models and QSAR in estimating species sensitivity to pesticides. SAR and OSAR in
Environmental Research. 31(1): 1-18.

Raimondo, S., C.R. Lilavois, andM. Elias. . In prep Application of Interspecies Correlation
Estimation models for derivation of water quality standards of PFAS chemicals.

Raimondo, S., C. R. Jackson and M. G. Barron. 2010. Influence of taxonomic relatedness and
chemical mode of action in acute interspecies estimation models lor aquatic species.
Environmental science & technology. 44( N). 771 1-7716.

Raimondo, S., D. N. Vivian and M. G. Barron. (U.S l-n\ ironmental Protection Agency). 2015.
Web-based interspecies correlation estimation (Web-ICE) for acute toxicity: user manual.
Version 3.3. Office of Research and Development, Gulf Ecology Division. Gulf Breeze, FL.
EPA/600/R-15/192

Rainieri, S.,N. Conlledo. T l.anuerholc. I:. Madorran, M. Sala and A. Barranco. 2017. Toxic
Effects of Perfluorinated Compounds at I luman Cellular Level and on a Model Vertebrate. Food
Chem. Toxicol. I<>4 14-25 (Supplemental journal materials).

Remde. A and R. Debus I liiodeuradability of fluorinated surfactants under aerobic and
anaerobic conditions. Chemosphere. 32(8): 1563-1574.

Remucal, C k 2<) I ^ Spatial and temporal variability of perfluoroalkyl substances in the
Laurentian Great l.akes F.m iron Sci Process Impacts. 21(11): 1816-1834.

Research, N. I. f. A. 2007a. Brominated flame retardants and perfluorinated substances in air.
OR 16/2008. Norwegian Institute for Air Research. ISBN 978-82-425-1962-7. www.nilu.no

Research, N. I. f. A. 2007b. Brominated flame retardants and perfluorinated substances in air.
OR28/2008. Norwegian Institute for Air Research. ISBN 978-82-425-1984-9. www.nilu.no

Rewerts, J. N., E. C. Christie, A. E. Robel, T. A. Anderson, C. McCarthy, C. J. Salice and J. A.
Field. 2021. Key considerations for accurate exposures in ecotoxicological assessments of
perfluorinated carboxylates and sulfonates. Environmental toxicology and chemistry. 40(3): 677-
688.

185


-------
Rhoads, K. R., E. M.-L. Janssen, R. G. Luthy and C. S. Criddle. 2008. Aerobic biotransformation
and fate of N-ethyl perfluorooctane sulfonamidoethanol (N-EtFOSE) in activated sludge.
Environmental science & technology. 42(8): 2873-2878.

Rice, T. 2008. A review of methods for maintaining odonate larvae in the laboratory, with a
description of a new technique. Odonatologica. 37(1): 41-54.

RIVM (National Institue for Public Health and the Envrionment). 2010. Environmental risk
limits for PFOS: A proposal for water quality standards in accordance with the Water
Framework Directive. Bilthoven, the Netherlands Report 601714<~> 13 '2010: 1 -70.

Rodea-Palomares, I., F. Leganes, R. Rosal and F. Fernandez-Pinas 2<) 12 Toxicological
interactions of perfluorooctane sulfonic acid (PFOS) and pcilluorooclanoic acid (PFOA) with
selected pollutants. J Hazard Mater. 201-202: 209-218.

Rodea-Palomares, I., M. Makowski, S. Gonzalo, M. Gonzalez-PIeiter, I' l.cgancs and F.
Fernandez-Pinas. 2015. Effect of PFOA/PFOS Pre-Exposure on the Toxicity of the Herbicides
2,4-D, Atrazine, Diuron and Paraquat to a Model Aquatic Pholosynthetic Microorganism.
Chemosphere. 139: 65-72.

Roland, K., P. Kestemont, R. Loos, S. Ta\ azzi, B. Paracchini, ('. Ik-lpaire, M. Dieu, M. Raes and
F. Silvestre. 2014. Looking for Protein Expression Signatures in European Eel Peripheral Blood
Mononuclear Cells After In Vivo Exposure to Perfluorooctane Sulfonate and a Real World Field
Study. Sci. Total Environ. 468/469: 958-967.

Saez, M., P. de Voogt and J. R. Parsons. 2008. Persistence ol'perfluoroalkylated substances in
closed bottle tests with municipal sewage sludge. Environ Sci Pollut Res Int. 15(6): 472-477.

Sakurai, T., J. Kobayashi, N. Ito, S. Serizawa, H. Shiraishi, T. Yabe, Y. Ishii andN. Suzuki.
2017. Respiratory Uptake and Depuration Kinetics of Perfluorooctanesulfonate (PFOS) in a
Marine Sandworm Species Ikill I jniron Contain Toxicol. 99(2): 203-207.

San-Seuundo, L., L. Guimaracs. (' I' Toi ija, E. M. Beltran, L. Guilhermino and M. V. Pablos.
2016. Alterations in Gene l-xpression Levels Provide Early Indicators of Chemical Stress During
Xenopus lac\ is l-mbryo Dc\ clopment: a Case Study with Perfluorooctane Sulfonate (PFOS).
Ecotoxicol. I ji\ iron Saf. 127 51-60.

Sanderson, H., T. M lioudreau, S. A. Mabury, W. J. Cheong and K. R. Solomon. 2002.
Ecological Impact and Environmental Fate of Perfluorooctane Sulfonate on the Zooplankton
Community in Indoor Microcosms.Environ. Toxicol. Chem., 21: 1490-1496.

Sant, K. E., K. Annunziato, S. Conlin, G. Teicher, P. Chen, O. Venezia, G. B. Downes, Y. Park
and A. R. Timme-Laragy. 2021. Developmental exposures to perfluorooctanesulfonic acid
(PFOS) impact embryonic nutrition, pancreatic morphology, and adiposity in the zebrafish,

Danio rerio. Environ Pollut. 275: 116644.

186


-------
Sant, K. E., H. M. Jacobs, K. A. Borofski, J. B. Moss and A. R. Timme-Laragy. 2017.

Embryonic exposures to perfluorooctanesulfonic acid (PFOS) disrupt pancreatic organogenesis
in the zebrafish, Danio rerio. Environ Pollut. 220(Pt B): 807-817.

Sant, K. E., P. P. Sinno, H. M. Jacobs and A. R. Timme-Laragy. 2018. Nrf2a modulates the
embryonic antioxidant response to perfluorooctanesulfonic acid (PFOS) in the zebrafish, Danio
rerio. Aquat Toxicol. 198: 92-102.

Sasaki, K., K. Harada, N. Saito, T. Tsutsui, S. Nakanishi, H. Tsuzuki and A. Koizumi. 2003.
Impact of airborne perfluorooctane sulfonate on the human body burden and the ecological
system. Bull. Environ. Contam. Toxicol. 71: 408-413.

Schroder, H. F. 2003. Determination of fluorinated surfactants and their metabolites in sewage
sludge samples by liquid chromatography with mass spectrometry and tandem mass
spectrometry after pressurised liquid extraction and separation on fluorine-modified reversed-
phase sorbents. Journal of Chromatography A. 1020( 1): 131-151.

Scott, B. F., A. O. De Silva, C. Spencer, E. Lopez, S. M Backus and D. C. G. Muir. 2010.
Perfluoroalkyl acids in Lake Superior water: Trends and sources. Journal of Great Lakes
Research. 36(2): 277-284.

Sedlak, M. D., J. P. Benskin, A. Wong, R Grace and I) .1 Greig. 2d I 7 Per- and polyfluoroalkyl
substances (PFASs) in San Francisco Bay wildlife Temporal trends, exposure pathways, and
notable presence of precursor compounds ('hcmosphcre 185' 1217-1226.

Seyoum, A., A. Pradhan. .1 .lass and P E. Olsson. 2020. Perilnorinated Alkyl Substances Impede
Growth, Reproduction. Lipid Metabolism and Lifespan in Daphnia magna. Sci. Total Environ.
737: 12 p. (139082)

Sharpe, R. T,. J. P lienskin. A 11 l.aarman, S. L. Macleod, J. W. Martin, C. S. Wong and G. G.
Goss. 2010. Periluorooctane sulfonate toxicity, isomer-specific accumulation, and maternal
transfer in zebralish (Danio rerio) and rainbow trout (Oncorhynchus mykiss). Environ Toxicol
Chem 21->(9V 1957-1960

Shi, \ . Y Du. P K. Lam, R S. Wu and B. Zhou. 2008. Developmental toxicity and alteration of
gene expression in zebrafish embryos exposed to PFOS. Toxicol Appl Pharmacol. 230(1): 23-32.

Shi, X., C. Liu, G. Wu and B. Zhou. 2009. Waterborne exposure to PFOS causes disruption of
the hypothalamus-piluilary-thyroid axis in zebrafish larvae. Chemosphere. 77(7): 1010-1018.

Shi, X. and B. Zhou. 2010. The Role of Nrf2 and MAPK Pathways in PFOS-Induced Oxidative
Stress in Zebrafish Embryos. Toxicol. Sci. 115: 391-400.

Simcik, M. F. and K. J. Dorweiler. 2005. Ratio of perfluorochemical concentrations as a tracer of
atmospheric deposition to surface waters. Environmental science & technology. 39(22): 8678-
8683.

187


-------
Simpson, S. L., Y. Liu, D. A. Spadaro, X. Wang, R. S. Kookana and G. E. Batley. 2021. Chronic
effects and thresholds for estuarine and marine benthic organism exposure to perfluorooctane
sulfonic acid (PFOS)-contaminated sediments: Influence of organic carbon and exposure routes.
Science of The Total Environment. 776: 146008.

Sinclair, E., D. T. Mayack, K. Roblee, N. Yamashita and K. Kannan. 2006. Occurrence of
perfluoroalkyl surfactants in water, fish, and birds from New York State. Arch Environ Contam
Toxicol. 50(3): 398-410.

Smithwick, M., R. J. Norstrom, S. A. Mabury, K. Solomon, T. J. F.\ ans, I. Stirling, M. K. Taylor
and D. C. Muir. 2006. Temporal trends of perfluoroalkyl contaminants in polar bears (Ursus
maritimus) from two locations in the North American Arctic, 1^72 2< >02. Environmental
science & technology. 40(4): 1139-1143.

Spachmo, B. and A. Arukwe. 2012. Endocrine and developmental effects in Atlantic salmon
(Salmo salar) exposed to perfluorooctane sulfonic or perfluorooctane carboxvlie acids. Aquat
Toxicol. 108: 112-124.

Spence, R., M. Fatema, M. Reichard, K. Huq, M. Wahak / Ahmed and C. Smith. 2006. The
distribution and habitat preferences of the zebrafish in Bangladesh Journal of fish biology.
69(5): 1435-1448.

Spulber, S., P. Kilian, W. N. Wan Ibrahim. \ Onishchenko. M. L'lhaq. L. Norrgren, S. Negri,
M. Di Tuccio and S. Ceccatelli. 2014. PFOS induces h>elia\ioral alterations, including
spontaneous hyperactivity that is corrected by dexamfetamine in zebrafish larvae. PLoS One.
9(4): e94227.

SRC (Syracus Research Corporation). 2016. PHYSPROP Database. SRC, Inc. North Syracuse,
NY. Accessed May 2016.	;;v,.:,;; -we-do/environmental/scientific-

database

Stahl. I. L, B. D. Snyder. A R. Olsen. T M. Kincaid, J. B. Wathen andH. B. McCarty. 2014.
Perflnoi inated compounds in fish from I S. urban rivers and the Great Lakes. Sci Total Environ.
499: 1X5-1

Stefani, F., M Rusconi, S. \ alsecchi and L. Marziali. 2014. Evolutionary ecotoxicology of
perfluoralkyl substances (PFASs) inferred from multigenerational exposure: a case study with
Chironomus riparius (Diptera, Chironomidae). Aquat Toxicol. 156: 41-51.

Stengel, D., F. Zindler and T. Braunbeck. 2017. An Optimized Method to Assess Ototoxic
Effects in the Lateral Line of Zebrafish (Danio rerio) Embryos. Comp. Biochem. Physiol. C
Toxicol. Pharmacol. 193: 18-29.

Stock, N. L., V. I. Furdui, D. C. G. Muir and S. A. Mabury. 2007. Perfluoroalkyl contaminants in
the Canadian Arctic: evidence of atmospheric transport and local contamination. Environ. Sci.
Technol. 41: 3529-3536.

188


-------
STS/MPCA. (Minnesota Pollution Control Agency). 2007. Surface Water Quality Criterion for
Perfluorooctane Sulfonic Acid St. Paul, MN. STS Project 200604796

https://www.pca.state.mn.iis/sites/defaiilt/files/pfos-report.pdf

Suski, J. G., C. J. Salice, M. K. Chanov, J. Ayers, J. Rewerts and J. Field. 2021. Sensitivity and
Accumulation of Perfluorooctanesulfonate and Perfluorohexanesulfonic Acid in Fathead
Minnows (Pimephales promelas) Exposed over Critical Life Stages of Reproduction and
Development. Environ Toxicol Chem. 40(3): 811-819.

Sutherland, C. A. and H. O. Krueger. 2001. PFOS: A 96-Hour Toxicity Test with the Freshwater
Diatom (Navicula pelliculosa).Project 454A-112, Wildlife International Ltd., Easton, MD. 56 p.
(OPPTS 850-5400).

Thompson, J., A. Roach, G. Eaglesham, M. E. Bartkow, K. Edge and .1 I- Mueller. 2011.
Perfluorinated alkyl acids in water, sediment and wildlife from Sydney Harbour and
surroundings. Marine Pollution Bulletin.(62): 2869-2875.

Tomy, G. T., S. A. Tittlemier, V. P. Palace, W. R. Budakouski. E. Braekevelt, I. lirinkworth
and K. Friesen. 2004. Biotransformation of A-Ethyl Peril uorooctanesulfonamide by Rainbow
Trout (Onchorhynchus mykiss) Liver Microsomes. Environ Sci Technol. 38: 758-762.

Tornabene, B. J., M. F. Chislock, M. E. Gannon. M S SepuKcda and J. T. Hoverman. 2021.
Relative acute toxicity of three per- and polylluoioalkyl substances on nine species of larval
amphibians. Integr Environ Assess Manau I 7( 4) (•>X4-(->l)<)

Truong, L., D. M. Reif, L. St Mary, M. C. Geier, H. D. Tinong andR. L. Tanguay. 2014.
Multidimensional In Vivo Hazard Assessment Using Zebrafish. Toxicol. Sci. 137: 212-233
(Supplemental Journal Materials)

Tse, W. K. F . .1 W Li, A. C. K. Tsc. T. I Chan. J. C. H. Ho, R. S. S. Wu, C. K. C. Wong and
K. P. Lai. 2016. Fatly Liver Disease Induced by Perfluorooctane Sulfonate: Novel Insight from
Transci iptonie Analysis Chcmospherc I 59: 166-177 (Supplemental journal materials).

U.S.EPA (I S Environmental Protection Agency). 1985. Guidelines for deriving numerical
national water quality criteria lor the protection of aquatic organisms and their uses. Office of
Research and Development PB85-227049. https://www.epa.gov/sites/default/files/:;

02/docum ents/guid"i:— " -qualitv-criteria.pdf

U.S.EPA. (U.S. Environmental Protection Agency). 1991. Technical Support Document for
Water Quality-based Toxics Control. Office of Water. Washington, DC. EPA/505/2-90-001.
pp.143, https://www3.epa.eov/npdes/pubs/owm0264.pdf

U.S.EPA. (U.S. Environmental Protection Agency). 1995. Short-term methods for estimating the
chronic toxicity of effluents and receiving waters to west coast marine and estuarine organisms.
Environmental Monitoring and Support Laboratory. Cincinnati, Ohio. EPA/600/R-95/136.

U.S.EPA. (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk
Assessment. EPA, U. S.,Washington, DC. Federal Register 63(93):26846-26924.

189


-------
U.S.EPA (U.S. Environmental Protection Agency). 2000a. 3M Phase-Out Plan for POSF-Based
Products. Accessed November 2019.

https://archive.epa.eov/epapaees/newsroom archive/newsreleases/33aa946ef ' ~	58elO

05246b4.html.

U.S.EPA. (U.S. Environmental Protection Agency). 2000b. Methods for measuring the toxicity
and bioaccumulation of sediment-associated contaminants with freshwater invertebrates. Office
of Research and Development. Washington, DC. EPA-600-R99-064.

U.S.EPA. (U.S. Environmental Protection Agency). 2002a. Methods for measuring the acute
toxicity of effluents and receiving waters to freshwater and marine organisms. Office of Water.
Washington, DC. EPA/600/4-90/027F. https://www.epa.eov/sites/production/files/2015-
08/documents/acute-fresh'water-and-marine-wet-manual 2002.pdf

U.S.EPA. (U.S. Environmental Protection Agency). 2002b. Methods for measuring the acute
toxicity of effluents and receiving waters to freshwater and marine organisms LPA/600/4-
90/027F.

U.S.EPA. (U.S. Environmental Protection Agency). 2013 \<.|uatic Life Ambient Water Quality
Criteria for Ammonia - Freshwater. Office of Water, Office of Science and Technology.
Washington. DC. EPA 822-R-18-002. httos -.epa.ee ปroduction/files/2015-

08/documents/aquatic-life-ambient-watei	-for-a	-freshwati	f

U.S.EPA. (U.S. Environmental Protection Agency). 2d Km Aquatic Life Ambient Water Quality
Criterion for Selenium Freshwater. Office of Water. Washington, DC. EPA 822-R-16-006.

https://www. epa. gov/ si	3s/2016-

07/docum ents/aquati c	lenium - freshwa. I |;df

U.S.EPA. (U.S. Environmental Piolection Agency). 2016b. Ecological Effects Test Guidelines
OCSPP 850 irr5 Ficshualcr and Saltwater l-'ish Acute Toxicity Test. Office of Chemical Safety
and Pollution Pre\ union Washington. DC. EPA 712-C-16-007.

https://nepis.epi				>SH65.PDF?Dockev ffl.65.PDF

U.S.EPA (I S. Environmental Protection Agency). 2016c. Series 850 - Ecological Effects Test
Guidelines. Office of Chemical Safety and Pollution Prevention, Washington, DC. Accessed
March 2021.	www >ov/test-guidelines-pesticides-and-toxic-substances/series-850-

ecoloeical-e	 st~guii is.

U.S.EPA. (U.S. Environmental Protection Agency). 2018. Application of systematic review in
TSCA risk evaluations. Prevention, O. o. C. S. a. P.,Washington, DC. EPA Document# 740-P1-
8001. https://www.epa.gov/sites/production/files/2018-

06/documents/final application of sr in tsc	if

U.S.EPA. (U.S. Environmental Protection Agency). 2020. Long-Chain Perfluoroalkyl
Carboxylate and Perfluoroalkyl Sulfonate Chemical Substances; Significant New Use Rule.
pp.45109-45126.

190


-------
U.S.EPA 2021a. CompTox Chemicals Dashboard. Perfluorooctanesulfonate. Accessed March 5,
2021. https://comptox.epa.eov/dashboard/dsstoxdb/results?search=DTXSID80108992.

U.S.EPA. (U.S. Environmental Protection Agency). 2021b. National PFAS Testing Strategy:
Identification of Candidate Per- and Polyfluoroalkyl Substances (PFAS) for Testing.

https://www.epa.eov/sYStem/files/documents/20	>fas-natl-test-strategy.pdf

U.S.FWS. (U.S. Fish and Wildlife Service). 2018. Zebra Danio (Danio rerio) Ecological Risk
Screening Summary. https://www.fws.gov/fisheries/ans/erss/uncertainrisk/ERSS-Danio-rerio-

2018.pdf

Ulhaq, M., G. Carlsson, S. Orn and L. Norrgren. 2013. Comparison of developmental toxicity of
seven perfluoroalkyl acids to zebrafish embryos. Environ Toxicol Pharmacol 36(2): 423-426.

UNEP (United Nations Environmental Program). 2006. Report of the Persistent Organic
Pollutants Review Committee on the Work of its Second Meeting. Addendum Risk profile on
perfluorooctane sulfonate. UNEP/POPS/POPRC.2/1 7. Add 5 I nited Nations. Accessed May
2016. http://chm.pop.int/Default.aspx/tabid=2301.

Van Gossum, H., J. Bots, T. Snijkers, J. Meyer, S. Van \Yassenbergh, W. De Coen and L. De
Bruyn. 2009. Behaviour of damselfly Ian ae (I jiallagma cyalhiuci iim) (Tnsecta, Odonata) after
long-term exposure to PFOS. Environ Pol I nl 157(4) 1332-1330.

Vedagiri, U. K., R. H. Anderson, H. M. Loso and C M Sclmach. 2018. Ambient levels of PFOS
and PFOA in multiple environmental media. Remediation Journal. 28(2): 9-51.

Vogs, C., G. Johanson. VI. Naslund. S Wulff, M. Sjodin, M. Hellstrandh, J. Lindberg andE.
Wincent. 2019. Toxicokinetics of Perlluorinated Alkyl Acids Influences Their Toxic Potency in
the Zebrafish Embryo (Danio ici io) Ijniron Sci Technol. 53: 3898-3907 (Supplemental
Journal Materials).

Vogt. C . D. Belz, S Galluha. C \ouak. M. Oetken and J. Oehlmann. 2007a. Effects of
cadmium and tributyltin on development and reproduction of the non-biting midge Chironomus
riparius (Diptera)—baseline experiments for future multi-generation studies. Journal of
Environmental Science and I lealth, Part A. 42(1): 1-9.

Vogt, C., M. Hel.V (' Nowak. J B. Diogo, J. Oehlmann and M. Oetken. 2010. Effects of
cadmium on life-cycle parameters in a multi-generation study with Chironomus riparius
following a pre-exposure of populations to two different tributyltin concentrations for several
generations. Ecotoxicology. 19(7): 1174-1182.

Vogt, C., C. Nowak, J. B. Diogo, M. Oetken, K. Schwenk and J. Oehlmann. 2007b. Multi-
generation studies with Chironomus riparius - Effects of low tributyltin concentrations on life
history parameters and genetic diversity. Chemosphere. 67(11): 2192-2200.

Wang, M., J. Chen, K. Lin, Y. Chen, W. Hu, R. L. Tanguay, C. Huang and Q. Dong. 2011.
Chronic zebrafish PFOS exposure alters sex ratio and maternal related effects in F1 offspring.
Environ Toxicol Chem. 30(9): 2073-2080.

191


-------
Wang, S., J. Huang, Y. Yang, Y. Hui, Y. Ge, Larssen, T, G. Yu, S. Deng, B. Wang and C.
Harman. 2013a. First Report of a Chinese PFOS Alternative Overlooked for 30 Years: Its
Toxicity, Persistence, and Presence in the Environment.Environ. Sci. Technol., 47: 10163-10170
(Supplemental Journal Materials).

Wang, T. T., G. G. Ying, L. Y. He, Y. S. Liu and J. L. Zhao. 2020. Uptake Mechanism,
Subcellular Distribution, and Uptake Process of Perfluorooctanoic Acid and Perfluorooctane
Sulfonic Acid by Wetland Plant Alisma orientale.Sci. Total Environ., 733: lip. (Supplemental
journal materials).

Wang, Z., I. T. Cousins, M. Scheringer and K. Hungerbuehler. 2015. Hazard assessment of
fluorinated alternatives to long-chain perfluoroalkyl acids (PFAAs) and their precursors: status
quo, ongoing challenges and possible solutions. Environ Int. 75: 172-1 7l).

Wang, Z., I. T. Cousins, M. Scheringer and K. Hungerbuhler. 2013b. Fluoiinaled alternatives to
long-chain perfluoroalkyl carboxylic acids (PFCAs), perfluoroalkane sulfonic acids (PFSAs) and
their potential precursors. Environ Int. 60: 242-248.

Wang, Z., J. C. DeWitt, C. P. Higgins and I. T. Cousins. 2017. A Never-Ending Story of Per- and
Polyfluoroalkyl Substances (PFASs)? Environ Sci Technol. 51(5)' 2508-2518.

Warne MSt.J., B. G. E., van Dam R.A., Chapman .1 C., Fox DR. I lickey C.W. and Stauber .J.L.
2018. Revised Method for Deriving Australian and New Zealand \Yaler Quality Guideline
Values for Toxicants - update of 2015 version. Prepared for the Council of Australian
Government's Standing Council on Environment and Water (SCEW). Department of Science.
Information Technology and lnno\alion Brisbane, Queensland.

https://www.waterc	iult/files/documents/warne-wqg-derivation2018.pdf

Washington, J. W. and T M Jenkins 2<> I 5 Abiotic Hydrolysis of Fluorotelomer-Based
Polymers as a Source of Pci'lliioiocaiium lales al the Global Scale. Environ Sci Technol. 49(24):
14129-14135.

Washington. .1. W., T M Jenkins, k Rankin and J. E. Naile. 2015. Decades-scale degradation of
commercial, side-chain, fluorolelomer-based polymers in soils and water. Environ Sci Technol.
49(2): 915-^23

Wei, Y., X. Shi. 11 /hang. J W ang, B. Zhou and J. Dai. 2009. Combined Effects of
Polyfluorinated and Perilnorinated Compounds on Primary Cultured Hepatocytes from Rare
Minnow (Gobiocypris rarus) Using Toxicogenomic Analysis. Aquat. Toxicol. 95: 27-36
(Supplemental Journal Materials).

Wen, W., X. Xia, X. Chen, H. Wang, B. Zhu, H. Li and Y. Li. 2016. Bioconcentration of
perfluoroalkyl substances by Chironomusplumosus larvae in water with different types of
dissolved organic matters. Environ. Pollut. 213: 299-307.

Willming, M. M., C. R. Lilavois, M. G. Barron and S. Raimondo. 2016. Acute toxicity
prediction to threatened and endangered species using interspecies correlation estimation (ICE)
models. Environmental science & technology. 50(19): 10700-10707.

192


-------
Wixon, J. 2000. Featured organism: Danio rerio, the zebrafish. Yeast. 17(3): 225-231.

Wu, J., Z. Liu, Z. Yan and X. Yi. 2015. Derivation of water quality criteria of phenanthrene
using interspecies correlation estimation models for aquatic life in China. Environmental Science
and Pollution Research. 22(12): 9457-9463.

Wu, J., Z. Yan, X. Yi, Y. Lin, J. Ni, X. Gao, Z. Liu and X. Shi. 2016. Comparison of species
sensitivity distributions constructed with predicted acute toxicity data from interspecies
correlation estimation models and measured acute data for Benzo [a] pyrene. Chemosphere. 144:
2183-2188.

Wu, X., Q. Huang, C. Fang, T. Ye, L. Qiu and S. Dong. 2012. PFOs Induced Precocious
Hatching of Oryzias melastigma - from Molecular Level to Individual I .e\ el Chemosphere. 87:
703-708.

Xia, J.-G., Y.-J. Ma, W.-M. Guo, L. Huang and S.-J. Fu. 2015a. Temperalure-dependent effects
of PFOS on risk recognition and fast-start performance in juvenile Spinibarhns sinensis. Aquatic
Biology. 24(2): 101-108.

Xia, J., Z. Cao, J. Peng, S. Fu and C. Fu. 2014. The Use of Spontaneous Behavior, Swimming
Performances and Metabolic Rate to Evaluate Toxicity of PFOS on Topmouth Gudgeon
Pseudorasbora parva.ActaEcol. Sin., 34: 2K4-2X1)

Xia, J. and C. Niu. 2016. Acute toxicity effects of peilliioiooclanc sulfonate on sperm vitality,
kinematics and fertilization success in zebrafish. Chinese Journal of Oceanology and Limnology.
35(4): 723-728.

Xia, X., Z. Dai, A. 11 Rahearisoa. P Zhao and X. Jiang. 2015b. Comparing humic substance and
protein compound effects on the hioaccunuilation of perfluoroalkyl substances by Daphnia
magna in water Chemosphere I I ^ l)7K-l)X(-<

Xia, X . A. H. Rabearisoa. Z Dai. \ Jiang, P. Zhao and H. Wang. 2015c. Inhibition effect of
Na+ and Ca2+ on the bioaccuniulation of perfluoroalkyl substances by Daphnia magna in the
presence of protein. Environmental toxicology and chemistry. 34(2): 429-436.

Xia, X., A. 11 Rabearisoa, \ Jiang and Z. Dai. 2013. Bioaccumulation of perfluoroalkyl
substances by Daphnia magna in water with different types and concentrations of protein.
Environ Sci Technol 47(N): 10955-10963.

Xiao, F. 2017. Emerging poly- and perfluoroalkyl substances in the aquatic environment: A
review of current literature. Water Res. 124: 482-495.

Xu, D., X. Chen and B. Shao. 2017. Oxidative Damage and Cytotoxicity of Perfluorooctane
Sulfonate on Chlorella vulgaris.Bull. Environ. Contam. Toxicol., 98: 127-132.

Xu, J., C. S. Guo, Y. Zhang and W. Meng. 2014. Bioaccumulation and trophic transfer of
perfluorinated compounds in a eutrophic freshwater food web. Environ Pollut. 184: 254-261.

193


-------
Yamashita, N., S. Taniyasu, G. Petrick, S. Wei, T. Gamo, P. K. Lam and K. Kannan. 2008.
Perfluorinated acids as novel chemical tracers of global circulation of ocean waters.
Chemosphere. 70(7): 1247-1255.

Yang, H. B., Z. Ya-Zhou, Y. Tang, G. Hui-Qin, F. Guo, S. Wei-Hua, L. Shu-Shen, H. Tan and F.
Chen. 2019. Antioxidant Defence System is Responsible for the Toxicological Interactions of
Mixtures: A Case Study on PFOS and PFOA in Daphnia magna. Sci. Total Environ. 667: 435-
443 (Supplemental Journal Materials).

Yang, S., F. Xu, F. Wu, S. Wang and B. Zheng. 2014. Developmenl of PFOS and PFOA Criteria
for the Protection of Freshwater Aquatic Life in China. Sci. Total l-n\iron. 470/471: 677-683.

Ye, X., M. J. Strynar, S. F. Nakayama, J. Varns, L. Helfant, J. Lazorchak and A. D. Lindstrom.
2008. Perfluorinated compounds in whole fish homogenates from the Ohio. Missouri, and Upper
Mississippi Rivers, USA. Environ. Pollut. 156: 1227-1232.

You, C., C. Jia and G. Pan. 2010. Effect of salinity and sediment characteristics on the sorption
and desorption of perfluorooctane sulfonate at sedimenl-walci' interface. Environ Pollut. 158(5):
1343-1347.

Young, C. J. and S. A. Mabury. 2010. Atmospheric periluorinaled acid precursors: chemistry,
occurrence, and impacts. Rev Environ Contain Toxicol. 208: 1 -1 Ul->

Yuan, Z., J. Zhang, W. Meng and Y. Zhou. 2014. Effects of IV-i lluorooctane Sulfonate on
Behavioural Activity, Regeneration and Antioxidant Enzymes in Planarian Dugesia japonica.
Chem. Ecol. 30: 187-195.

Zareitalabad, P., J. Siemens. M I lamer and W Amelung. 2013. Perfluorooctanoic acid (PFOA)
and perfluorooctanesulfonic acid (PI OS) in surface waters, sediments, soils and wastewater - A
review on concentrations and distribution coefficients. Chemosphere. 91(6): 725-732.

Zhang. F . J. Wei, 0 li- R Jiang. \ Yu, J. Qin and L. Chen. 2015. Effects of Perfluorooctane
Sulfonate on the Immune Responses and Expression of Immune-Related Genes in Chinese
Mitten-Handed Crab Eriocheir sinensis.Comp. Biochem. Physiol. C Toxicol. Pharmacol.,
172/173: 13-IS

Zhang, L., J. Niu. Y. I ,i. Y Wang and D. Sun. 2013. Evaluating the Sub-Lethal Toxicity of
PFOS and PFOA Using Rotifer Brachionus calyciflorus. Environ. Pollut. 180: 34-40
(Supplemental Journal Materials).

Zhang, L., J. Niu, Y. Wang, J. Shi and Q. Huang. 2014. Chronic Effects of PFOA and PFOS on
Sexual Reproduction of Freshwater Rotifer Brachionus calyciflorus.Chemosphere. 114: 114-120
(Supplemental Journal Materials).

Zhang, S., L. Wang, Z. Wang, D. Fan, L. Shi and J. Liu. 2017. Derivation of freshwater water
quality criteria for dibutyltin dilaurate from measured data and data predicted using interspecies
correlation estimate models. Chemosphere. 171: 142-148.

194


-------
Zhang, X., R. Lohmann, C. Dassuncao, X. C. Hu, A. K. Weber, C. D. Vecitis and E. M.
Sunderland. 2016. Source attribution of poly- and perfluoroalkyl substances (PFASs) in surface
waters from Rhode Island and the New York Metropolitan Area. Environ Sci Technol Lett. 3(9):
316-321.

Zheng, X. M., H. L. Liu, W. Shi, S. Wei, J. P. Giesy and H. X. Yu. 2012. Effects of
Perfluorinated Compounds on Development of Zebrafish Embryos. Environ. Sci. Pollut. Res. 19:
2498-2505.

Zhu, Y., D. Yang, X. Duan, Y. Zhang, D. Chen, Z. Gong and C. Lin 2021. Perfluorooctane
sulfonate promotes doxycycline-induced liver tumor progression in male Kras(vl2) transgenic
zebrafish. Environ Res. 196: 110962.

195


-------
Appendix A Acceptable Freshwater Acute PFOS Toxicity Studies

A.l Summary Table of Acceptable Quantitative Freshwater Acute PFOS Toxicity Studies

Species (

Method''

Tesl
Dnnilion

( hemiciil /
I'uriU

pll

Temp.

<ฐC)

r.iTcci

AiiIIkii'
Keporied

l.riWi

(ttne.
(iiiji/l-l

l-'.PA
( iik'iiliiled

r.lTecl
(one.
(inu/l.)

I'iiiid
r.lTeel
(one.

(msi/l.)'

Speeies
Mesin
Aeule

\ illllC

Reference

l'l:ni:iri:i (<> K> am.

Dugesia japonica

S. 1

<>(. hr

H< )S-K



25

l.( 5(1

r

-

r

-

l.i (2()UX)

Planaria (0.9 ฑ0.1 cm),
Dugesia japonica

s,u

96 hr

PFOS-K

>98%

-

25

I.C50

23

22 (.S

22.68

-

(Li 2009)

Planaria (10-12 mm),
Dugesia japonica

R,U

96 hr

PFOS-K

>99%

-

20

LC5U

29.46

-

29.46

22.48

(Yuan et al. 2014)



Eastern elliptio (76.5 g,
48.7 mm).

Elliptio complanata
(formerly, Unio
complamatus)

R,M

96 hr

PFOS-K
90.49%

7.9-8.5

:i x-
"

l.( 50

59

64.35

64.35

64.35

Drottar and
Krueger (2000e)



Fatmucket (glochidia,
<24 hr),

Lampsilis siliquoidea

S,M

24 hr

PI ()S

S 4<>

20

EC50

(viability)

16.5

-

16.5

-

Hazelton (2013);
Hazelton et al.
(2012)

Fatmucket (juvenile, 4-
6 wks),

Lampsilis siliquoidea

R, M

96 hr

PI ()S

S 4(>

2d

I.C50

158.1

-

158.ld

16.5

(Hazelton 2013;
Hazelton et al.
2012)



Black sandshell
(glochidia, <24 hr),

Ligumia recta

S,M

24 hr

Pl< )S

>98%

8 4(>

20

EC50

(viability)

13.5

-

13.5

-

(Hazelton 2013;
Hazelton et al.
2012)

Black sandshell
(juvenile, 4-6 wk),
Ligumia recta

R, M

hr

PFOS

>98ฐ/.,

8.46

20

LC50

141.7

-

141.7d

13.5

Hazelton (2013),
Hazelton et al.
(2012)



Bladder snail (mixed
age),

Physella acuta
(formerly, Physa acuta)

S,U

96 hr

PFOS-k

>98%

-

25

LC50

178

183.0

183.0

183.0

(Li 2009)



A-l


-------
Species (lilVst;iiio>

Method''

Tesl

Dui'iilioii

( hemiciil /
PuriU

pll

Temp.

(ฐC)

r.iTcci

Aullior
Rcpnrk'd
r.lTecl
(one.
(111^/1.)

i:p\

( iileuliiled
r.lTeel
Cone.

I'iiiid
i:iTee(
(one.
(inu/l.)'

Speeies
Mesin
Aeule
\ idne

Reference

Snail (adult, 4 mo.),

Physella heterostropha
pomilia

(formerly, Physa
pomilia)

S,M

96 hr

PFOS-K

>98%



25

LC50

I(.| ""

-

161.8

161.8

(Funkhouser
2014)



Rotifer (<2 hr old
neonates),

Brachionus calyciflorus

s,ub

24 hr

PFOS-K

>98%

-

20

I.C50

61.8

-

61.8

61.8

(Zhang et al.
2013)



Cladoceran (6-12 hr),

Daphnia carinata

s,u

48 hr

PFOS-K

>98%

-

21

LC5n

8.8

11.56

11.56

11.56

Logeshwaran et
al. (2021)



Cladoceran (<24 hr),
Daphnia magna

S,M

48 hr

PFOS-K
90.49%

8.2-8.6

19.3-
20.2

IX'5(1

(.1

58.51

58.51

-

Drottar and
Krueger (2000b)

Cladoceran (<24 hr),
Daphnia magna

S,U

48 hr

PFOS-K
<>5".,

-

21

EC 50

(immobility)

67.2

-

67.2

-

(Boudreau 2002;
Boudreau et al.
2003a)

Cladoceran (<24 hr),
Daphnia magna

S,U

48 hr

PI (>S
Univpuried

-

21

EC50

(immobility)

37.36

35.46

35.46

-

(Ji et al. 2008)

Cladoceran (<24 hr),
Daphnia magna

S,U

48 hr

PF ()S-K
> W„

" x:-
"i>i

25

nc50

63

55.40

55.40

-

(Li 2009)

Cladoceran (<24 hr),
Daphnia magna

s,u

48 lvr

PT ()S-K
i

25

i:c50

63

72.70

72.70

-

(Li 2009)

Cladoceran (<24 hr),
Daphnia magna

s,u

4S hr

PI ()S-K
> w.,

".82-
7.91

25

EC50

63

64.60

64.60

-

(Li 2009)

Cladoceran (<24 hr),
Daphnia magna

S,M

4S hr

PFOS-K

99%

7

22

LC50

78.09

-

78.09

-

(Yang et al. 2014)

Cladoceran (<24 hr),
Daphnia magna

S,U

48 hr

PFOS

98%

"2

20

EC50

(death/immobility)

23.41

-

23.41

-

{Lu, 2015 #325}

Cladoceran (<24 hr),
Daphnia magna

s,u

48 lir

H< )S-K
ฆฆ'J8%

7

20

EC50

(death/immobility)

79.35

94.58

94.58

-

(Liang et al. 2017)

Cladoceran (12-24 hr),

Daphnia magna

s,u

48 hr

PFOS-K

98%

-

-

LC50

22.77

22.43

22.43

-

Yang et al. (2019)

Cladoceran (0-24 hr),

Daphnia magna

s,u

48 hr

PFOS-K
Unknown

7.6

22

EC50

27

-

27

48.87

3M Company
2000



A-2


-------
Species (lilVst;iiio>

Method''

Tesl
Dui'iilioii

( hcmic;il /
Piiriu

pll

Temp.

(ฐC)

r.lTeel

Author
Reported
1". fleet
(one.
(mji/l -'

i:p\

( iileuliiled
r.lTeel
Cone.
(ni"/l-l

liiiiil

i:iTcc(
(one.
diiji/l

Species
Mesin
Aciilc
\ iilne
(mii/l.)

Reference

Cladoceran (<24 hr),
Daphnia pulicaria

S,U

48 hr

PFOS-K

95%



21

EC50

(immobility)

1 U

-

134

134

(Lloudrcau 2002,
Boudreau et al.
2003a)



Cladoceran (<24 hr),
Moina macrocopa

s,u

48 hr

PFOS
Unreported

-

25

EC50

(immobility)

17.95

I" :n

17.20

17.20

(Ji et al. 2008)



Crayfish (juvenile, 2
wks, 0.041 g),
Procambarus fallax f.
virginalis

S,M

96 hr

PFOS-K

>98%

-

25

1.(

59.87

-

59.87

59.87

(Funkhouser
2014)



Japanese swamp
shrimp,

Neocaridina
denticulata

S,U

96 hr

PFOS-K

>98%

-

25

l.( 50

ID

12.91

12.91

-

(Li 2009)

Japanese swamp
shrimp,

Neocaridina
denticulata

S,U

96 hr

PF( )S-K

-

25

LC50

log

28.55

28.55

-

(Li 2009)

Japanese swamp
shrimp,

Neocaridina
denticulata

S,U

96 hr

PF( )S-K
-<)8"<,

-

25

I.C50

108

10.32

10.32

15.61

(Li 2009)



Rainbow trout
(juvenile),

Oncorhynchus mykiss

S,M

96 hr

Pi ( )S-k
8(.li"„

-

1 1 -
i:')

LC50

22

22.59

22.59

-

Palmer et al.
(2002a)

Rainbow trout (parr),

Oncorhynchus mykiss

R, M

l><> hr

PFOS-K

98%

-

10

LC50

2.5

-

2.5

7.515

Sharpe et al.
(2010)



Zebrafish (embryo),

Danio rerio

S,U

96 hr

I'POS-K

j.97%

7.2-7.5

26

LC50

58.47

-

58.47

-

Hagenaars et al.
(2010, 2011)

Zebrafish (adult),

Danio rerio

R, M

96 hr

PFOS-K

98%

-

26

LC50

22.2

-

22.2

-

Sharpe et al.
(2010)

A-3


-------
Species (lilVst;iiio>

Method''

Tesl

Dui'iilioii

( hemiciil /
PuriU

pll

Temp.

(ฐC)

r.iTcci

Author
Reported
r.lTecl
(one.
(inii/l.)

i:p\

( iileuliiled
r.lTeel
Cone.

I'iiiid
Klleel
(one.

(mii/l.)'

Speeies
Mesin
Aeule
\ idne

Reference

/.chialisli (^ mo . 2 2

cm),

Danio rerio

R, U

96 hr

PFOS-K
Unknown



23

LC50

To

-

17.0

-

Wang et al.
(2013a)

Zebrafish (embryo),

Danio rerio

S,U

96 hr

PFOS-K

98%

8.3

28.5

LC50

08

"i i:

71.12

-

(Li et al. 2015)

Zebrafish (embryo),

Danio rerio

S,U

96 hr

PFOS-K

98%

-

28

l.( 50

3.502

-

3.502



Du et al. (2016);
Du et al. (2017)

Zebrafish (embryo, 1
hpf),

Danio rerio

R,U

96 hr

PFOS
Unreported

-

26

I.C5U

34.2

s:

38.82

24.44

(Stengel et al.
2017)



Fathead minnow
(juvenile),

Pimephales promelas

S,M

96 hr

PFOS-K
90.49%

8.2-8.5

::

l.( 50

K> 5

9.020

9.020

-

Drottar and
Krueger (2000d)

Fathead minnow (79 d),

Pimephales promelas

S,U

96 hr

PFOS-Li
24.5%

8.0-8.4

i<> :-
l<> 5

l.( 50

4 (>55'

5.356f

5.356

6.950

3M Company
2000



American toad (larva,
Gosner stage 26),
Anaxyrus americanus

S,U

96 hr

PI OS

1 IlkllOW II

-

21

LC50

62g

63.41

63.41d

-

Tornabene et al.
(2021)

American toad (larva,
Gosner stage 41),

Anaxyrus americanus

S,U

96 hr

PI ()S
Unknow n

-

:i

I.C50

62g

56.49

56.49

56.49

Tornabene et al.
(2021)



Gray treefrog (larva,
Gosner stage 26),
Hyla versicolor

S. TT

<>(. III'

\>\( )S
U nknow ii

-

:i

LC50

79

78.33

78.33d

-

Tornabene et al.
(2021)

Gray treefrog (larva,
Gosner stage 40),
Hyla versicolor

S,U

l><> hr

PFOS
Unknown

-

21

LC50

24

19.88

19.88

19.88

Tornabene et al.
(2021)



American bullfrog
(tadpole, Gosner stage
25),

Lithobates catesbeiana
(formerly, Rana
catesbeiana)

S,U

96 hr

PFOS
Unknown

-

21

LC50

144

154.8

154.8d

-

Flynnetal. (2019)

A-4


-------
Species (lilVst;iiio>

Method''

Tesl

Dui'iilioii

( hemiciil /
PuriU

pll

Temp.

(ฐC)

r.iTcci

Author
Keporled
r.lTccl
(one.
(111^/1.)

i:p\

( iilciiliiled

r.lTecl
Cone.

I'iiiid
Klleel
(one.
(inu/l.)'

Speeies
Mesin
Aeule
\ idne

Referenee

American bulllrog
(larva, Gosner stage
26),

Lithobates catesbeiana

S,U

96 hr

PFOS
Unknown



21

LC50

I(.'

1^3.3

133.3

133.3

Tornabene et al.
(2021)



Green frog (larva,
Gosner stage 26),
Lithobates clamitans
(formerly, Rana
clamitans)

s,u

96 hr

PFOS
Unknown

-

21

I.C50

113

-

113

113

Tornabene et al.
(2021)



Northern leopard frog
(larva, Gosner stage
26),

Lithobates pipiens
(formerly, Rana
pipiens)

s,u

96 hr

PFOS
Unknown

-

:i

I.C5<>

73

72.72

72.72

72.72

Tornabene et al.
(2021)



Wood frog (larva,
Gosner stage 26),
Lithobates sylvatica
(formerly, Rana
sylvatica)

s,u

96 hr

I'l-'OS
Unknow ii

-

:i

LC50

130

-

130

130

Tornabene et al.
(2021)



African clawed frog
(embryos),

Xenopus laevis

R,M

l><> hr

\>\( )S-K
8o.9" „

7.3

24

LC50

13.8

15.53

15.53

-

(Palmer and
Krueger 2001)

African clawed frog
(embryos),

Xenopus laevis

R,M

l><> hr

PFOS-k
86.9'}. i

7.27

24

LC50

17.6

18.04

18.04

-

(Palmer and
Krueger 2001)

African clawed frog
(embryos),

Xenopus laevis

R, M

96 lir

I'POS-K

86.9%

7.26

24

LC50

15.3

14.6

14.60

15.99

(Palmer and
Krueger 2001)



A-5


-------
Species (

Method'

Tesl

Dui'iilioii

( hcmic;il /
Piiriu

|)ll

Temp.

(ฐC)

r.iTcci

Author
Reported
r.lTeel
(one.

i:p\

( iileuliiled
r.lTeel
Cone.

liiiiil
I-! I'l'eel
(one.

Speeies
Mesin
Aeule
\ iilne

Reference

Jefferson salamander
(larva, Harrison stage
40),

Ambystoma
jeffersonianum

S,U

96 hr

PFOS
Unknown



21

LC50

(4

51.71

51.71

51.71

Tornabene et al.
(2021)



Small-mouthed
salamander
(larva, Harrison stage
40),

Ambystoma texanum

s,u

96 hr

PFOS
Unknown

-

21

l.( 5(i

41-

4(. "1

46.71d

-

Tornabene et al.
(2021)

Small-mouthed
salamander
(larva, Harrison stage
46),

Ambystoma texanum

s,u

96 hr

PFOS
Unknown

-

21

LC50

41

30.00

30.00

30.00

Tornabene et al.
(2021)



Eastern tiger
salamander
(larva, Harrison stage
40),

Ambystoma tigrinum

s,u

96 hr

I'R )S

1 IlkllOW 11

-

21

LC50

73

68.63

68.63

68.63

Tornabene et al.
(2021)

a S=static, R=renewal, F=flow-through, U=unmeasured. M=nic:isured. T=total. D=dissolved. Diet=diclary, MT=maternal transfer

b Chemical concentrations made in a side-test represeiil;il in e of exposure and verified stability of concentrations of PFOS in the range of concentrations tested under similar

conditions. Daily renewal of test solutions.

0 Reported in moles converted to milligram based on a molecular weight of 500.13 mg/mmol.
dNot used in SMAV calculation; only the most sensitive life-stage used.
e Values in bold used the in the SMAV calculal ion

f Author-reported LC50 of 19 mg/L x 24.5% PI-'( )S = 4 655 mg/L I'R )S; EPA-calculated LC\(J of 21.86 mg/L x 24.5% PFOS = 5.356 mg/L PFOS.
g Author pooled tests or lifestages.

A-6


-------
A.2 Detailed PFOS Acute Freshwater Toxicity Study Summaries and

Corresponding Concentration-Response Curves (when calculated for
the most sensitive genera)

The purpose of this section was to present detailed study summaries for tests that were
considered quantitatively acceptable for criteria derivation, with summaries grouped and ordered
by genus sensitivity. C-R models developed by EPA that were used to determine acute toxicity
values used for criterion derivation are also presented for the most scnsili\ e genera when
available. C-R models included here with study summaries were those lor the live most sensitive
genera (consistent with Section 3.1.1.1). When required, EPA also included models for non-
resident species that were more sensitive than the fourth most sensitive North American resident
genus. In many cases, authors did not report concentration-response data in the
publication/supplemental materials and/or did not provide concentration-response data upon
EPA request. In such cases, EPA did not independently calculate a toxicity value and the author
reported effect concentrations were used in the derivation of the criterion.

A.2.1 Most Sensiti\ e I'rcshwalcr (ienus for Acute Toxicity: Pimephales (fathead minnow)

3MCom|)sinv (2000) |iro\ ides the results of a 96-hour static, unmeasured acute toxicity

test with the fathead mi nnou. I'micphalcs promelas, and PFOS-Li (perfluorooctanesulfonate

lithium salt. C AS # 29457-72-5). A stock solution was made with carbon-filtered well water at a

test sample concentration of 4<)Q mg/L and where the test sample was reported as a mixture of

PFOS-Li (24,5ฐ/..) in water (75.5%). Fish were obtained from a commercial supplier (Aquatic

Biosystems, Fort Collins, CO) and were 79 days old at test initiation with an average length of

2.1 cm and weight of 0.069 g. Exposure vessels were 2 L glass beakers containing 1 L of

solution and 10 fish per beaker (0.69 g fish/L). Each test treatment was replicated twice with

nominal test concentrations (control, 3.2, 5.6, 10.0, 18.0, 32.0 and 56.0 mg/L test sample).

Throughout the experiment the D.O. ranged from 4.8 - 7.9 mg/L, pH 8.0 - 8.4 and a test

A-7


-------
temperature of 19.2 - 19.5ฐC. The low D.O. of 4.8 mg/L was only observed in one replicate of
the highest test concentration at 96 hours; D.O. was >6.0 mg/L for all other treatments and
replicates. No mortality occurred in the control treatment and 100% was observed in the highest
treatment (56 mg/L). The study authors reported that the test sample containing 24.5% PFOS-Li
exhibited a 96-hour LCso of 19 mg/L, which equates to 4.655 mg/L as PFOS. The independently-
calculated 96-hr LCso value was 21.86 (17.63 - 26.08) mg/L, u liich equates to 5.356 mg/L as
PFOS and is acceptable for quantitative use in the derivation of the acute fresh water criterion for
PFOS.

Drottar and Krueger (2000d) evaluated the acute effects of PFOS-K (CAS 2795-39-3,
Lot # 217 (T-6295) obtained from the 3M Company. l)<) 4l)'\> purity, stored at ambient room
temperature) on juvenile fathead minnows (1'imcphalcs promdas) during a 96-hour measured,
static study. Researchers stated they followed protocols U.S. EPA Scries 850 (OPPTS 850.1075),
OECD Guideline 203. and ASTM E729-88a. A primary stock solution was prepared at 27 mg/L
and mixed with an electric mixer lor 22 hours prior to use in testing to ensure solubilization of
the test substance After mixing, the primary stock solution was proportionally diluted with
dilution water to prepare the four additional test concentrations. Test fish were obtained from
cultures at Wildlife International J .id in Easton, Maryland. The minnows were held for
approximately 12o days prior 10 testing and were acclimated to test conditions for 48 hours prior
to test initiation. l-'ish were fed a commercially-prepared diet prior to the 48-hour acclimation
period. All fish used in the test were from the same source and year class, and the total length of
the longest fish was no more than twice the length of the shortest. Fathead minnows were
randomly distributed among mean measured test concentrations of 0 (control), 3.3, 5.6, 9.5, 17
and 28 mg/L, with 10 fish per 25-L polyethylene aquarium provided in duplicate. Aquaria were

A-8


-------
filled with 15 L of test solution with an observed dissolved oxygen of 7.7 - 8.4 mg/L,
temperature of 22 ฑ 2ฐC, pH of 8.2 - 8.5 and a total hardness of 131 mg/L as CaCCb. Fathead
minnows were subjected to a 16-hr:8-hr photoperiod at 391 lux. Sand and 0.45 um filtered well
water from a 40 m deep well on site served as both the culture water and the testing media. The
authors reported an LCso of 9.5 mg/L PFOS. EPA's independently-calculated 96-hour LCso was
9.0196 (7.146 - 10.89) mg/L and was used quantitatively to derive the draft acute water column
criterion for freshwater.

A.2.1.1 3M Company (2000) Concentration Response Carve - Pimephales (fathead minnow)

Publication: 3M Company (2000)

Species: Fathead minnow, Pimephalespromelas

Genus: Pimephales

EPA-Calculated LCso: 21.86 (95% C.I. 17.63 - 26.08) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

1.8756

0.3016

6.2191

4.999 e"10

e

26.5720

2.5673

10.3501

< 2.2 e"16

Concentration-Response Model Fit:

Pimephales promelas
Weibull type 1, 2 para

PFOS (mg/L )

A-9


-------
A.2.1.2 Drottar andKrueger (2000d) Concentration Response Curve - Pimephales (fathead
minnow)

Publication: Drottar and Krueger (2000d)

Species: Fathead minnow, Pimephalespromelas
Genus: Pimephales

EPA-Calculated LC50: 9.012 (95% C.I. 7.146315 - 10.892956) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

2.9683

0.4433

6.6965

2.134 e"11

d

1.0503

0.0174

60.4403

< 2.2 e"16

e

9.0196

0.8923

10.1084

< 2.2 e"16

Concentration-Response Model Fit:

Drottar and Krueger 2000

Pimephales promelas
Log Logistic type 1, 3 para

PFOS (mg/L )

A.2.2 Second Sensitive Freshwater Genus for Acute Toxicity: Oncorhynchus (trout)

(Sharpe et al. 2010) evaluated the acute effects of perfluorooctane sulfonate (PFOS,

potassium salt, CAS #2795-39-3, 98% purity) to Oncorhynchus mykiss, rainbow trout, via a 96-

hour renewal measured exposure (renewal not stated in paper, but assumed based on other

information provided, including the test Guideline protocol). Limited details about the test

A-10


-------
protocol were provided in the publication, but the authors noted they followed OECD Guideline
203, and did not identify any deviations from these test guidelines. Trout eggs were obtained
from Raven Trout Hatchery, transported immediately postfertilization to the University of
Alberta aquatics facility, and kept in dechlorinated City of Edmonton water. The eggs were
reared until hatching in Heath trays in a recirculating, temperature-controlled system at 10ฐC
with a 12-hr: 12-hr, light:dark photoperiod (the same conditions arc assumed for the toxicity test).
The rainbow trout used in the study were parr (2-3 g, the fourth stage of llie salmon life cycle) at
test initiation. The dilution water was dechlorinated City of Edmonton water, and dissolved
oxygen content and temperature were monitored daily, but physico-chemical results were not
reported. PFOS was dissolved in MeOH, and all vehicle controls received a volume of MeOH
equal to that present in the highest PFOS dose of that experiment (final MeOH content 0.2%
v/v). The concentration of PFOS in any experiment was always well below its reported solubility
in water (-500 mg I.) Trout toxicity tests were performed using food-grade 2 L plastic tanks
with four fish per tank, and two tanks per dose N\\ obtained clarification from the study
authors regarding the experimental set-up pertaining to the biomass loading rate, which was 1 to
1.5 g'l. (based on lour lisli weighing a total of 2 to 3 g per 2 L tank (personal communication
with (ireg (ioss and Rainie Sliarpe. March 2021). This biomass loading rate was nearly two-fold
higher than that stated in 01XI) Guidelines of 0.8 g/L (OECD 1992). The trout were randomly
assigned to doses defined as control (0 mg/L PFOS); vehicle control (0 mg/L PFOS, 0.2%

MeOH v/v); and 0.78, 1.56, 3.12, 6.25, and 12.5 mg/L PFOS. Authors indicated that measured
PFOS concentrations averaged 88% of nominal, but did not indicate whether LCsos were based
on measured or nominal concentrations. Given the clarifications regarding the biomass loading,
this study was considered for quantitative use in the derivation of the acute PFOS freshwater

A-ll


-------
criterion. The author-reported 96-hour LCso for the study of 2.5 mg/L (authors did not specify if
this concentration was nominal or measured) was acceptable for quantitative use and was among
other toxicity values used for this species to calculate the SMAV/GMAV (see further details at
the end of the study summaries in this section) that was utilized to derive the draft acute water
column criterion.

In addition to the two additional analyses presented in Section 4 2 1 above, a third
additional analysis was conducted to determine the influence of the I \ aluc from Sharpe et al.
2010. In this calculation, the toxicity data for aquatic insects (consisting of data lor yellow fever
mosquito (Aedes aegypti) and a chironomid (Chimnomiis phimostis) were not used and the LCso
for rainbow trout (Oncorhynchus mykiss) from Sharpe et al. (2< U 0) was also removed given that
the biomass loading rate (see Section 3.1 I 12 and A 2 2) was higher than that stated in OECD
Guidelines of 0.8 g/L (OECD 1992). In this scenario the alternate SMAV for O. mykiss (22.59
mg/L) would not rank among the lour most acutely sensiti\ e genera. The third analyses yielded a
freshwater FAY for PI-OS of 7 42') mg I.. again calculated following the procedures described in
the 1985 Guidelines (I S N\\ llM5) The l:AV was then divided by two (see Section 3.2.1.1) to
calculate an acute water column concentration, of 3.7 mg/L PFOS, which was 1.5 times higher
than the reported I.Cso from the Sharpe et al. (2010) rainbow trout study. EPA retained the
rainbow trout \ alue for the purposes of the current derivation, to ensure protection of sensitive
salmonid species as a group, which includes commercially and recreationally important species,
as well as endangered species. The availability of toxicity data for these taxa developed using
more standardized and better documented procedures would reduce the uncertainty in the
analysis. Additionally, this third analysis lacked the MDRs for aquatic insects.

A-12


-------
(Palmer et al. 2002a) evaluated the acute effects of PFOS (PFOS, potassium salt,
identified as FC-95 obtained from 3M Company) to Oncorhynchus mykiss, rainbow trout, via a
96-hour static exposure with measured concentrations. The test organisms were obtained from
Thomas Fish Company in Anderson, California and were reported as juveniles with a mean
weight of 0.34 g and total length of 3.6 cm. All test organisms were from the same source and
year class, and the length of the longest fish was no more than i\\ ice llie length of the shortest.
The fish were held for approximately five weeks prior to the initiation of the lest. This
acclimation was done in water from the same source and at the same temperature as the test.
During the acclimatation period, no mortalities or signs of disease were observed Test
organisms were only fed a commercially-prepared diet (reported from Zeigler Brothers Inc.)
during a 14-day holding period after which point lish were no longer fed through the acclimation
period (at least 48 hours prior to the test) or din ing the test The test water was obtained from a
well located near the testing facility and was characterized as moderately-hard water. The target
test temperature was 12 I (' and a I (->-hr8-hr, light:dark photoperiod were maintained through
the holding, acclimatation. and testing periods Dissolved oxygen and pH measurements were
made on water samples collected at test initiation followed by 24-hour intervals for each
replicate test chamber of each treatment and control. Test chambers were 25-L polyethylene
aquaria containing I 5 T. of test solution. At the initiation of the test, rainbow trout were
indiscriminately mo\ ed from the acclimation tank and distributed two at a time to the test
chambers until each contained ten fish. The resulting biomass loading rate was 0.23 g fish/L of
test water. A 40-L stock solution was prepared in dilution water at a concentration of 150 mg
PFOS/L. Nominal concentrations were 3.1, 6.3, 13, 25, and 50 mg/L. Two replicates of each test
solution were prepared at nominal concentrations by adding the appropriate volume of stock

A-13


-------
solution to dilution water in the test aquaria to achieve the final volume of 15 L. Measured test
concentrations at the end of the test ranged from 97 to 100% of nominal with concentrations of
3.0, 6.3, 13, 25, and 50 mg/L. Results from this study were based on measured concentrations.
Mortality and other signs of toxicity were observed daily. Trout in the control group appeared
normal and healthy throughout the test period. Additionally, test organisms in the lowest
treatment groups (3.0 and 6.3 mg/L) appeared healthy with no mortalities or other signs of
toxicity. After 96-hours of exposure, mortality in the 13, 25, and 50 mu I. treatment groups was
20, 50, and 100%, respectively. The author-reported 96-hour LCso for the study was 22 mg/L.
This study was considered acceptable for quantitative use in the derivation of the acute PFOS
freshwater criterion. The independently-calculated 96-hour LCso value was 22.59 mg/L. This
independently-calculated LCso value of 22 5l> (14 53 30.65) mu I. was used quantitatively and
was among other toxicity values used for this species to calculate the SMAV/GMAV (see further
details at the end of the study summaries in this section) that was utilized to derive the draft acute
water column criterion

A.2.2.1 Shar/Kฆ el al. (20/0) ('oiiceniraiion Response Curve - Oncorhynchus (trout)

Publication Sliarpe et al (2<)|<))

Species. Rainbow trout, (hicorhynchiis mykiss

(ien iis: Oncorhynchus

KIW-Csilculated I.On Not calculable, concentration-response data not available

A-14


-------
A.2.2.2 Palmer el al. (2002) Concentration Response Curve - Oncorhynchus (trout)

Publication: Palmer et al. (2002a)

Species: Rainbow trout, Oncorhynchus mykiss
Genus: Oncorhynchus

EPA-Calculated LCso: 22.59 (95% C.I. 14.53 - 30.65) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

2.3775

0.5634

4.2189

2.455 e5

d

1.0339

0.0140

73.6910

< 2.2 e"16

e

26.3557

5.7449

4.5877

4.482 e6

Concentration-Response Model Fit:

Palmer et al. 2002

Oncorhynchus mykiss
Weibulltype 1, 3 para

PFOS(mgL)

A.2.3 Third Most Sensitive Freshwater Genus for Acute Toxicity: Lisumia (mussel)

Hazelton (2013); Hazelton et al. (2012) evaluated the acute effects of PFOS (acid form,

> 98% purity) on two freshwater mussels: Ligumia recta and Lampsilis siliquoidea. The tests

yielded the 3rd and 6th most sensitive genera values (respectively) in the PFOS freshwater acute

toxicity database (The L. siliquoidea results are reported below). Acute toxicity was observed

under static conditions over a 24-hour period (< 24-hour old glochidia) or a 96-hour period (4-6-

A-15


-------
week-old juveniles). The tests followed the ASTM E2455-06 (2006) test method. Dilution water
was hard reconstituted water (total hardness typically 160-180 mg/L as CaCCte). Photoperiod and
light intensity were not reported. No details were provided regarding primary stock solution and
test solution preparation. Experiments were conducted in 3.8 L glass jars of unspecified fill
volume. The test employed three replicates of 150 glochidia or seven juvenile mussels each in
six measured test concentrations plus a negative control (10 ju\ cnilcs lor the control treatment).
Nominal concentrations were 0 (negative control), 0.005, 0.05, 0.5, 5, 5<). and 500 mg/L;
respective measured concentrations were < LOQ (specifics not provided), U.oo54. 0.0514, 0.456,
4.68, 47.2, and 490 mg/L. Recovery of PFOS standards ranged from 85.3-123" n o\ or all
experiments. For all acute tests, alkalinity ranged from 97 to 1 10 mg CaCCte/L with a mean of
104.4 mg CaCCte/L; total hardness ranged from 132 to 162 mu CaCO ;'L with a mean of 149.6
mg CaCCte/L; conductivity ranged from 5 14 lo M3 uS cm with a mean of 556.5 |is/cm; pH
ranged from 8.05 lo S 5o with a mean of 8.46; and dissoK ed oxygen ranged from 8.16 to 9.46
mg/L with a mean of X (->2 mu I. (n I 2 for alkalinity and total hardness, n = 55 for all other
parameters) Exposures were conducted in en\ ironmental chambers set at a temperature of 20ฐC
(glochidia tests), or in dilution water maintained at20ฐC (juvenile tests). Survival of mussels in
the neuati\ e control was > Wo in all exposures. The 24-hour ECso reported by the study authors
for glochidia of/, recta was 13.5 mg/L (C.I. 5.7-31.8). The 96-hour LCso reported by the study
authors for juvenile /. recta was 141.7 mg/L (C.I. 80.4-249.6). The 24-hour ECso forZ. recta
glochidia represented an acute value acceptable for quantitative use. The juvenile life stage was
less sensitive, such that its LCsos are not used quantitatively in species mean acute values
(SMAVs). The independently-calculated toxicity value could not be calculated at this time given

A-16


-------
the lack of data presented in the paper. The study author reported values are currently used
quantitatively to derive the draft acute water column criterion.

A.2.3.1 Hazelton et al. (2012) Concentration Response Curve - Ligumia (mussel)

Publication: Hazelton et al. (2012)

Species: Black sandshell, Ligumia recta
Genus: Ligumia

EPA-Calculated LCso: Not calculable, concentration-response data not available

A.2.4 Fourth Most Sensitive Freshwater Genus for Acute Toxicity Seocaridina (shrimp)
Li (2009) conducted three independent repeats of a 96-hour static test on PFOS

(potassium salt, >98% purity) with the freshwater shrimp species, Neocaridma dcmiculata (a

non-North American species). Test organisms were obtained from an unspecified local supplier

and acclimated in the laboratory for at least seven days prior to the experiments. N. denticulata

of unspecified age were used at test initiation. Dilution water was dechlorinated tap water. The

photoperiod consisted of 12 hours of illumination al an unreported light intensity. A primary

stock solution was prepared in dilution water. Exposure \ essels were polypropylene beakers of

unreported dimensions and I I. lill \ olunie The test employed five replicates of six shrimp each

in at least fi\ e test concentrations (the lu st repeated experiment had one additional PFOS

treatment group at l<> nig I. compared to the other two experimental repeats) plus a negative

control l -ach treatment was tested three independent times. Nominal test concentrations were in

the range of 5-2< >< > mg/L PI ( )S The test temperature was maintained at 25ฑ2ฐC. Water quality

parameters including pi I. conductivity, and D.O. were reported as having been measured at the

beginning and end of each test, but the information was not reported. Survival of negative

control animals was 90%. The study reported 96-hour LCso was 10 mg/L (C.I. 9-12). The

toxicity test was acceptable for quantitative use. The independently-calculated LCso values for

the three independent experimental repeats were 12.91 (10.29 - 15.53), 28.55 (15.05 - 42.05),

A-17


-------
10.32 (7.788 - 12.85) mg/L, respectively. These independently-calculated LCso values were used
to calculate the GMAV value (as the geometric mean of the three LCso values previously
mentioned) of 15.61 mg/L and was used to derive the freshwater aquatic life criteria.

A.2.4.1 Li (2009) Concentration Response Carve - Neocaridina (shrimp)

Publication: Li (2009)

Species: Japanese Swamp Shrimp, Neocaridina denticulata
Genus: Neocaridina

EPA-Calculated LCsos: 12.91 (95% C.I. 10.29 - 15.53), 28.55 (95% C.I. 15.05 -42.05),
10.32 (95% C.I. 7.788 - 12.85) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

-1.5141

0.1920

-7.8879

3.091 e15

e

10.1360

1.0252

9.8865

< 2.2 e"16

Concentration-Response Model Fit: In order of LCsos listed immediately above

Li 2009

Neocaridina denticulate
WeibuD type 2,2 para

PFOS ( mg'L )

A-18


-------
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

1.0404

0.2369

4.3919

1.124 e"5

d

0.8880

0.0571

15.5493

< 2.2 e16

e

40.6105

7.5714

5.3637

8.156 e8

Li 2009

Neocaridina denticulate
Wefijull type 1, 3 para

PFOS (mg L )

Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

0.7749

0.1332

5.8195

5.903 e"9

e

16.5563

3.3654

4.9196

8.672 e7

Li 2009

Neocaridina denticulate
Weibull t\pe 1, 2 para

PFOS ( mg'L )

A-19


-------
A.2.5 Fifth Most Sensitive Freshwater Genus for Acute Toxicity: Xenopus (frog)

Palmer and Krueger (2001), associated with Wildlife International, conducted three

GLP renewal definitive assays with the potassium salt of PFOS (86.9% purity) using the frog

embryo teratogenesis assay - Xenopus (FETAX) with Xenopus laevis. A primary PFOS stock

solution was prepared in FETAX solution at a concentration of 48 mg/L, and subsequently

diluted with FETAX solution to prepare the six nominal test concentrations (1.82, 3.07, 5.19,

8.64, 14.4 and 24.0 mg PFOS/L). Eggs were obtained from luccding colonics ofX. laevis at the

University of Maryland Wye Research and Education Center. Adults were held in llow-through

polyethylene aquaria with 10 cm of dechlorinated tap water (23.5ฑ0.5ฐC) and a maximum of 10

adults/chamber and photoperiod of 16-hr:8-hr (light:dark) They were bred in the dark following

injection of human chorionic gonadotropin lo dorsal lymph sac of males and females. During

each assay, X laevis embryos (between IS I' stages S-l I) were exposed to PFOS for 96 hours.

Two replicate test chambers were maintained in each treatment group and four replicates were

maintained in each control group from the three separate assays. Each test chamber contained 25

embryos for a total of 50 embryos per treatment group and 100 embryos per control group. Tests

were conducted at 24 C, pH of 7.26-7.30, estimated total hardness of 75 mg/L as CaC03,

dissoK ed oxygen of 7.8-8.1 mg/L and a 12-hr: 12-hr light:dark photoperiod (60-85 foot candles).

PFOS concentrations were measured at the initiation and termination of all three assays. The

authors reported 96-hour I values for mortality of 13.8, 17.6 and 15.3 mg/L PFOS,

teratogenesis ECsos of 12 1, 17.6 and 16.8 mg/L PFOS, and minimum concentration to inhibit

growth values (effectively a LOEC) of >14.7, 7.97 and 8.26 mg/L for the same three tests,

respectively. Independently-calculated 96-hour LCso values for mortality were 15.53 (13.86 -

17.21), 18.04 (15.33 - 20.74), and 14.60 (12.65 - 16.55) mg/L for the three assays, respectively.

All data are deemed quantitative and the independently-calculated toxicity values were utilized

A-20


-------
to derive the PFOS aquatic life criteria. No additional quantitative, acute toxicity data were
available for this species. Therefore, these independently-calculated LCso values were used to
calculate the GMAV value (as the geometric mean of the three LCso values previously
mentioned) of 15.99 mg/L and was used to derive the freshwater aquatic life acute criterion.

A.2.5.1 Palmer andKrueger (2001) Concentration Response Carve -Xenopus (frog)

Publication: (Palmer and Krueger 2001)

Species: Frog, Xenopus laevis
Genus: Xenopus

EPA-Calculated LCso: 15.53 (95% C.I. 13.86- 17.21), 18.04 (95% C.I. 15.33 -20.74),
14.60 (95% C.I. 12.65 - 16.55) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

4.1306

1.0464

3.9475

7.897 e5

d

0.9633

0.0149

64.7242

< 2.2 e"16

e

16.9770

0.7991

21.2452

< 2.2 e"16

Concentration-Response Model Fit: In order ofLCsos listed immediately above

Palmer and Krueger 2001
Xenopus laevis
Weibull type 1, 3 para

PFOS ( mg'L )

A-21


-------
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

1.8800

0.3458

5.4366

5.431 e"8

d

0.9868

0.0127

77.7694

< 2.2 e16

e

21.9190

1.9259

11.3812

< 2.2 e"16

Palmer and Krueger 2001
Xenopus laevis
Weibull type 1; 3 para

0.1	1.0	10.0

PFOS ( mg'L )

Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

-1.9934

0.2667

-7.4757

7.681 e14

e

12.1461

0.7197

16.8763

< 2.2 e"16

Palmer and Krueger 2001
Xenopus laevis
Weibull type 2,2 para

PFOS ( mg'L )

A-22


-------
A.2.6 Sixth Most Sensitive Freshwater Genus for Acute Toxicity: Lampsilis (mussel)

Hazelton (2013); Hazelton et al. (2012) evaluated the acute effects of PFOS (acid form,

> 98% purity) on two freshwater mussels, as noted above: Lampsilis siliquoidea and Ligumia

recta. The L. siliquoidea studies yielded the 6th most sensitive genus value in the PFOS

freshwater acute toxicity database. Hazelton et al.'s experimental design and study conditions for

L. siliquoidea were reported above under the description of the third most sensitive taxa,

Ligumia. The 24-hour EC so reported by the study authors for ulochidia of /. siliquoidea was 16.5

mg/L (C.I. 8.0-33.9). The 96-hour LCso reported by the study authors lor ju\ enile L. siliquoidea

was 158.1 mg/L (C.I. not calculable). The 24-hour ]ฆ("*ฆฆ for L. siliquoidea glochidia represented

an acute value acceptable for quantitative use for the mussel species. Because the juvenile life

stage was less sensitive, only the glochidia I .("*ฆฆ was used to calculate the SMAV. The

independently-calculated toxicity values could not he calculated al this time given the data

presented in the paper. The study author reported \ alues were used quantitatively to derive the

draft acute water column criterion.

A.2.7 Seventh Most Sensitive Freshwater Genus for Acute Toxicity: Moina (cladoceran)

Ji ol al. (200S) performed a 48-hour static, unmeasured acute test of PFOS (acid form,

CAS I703-23-1, purity unreported) with Moina macrocopa. The test followed EPA's Methods

for measuring the acute toxicity of effluents and receiving waters to freshwater and marine

organisms (1 S N\\ (•>dO/4-^d'027F; U.S. EPA 2002). M macrocopa used for testing were

obtained from brood stock cultured at the Environmental Toxicology Laboratory at Seoul

National University (South Korea). Test organisms were less than 24 hours old at test initiation.

Dilution water was moderately hard reconstituted water (hardness typically 80-100 mg/L as

CaC03). Experiments were conducted in glass jars of unspecified size and fill volume.

Photoperiod for M. macrocopa was assumed by the reviewers to have been 16-hr: 8-hr

A-23


-------
(light: dark), as was used for daphnid culture in tests by the same authors. Preparation of test
solutions was not described. The test involved four replicates of five neonates each in five
nominal test concentrations plus a negative control. Nominal concentrations were 0 (negative
control), 6.25, 12.5, 25, 50 and 100 mg/L. Test temperature was maintained at 25 ฑ 1ฐC. Authors
note water quality parameters (pH, temperature, conductivity, and dissolved oxygen) were
measured 48 hours after exposure, but the information was not reported. Survival of organisms in
the negative control was not reported in the paper. However, raw data were obtained by EPA
from the study authors and control survival was 100% in the acute test, meeting the EPA/600/4-
90/027F requirement of at least 90% survival for test acceptability. The study authors reported a
48-hour ECso value of 17.95 mg/L (C.I. 14.72-21.18) for XI. macrocopa. The 48-hour EC50 value
was independently-calculated by EPA as I 7 2<) (1.1 73 20.(•>(•>) mu I. forM macrocopa. The
independently-calculated acute toxicity value was qiianlilali\ ely used in the derivation of the
PFOS acute crilerion

A.2.8 Eighth most Sensiti\ e 1'ieshwater (ienus for Acute Toxicity: Hyla (frog)

Tornabono ol al. (2021) conducted acute toxicity tests with the gray treefrog, Hyla

versicolor, and PTOS (purchased from Sigma Aldrich, Catalog # 77282-10G; purity not

provided) The acute tests followed standard 96-hour acute toxicity test guidance (U.S. EPA

2002; ASTM 2d I 7) The frog was collected from the field in the wetlands of Indiana near the

campus of Purdue I ni\ eisity. Collected egg masses were raised outdoors in 200 L polyethylene

tanks filled with well water. Experiments began when frogs reached Gosner stage 26, defined as

when larvae are free swimming and feeding. An additional acute test with Gosner stage 40 was

run to determine if toxicity varied between life stages. Before test initiation larvae were

acclimated to test conditions (21ฐC and 12-hr: 12-hr light:dark photoperiod) for 24 hours. A stock

solution of PFOS (500 mg/L) was made in UV-filtered well water and diluted with well water to

A-24


-------
reach test concentrations (ranged from 0 - 500 mg/L PFOS). Test concentrations were not
measured in test solutions, based on previous research that demonstrated limited degradation
under similar conditions. Larva were transferred individually to 250 mL plastic cups with 200
mL of test solutions and were not fed during the exposure period. The number of replicates
varied by life stage, and treatment; 10 replicates for each treatment for Gosner stage 26 larva,
and five to six replicates for each treatment for Gosner stage 4<~> iVous No mortality occurred in
the controls of the Gosner stage 26 test and two of the six frogs died in the controls of the Gosner
stage 40 test. The author reported 96-hour LCsos were 79 and 24 mg/L IM OS lor Gosner Stage
26 and 40, respectively. The independently-calculated 96-hr I .Cso values were 7K 33 and 19.88
mg/L and are acceptable for quantitative use. Given that GS 4<) appear to be a more sensitive
life-stage the LCso of 19.88 (13.80 - 25.95) mu I. was utilized in the derivation of the acute
freshwater criterion for PFOS.

A.2.9 Ninth Most Sensitive Freshwater Genus for Acute Toxicity: Dugesia (planarian)

Li (2008) conducted three independent repeats of a 96-hour static, unmeasured acute

toxicity test on the potassium salt of PI-OS (CAS # 2795-39-3, > 98% purity) with the planarian,

Dugesia /apoiiica (a non-North American species). The test organisms were originally collected

from Nan-shi stream located in Wu-lai. Taipei County, Taiwan in 2004 and maintained in the

laboratory in dechlorinated tap water. The planarians had a body length of 0.9 ฑ0.1 cm attest

initiation. Dilution water was dechlorinated tap water. The photoperiod consisted of 12 hours of

illumination at an unreported intensity. A primary stock solution was prepared in dilution water.

Exposure vessels were polypropylene beakers of unreported dimensions and 50 mL fill volume.

The test employed five replicates of five planarians each in at least five test concentrations plus a

negative control. Nominal test concentrations were in the range of 10-200 mg/L PFOS. The test

temperature was maintained at 25 ฑ1ฐC. No other water quality parameters were reported for test

A-25


-------
solutions. Survival of negative control animals was not reported. The study author reported 96-
hour LCso was 17 mg/L (C.I. 16-18). The independently-calculated toxicity value could not be
calculated at this time given the level of data that was presented in the paper. The study author
reported value was used quantitatively to derive the draft acute water column criteria.

Li (2009) conducted three independent repeats of a second 96-hour static, unmeasured
acute test of PFOS (potassium salt, > 98% purity) with Dugesia /apomca. The tested individuals
were originally collected from Nan-shi stream located in Wu-lai, Taipei County, Taiwan in 2004
and maintained in the laboratory in dechlorinated tap water. The planarians had a body length of
0.9ฑ0.1 cm at test initiation. Dilution water was dechlorinated tap water. The photoperiod
consisted of 12 hours of illumination at an unreported intensity. A primary stock solution was
prepared in dilution water. Exposure vessels were polypropylene beakers of unreported
dimensions and 50 mL fill volume. Each of the three independent repeats employed three
replicates of 10 planarians each in at least five test concentrations plus a negative control.
Nominal test concentrations were in the range of 5-200 mg/L PFOS. The test temperature was
maintained at 25 I C \Y'titer quality parameters including pH, conductivity, and D.O. were
reported as lui\ inu been measured at the beginning and end of each test, but the information is
not reported. Sur\ ival of negative control animals was also not reported. The study author
reported 96-hour I .("so was 23 mg/L (C.I. 20-25). The independently-calculated LC50 could not
be estimated for the first and second independent tests (as EPA was unable to fit a model with
significant parameters), but was estimated for the third independent test as 22.68 (18.27 - 27.10)
mg/L. This acute value was acceptable for quantitative use and was used to derive the PFOS
acute water column criterion.

A-26


-------
Yuan et al. (2014) also conducted a 96-hour unmeasured acute test on PFOS (potassium
salt, > 99% purity) with Dugesia japonica, under daily renewal conditions. D. japonica used for
testing were originally collected from a fountain in Quan HetouBoshan, China, and cultivated in
the laboratory for an unspecified time period before use. The planarians had a body length of 10-
12 mm at test initiation. Dilution water was aerated tap water. No details were provided
regarding photoperiod or light intensity. A primary stock solution was prepared by dissolving the
salt in DMSO. The control and exposed planarians received 0.005% DMSO (\ v). Exposure
vessels were beakers of unreported material type and dimensions with 50 inL till \ olume. The
test employed three replicates of 10 planarians each in six test concentrations plus a solvent
control. Nominal test concentrations were 0 (solvent control), 10, 30, 35, 37.5, 40 and 45 mg/L
PFOS. The test temperature was reported as 2<) (' \'o other water quality parameters were
reported. Survival of solvent control animals was not reported The study author reported 96-
hour LCso was 2^ 4o mu I. (C I 25 K< 1-33.12). The independently-calculated toxicity value could
not be calculated at this time ui\ en the level of data that was presented in the paper. The study
author reported \ nine was used <.|iianlilali\ ely to derive the draft acute water column criterion.

The noted toxicity \ allies pimided in each study summary above (17.00, 22.68, and
29.4<-> mu I.). comprising of both author-reported and independently-calculated LCso values,
were used to calculate the (iM.VV value (as the geometric mean of the three LCso values
previously mentioned) of 22 4S mg/L, which was used to derive the freshwater aquatic life
criterion.

A.2.10 Tenth Most Sensitive Freshwater Genus for Acute Toxicity: Danio (zebrafish)

The acute effects of PFOS on the zebrafish, Danio rerio, have been reported by numerous

researchers. Hagenaars et al. (2011) exposed/), rerio embryos to the potassium salt of PFOS
(CAS # 2795-39-3, purity >97%) under static unmeasured conditions for 96 hours. The PFOS

A-27


-------
was dissolved in medium-hard reconstituted laboratory water, which was aerated and kept at
26ฐC until use (no solvent). Adult wild-type zebrafish (breeding stock) were obtained from a
commercial supplier (Aqua hobby, Heist-op-den-berg, Belgium) and kept in aerated and
biologically filtered medium-hard reconstituted freshwater. Four males and four females were
used for egg production. Fertilized eggs were collected in egg traps within 30 minutes of
spawning. Eggs were transferred to the test solutions (nominal PTOS concentrations of 0.1, 0.5,
1, 5, 10 mg/L in the ELS test and 1, 5, 10, 25, 50 and 100 mg/L in the range finding test) within
60 minutes after spawning. Eggs with anomalies or damaged membranes were discarded and
fertilized eggs were separated from the non-fertilized eggs using a stereomicroscope Twenty
normally shaped fertilized eggs per exposure concentration were divided over a 24-well plastic
plate and each egg was placed individually in 2 nil. of the lest solution The remaining four wells
were filled with clean water and used for the control euus Two replicate plates were used for
each exposure concentration resulting in 40 embryos per exposure condition at the beginning of
the experiment. The 24-well plates were covered with a self-adhesive foil, placed in an
incubation chamber at 2o n 3 (' and subjected to a 14-hr: 10-hr (light:dark) cycle. Atestwas
considered \alid if more than W,, of the controls successfully hatched and showed neither
sublethal nor lethal effects The study authors reported a 96-hour LCso of 58.47 mg/L PFOS,
based on the results of the range finding test. The study author reported value was used
quantitatively to clei i\ e the draft acute water column criterion.

Sharpe et al. (2010) examined the toxicity and bioaccumulation of PFOS isomers on
Danio rerio through three different tests, a 96-hour renewal toxicity test on adults, a 48-hour
renewal toxicity test on embryos, and a chronic exposure test that evaluated maternal transfer
and fecundity of PFOS isomers. The 96-hour test is described in this present section, as these

A-28


-------
results were used quantitatively to derive the acute water column criterion. The 48-hour tests
were used qualitatively (see G.2.2.3) and the chronic toxicity tests were used qualitatively and
are summarized below in Section 4.4.2.1.4. and G.3.2.4 The authors provided little detail about
the test protocol, other than following OECD Guideline 203. Adult zebrafish were obtained from
a local pet store. They were acclimated and held in 70 L glass aquaria in an environmental
chamber set to 26ฐC and a 14-hr: 10-hr (light:dark) photoperiod lor six to 1 months prior to use
in experiments. Conditioned zebrafish water (ZF water) was obtained from the Biological
Sciences Zebrafish Facility at the University of Alberta, where an automated re\ u se osmosis
system regulated both salinity and hardness (160 mu I. total hardness and 20 mg I. calcium
carbonate hardness) of the water. A stock solution of 25 mg/inl PFOS in MeOH was used for
dosing in all experiments. All vehicle controls received a volume of MeOH equal to that present
in the highest PFOS dose of that experiment (linal MeOI I content 0.65% v/v). The concentration
of PFOS in any experiment was always well below its reported solubility in water (approx. 500
mg/L). Zebrafish toxicity tests were performed using food grade 2 L plastic tanks with four fish
per tank, and two tanks per dose I 'ish were randomly assigned to nominal doses defined as
control mu I. PI-OS): \ chicle control (<) mu I. PI OS, 0.4% MeOH v/v); and 6.25, 12.5, 2, 50
and 1 mi mu I. PI OS. Authors indicated that measured PFOS concentrations averaged 88% of
nominal, hut did not indicate w hether LCso was measured or nominal. The adult 96-hour acute
test followed OECI) 2<)3 protocol and was acceptable for quantitative use. The study author
reported LCso was 22.2 ฑ 4.6 mg/L for PFOS. The study author reported value was used
quantitatively to derive the draft acute water column criterion.

Wang et al. (2013a) evaluated the acute effects of perfluorooctane sulfonate, potassium
salt (PFOS-K, CAS# 2795-39-3 purchased from Wellington Laboratories Inc., Ontario, Canada)

A-29


-------
on zebrafish (Danio rerio) during a 96-hour unmeasured, static-renewal study. Zebrafish were
purchased from a local market at approximately three months in age and 2.2 cm in length. Fish
were allowed to acclimate for seven days and were fed three times per week until 24 hours
before the test was started. Water used for the testing was aerated for 48 hours before testing
began, and testing followed OECD Guideline 203. Observed exposure water characteristics were
total hardness of 180-220 mg/L as CaCCte, temperature of 23 1 1 ฐC. dissolved oxygen of 7.0 -
8.6 mg/L and a photoperiod of 12-hr: 12-hr light:dark. Each 2-L beaker was 111 led with 1,500 mL
of test solution at nominal concentrations of 0 (control), 2.87, 5, 8.7, 15.14. 2(-> 34. 45.83 and
79.74 mg/L PFOS. There were three replicates per concentration, and seven fish per heaker. Test
solutions were renewed at 48 hours. The author-reported 96-hour LCso was 17.0 mg/L PFOS
based on a sigmoidal three-parameter regression I-PA was unable to independently calculate a
96-hour LCso value based on the level data pn>\ ided in the paper by the study authors. Therefore,
the author-reported I .("*ฆฆ \ alue of I 7 1) mg/L PFOS was used quantitatively to derive the draft
acute freshwater criterion

Li et al. (2015) e\ aluated the acute effects of PFOS (CAS # 2795-39-3, 98% purity) to
Danio rcno \ ia a lH->-hour static unmeasured exposure. Solutions for waterborne exposure were
formulated with medium used to rear embryos (reconstituted laboratory water). Adult, wild-type
zebrafish were obtained from the Institute of Hydrobiology, at the Chinese Academy of Sciences
(Wuhan, China), and kept in treated tap water at 26-29ฐC. Fish were reared with a female/male
ratio of 1:2 under 14-hr: 10-hr light:dark photoperiod, with 1/3 of the water exchanged daily.
Spawning and fertilization took place within 30 minutes after the lights were turned on in the
morning, with fertilized embryos collected and cleaned with embryo rearing water. Immediately
after fertilization, embryos were examined, and damaged or unfertilized embryos were discarded.

A-30


-------
Zebrafish embryos were exposed in 24-well cell culture plates (assume plastic) to a series of
PFOS concentrations (6.25, 12.5, 25.0, 50.0, 100.0 and 200.0 mg/L). Twenty, normally shaped,
fertilized exposed embryos were assigned to each treatment (three replicates) and 2 mL
corresponding solution per well; the four remaining wells were assigned with control embryos.
Embryos were cultured in an incubator at 28.5ฐC, pH of 8.3 and a 14-hr: 10-hr light:dark
photoperiod. Toxicological endpoints included whether embryos were clear or opaque, have
edema at 4, 8, 24, 48, 72, or 96 hpf, or have structural malformations ill 72 or % hpf until
hatching. Coagulated embryos before hatching are opaque, milky white, and appear dark under
the microscope. Coagulation of embryos was recorded and used for the calculation of an LCso
value. The author reported 96-hour LCso was 68.0 mg/L PFOS. The independently-calculated
LCso value was 71.12 (56.82 - 85.42) mg I. Pl'OS and this toxicity value is acceptable for
quantitative use and was used to derive the freshwater water criterion for PFOS.

In a later study. I)u cl ;il. (2016;2017) exposed / kinio rcrio to
heptadecafhiorooctanesulIonic acid (Pl'OS, potassium salt, CAS# 2795-39-3, 98% purity) using
static unmeasured procedures Although the study focused on the protective effects of zinc
nanoparticles (ZnO-\Ps) 011 Pl 'OS toxicity (development and damage to DNA), data were also
reported for PI OS-only exposures. Adult AB strain zebrafish were purchased from State Key
Laboratory of freshwater Ecology and Biotechnology, Chinese Academy of Sciences (Wuhan,
China). Fish were maintained and tested at 28ฐC under a 14-hr: 10-hr light:dark cycle. Male and
female fish were paired in spawning boxes overnight in rearing water and spawning was
completed at the beginning of the light cycle. Eggs were collected from the spawn traps and
transferred to clean rearing water prior to testing. The embryos were exposed to PFOS (1, 2, 4, 8
and 16 mg/L in a preliminary test to determine the LC50, and at 0.4, 0.8 and 1.6 mg/L in later test

A-31


-------
with ZnO-NPs) solutions to evaluate mortality (at 96 hours), body length (at 96 hours), hatch rate
(at 72 hours), heart rate (at 48 hours), and malformation rate (at 96 hours). Embryos were kept in
24-well multi-plates at two embryos/well, in which 20 wells contained 2 mL reconstituted water
test solution and four wells contained 2 mL of culture medium as the control; each plate
contained 40 embryos for exposure testing and eight embryos as internal water controls. For each
concentration and water control, three 24-well plates (replicates) were included. The study
authors reported a 96-hour LCso of 3.502 mg/L for PFOS. EPA was unable lo independently
calculate a 96-hour LCso value based on the level data pro\ ided in the paper In the study authors.
Therefore, the author-reported LCso value of 3.502 mg I. PI'OS was used quanlilali\ ely to derive
the draft acute freshwater criterion.

Stengel et al. (2017) exposed 1 hpf / kinio rcrio emlnyos lo PFOS for 96 hours using
renewal unmeasured procedures as specified in (OI X'I) 2d 13) guidelines. PFOS stock and
exposure solutions were prepared in reconstituted laboratory water. All adult zebrafish used for
breeding were wild-type descendants of the "Westaquarium" strain and obtained from the
Aquatic Ecology and Toxicology breeding facilities at the University of Heidelberg. Details of
zebrafish maintenance, egg production and embryo rearing as described previously (Kimmel et
al. 19l)5. kimmel et al. llMX. \agel 2<)(J2; Spence et al. 2006; Wixon 2000) and have been
updated for the purpose of the zebrafish embryo toxicity test by (Lammer et al. 2009). Embryos
no older than 1 hpf were exposed in glass vessels, which had been preincubated (saturated) for at
least 24 hours, to a series of PFOS dilutions (0, 3.125, 6.25, 12.5, 25, 50 mg/L). After control of
fertilization success, embryos were individually transferred to the wells of 24-well plates, which
had been pre-incubated with 2 mL of the test solutions per well for 24 hours prior to the test start,
and kept in an incubator at 26.0 ฑ1,0ฐC under a 14-hr: 10-hr light:dark regime. In order to prevent

A-32


-------
evaporation or cross-contamination between the wells, the plates were sealed with self-adhesive
foil. Embryo tests were classified as valid if the mortality in the negative control was < 10%, and
if the positive control (3,4-dichloroaniline) showed mortalities between 20 and 80%. All fish
embryo tests were run in three independent replicates. Both lethal and sublethal effects were
used for the determination of EC values. The author reported 96-hour LCso and EC so
(combination of lethal and sublethal effects) values were 34.2 and 21.4 mg/E PFOS,
respectively. The independently-calculated LCso was 38.82 (36.67 - 4<) lW) mu/L PFOS. The
independently-calculated LCso were considered quantitative and were used lo deri\ e the PFOS
acute water criteria.

A.2.11 Eleventh Most Sensitive Freshwater Genus for Acute Toxicity: Dayhnia (cladoceran)
Logeshwaran et al. (2021) conducted acute and chronic toxicity tests with the

cladoceran, Daphnia carinata, and PFOS-k (perlluoiooclancsulfonalc potassium salt, > 98%

purity, purchased from Sigma-Aldrich Australia) In-house cultures of daphnids were maintained

in 2 L glass bottles with 3<)"() natural spring water in deionized water, 21ฐC and a 16-hr:8-hr

light:dark photoperiod The acute test protocol followed OECD guidelines (2000) with slight

modifications A PI-OS stock solution (20 mg/mL) was prepared in dimethylformamide and

diluted with deionized water to achieve a concentration of 200 mg/L PFOS. Cladoceran culture

medium was used to prepare the PFOS stock and test solutions. Ten daphnids (6-12 hours old)

were transferred to polypropylene containers containing one of 14 nominal test concentrations

(0, 0.5, 1, 2.5, 5, 10, I 5. 20, 25, 30, 35, 40, 45 and 50 mg/L PFOS). Each test treatment was

replicated three times and held under the same conditions as culturing. At test termination (48

hours) immobility was determined after 15 seconds of gentle stirring. No mortality occurred in

the controls. The authors reported 48-hour EC so was 8.8 mg/L PFOS. The independently-

A-33


-------
calculated 48-hour ECso value was 11.56 (10.06 - 13.07) mg/L and is acceptable for quantitative
use in the derivation of the acute freshwater criterion for PFOS.

3MCompany (2000) provides the results of a 48-hour static, unmeasured acute toxicity
test completed with the cladoceran, Daphnia magna, and PFOS-K (perfluorooctancesulfonate
potassium salt, CAS # 2795-39-3, unknown purity) in 1984. In-house culture of daphnids were
tested in unchlorinated, carbon filtered well water under a 16-hr S-lir light:dark photoperiod, pH
7.6, total hardness of 256 mg/L as CaCCte, dissolved oxygen >70% saturation and average
temperature of 22ฐC. Twenty daphnids (12 ฑ 12 hour) were placed in 250 ml. beakers with 200
mL of test solution and exposed to one of six nominal test treatments (specifics not provided) for
48 hours; each test treatment was duplicated. The author reported ECso, based on immobility,
was 27 mg/L PFOS. EPA was unable to independently calculate a 48-hour ECso value based on
the level data provided in the paper by the study authors Therefore, the author-reported ECso
value of 27 mg/l. PI-OS was used quantitatively to derh e the draft acute freshwater.

Drottar and krucgcr (2000h) reported the results of a 48-hour static, measured acute
toxicity test on PI-OS (potassium salt. CAS 27l)5-39-3, 90.49% purity) with Daphnia magna.
The Gl.P test was conducted at Wildlife International, Ltd. in Easton, MD in February, 1999.
The lest followed OECD 2<>2 (1994) and U.S. EPA OPPTS Number 850.1010 (1996). The test
organisms were less than 24 hours old at test initiation. Dilution water was 0.45 jam filtered well
water [hardness: 132 (12S-136) mg/L as CaC03; alkalinity: 178 (176-178) mg/L as CaC03; pH:
8.3 (8.2-8.3); TOC: l.<> mg/L; conductivity: 313 (310-315) |imhos/cm]. Photoperiod was 16-
hr: 8-hr (light:dark) with a 30 minute transition period. Light was provided at an intensity of
approximately 359 lux. A primary stock solution was prepared in dilution water at 91 mg/L. It
was mixed for -19.5 hours prior to use. After mixing, the primary stock was proportionally

A-34


-------
diluted with dilution water to prepare the four additional test concentrations. Exposure vessels
were 250 mL plastic beakers containing 240 mL of test solution. The test employed two
replicates of 10 daphnids each in five measured test concentrations plus a negative control.
Nominal concentrations were 0 (negative control), 12, 20, 33, 55, 91 mg/L. Mean measured
concentrations were  lo 106% of
nominal. Dissolved oxygen in control and highest test concentration (91 mg/L) ranged from 8.6-
8.9 mg/L and 8.6-9.1 mg/L; pH ranged from S 2-S 5 and 8.5-S (•>. respectively. Test temperature
ranged from 19.5-20.2ฐC and 19.3-20.1ฐC. respecti\ ely Daphnids in the negative control, and
the 11 and the 20 mu I. treatments appeared healthy and normal throughout the test with no
mortality, immobility or o\ ert clinical signs of toxicity. Five percent mortality was observed at
48 hours in the neuati\ e control The study author reported 48-hour ECso was 61 mg/L (C.I. 33-
91). The independently-calculated toxicity value was 58.51 (53.59 - 63.43) mg/L and was used
quantitiiti\ ely to derive the draft acute water column criterion.

Bourircsui (2002) performed a 48-hour static test on PFOS (potassium salt, CAS # 2795-
39-3, 95% purity) with / ktphnia magna and Daphnia pulicaria as part of a Master's thesis at the
University of Guelph, Ontario, Canada. The results were subsequently published in the open
literature (Boudreau et al. 2003a). The test followed ASTM E729-96 (1999). Daphnids used for
testing were less than 24 hours old at test initiation. D. magna were obtained from a brood stock
(Dm99- 23) at ESG International (Guelph, ON, Canada). D. pulicaria were acquired from a

A-35


-------
brood stock maintained in the Department of Zoology at the University of Guelph. Dilution
water was clean well water obtained from ESG International. Hardness was softened by addition
of distilled deionized water to achieve a range of 200-225 mg/L of CaCCte. Photoperiod was 16-
hr: 8-hr (light:dark) under cool-white fluorescent light between 380 and 480 lux. Laboratory-
grade distilled water was used for all solutions with maximum concentrations derived from stock
solutions no greater than 450 mg/L. Test vessels consisted of 225 m I. polypropylene disposable
containers filled with 150 mL of test solution. All toxicity testing invol\ ed four replicates of 10
daphnids each in five nominal test concentrations plus a negative control. Nominal
concentrations were 0 (negative control), 31, 63. 125. 25<~> and 450 mg/L. Experiments were
conducted in environmental chambers at a test temperature of 21 ฑ1 ฐC. Authors note temperature
and pH were measured at beginning and end of study, but the information was not reported.
Survival of daphnids in the negative control was also not reported, although ASTM E729-96
requires at least 90".. sur\ i\ al for test acceptability The study author reported 48-hour ECso for
D. magna was 67 2 mu I. (C I 31 3-XX 5) The study author reported 48-hour ECso for D.
pulicaria was 134 mu I. (C I 1 <ป3-1 75). The independently-calculated toxicity values could not
be calculated at this time ui\ en the le\ el of data that was presented in the paper. The study author
reported \ allies u ere used quantitati\ ely to derive the draft acute water column criterion.

Ji el ;il. (2008) performed a 48-hour static, unmeasured acute test of PFOS (acid form,
CAS # 1763-23-1. purity unreported) with Daphnia magna. The test followed EPA's Methods
for measuring the acute toxicity of effluents and receiving waters to freshwater and marine
organisms (U.S.EPA 2002b). D. magna used for testing were obtained from brood stock cultured
at the Environmental Toxicology Laboratory at Seoul National University (South Korea). Test
organisms were less than 24 hours old at test initiation. Dilution water was moderately hard

A-36


-------
reconstituted water (hardness typically 80-100 mg/L as CaCCte). Experiments were conducted in
glass jars of unspecified size and fill volume. Photoperiod was assumed by the reviewers to have
been 16-hr:8-hr (light:dark). Preparation of test solutions was not described. The test involved
four replicates of five daphnids each in five nominal test concentrations plus a negative control.
Nominal concentrations were 0 (negative control), 6.25, 12.5, 25, 50 and 100 mg/L. Test
temperature was maintained at 21 ฑ 1ฐC. Authors note water quality parameters (pH,
temperature, conductivity, and dissolved oxygen) were measured 48 hours after exposure, but the
information is not reported. Survival of daphnids in the negative control was not reported in the
paper. However, raw data were obtained by EPA from the study authors and control survival was
100% and therefore meets the EPA/600/4-90/027F requirement of at least 90% survival for test
acceptability. The study author reported 4S-hour I ฆ("*<) value lor the study was 37.36 mg/L (C.I.
30.72-43.99) for D. magna. The 48-hour ECso value was independently-calculated by EPA and
equaled 35.46 (2S 2o 42 (•>(•>) mu I. IbrZ). magna. This independently-calculated acute toxicity
value was included in the dominion of the PFOS acute criterion.

Li (2009) conducted three independent repeats of a 48-hour static acute test on PFOS
(potassium salt, lW() purity) with / ktplmia magna. The test followed OECD 202 (1984). D.
magna used for the test were less than 24 hours old at test initiation. Dilution water was
dechlorinated tap water. The photoperiod consisted of 12 hours of illumination at an unreported
light intensity. A primary stock solution was prepared in dilution water and did not exceed 400
mg/L. Exposure vessels were polypropylene of unreported dimensions and 50 mL fill volume.
The test employed five replicates of six daphnids each in at least five test concentrations plus a
negative control. Each treatment was tested three independent times. Based on water solubility of
test chemicals and preliminary toxicity results, nominal test concentrations were in the range of

A-37


-------
10-400 mg/L for PFOS. Water quality parameters including water pH and conductivity and
dissolved oxygen were measured at the beginning and at the end of each test. Initial values of pH
were 7.82ฑ0.12 and 7.91ฑ0.03 after 48 hours. At the start of the bioassays, dissolved oxygen and
specific conductivity were 67.7ฑ6.8% and 101.8ฑ6.8 |iS/cm. After the 48-hour testing period,
dissolved oxygen and specific conductivity were 55.6ฑ1.26% and 109.1ฑ3.5 |iS/cm. The
dissolved oxygen after 48-hours of testing was lower than the lest guideline recommendation of
>60% (ASTM 2002; U.S. EPA 2016a and b); however, it was not low enough to change the use
of the study. The test was conducted in a temperature incubator at 25 ฑ2ฐC. None of the control
animals became immobile at the end of the test. The study author reported 48-hour l -Cso was 63
mg/L (C.I. 58-69) which was an average LCso of the three tests. The independently-calculated
LCso values for the three independent experimental repeats were 55 4<~> (45.97), 72.70 (61.63 -
83.77) and 64.60 (49.53 - 79.66) mg/L, respecli\ ely These independently-calculated LC50
values were used in the (i\l.\V calculation.

Yang et al. (2014) conducted a 48-hour static acute test of PFOS (potassium salt, CAS #
2795-39-3. 00ฐ„ puriiy) with / kiphma magna, following ASTM E729 (1993). D. magna used for
the test were donated In the Chinese Research Academy of Environmental Sciences. The
daphnids were less than 24 hours old at test initiation. Dilution water was dechlorinated tap water
(pH, 7.0ฑ0 5. dissoK ed oxygen, 7.0ฑ0.5 mg/L; total organic carbon, 0.02 mg/L; and hardness,
190.0ฑ0.1 mg/L as CaCO;) Photoperiod was 12-hr:12-hr (light:dark) at an unreported light
intensity. A primary stock solution was prepared by dissolving PFOS in deionized water and
cosolvent DMSO. The primary stock was proportionally diluted (0.56x) with dilution water to
prepare the test concentrations. Exposure vessels were 200 mL beakers of unreported material
type containing 200 mL of test solution. The test employed three replicates of 10 daphnids each

A-38


-------
in six test concentrations (measured in low and high treatments) plus a negative and solvent
control. Nominal concentrations were 0 (negative and solvent controls), 20.00, 36.00, 64.80,
116.64, 209.95 and 377.91 mg/L. Mean measured concentrations before and after renewal were
respectively 18.43 and 19.80 and mg/L (lowest concentration) and 341.74 and 372.35 mg/L
(highest concentration). Analyses of test solutions were performed using HPLC/MS and negative
electrospray ionization. The concentration of PFOS was calculated lVom standard curves (linear
in the concentration range of 1-800 ng/mL), and the average extraction efficiency was in the
range of 70-83%. The concentrations and chromatographic peak areas exhibited a significant
positive correlation (r=0.9987, p<0.01), and the water sample-spiked recovery was 105%. The
temperature, D.O., and pH were reported has having been measured every day during the acute
test, but results are not reported. Negative control survival was lH->% Solvent control survival
was 100%. The study author reported 48-hour I.Cs" was 7S <)0 m^/L (C.I. 54.38-112.13). The
study author reported \ alue was used quanti1ati\ ely to deri\ e the draft acute water column
criterion.

Lu et al. (2015) conducted a 4S-hour static test on PFOS (purity 98%) with Daphnia
magna, following Ol-('l) 2"2 (2<>(>4c) A magna used for the test were originally obtained from
the Chinese Center for Disease Control and Prevention (Beijing, China) and cultured in the
laboratory according to the International Organization for Standardization (ISO 1996). Daphnids
were less than 24 hours old at lest initiation. Dilution water was the same used for daphnid
culture and was reconstituted according to OECD (2004c) with a hardness of 250 mg/L as
CaC03, as calculated based on the recipe provided, and pH ranging from 7.7 to 8.4. Photoperiod
was 16-hr:8-hr (light:dark) at an unreported light intensity. The test solution was prepared
immediately prior to use by diluting the stock solution with a daphnia culture medium. Exposure

A-39


-------
vessels were 100 mL glass beakers containing 45 mL of test solution. The test employed three
replicates of 10 daphnids each in six nominal test concentrations plus a negative control.

Nominal concentrations were 0 (negative control), 1,3, 10, 30, 100, and 300 mg/L. Exposure
water quality was checked daily and maintained at a temperature of 20 ฑ1ฐC, pH of 7.2 ฑ0.3, and
dissolved oxygen of 6.5 ฑ0.5 mg/L. 100% survival was observed at 48 hours in the negative
control. The author reported 48-hour EC so was 23.41 mg/L (I.C4l) 27) The study author
reported value was used quantitatively to derive the draft acute water column criterion.

Liang et al. (2017) conducted a 48-hour static test on PFOS (potassium sail, CAS #
2795-39-3, >98% purity) with Daphnia magna. The test followed OECD 202 (ป(>4c). D. magna
used for the test were originally obtained from State Key Laboratory of Environmental Aquatic
Chemistry (Eco-Environmental Sciences of Chinese Academy of Sciences, Beijing) and cultured
in the laboratory according to Revel et al. (2') I 5) Daphnids were less than 24 hours old at test
initiation. Dilution water was artificial medium (M4) at 2d C and pH 7 (Revel et al. 2015).
Photoperiod was Io-hr X-lir (light dark) at an unreported light intensity. The test solution was
prepared immediately prior to use In diluting the stock solution with M4 medium. Exposure
vessels were So ml. honkers of unreported material type containing an unspecified volume of test
solution The test employed li\ e replicates of five daphnids each in six nominal test
concentrations plus a negati\ e control. Nominal concentrations were 0 (negative control), 30, 44,
66, 100, and 150 mu I. No mention was made of water quality being checked during the
exposure. 100% survival was observed at 48 hours in the negative control. The study author
reported 48-hour ECso was 79.35 mg/L. The independently-calculated toxicity value was 94.58
(94.20 - 94.96) mg/L and was used quantitatively to derive the draft acute water column
criterion.

A-40


-------
Yang et al. (2019) evaluated the acute effects of perfluorooctane sulfonate, potassium
salt (PFOS-K, CAS# 2795-39-3, 98% purity, purchased from Sigma-Aldrich in St. Louis, MO)
on Daphnia magna via a 48-hour unmeasured, static mortality test. D. magna cultures were
obtained from the Institute of Hydrobiology of Chinese Academy of Science in Wuhan, China.
Organisms were cultured in Daphnia Culture Medium according to the parameters laid out in
OECD Guideline 202 and all testing followed OECD Guideline 2'>2 Cultures were fed green
algae daily and were acclimated for two to three weeks before testing Acute lest concentrations
included 0 (control), 0.0000156, 0.0000234, 0.0000349, 0.0000788 and O.i.inn I IS niol/L (or 0
(control), 8.396, 12.59, 18.78, 28.31, 42.41, and 63.51 mg/I. given the molecular weight of the
form of PFOS used in the study, CAS # 2795-39-3, of 538.22 g/mol). Five neonates (12-24 hours
old) were placed randomly in 100 mL glass honkers filled with (ฆ><> ml. test solution, with four
replicates per concentration. Organisms ^ere ohser\ ed lor mortality at 48 hours, and the authors
reported aLCso of 22 77 mg I. l-IWs independently-calculated 48-hour LCsowas 22.43 (15.74
- 29.12) mg/T, PFOS and was used quantitatively to derive the draft acute water column criterion
for freshwater

A.2.12 Twelfth Most Sensili\ e I'reshwaler Genus for Acute Toxicity: Ambystoma (salamander)
Tornahcnc ol al. (2021) conducted acute toxicity tests with three species of salamanders

in the genus .\mbysioma and PI OS (purchased from Sigma Aldrich, Catalog # 77282-10G;

purity not provided) Acute tests followed standard 96-hour acute toxicity test guidance (U.S.

EPA 2002; ASTM 2<) I 7) The three test species (Jefferson salamander, Ambystoma

jeffersonianum\ small-mouthed salamander, A. texanum\ eastern tiger salamander, A. tigrinum)

were collected from a field in the wetlands of Indiana near the campus of Purdue University.

Collected egg masses were raised outdoors in 200 L polyethylene tanks filled with well water.

Experiments began when salamanders reached Harrison stage 40, defined as when larvae are free

A-41


-------
swimming and feeding. Before test initiation larvae were acclimated to test conditions (21ฐC and
12-hr: 12-hr light:dark photoperiod) for 24 hours. An additional acute test with Harrison stage 46
small-mouthed salamanders was run to determine if toxicity varied between life stages. A stock
solution of PFOS (500 mg/L) was made in UV-filtered well water and diluted with well water to
reach test concentrations (ranged from 0-500 mg/L PFOS). Test concentrations were not
measured in test solutions, based on previous research that demonstrated limited degradation
under similar conditions. Larva were transferred individually to 250 nil. plastic cups with 200
mL of test solutions and were not fed during the exposure period. The number of replicates
varied by species, life stage and treatment; five replicates per treatment for Jefferson salamander
and Harrison stage 46 small-mouthed salamander, seven replicates per treatment for Harrison
stage 40 small-mouthed salamander, and 2<> replicates in the control and 10 replicates in each
treatment for eastern tiger salamander. No mortality occurred in any of the control groups.
Author-reported lH->-hour I .("*ฆ* were M. 41 and 73 mg/1. PTOS for the Jefferson salamander,
small-mouthed salamander and eastern tiger salamander, respectively. The authors did not find a
significant difference between the life stages of sinall-mouthed salamander so results of the two
tests were pooled The independently-calculated 96-hour LCso values were 51.71 (40.84 - 62.58)
and 4o 71 (34 33 59.09) for I larrison stage 40, 30.00 (27.14 - 32.86) for Harrison stage 46, and
68.63 (61.^i) 75.37) mg/l. lor the Jefferson salamander, small-mouthed salamander (both life
stages tested) and eastern tiger salamander, respectively. In general, the independently-calculated
toxicity values were acceptable for quantitative use and were utilized to derive the acute
freshwater criterion for PFOS. Specifically, the LCso value of Harrison stage 46 of 30.00 mg/L
for small-mouthed salamander was used for this species alone (as opposed to both LCso values)
as this life stage was determined to be the most sensitive.

A-42


-------
A.2.13 Thirteenth Most Sensitive Freshwater Genus for Acute Toxicity: Anaxyrus (toad)

Tornabene et al. (2021) conducted acute toxicity tests with the American toad, Anaxyrus

americanus, and PFOS (purchased from Sigma Aldrich, Catalog # 77282-10G; purity not

provided). The acute tests followed standard 96-hour acute toxicity test guidance (U.S. EPA

2002; ASTM 2017). The frog was collected from a field in the wetlands of Indiana near the

campus of Purdue University. Collected egg masses were raised outdoors in 200 L polyethylene

tanks filled with well water. Experiments began when toads readied (iosner stage 26, defined as

when larvae are free swimming and feeding. An additional acute test with (iosner stage 41 was

run to determine if toxicity varied between life stages Before test initiation lar\ ae were

acclimated to test conditions (21ฐC and 12-hr:12-hr liuhcdark photoperiod) for 24 hours. A stock

solution of PFOS (500 mg/L) was made in I'V-filiered'well water and diluted with well water to

reach test concentrations (ranged from 0 - 5<)i) nig I. PI OS) Test concentrations were not

measured in test solutions, based on previous research that demonstrated limited degradation

under similar conditions I ,ar\ a were transferred individually to 250 mL plastic cups with 200

mL of test solutions and were not led during the exposure period. The number of replicates

varied In lile stage, and treatment, in replicates for each treatment for Gosner stage 26 larva,

and six to l<> replicates lor each treatment for Gosner stage 41 toads. No mortality occurred in

any of the control groups. The author reported 96-hour LCso was 62 mg/L PFOS. The authors did

not find a significant difference between the life stages of the American toad, so results of the

two tests were pooled. The independently-calculated 96-hour LCso values were 63.41 (62.32 -

64.51) mg/L for the GS 26 frogs and 56.49 (49.10 - 63.90) mg/L for GS 41 toads. Given that the

GS 41 appear to be a more sensitive life-stage the LCso of 56.49 mg/L was considered acceptable

for quantitative use and was utilized in the derivation of the acute freshwater criterion for PFOS.

A-43


-------
A.2.14 Fourteenth Most Sensitive Freshwater Genus for Acute Toxicity: Procambarus (crayfish)
Funkhouser (2014) conducted a 7-day static acute test on PFOS (potassium salt, >98%

purity) with the crayfish species, Procambarus fallax (f. virginalis), as part of a Master's thesis

at the Texas Tech University, Lubbock, TX. Juvenile P. fallax (2-week old, 0.041 g) used for the

test were originally purchased from a private collector. The crayfish reproduced for several

generations before being used for experiments. Based on an average reproductive age of 141-255

days, an interclutch period of 50-85 days, and a brooding time of 22-42 days, the author

estimated the experimental animals to be F4-F6 (Seitz et al. 2005). Dilution water was

moderately hard reconstituted laboratory water (3 <) g CaSC>4, 3.0 g MgSC>4, 0.2 g KCI. and 4.9 g

NaFtCCte added to 50 L deionized water). Photoperiod was 14-hr: 10-hr (light:dark) at an

unreported light intensity. PFOS was dissoK ed in dilution water to prepare the test

concentrations. Exposure vessels were 1 I. polypropylene containers containing 500 mL of test

solution. The test employed two replicates of three snails each in five test concentrations plus a

negative control Nominal concentrations were 0 (negative control), 40, 80, 120, 160, and 200

mg/L. Exposure concentrations were reportedly measured, but concentrations were not reported.

Analyses of test solutions were performed using LC-MS/MS. Standards were used as part of the

analytical method, but details were not reported. The reporting limit was 0.010 mg/L.

Experiments were conducted in an incubator at 25 ฑ1ฐC and covered with plastic opaque

sheeting to limit e\ aporation No other water quality parameters were reported as having been

measured in test solutions. Negative control survival was 100% after seven days. The study

author reported 96-hour LCso was reported as 59.87 mg/L. For comparison, the 7-day LCso was

39.71 mg/L. The 96-hour study author-reported value was used quantitatively to derive the draft

acute water column criterion.

A-44


-------
A.2.15 Fifteenth Most Sensitive Freshwater Genus for Acute Toxicity: Brachionus (rotifer)
Zhang et al. (2013) performed a 24-hour static test of PFOS (potassium salt, CAS #

2795-39-3, >98% purity) with Brachionus calyciflorus. Organisms were less than two hours old

at test initiation. All animals were parthenogenetically-produced offspring of one individual from

a single resting egg collected from a natural lake in Houhai Park (Beijing, China). The rotifers

were cultured in an artificial inorganic medium at 20ฐC (16-hr:8-hr. liuht:darl<; 3,000 lux) for

more than six months before toxicity testing to acclimate to the experimental conditions. All

toxicity tests were carried out in the same medium and under the same conditions as during

culture (i.e., pH, temperature, illumination). Solvent-free stock solutions of PI OS (1 .<)<><) mg/L)

were prepared by dissolving the solid in deionized water \ ia sonication. After mixing, the

primary stock was proportionally diluted with dilution water to prepare the test concentrations.

Exposures were in 15 mL, 6-well cell cull lire pi ales (assumed plastic) each containing at total of

10 mL of test solution. The test employed se\ en measured test concentrations plus a negative

control. Each treatment consisted of one replicate plate of 10 rotifers each in individual cells and

repeated six times. Nominal concentrations were 0 (negative control), 40, 50, 60, 70, 80, 90, 100

mg/L. PFOS concentrations were not measured in the rotifer exposures, but rather, in a side

experiment using HPLC MS The side experiment showed that the concentration of PFOS

measured e\ cry eight hours o\ er a 24-hour period in rotifer medium with green algae incurs

minimal change in the concentration range 0.25 to 2.0 mg/L. The acute test was conducted

without green algae added to the exposure medium. 100% survival was observed at 24 hours in

the negative control. The study author reported 24-hour LCso was 61.8 mg/L. The study author-

reported value was used quantitatively to derive the draft acute water column criterion.

A-45


-------
A.2.16 Sixteenth Most Sensitive Freshwater Genus for Acute Toxicity: Elliptio (mussel)

Drottar and Krueger (2000e) reported the results of a 96-hour renewal, measured test

on the effects of PFOS (potassium salt, CAS # 2795-39-3, 90.49% purity) on Elliptio

complanata (formerly known as Unio complanatus). The good laboratory practice (GLP) test

was conducted at Wildlife International, Ltd. in Easton, MD in August, 1999, using a protocol

based on procedures outlined in U.S. EPA, OPPTS Number 850.1 <>75. OECD 203, and ASTM

E729-88a (1988). E. complanata (76.5 g and 48.7 mm) used lor the test were purchased from

Carolina Biological Supply Company in Burlington, NC, after being caught in the wild. They

were of an unspecified age at test initiation. Dilution water was 0.45 |im filtered u el I w ater [total

hardness: 126 (120-132) mg/L as CaCCte; alkalinity: 174 (1 7<)-178) mg/L as CaCOj; pH: 8.3

(8.1-8.5); TOC: <1.0 mg/L; conductivity 21 (3 I n-330) |imhos cm | Photoperiod was 16-hr:8-hr

(light:dark) with a 30-minute transition period I .ighl uas provided at an intensity of

approximately 369 Ui\ \ primary slock solution was prepared in dilution water at 91 mg/L. It

was mixed for - 24 hours prior lo use After mixing, the primary stock was proportionally diluted

with dilution water to prepare the four additional test concentrations. Exposure vessels were 25 L

polyethylene aquaria containing 20 L of test solution. The test employed two replicates of 10

mussels each in five measured test concentrations plus a negative control. Nominal

concentrations were 0 (negati\ e control), 5.7, 11, 23, 46, and 91 mg/L. Mean measured

concentrations were less n 115 mg/L, 5.3, 12, 20, 41, and 79 mg/L, respectively. Analyses of

test solutions were performed at Wildlife International, Ltd. using high performance liquid

chromatography with mass spectrometric detection (HPLC/MS). The mean percent recovery of

matrix fortifications analyzed concurrently during sample analysis was 94.7%. Concentrations

measured at test initiation averaged 86% of nominal. Concentrations measured prior to renewal

at 48 hours averaged 89% of nominal. Concentrations measured at 96 hours averaged 100% of

A-46


-------
nominal. Dissolved oxygen in control and the high-test concentration (79 mg/L) respectively
ranged from 5.8-8.5 mg/L and 5.0-8.6 mg/L; pH ranged from 8.0-8.4 and 7.9-8.5. Test
temperature ranged from 21.4-21.8ฐC and 21.8-23.7ฐC. Mussels in the negative control, the 5.3,
12, and 20 mg/L treatments appeared healthy and normal throughout the test with no mortality,
immobility or overt clinical signs of toxicity. The author reported 96-hour LCso was 59 mg/L
(C.I. 51-68). The independently-calculated LCso value was 64 35 (5(-> 22 72.48) mg/L. This
independently-calculated acute value was acceptable for quantitative use and utilized in the
derivation of the acute PFOS aquatic life criteria.

A.2.17 Seventeenth Most Sensitive Freshwater Genus for Acute Toxicity: Lithohalcs (frog)
Flynn et al. (2019) evaluated the acute effects of peril uorooctanesulfonic acid (PFOS,

CAS# 1763-23-1, purchased from Sigma-.\klrich) on the American bullfrog (Lithobates

catesbeiana, formerly, Rana catesbeiana) during a lH->-lioiir unmeasured, static study. Testing

followed Purdue University's Institutional Animal Care and I se Committee Guidelines Protocol

#1601001355 1. American bullfrog eggs were taken from a permanent pond in the Martell Forest

outside of West Lafayette, Indiana. The eggs from a single egg mass were acclimated in 100-L

outdoor tanks tilled with 70 L of aged well water and covered with a 70% shade cloth. Once

hatched, tadpoles were ted rahhit chow and TetraMin ad libitum and were acclimated to

laboratory conditions for 24 hours before testing. A 500 mg/L PFOS stock solution was prepared

with RO water to create nominal test concentrations of 0 (control), 10, 25, 50, 75, 100, 150, 300

and 500 mg/L. Each treatment contained 10 replicates with one Gosner Stage 25 tadpole in each

250 mL plastic tub maintained at 21ฐC and a 12-hr: 12-hr light:dark photoperiod. Mortality was

monitored twice daily. The author reported LCso value was 144 mg/L PFOS. EPA's

independently-calculated 96-hour LCso was 154.8 (108.7 - 200.9) mg/L PFOS and was

A-47


-------
considered acceptable to be used quantitatively to derive the draft acute water column criterion
for freshwater.

Tornabene et al. (2021) conducted acute toxicity tests with four species of frogs in the
genus Lithobates (formerly, Rana) and PFOS (purchased from Sigma Aldrich, Catalog # 77282-
10G; purity not provided). Acute tests followed standard 96-hour guidance (U.S.EPA 2002a);
ASTM 2017). The four test species (American bullfrog, Lithobatcs caicsbeiana\ green frog, L.
clamitans; northern leopard frog, L. pipiens; wood frog, A. sy/vattea) were collected from a field
in the wetlands of Indiana near the campus of Purdue University. Collected euu masses were
raised outdoors in 200 L polyethylene tanks filled with well water. Experiments began when
frogs reached Gosner stage 26, defined as when larvae are free swimming and feeding. Before
test initiation larvae were acclimated to test conditions (21ฐC and 12-hr 12-hr light:dark
photoperiod) for 24 hours. A stock solution of PI-OS (5<)<) mu I.) was made in UV-filtered well
water and diluted with well water to reach test concentrations (ranged from 0 - 500 mg/L PFOS).
Test concentrations were not measured in test solutions, based on previous research that
demonstrated limited degradation under similar conditions. Larva were transferred individually
to 25') ml. plastic cups with 2<)() ml. of test solutions and were not fed during the exposure
period The number of replicates varied by species, and treatment; 20 replicates in the control
and five to 10 replicates in each treatment for American bullfrog, 10 replicates for each treatment
for green frog, northern leopard frog and wood frog. No mortality occurred in any of the control
groups. Author reported 96-hour LCsos were 163, 113, 73 and 130 mg/L PFOS for the American
bullfrog, green frog, northern leopard frog, and wood frog, respectively. The independently-
calculated 96-hr LCso values for American bullfrog and northern leopard frog were 133.23
(95.75 - 170.8), and 72.72 (63.88 - 81.55) mg/L, respectively. EPA was unable to independently

A-48


-------
calculate LCso values for green frog and wood frog as a curve could not be fit with significant
parameters. Therefore, the independently-calculated LCso values for American bullfrog (133.3
mg/L) and northern leopard frog (72.72 mg/L) were used quantitatively to derive the acute
freshwater criterion for PFOS. The author-reported LCso values for green frog (113 mg/L) and
wood frog (130 mg/L) were used quantitatively to derive the acute freshwater criterion for PFOS
as the author-reported toxicity values were consistent with the independently-calculated LCso
values for other species included in the study.

A.2.18 Eighteenth Most Sensitive Freshwater Genus for Acute Toxicity: I'/iysc/la (snail)

Li (2009) conducted three independent repeals of a 96-hour static acute lest on PFOS

(potassium salt, > 98% purity) with the bladder snail species. I'liysella acuta (Nole. formerly

known as Physa acuta). The test organisms were collected from a ditch located in Shilin of

Taipei City in June 2005. Snails were fed with lellnce and half of llie culture medium was

changed with deehlorinated water every two weeks, implyinu a holding time of greater than two

weeks. P. acuta of mixed ages were used at lest initiation. Dilution water was deehlorinated tap

water. The photoperiod consisted of 12 hours of illumination at an unreported light intensity. A

primary stock solution was prepared in dilution water. Exposure vessels were polypropylene

beakers of unreported dimensions and 1 L fill volume. The test employed 5-6 replicates of six

snails each in at least five lest concentrations plus a negative control. Each treatment was tested

three independent limes Nominal test concentrations were in the range of 25-300 mg/L PFOS.

The test temperature was maintained at 25ฑ2ฐC. Water quality parameters including pH,

conductivity, and DO were reported as having been measured at the beginning and end of each

test, but the information was not reported. Survival of negative control animals was also not

reported. The study author reported 96-hour LCso was 178 mg/L (C.I. 167-189) and represented

an average of the LCsos for each test. Only one of three independent experiments could be fitted.

A-49


-------
The independently-calculated LCso value was 183.0 (161.4 - 204.7) mg/L and was used
quantitatively to derive the draft acute water column criterion.

Funkhouser (2014) conducted a 96-hour static test on PFOS (potassium salt, >98%
purity) with the physid snail, Physella heterostrophapomilia (Note: formerly known as Physa
pomilia), as part of a Master's thesis at the Texas Tech University, Lubbock, TX. Adult P.
pomilia (4 month old) used for the test were field collected from two different collections from
the North Fork of the Double Mountain Fork of the Brazos River near I .uhhock, TX. Offspring
from both collections were reared in 12, 10-gallon aquaria with lab water for se\ eral generations
prior to use in the test. Dilution water was moderately hard reconstituted laboratory water (3.0 g
CaSC>4, 3.0 g MgSC>4, 0.2 g KC1, and 4.9 g NaFtCCb added to 50 L deionized water).

Photoperiod was 12-hr:12-hr (light:dark) al an unreported light intensity PFOS was dissolved in
dilution water to prepare the test concentrations Exposure \ essels were 400 mL polypropylene
containers containing Zoo ml. of test solution. The test employed two replicates of four snails
each in six test concentrations plus a negative control. Nominal concentrations were 0 (negative
control), 100. I5<). 2<)(). 25<). 3<)(). and 375 nig I. Exposure concentrations were reportedly
measured. Init concentrations were not reported. Analyses of test solutions were performed using
liquid chromatography/ tandem mass spectrometry (LC-MS/MS). Standards were used as part of
the analytical method, but detai is were not reported. The reporting limit was 0.010 mg/L.
Experiments were conducted in incubators set to 25ฐC, which did not vary more than 1ฐC during
the course of the studies. No other water quality parameters were reported as having been
measured in test solutions. Negative control survival was not reported specifically for the test,
but was reported to be 85-100% across all experiments. The author reported 96-hour LCso was
reported as 161.77 mg/L. The independently-calculated toxicity value could not be calculated at

A-50


-------
this time given the level of data that was presented in the paper. The study author reported value
was used quantitatively to derive the draft acute water column criterion.

A-51


-------
Appendix B Acceptable Estuarine/Marine Acute PFOS Toxicity Studies

B.l Summary Table of Acceptable Quantitative Estuarine/Marine Acute PFOS Toxicity Studies

Species (life's!;iiicI

Method'

Test
Diimlion

( licniiciil /
I'llril

pll

Temp

(ฐC)

Sidinilt
(ppl)

ll'lccl

Anllior
Kcporicd
i riw-i
Cone.

diiii/l.)

i:p\

( iilculiilod
ll'lccl
(one.

I'iiiiil

i.iivci

Cone.
(ill"/!.)1'

Spocics
Mean
Acnlc
VidllC

Reference

Sea urchin (larvae),

Paracentrotus lividus

s,u

72 hr

PFOS
Unreported



18

35

LC50

(malformation)

1 "ซJ5

-



1 _ซ>5

((Hindu/ el

al. 2013;



Purple sea urchin
(embryo),

Strongylocentrotus
purpuratus

S, M

96 hr

PFOS-K

98%

-

15

30

1!( 5u

(normal
development)

1.7

-

1.7

1.7

Hayman et
al. (2021)



Mediterranean mussel
(larva),

Mytilus galloprovincialis

S,U

48 hr

PFOS
Unreported

7.9-
8.1

l(.



i:( 5ii

(malformation)

>1

-

>lc

-

(Fabbri et
al. 2014)

Mediterranean mussel
(embryo),

Mytilus galloprovincialis

S, M

4X hr

\>\( >S-k

-

15

3d

IX '50

(normal
development)

1.1

-

1.1

1.1

Hayman et
al. (2021)



Mysid (3 d),

Americamysis bahia

S, M

<>(. hr

H< )S-K

-

:u

}(. hr

H'< )S

-

:u

-

LC50

6.9

-

6.9

6.9

(Mhadhbi
etal. 2012)



Sheepshead minnow
(3.0 cm, 0.44 g),

Cyprinodon variegatus

R, M

hi'

I'FOS-K
86.9".,

-

22

20

LC50

>15

-

>15

>15

Palmer et
al. (2002b)

a S=static, R=renewal, F=flow-through, U=unmeaMiicd. \1 =mca.surcd, T=total, D=dissolved, Diet=dietary, MT=maternal transfer
b Values in bold used the in the SMAV calculation
0 Not used in SMAV calculations, because a definitive \ aluc is ;i\ ;iilable

B-l


-------
B.2 Detailed PFOS Acute Saltwater Toxicity Study Summaries and

Corresponding Concentration-Response Curves (when calculated for
the most sensitive genera)

The purpose of this section was to present detailed study summaries for acute saltwater
tests that were considered quantitatively acceptable for criteria derivation, with summaries
grouped and ordered by genus sensitivity. The data available for saltwater invertebrates fulfilled
three of the eight MDRs. EPA could not, therefore, develop acute estuari lie/marine criteria
following the 1985 Guidelines methods. In the interest of providing recommendations to
states/tribes on protective values, EPA developed an estuarine/marine acute benchmark using the
available empirical data supplemented with toxicity values generated through the use of New
Approach Methods, specifically through the use of the EPA Office of Research and
Development's peer-reviewed publicly-a\ailaMe uchTCF. tool (Raimondo et al. 2010). These
benchmarks are provided in Appendix L.

B.2.1 Most Sensili\ e l-suiarine Marine (ienns lor Acute Toxicity: Mytilus (mussel)

The acute toxicity of peril uorooclane sulfonate (PFOS, purity not provided) on the

Mediterranean mussel. Myiilns^alloprovmcialis was evaluated by Fabbri et al. (2014). This

species is not resident to North America, but is a surrogate for North American mussel species,

including the widespread, commercially and ecologically important blue mussel, Mytilus edulis.

Sexually mature mussels were purchased from an aquaculture farm in the Ligurian Sea (La

Spezia, Italy) and held lor two days for gamete collection. Gametes were held in artificial sea

water (ASW) made of analytical grade salts and at a constant temperature of 16 ฑ1ฐC. It was

assumed that the gametes were held at the same environmental conditions as the adults, so test

salinity was assumed to be 36 ppt with a pH of 7.9-8.1. Embryos were transferred to 96-well

microplates with a minimum of 40 embryos/well. Each treatment had six replicates. Embryos

were incubated with a 16-hr: 8-hr light: dark photoperiod for 48 hours and exposed to one of six

B-2


-------
nominal PFOS concentrations (0.00001, 0.0001, 0.001, 0.01, 0.1, 1 mg/L) or controls. The PFOS
stock was made with ethanol, and ASW control samples run in parallel included ethanol at the
maximal final concentration of 0.01%. Each experiment was repeated four times. At test
termination (48 hours), the endpoint was the percentage of normal D-larvae in each well,
including malformed larvae and pre-D stages. The acceptability of test results was based on
controls for a percentage of normal D-shell stage larvae, >75% (ASTM 2004a). Authors noted
that controls had >80% normal D-larvae across all tests. PFOS was only measured once in one
treatment which was similar to the nominal concentration; that is, 0.000085 mg I. \ ersus the
nominal concentration of 0.0001 mg/L. PFOS was below the limit of detection in the control
ASW (0.06 ng/L or 0.00000006 mg/L). The percentage of normal D-larva decreased with
increasing test concentrations. The NOEC and I .OliC reported lor the study were 0.00001 and
0.0001 mg/L, respectively. However, the lest concentrations failed to elicit a 50% reduction in
malformations in the highest test concentration, and an l-(';.. was not determined. Therefore, the
ECso for the study was greater than the highest test concentration (1 mg/L). The 48-hour EC50
based on malformation of I mg I. was acceptable for quantitative use.

I l:i\ 111:111 ol ;il. (2021) report the results of a 48-hour static, measured test on the effects
of PFOS-k (potassium salt. CAS # 2795-39-3, 98% purity, purchased from Sigma-Aldrich, St.
Louis, MO) 011 the Mediterranean mussel, Mytilus galloprovincialis. Authors note tests followed
U.S. EPA (1995) and ASTM (2004a) protocols. Mussels were collected in the field (Sand Diego
Bay, CA) and conditioned in flow-through system at 15ฐC. Mussels were induced to spawn by
heat-shock and approximately 250 embryos (2-cell stage) were added to 20 mL borosilicate glass
scintillation vials with 10 mL of test solution. There were five replicates per test concentration.
Test conditions were 30 ppt, 15ฐC and a 16-hr:8-hr light:dark photoperiod. Six test solutions

B-3


-------
were made in 0.45 jam filtered seawater (North San Diego Bay, CA) with PFOS-K dissolved in
methanol. The highest concentration of methanol was 0.1% (v/v). Measured test concentrations
ranged from 0.52 - 2.50 mg/L. Controls were made in the same seawater and the acute test also
included a solvent control. At test termination (48 hours), larvae were enumerated for total
number of larvae that were alive at the end of the test (normally or abnormally developed) as
well as number of normally-developed (in the prodissoconch "D-shaped" stage) larvae. There
were no significant differences between solvent control and filtered seawater. suggesting no
adverse effects of methanol. The author reported 48-hr ECso, based on normal de\ clopment, is
1.1 mg/L PFOS. EPA was not able to independently calculate a 48-hour ECso \alue as the curve
fitted model did not result in a good fit. Therefore, the author-reported ECso 1.1 mg/L mg/L was
considered for quantitative use.

B.2.2 Second Most Sensitive Estuarine/Maiine (ienus for Agile Toxicity: Stronevlocentrotus

(sea urchin)

Hayman ol al. (2021) report the results of a 96-hour static, measured test on the effects
of PFOS-K (potassium sail. CAS	lW(. purity, purchased from Sigma-Aldrich, St.

Louis, MO ) on the purple sea urchin. Simiii*vloccinmiiispurpuratus. Authors note that tests
followed I S l-IW (Iiw5)and ASTM (2<)i)4) protocols. Sea urchins were collected in the field
(Sand Dieuo Bay, CA) and conditioned in flow-through system at 15ฐC. They were induced to
spawn by KC1 injection and approximately 250 embryos (2-cell stage) were added to 20 mL
borosilicate glass scintillation vials with 10 mL of test solution. There were five replicates per
test concentration. Test conditions were 30 ppt, 15ฐC and a 16-hr:8-hr light:dark photoperiod.
Seven test solutions were made in 0.45 |im filtered seawater (North San Diego Bay, CA) with
PFOS dissolved in methanol. The highest concentration of methanol was 0.1% (v/v). Measured
test concentrations ranged from 0.52 - 10.0 mg/L. Controls were made in the same seawater and

B-4


-------
the acute test also included a solvent control. At test termination (96 hours), the first 100 larvae
were enumerated and observed for normal development (4-arm pluteus stage). There were no
significant differences between solvent control and filtered seawater, suggesting no adverse
effects of methanol. The author reported 96-hour ECso, based on normal development, is 1.7
mg/L PFOS. EPA was not able to independently calculate a 96-hour EC?o value as the curve
fitted model did not result in a good fit. Therefore, the author-repoi'icd F.C-> of 1.7 mg/L mg/L
was considered for quantitative use.

B.2.3 Third Most Sensitive Estuarine/Marine Genus for Acute Toxicity: I'amcciiirolus (sea

urchin)

A 72-hour static, unmeasured PFOS (purity not pro\ided) toxicity test \\\ill the sea
urchin, Paracentrotus lividus (a non-North American species) was conducted by Gunduz et al.
(2013) Adult sea urchins were collected from the Aegean coast of Turkey, in an area the authors
noted as clean and lacking domestic and industrial wastewater inputs. Filtered natural seawater
from the same area was used as the dilution water. Adult sea urchins were cultivated in the same
filtered natural sea water with a salinity of .15 ppt and 18ฐC. Zygote suspensions (1 mL) were
added to the controls or ^ nil. of the \ arious PI'OS treatments. This ensured that there were about
30 fertilized embryos ml. or approximately 3<)o cmlnyos per treatment. The experiments were
conducted in six-well TPP culture plates with six replicates per treatment. PFOS stock solutions
were made ^ ith dimethyl sulfoxide (DMSO) and diluted with seawater to obtain five nominal
treatments (0.5, I <). o. 50 and 10 mg/L PFOS). In addition to a natural seawater control,
experiments also included a DMSO solvent control equal to the amount in the highest test
concentration. The embryos were incubated in a growth chamber at 18 ฑ2ฐC for 10 minutes after
fertilization up to 72-hour pluteus larval stage. At test termination, 100 individuals were selected
randomly from each treatment and evaluated for normal plutei, retarded plutei, pathologic

B-5


-------
malformed plutei, pathologic embryos unable to differentiate up to the pluteus larval stages and
dead embryos/larvae. There was 97.75% and 91% frequency of normal larvae in the control and
solvent control, respectively with no deaths reported in the controls or any PFOS treatments. The
72-hour ECso based on normal development to the pluteus stage was 1.795 mg/L PFOS and was
acceptable for quantitative use; however, additional consideration needs to be given to the short
test duration.

B.2.4 Fourth Most Sensitive Estuarine/Marine Genus for Acute Toxicity . \111cricamvsis

(mysid)

Hayman et al. (2021) report the results of a 96-hour static, measured test 011 the effects
of PFOS (potassium salt, CAS # 2795-39-3, 98ฐ/. puriiy. purchased from Sigma-A Id rich, St.
Louis, MO) on the mysid, Americamysis bahia. Authors note that tests followed U.S. EPA
(1995;, 2002) and (ASTM 2004a) protocols \l\ sids were purchased iVoin a commercial supplier
(Aquatic Research Organisms, Hampton, IS 11) and acclimated to test conditions (30 ppt, 20ฐC
and a 16-hr:8-hr light dark phoiopcriod) I'ive test solutions were made in 0.45 |im filtered
seawater (North San Dieuo Bay. ( A) with PI'OS-K dissolved in methanol. The highest
concentration of methanol was 0 l"n (\ \ ) Measured test concentrations ranged from 0.95 - 16
mg/L. Controls were made in the same seawater and the acute test also included a solvent
control I i\ e m\ sids (3 days old, which is older than the typical age of < 24 hours at test
initiation) were added to 12<> 111L polypropylene cups and 100 mL of test solutions with six
replicates per treatment I ,i\ ing mysids were counted and dead organisms were removed daily.
There were no significant differences between solvent control and filtered seawater, suggesting
no adverse effects of methanol. Only two organisms were found dead in the controls at test
termination. The author reported 96-hour LC50 is 5.1 mg/L PFOS. The independently-calculated
96-hr LC50 value was 4.914 (3.578 - 6.250) mg/L and is acceptable for quantitative use.

B-6


-------
B.2.5 Fifth Most Sensitive Estuarine/Marine Genus for Acute Toxicity: Siriella (mysid)

Mhadhbi et al. (2012) performed a 96-hour static, unmeasured acute test with PFOS

(98% purity) and the mysid, Siriella armata. A stock solution of PFOS was made either with
filtered sea water from the Ria of Vigo (Iberian Peninsula) for low exposure concentrations, or
with DMSO for high PFOS concentrations (a final maximum DMSO concentration of 0.01%
(v/v) in the test medium). However, the authors did not indicate w hat was considered a high-test
concentration. If DMSO was used, a solvent control was also included M\ sids were exposed to
one of five nominal PFOS treatments (1.25, 2.5, 5, 10 and 20 mg/L). Mysids were also collected
from the same source as the dilution water and quarantined before use in 100 I. plastic tanks with
circulating sand-filtered seawater. The adult stock was led daily and maintained al laboratory
conditions (17-18ฐC, salinity between 34 4-35 ^ ppt, and oxygen (•> mg/L). Twenty neonates (<24
hours old) were used per each treatment. To |iiv\ cut cannibalism, a single individual was added
to each glass vial with 2-4 niT. of test solution. Vials were incubated at 20ฐC with a 16-hour light
period. Neonates were led I <>-1 5 . \ricmia salina nauplii daily and mortality was recorded after
96 hours. The 90-hour I.O.. was(vl) muLPFOS and was acceptable for quantitative use.

B.2.6 Sixth Most Scnsili\c l-stuaiine Marine Genus for Acute Toxicity: Cyprinodon

(sheepshead minnow)

Palmer et al. (2UU2h) conducted a 96-hour static-renewal measured acute test with
PFOSK (peril uorooctanesuI Ibnate potassium salt, 86.9% purity from the 3M Company) on the
sheepshead minnow. ('vprinodon variegatus. The test followed standard guidance for acute
toxicity tests outlined in U.S. EPA (1985, 1996) and (ASTM 1994). Sheepshead minnows were
purchased from a commercial supplier (Aquatic Biosystems, Fort Collins, CO) and held for
several weeks prior to testing. Fifty-one hours before testing fish were acclimated to test
conditions (16-hr:8-hr light:dark photoperiod, salinity of 20 ppt and 22ฐC). Natural seawater
(Indian River Inlet, Delaware) was filtered and diluted with well water to 20 ppt and was used

B-7


-------
for culturing and testing. A nominal PFOS stock solution (40 mg/L) was made by dissolving
PFOS in methanol and diluting it with seawater to achieve the nominal test concentration (20
mg/L). A solvent control (0.5 mL/L methanol) and a sea water control were also included. Ten
minnows (3.0 cm, 0.44 g) were added to 25 L polyethylene aquaria with 15 L of test solution
(loading was 0.30 g fish/L of test water). Test treatments were replicated three times. PFOS
concentrations were measured daily at test solution renewal with a\ eraged measured
concentrations in the control and solvent control less than the limit of <.|iianlilieation (5 mg/L)
and PFOS-spiked seawater, 15 mg/L. At test termination (96 hours) none of llie minnows died in
any of the test treatments, therefore the author reported I C™ was > 15 mg/L. I ฆ l\\ u as unable to
independently calculate the LCso value as this test only consisted of one treatment group. As such
the author-reported LCso >15 mg/L is acceptable lor quantitali\ e use based on the 2013
Ammonia rule which states that greater than high \ allies can be used in the derivation of criteria.

B-8


-------
Appendix C Acceptable Freshwater Chronic PFOS Toxicity Studies

C.l Summary Table of Acceptable Quantitative Freshwater Chronic PFOS Toxicity Studies

Species (

Method'

1 osl
Dui'iilioii

Chciniciil /
PuriU

nil

Temp.

(ฐC)

('limine Yiiluc
llndpoinl

Author
Kcpnrled
Chronic
Value
(inii/l.)

i:p\

( iilculiilcd
Chronic

\ illllC

(iiiU/l.)

l-'iiiiil
(hidiiic
N'iiluc

(ill"/!.)*

Species
Mesin
Chronic
\ iilue
(mป/l.)

Reference

Fatmucket (adult),

Lampsilis siliquoidea

R,M

36 d

PFOS

>98%

,r

00

14.6-
16.1

\1 VTC

(metamorphosis success)

0.017(>S

98%

-

25

EClu

(clutch size)

14.14

8.831

8.831

8.831

(Funkhouser
2014)



Rotifer (<2 hr old neonates),

Brachionus calyciflorus

R,Ub

Up to
158 hr

PFOS
>98",,

-

20

LOIX

(reduced net reproductive
rate)

0.25

-

0.25

0.2500

(Zhang et al.
2013)



Cladoceran (6-12 hr),

Daphnia carinata

R, U

21 d

H< >S-k

W„

-

21

MATC

(days to first brood)

0.003162

-

0.003162

0.003162

Logeshwaran
et al. (2021)



Cladoceran (<24 hr),
Daphnia magna

R, M

:i d

PFOS-K
'ซ> 4<>%

S 1-
S 5

l<> 4-

2U.1

EC10

(survival)

16.97

11.19

11.19

-

Drottar and

Krueger

(2000i)

Cladoceran (<24 hr),
Daphnia magna

R,U

:i d

PI ()S-K

95ฐ,,

-

:i

EC10

(survival)

35.36

16.35

16.35

-

(Boudreau
2002;

Boudreau et al.
2003a)

Cladoceran (<24 hr),
Daphnia magna

R, U

:i d

PFOS
Unreported

-

21

EC10

(number of young/adult)

1.768

0.7885

0.7885

-

(Ji et al. 2008)

Cladoceran (<24 hr),
Daphnia magna

R, U

21 d

PFOS-K

>98%

-

20

EC10

(total neonates/female)

2.236

2.919

2.919

-

(Li 2010); {Lu,
2015 #325}

Cladoceran (<24 hr),
Daphnia magna

R, M

21 d

PFOS-K

99%

7

22

EC10

(reproduction)

2.26

-

2.26

-

(Yang et al.
2014)

Cladoceran (<24 hr),
Daphnia magna

R, U

21 d

PFOS
98%

7.2

20

EC10

(number of
offspring/brood/female)

0.0179

0.001712

0.001712

-

Li (2009)

C-l


-------
Species (lilVst;iiio>

Method'

1 osl
Dui'iilioii

( hcmiciil /
Piiriu

pll

Temp.

(ฐC)

Chronic \ iilue
I'lndpoim

Author
Reported
Chronic
Yiiluc

(in Si/I.)

i:p\

( iilculiilcd
Chronic
Yiiluc
(iiiU/l.)

liiiiil
Chronic
Yiiluc

(msi/l.)*

Species
Mcsin
(hidiiic
Yiiluc
(mป/l.)

Reference

( ladncciaii ( 24 hi ).

Daphnia magna

k. 1

21 d

\'\:( >S-k

"

:u

i:( iu

(survival)

5.t>57

3.596

3.596

-

( Liang el al

2017)"

Cladoceran (12-24 hr),

Daphnia magna

R, U

21 d

PFOS-K

98%

-

20

EC10

(growth-length)

(is: 183

0.9093

0.9093

-

Yang et al.
(2019)

Cladoceran (<24 hr),
Daphnia magna

R, U

21 d

PFOS

>99%

7.5

23

MATC

(number of young)

1.58 1

-

1.5815

1.267

Seyoum et al.
(2020)



Cladoceran (<24 hr),
Moina macrocopa

R, U

7 d

PFOS
Unreported

-

25

IC10

(number of young starling
adult)

<0.3125

u. 1789

0.1789

0.1789

(Ji et al. 2008)



Crayfish

(4 wk juvenile, 0.056 g),

Procambarus fallax f.
virginalis

R, M

28 d

PFOS-K

>98%

-

25

i.( :u

0.1670

-

0.1670

0.1670

(Funkhouser
2014)



Blue damselfly (nymph),

Enallagma cyathigerum

R, U

320 d

Perfluorooctanes

1 r '1

tetraethylammoni

" 5

:i

MATC

(survival at 150 days)

0.03162

-

0.03162

0.03162

(Bots et al.
2010)



Midge

(newly hatched larva),
Chironomus dilutus

R, M

10 d

\>\( >S-k
95".,

-

:i-:^

EC10

(growth at 10 days)

0.04920

0.05896

0.05896

-

(MacDonald et
al. 2004)

Midge (4-day old larvae),
Chironomus dilutus

R, M

16 d

H'< )S
 d

PFOS
98ฐ.,

-

5.U-7.0

LOEC

(growth - weight and
length)

>0.1

-

>0.1

>0.1

(Spachmo and
Arukwe 2012)



Zebrafish (8 hpf),

Danio rerio

R, U

Life-cycle

I'FOS

96%

7.0-
7.5

28

EC10

(F1 offspring: % survival)

0.01581f

0.01650

0.01650

-

(Wang et al.
2011)

Zebrafish (male, 3-5 mo),
Danio rerio

R, U

21 d

PFOS
Unknown

7.0-
7.4

28

EC10

(mean body length)

0.05657

0.06274 0.06274

0.03217

Guo et al.
(2019)



C-2


-------
Species (lilVst;iiio>

Method'

1 osl
Dui'iilioii

Chcniiciil /
Piiriu

I'"

Temp.

(ฐC)

('limine Yiiluc
lliidpoint

Author
Reported
Chronic
Yiiluc

i:p\

( iileuliiled
( hrnnic
Value
(inii/l.)

l-'iiiiil
Chronic
Value

I mg/1.)*'

Species
Mesin
Chronic
Value

Reference

Falhead imimow
(embryo, 48 hpf),

Pimephales promelas

F, M

33 d

PFOS-K
Unknown

6.6-
7.3

22-26

EC10

(survival)

1

0.4408

0.4408

-

3MCompany
(2000)

Fathead minnow
(embryo, <24 hpf),

Pimephales promelas

F, M

47 d

PFOS-K
90.49%

8.2

24.5

EC 10

(survival)

04:4^

u.4732

0.4732

-

Drottar and

Krueger

(2000j)

Fathead minnow (adult),

Pimephales promelas

F, M

21 d

PFOS

>98%

7.3

25

1C10

(fecundity)

0.4794

o ()5 101

0.05101

-

(Ankley et al.
2005)

Fathead minnow
(adult, 5 mo.),

Pimephales promelas

R, M

42 d

PFOS-K

>98%

7.9

24.96

i:( iu

(F1 larval growth - weight)

0.06223

0.0549

0.0549

0.1555

Suski et al.
(2021)



Swordtail fish
(juvenile female),
Xiphophorus helleri

R, U

90 d

PFOS-K

>98%

-

27

EC 10

(female survival)

>0.1

0.5997

0.5997

0.5997

(Han and Fang
2010)



Northern leopard frog (stage
8/9 embryo),

Lithobates pipiens

F, M

35 d

PFOS-K

-

:u

LC50

6.210

-

6.21



(Ankley et al.
2004)

Northern leopard frog (stage
8/9 embryo),

Lithobates pipiens

F, M

112 d

\>\( >S-k

-

:u

MATC

(growth - length)

1.732

-

1.732

-

(Ankley et al.
2004)

Northern leopard frog (larva,
Gosner stage 26),

Lithobates pipiens

R, M

40 d

H '< )S

_W„

-

:u

MATC

(Gosner stage at 40 d)

0.0316

-

0.03162

-

(Hoover et al.
2017)

Northern leopard frog (larva,
Gosner stage 26),

Lithobates pipiens

R, M

40 d

I'lOS

_W„

-

:o

LOEC

(growth - snout-vent
length)

>1

-

>1

1.3161

(Hoover et al.
2017)



African clawed frog
(larvae, NF 46/47 - 5 dpf),

Xenopus laevis

R, M

4 mo

PFOS
98ฐ,,

6.5-
7.0

22

LOEC

(survival, weight, sex
ratio/intersex)

>1

-

>1

>1

(Lou et al.
2013)



C-3


-------
Species (

Method'

1 osl
l)n ration

( hornic;i 1 /
PuriU

pll

Temp.

(ฐC)

('limine Yiiluc
llndpoint

Author
Reported
Chronic
Ysilue

i:p\

( iilculiiled
Chronic
Ysilne

l-'iiiiil
Chronic
Ysilne

I mg/l.)''

Species
Mesin
( hronic
Ysilne
(mป/l.)

Reference

Clawed frog
(embryo, NF 10),
Xenopus tropicalis
(formerly, Silurana
tropicalis)

F, M

150 d post
metamorp
ho sis

PFOS

>98%

7.5

26

MATC

(weight at metamorphosis)

(1 "S_l

-

0.7871

0.7871

Fort et al.
(2019)

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D=dissolved. Diei-dioiary, MT=matcrnal uanslcr

b Chemical concentrations made in a side-test representative of exposure and verified stability of concentrations of PFOS in I lie ranue of concentrations tested under similar
conditions. Daily renewal of test solutions.

0 Values in bold used in SMCV calculation. SMCVs are calculated as the geometric mean of all Md-I'acal \ allies for a species. Sec s>eclion 2.10.3.2 (Chronic Measures of Effect)
for decision rules regarding use of greater (>) and less than (<) values in SMCV calculations

C-4


-------
C.2 Detailed PFOS Chronic Freshwater Toxicity Study Summaries and
Corresponding Concentration-Response Curves (when calculated for
the most sensitive genera)

The purpose of this section was to present detailed study summaries for tests that were
considered quantitatively acceptable for criteria derivation, with summaries grouped and ordered
by genus sensitivity. C-R models developed by EPA that were used to determine chronic toxicity
values used for criterion derivation are also presented for the most scnsili\ e genera when
available. C-R models included here with study summaries were those lor the lour most sensitive
genera (consistent with Section 3.1.1.3). When required, EPA also included models for non-
resident species that were more sensitive than the fourth most sensitive North American resident
genus. In many cases, authors did not report concentration-response data in the
publication/supplemental materials and/or did not provide concentration-response data upon
EPA request. In such cases, EPA did not independently calculate a toxicity value and the author
reported effect concentrations were used in the derivation of the criterion.

C.2.1 Most Sensili\c I'leshwater (ienus for Chronic Toxicity: Chironomus (midge)

MacDonsild ol ;il. (2004) conducted chronic larval and life-cycle tests to determine the

effects of PI OS (potassium salt. l>5"n purity) on the midge, Chironomus dilutus (formally known

as Chironomus ten tans). The test was performed under renewal conditions over 10 days for the

larval test and Mi days for the life-cycle test. The tests followed the general guidance given by

EPA-600-R99-0M (1" S.I VA 2002) and ASTM E 1706-00 (ASTM 2002). These methods are for

measuring the toxicity and bioaccumulation of sediment-associated contaminants with

freshwater invertebrates and have different exposure durations than those typically considered

for invertebrate aqueous exposures, as well as different control survival requirements and

recommendations. C. dilutus used for the tests were 10-day old larvae for the 10-day exposure

and newly-hatched larvae at test initiation for the 20-day exposure in the life-cycle test. Dilution

C-5


-------
water was reconstituted hard water consistent with ASTM (2002) with unspecified total
hardness, but typically 160-180 mg/L as CaCCte, with alkalinity 110-120 mg/L as CaCCte, and
pH 7.6-8.0. The photoperiod was 16-hrs:8-hrs, light:dark. Light intensity was not reported. A
primary stock solution was proportionally diluted with dilution water to prepare the test
concentrations. Exposure vessels were 250 mL polypropylene beakers containing 240 mL of test
solution and a sediment substrate. The 10-day exposure test employed at least two replicates with
10 individuals all of which were obtained from four large C-shaped euu eases that were
distributed among seven test solutions plus a negative control. The life-cycle lest (20-day
exposure) employed 12 replicates of 12 midges each in five measured test solutions plus a
negative control. Nominal test concentrations for the 10-day test were 0 (negative control),
0.001, 0.005, 0.010, 0.020, 0.040, 0.080, n I5<) mu I. The nominal test concentrations for the 20-
day exposure were 0 (negative control), 0.001, 0.005. nolo, n < >50, and 0.100 mg/L. Mean
measured concentrations for the I "-day test were 0 (LOO not reported), 0.0008, 0.00460, 0.0115,
0.241, 0.0491, () " I 5<)| mu I., respectively. Mean measured concentrations for the 20-day
exposure were 'Ml.OO not reported). <> oiP.v 0 n 144, 0.0217, 0.0949, and 0.149 mg/L,
respecti\ ely Analyses of test solutions were performed using LC-MS. The mean percent
reco\ cry and detection limits were not reported. Measured values of test concentrations in the
20-day exposure w ere 2 to 2 5-fold higher than nominal concentrations. Temperature and D.O.
concentrations were measured in at least two replicates per treatment on a daily basis for the 10-
day test and up to day 20 in the life-cycle test. Afterwards they were measured every other day
(on alternate days from test solution renewal) from days 21 to 60 for the life-cycle test. The
frequency of monitoring was reduced during this period, because both parameters consistently
remained within acceptable ranges (21.0-23.0ฐC; D.O. >5.0 mg/L). Survival of negative control

C-6


-------
animals was >75%, which was considered acceptable for a full life-cycle exposure per ASTM
(2002). The study authors reported ECios and NOECs; however, specific details pertaining to the
curve fitting process (including statistical output from the models and the curves) were not
provided in the paper and therefore, limit independent interpretation of the toxicity values.

The observed effects of PFOS on C. dilutes reported in the paper by the study authors
include survival and growth as weight (measured as mg of ash-five dry mass per individual) for
both the 10-day and 20-day exposure durations and emergence and reproduction over the 20-day
exposure duration. Significant reductions in larval weight were observed alter l<> days of
exposure to PFOS in the 0.0962 and 0.1501 mg/l. treatment groups (roughly 0.38 and 0.19 mg,
respectively) compared to control (roughly 0.88 mg). These differences resulted in roughly a
56.8 and 78.4% decline in midge weight in these treatment groups compared to those observed in
the control. In contrast, there were no significant differences reported for survival between any of
the PFOS treatments (with percent survn al ranging between roughly 69.1% in the highest
treatment group and l<)<)"„in the lowest) and the control (with roughly 100%) survival).

However, the authors noted that there was a >)ฐn decline of midge survival in the highest PFOS
treatment group with a measured concentration 0.1501 mg/L. The author reported 10-day growth
and sur\ i\ al lT.Ci-s for the study were 0.0492 and 0.1079 mg/L, respectively. The study authors
also reported \OI X's of 0.<)4l) I mg/L, LOECs of 0.0962 mg/L, and MATCs of 0.0687 mg/L for
both endpoints.

Similar to the 10-day exposure results summarized above, there was a general decline in
growth (as ash-free dry mass per individual) across the PFOS treatment groups (ranging roughly
between 29.2 and 47.2%> reduction compared to controls) in the 20-day exposure. However, only
the decline in the 0.0949 mg/L treatment group was significantly different (roughly 0.29 mg)

C-7


-------
compared to the control (roughly 0.89 mg) and there was not a concentration-response
relationship across the PFOS treatment groups. Additionally, midge survival was reduced after
20 days of exposure to PFOS in the 0.0949 and 0.149 mg/L treatment groups (29.2 and 0%
survival, respectively) compared to the control (75% survival). Survival was determined to be
not significantly different across the rest of the PFOS treatment groups (ranging roughly between
56.5 and 75% survival) compared to the control. However, it should be noted that there was a
25% decline in survival in the 0.0217 mg/L PFOS treatment group compared lo the control that
was determined not to be significantly different. The author reported 20-day l -Ci-s for growth,
survival, and total emergence were 0.0882, 0.0864, and <">893 mg/L, respectively. and the study
authors also reported NOECs of 0.0217 mg/L for growth and survival and < 0.0023 mg/L for
emergence, LOECs of 0.0949 mg/L for growth and survival and <> (>217 mg/L for emergence,
and MATCs of 0.0454 mg/L for growth and sui \ i\ al and <> <><>71 mg/L for emergence. Also, it
should be noted, the paper reports contrasting NOECs for 2'i-day survival. The text in the paper
stated that the "NOI-C was n <>2 I 7 mu I. and Table 2 of the paper stated 0.0949 mg/L. EPA
assumed the NOI-C in Table 2 of the paper was not correct and that 0.0217 mg/L was the correct
NOEC based on the data presented in l7igure 3 A of the paper. This assumption was applied to the
summary of the study results presented in this PFOS draft criteria.

Independent statistical analyses were conducted for both the 10-day and 20-day exposure
durations using data that were estimated (using Web plot digitizer) from the figures presented in
the paper. EPA could not fit a curve to independently verify the 10-day survival (due to a lack of
a specific sample size for this endpoint as the number of replicates was not stated in the paper;
however, the number of replicates was between two and four and EPA sought to obtain
clarification and treatment level data from the study authors) or the 20-day growth toxicity

C-8


-------
values (due to a lack of an observed concentration response for this endpoint). However, the
EPA-calculated 10-day ECio for growth was 0.05896 mg/L, which was slightly higher than the
growth-based ECio of 0.0492 mg/L reported in the paper. The 20-day ECios for larval survival
and emergence were 0.0171 and 0.0102 mg/L, respectively. The 20-day ECios were much lower
than those reported in the paper of 0.0864 and 0.0893 mg/L, respectively. The 20-day ECios for
survival and emergence were not considered to be reliable end points al this time given the
disparities in the calculated ECios and the level of data that was presented in the paper, which
made independent verification of the toxicity values less accurate. Specifically, lor the 20-day
survival endpoint, there appeared to be overdispersion (\ e . observed data display a larger
variability than would be expected given an assumed statistical distribution about the mean
response) in the data as it was presented in the paper (in Figure 3.\ of the paper), which adds
uncertainty around the independently-calculated I X'i- of <) <>171 mg/L and may explain the
disparity between the reported I X"i- and l-PA's independently-calculated value. As for the
emergence endpoint. there was a lack of a concentration-response relationship and there were
very similar levels of ohser\ ed effects (which ranged between 42.6 and 50.1%) despite the more
than nine-lbkl increase in the mid-range treatment concentrations (0.0023, 0.0144, 0.0217 mg/L,
respecti\ ely). I .astly, the toxicity \ allies from the observed effects from the 20-day exposure
were considered to he less certain given the relatively large difference between the nominal and
measured concentrations lor this test. The dosing of the 20-day exposure was more of a concern
than the 10-day exposure, which had measured concentrations that were much more in line with
the expected nominal concentrations. Thus, the 20-day survival and emergence endpoints were
not considered for quantitative use in the derivation of the chronic criterion. Instead, these

C-9


-------
endpoints were considered as supporting information until detailed replicate level data can be
obtained from the study authors.

The most sensitive endpoint from the remaining toxicity values that could be
independently-calculated was for 10-day growth with an ECio of 0.05864 mg/L. As mentioned in
the Bots et al. (2010) summary and in Section 4.1.1, the observed effects of PFOS on aquatic
insects appears to be consistent across the available data for chironomids and odonates.

However, Bots et al. (2010) did not measure the effects of PFOS on nymph growth and
therefore, the observed effects in MacDonald et al. (2004) on larval weight cannot he compared
across the two studies. The ECio of 0.05896 (0.0581 - 0 0612) mg/L for 10-day grow th was used
quantitatively to derive the chronic aquatic life criterion. The remainder of the toxicity values
were used as supporting information to corroborate the toxicity \ nine used to derive the
freshwater chronic criterion and to better understand the effects of PFOS on aquatic insects.

McCarthy ol al. (2021) conducted a 1'i-day sub-chronic toxicity test and a separate 20-
day (note, based on age of starting organisms, this test was actually 16 or 19 days of exposure)
toxicity test with PI OS (l)X"o purity, purchased from Sigma-Aldrich) on the midge, Chironomus
dilutns PI-OS stock solution was dissolved in reconstituted moderately hard water without the
use of a sol\ ent and stored in polyethylene at room temperature until use. Two chronic exposures
with PFOS were run. a 1'i-day and a 20-day exposure, following standard protocols (U.S.EPA
2000b) with slight modifications. The 10-day exposure was considered a range finding test, with
concentrations spaced by ~100x and only mortality measured, whereas the 20-day exposure
measured both survival and growth. The 20-day exposure is less than the recommended 65-day
full-life cycle method outlined in USEPA (2000b) and since exposures of midges started on day
two or four the actual exposure duration is only 16 or 19 days long. Exposure vessels for both

C-10


-------
experiments were 1 L high-density polyethylene beakers containing natural-field collected
sediment. The 10-day exposure had 60 mL of sediment and 105 mL of test solution and the 20-
day exposure had 100 mL of sediment and 175 mL of test solution. PFOS in test solutions was
added via pipette to the beakers with the tip just above the sediment substrate. Nominal test
concentrations for the 10-day and 20-day exposure were 0, 0.0004086, 0.33, 33, 100 and 350
mg/L PFOS and 0, 0.001, 0.005, 0.01, 0.05 and 0.1 mg/L PFOS. respectively. Egg cases were
obtained from outside suppliers (Aquatic Biosystems or USGS Columbia I ji\ ironmental
Research Center) and held for 10 days in the 10-day test or held for four days before testing in
the 20-day exposure (in test vessels). In the 20-day exposure the test organism age (Ibur-day old
larvae) was greater than the protocol recommendation ( 24 hour) because earlier experiments
had control survival issues (<70%). In both tests each beaker held 12 organisms with five
replicates per exposure treatment. Solutions were renewed e\ cry 48 - 72 hours in the 10-day
exposure and daily lor the 2<)-day exposure. Water samples of test concentrations were measured
on day one and day I') in the I "-day exposure and day 10, 15 and 20 in the 20-day exposure. In
the 10-day exposure measured test concentrations ranged from 7 - 62% of nominal. In the 10-day
exposure, the author-reported I.OI-C. based on mortality, of 0.4086 |ig/L (0.0004086 mg/L
PFOS) is reported as a nominal concentration. Mean PFOS concentrations in the 20-day
exposure ^ere " (control), <> <)<)0447, 0.00209, 0.0042, 0.0231 and 0.0463 mg/L PFOS. Percent
survival in the control and lowest test concentration were 77% with no survivors reported in the
highest two test concentrations. The most sensitive endpoint appeared to be survival with an
author-reported 16-day reported ECio of 1.36 |ig/L (0.00136 mg/L PFOS). Additionally, the
study authors reported ECios of 1.62 |ig/L (0.00162 mg/L PFOS) and 3.23 |ig/L (0.00323 mg/L
PFOS) for growth as mean biomass and mean weight, respectively. EPA was unable to

C-ll


-------
independently calculate ECios for survival and mean weight. However, the 16- to 19-day
independently-calculated ECio value for mean biomass was 0.0015879 (0.00118 - 0.00200)
mg/L PFOS. This independently-calculated ECio was acceptable for quantitative use and was
utilized in the derivation of the chronic freshwater criterion for PFOS.

ฃ.2.1.1 MacDonald et al. (2004) Concentration Response Curve - Chironomus (midge)

Publication: (MacDonald et al. 2004)

Species: Midge (Chironomus dilutus)

Genus: Chironomus

EPA-Calculated ECio: 0.05896 (95% C.I. 0.0577 - 0.0602) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

-2.6770

0.6384

-4.1933

0.0057

e

0.0805

0.0090

8.9243

0.0001

Concentration-Response Model Fit:

MacDonald et al. 2004
Chironomus dflutus

C-12


-------
C.2.1.2 McCarthy el al. (2021) Concentration Response Curve - Chironomus (midge)

Publication: (McCarthy et al. 2021)

Species: Midge Chironomus dilutus
Genus: Chironomus

EPA-Calculated ECio: 0.0015879 (95% C.I. 0.00118 - 0.00200) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

5.2881

1.0432

5.0693

0.0148

d

1.0372

0.0238

43.4942

2.675 e5

e

0.0024

0.0001

21.9936

0.0002

Concentration-Response Model Fit:

Chironomus dilutus
Log Logistic type 1,3 para

le-05	le-04	le-03	le-02

PFOS(mgL)

C.2.2 Second Most Sensitive Freshwater Genus for Chronic Toxicity: Lamysilis (mussel)
Hazelton (2013); Hazelton et al. (2012) conducted a test of the long-term effects of

PFOS (acid form, > 98% purity) on glochidia and juvenile life stages from the mussel Lampsilis

siliquoidea. To initiate the PFOS partial life-cycle test, brooding females were collected from

Perche Creek, Missouri and shipped over night to the test laboratory. The length of time between

collection from Perche Creek and shipment was not reported and authors were unable to recall

such details (R. Bringolf, personal comm.); however, EPA did not believe storage, shipping, and

handling compromised test results since study authors only relied on those mussels with >70%

glochidia viability. Dilution water was dechlorinated tap water. Mean total hardness (47.5 ฑ 9.2

C-13


-------
mg CaCCte/L) and alkalinity (34.8 ฑ4.1 mg CaCCb/L) were measured by titration twice weekly
(n = 8) prior to water changes. Replicates used for water quality measurements were changed
daily to allow measurements from all four replicates every four days. For all treatments, water
temperature ranged from 14.6 to 16.1ฐC, dissolved oxygen ranged from 6.1 to 7.3 mg/L, and pH
ranged from 7.6 to 8.5, but did not differ across treatments. Photoperiod and light intensity were
not reported. No details were provided regarding primary stock solution and test solution
preparation. The test exposed brooding glochidia (in marsupia) for 36 days lb I lowed by a 24-
hour exposure of free glochidia. Experiments were conducted in 3.8 L glass jars of unspecified
fill volume. The 36-day in marsupia exposure test employed four replicates incli\ idnally
containing single brooding females for each of the two PI OS treatment groups plus the control.
The in marsupia exposure was followed In a 24-hour free glochidia exposure consisting of a
factorial design, such that free glochidia from the control group of the in marsupia exposure
were divided between a control and the two PFOS treatments and the PFOS treatments were split
into control and the same PTOS treatment group as the in marsupia exposure. This factorial
design allowed for the comparison of PFOS effects in two different life stages. However, it
should he noted that glochidia were pooled from females within each in marsupia treatment
group, and thus the influence of parental effects could be a confounding factor that cannot be
separated from the PI'OS effects. Nevertheless, the influence of the potential parental
confounding factor was likely to be minimal compared to the effects of the PFOS exposures.

Nominal concentrations throughout the exposures were 0 (negative control), 0.001 and
0.100 mg/L. Mean measured concentrations were 0.00211 (negative control), 0.00452 and
0.0695 mg/L. Analyses of test solutions were performed at the U.S. EPA National Exposure
Research Laboratory in Research Triangle Park, NC using HPLC/MS. Two standard curves were

C-14


-------
used to quantify PFOS water concentrations during the experiment: low range (0.00005,

0.00025, 0.0005, 0.00075, 0.001, 0.0025, 0.005 mg/L) and high range (0.001, 0.005, 0.010,
0.025, 0.050, 0.100, 0.150 mg/L). Two replicate samples were measured at each standard
concentration. Accuracy (recovery) of PFOS in the low-range standard curve ranged from 89.5
to 123% (n = 7) and for the high-range standard curve accuracy was 85.3 to 123% (n = 7). Adult
mussel and glochidia survival in the negative control was 100% and 90%, respectively. The
study authors determined that the in marsupia exposure held the greatest weight of evidence and
explained 78% of the variability in the glochidia viability (A1C = 22843 . ir <) 7S) and 83% of
the metamorphosis success (AIC = 21955, wt = 0 S3). and therefore it appeared that the data
presented in the study are in terms of the in marsupia exposure alone and there are no data
presented in terms of the factorial design during the 24-hour live glochidia exposure.
Additionally, the specific treatment groups of the data presented in the paper are unclear in terms
of the factorial design dining the 24-hour free glochidia exposure (e.g., it is unclear if the data
presented in Figure 2 of the paper are lumped according to marsupial exposure, reducing seven
treatments to three, or if only the data in u hi ch the in marsupia and free glochidia exposures
were the same are presented)

The test resulted in an author-reported NOEC of 0.0695 mg/L, which was associated with
a 38%) reduction in the viability of free glochidia at 24 hours post removal from females, a point
when control viability of free glochidia was > 80% (author reported LOEC and MATC > 0.0695
mg/L). While a 38% reduction was observed at the NOEC (0.0695 mg/L) treatment group
compared to controls, authors reported this reduction was not statistically different from the
control. Over time, the study authors reported significant reductions in free glochidia survival
between three- and seven-days post removal from females, indicating a potential LOEC < 0.0045

C-15


-------
mg/L. However, it should be noted that the observed level of effect between the two PFOS
treatment groups (0.0045 and 0.0695 mg/L) were extremely similar despite the 15-fold
difference between treatment groups. Additionally, in accordance with the 2013 Ammonia
Aquatic Life Criteria and a study by Bringolf et al. (2013), only glochidia toxicity data within 24
hours and with survival of at least 80% in the control treatment would be considered (U.S. EPA
2013). These specific data requirements ensured that the related effects of PFOS exposure to the
viability of glochidia were consistent with environmental exposures during this short life stage
and also take the unique life cycle of mussels into account. Therefore, the chronic toxicity value
for viability of free glochidia at 24 hours following removal from females resulted in a NOEC of
>0.0695 mg/L, which is an uncertain value and indicated thai \ i ability of free glochidia at 24
hours was a less sensitive endpoint.

In contrast, the data presented in the paper lor metamorphosis success suggest a NOEC
of 0.0045 mg/L and a I.OI-C of <> <>(->l)5 mg I. The reduction in metamorphosis success at the
LOEC was estimated to lx- 35 4".. I lowever. as there were only two PFOS treatment groups and
the gap in these exposure concentrations is large (about 15-fold), EPA was not able to fit a
cun e to estimate an I X'iฆ ฆ in a manner similar to the other toxicity studies used to derive this
criterion Instead, both the use of an \1ATC and an estimated ECio were considered for the
chronic \alue An l-Cio was estimated by assuming the 0.0695 mg/L treatment represents an
EC35.4 and estimating the l-Cio using the exposure response slope from another PFOS toxicity
study focused on another mussel species (Perna viridis). Specifically, the chronic exposure of
Perna viridis reported by Liu et al. (2013), which is summarized in Section 3.1.1.4.1 and D.2.1,
was used to derive a ratio of EC10/EC35.4 levels from that study, which was: EC10/EC35.4 =
0.0033/0.0186 = 0.1774. Applying this ratio to Hazelton et al. (2012) yields an estimated EC10

C-16


-------
of 0.0123 mg/L. Given the similarity between this ECio and the author-reported MATC for
Hazelton et al. (2012), the MATC of 0.01768 mg/L was used to derive the chronic criterion for
PFOS. This MATC is currently used quantitatively to derive the draft chronic water column
criterion, and EPA hopes to further refine this estimated ECioby obtaining the treatment level
data from the study authors and exploring additional exposure response slopes from the PFOS
dataset.

C.2.2.1 Hazelton et al. (2012) Concentration Response ("nrve - I ani/>si/is lmussel)

Publication: (Hazelton et al. 2012)

Species: Fatmucket, Lampsilis siliquoidea
Genus: Lampsilis

EPA-Calculated ECio: 0.0123 mg/L

Concentration-Response Model Fit: Concentmiion-response data not available
Value used Quantitatively in Criterion: Author-reported MATC of 0.0177 mg/L

C.2.3 Third Most Sensitive Freshwater (ieiuis for Chronic Toxicity. Enallasma (damselflv)
Bots et al. (2010) conducted a 320-day purl in I lilc-cycle study under renewal test

conditions to look al the effects of PI-OS (leliaethylammonium salt, 98% purity) on the

damselfly Enallagma cyaihigernm. Test organisms were obtained by collecting mature female E.

cyathigemiii all from the same location near the edge of a fen (a groundwater fed wetland) in

northern Ik-luium. After collection, females were transported to the laboratory in small cages and

housed in opposition chambers for 24 hours before eggs were collected. E. cyathigerum used for

the test were new l\ -hatched nymphs at test initiation. Dilution water was dechlorinated tap water

that contained only a negligible concentration of PFOS (2.64 ng/L) and no other water quality

parameters from the tap water were provided other than pH >7.5. Photoperiod was 16-hrs:8-hrs,

light:dark in a climate room. Light intensity was not reported. Test solutions were prepared

taking purity into account. To start the test, a total of 18,552 eggs were distributed amongst 150

exposure chambers (i.e., petri dishes of unreported size and material type). The distribution of

C-17


-------
the total number of eggs consisted of the entire clutch from each of the 30 females being divided
into five subsamples, which were then randomly allotted to the various test solution; thereby
ensuring that each treatment group consisted of an even distribution of test organisms from the
30 females. After hatching, a total of 7,938 nymphs continued to be exposed (10 individuals per
cup of unreported size and material type). After 10 days, seven nymphs for every female and
treatment were monitored (resulting in a total of 741 nymphs) Nominal concentrations were 0
(negative control), 0.01, 0.1, 1.0, and 10 mg/L and the test concentrations were not measured. All
nymphs were housed (and presumably tested) in a climate room at 21 ฐC. Water quality (pH,
carbonate and total water hardness, O2, NO2, and NO; levels) was checked weekly using
standard aquarium tests, but values are not reported. Approximately 40% of the nymphs in the
control treatment died during the first 60 days and similar mortality levels were observed in the
other treatments. Additionally, it appears that control sui \ i\ al plateaued between 60 and 200
days, with 82.57'\. of the remaining nymphs in the control treatment surviving during this time,
indicating that sur\ i\ a I settled out during this phase of the experiment. The initial drop in nymph
survival can likely he attributed to the handling of the test organisms between the various phases
of the experiment This would explain the observed plateau between 60 and 200 days, as the
nymphs were not handled during this time. The observed control mortality in this test was
consistent with other odonate tests and excessive mortality of nymphs is typically expected
within the first 20<) days ui\ en the difficulty in maintaining odonates in a lab setting (Abbot and
Svensson 2007; Rice 2008). Therefore, the observed control survival for this study was
considered within the acceptable range for this species up to the 200-day exposure duration.
Further, the control survival observed in this study was largely consistent with the toxicity
testing guidelines for chironomids (requiring 70% control survival; ASTM 2002; U.S. EPA

C-18


-------
2002), which was currently the only test guidelines for an emergent aquatic insect as there
currently was no test guideline for odonates. Therefore, considerations regarding the use of these
data for chronic criterion derivation was based on best scientific judgement and were restricted to
the first 200 days of the experiment. After 200 days, nymph survival in the control and the PFOS
treatments decreased. This drop in survival likely coincided with metamorphosis. However,
control survival at the end of the exposure duration was only roughly 40% of the starting nymphs
and therefore, survival after 200 days of exposure were not considered quantitatively in the
derivation of the freshwater chronic criterion.

The observed effects of PFOS on E. cyathigerum reported in the paper In I he study
authors include decreased survival over the exposure duration and decreased metamorphosis
success. Nymph survival after five days did not differ between llie control, 0.01 and 0.100 mg/L
treatments and was significantly lower in the I <> and the I <> <> mu L treatments. After 10 days of
exposure, 80% of the nymphs in the 1.0 mg/L treatment and all nymphs in the 10 mg/L treatment
died. After 20 days of exposure, all nymphs in the 1.0 mg/L treatment died. However, there was
no observed statistical difference between the control and any of the other treatment groups
during this exposure time through I 2d days Between 120 and 250 days of exposure there was
not an ohser\ ed difference in survi\al between the control and the lowest treatment group (0.01
mg/L). In contrast, nymph sur\ ival in the 0.100 mg/L treatment group started to decrease
compared to the control and the 0.01 mg/L treatment group, with 60% survival in the control
compared to 48.5% survival in the 0.100 mg/L treatment after 150 days of exposure. This
decrease was statistically significantly different from controls. All nymphs in the 0.100 mg/L
treatment group died within 250 days of exposure. While nymph survival in the control was
roughly 40% at the end of the 320-day exposure duration, there was no observed difference

C-19


-------
between the control and the lowest treatment group of 0.01 mg/L. Lastly, the paper also reported
observed effects of PFOS on metamorphosis success stating that metamorphosis success was
lower with 75.5% success in the 0.01 mg/L treatment (the only treatment group to have nymphs
survive to this life stage) compared to the control with 92.5%. However, data for this observed
endpoint was not provided in the paper beyond the percentages observed in the control and 0.01
mg/L PFOS treatment group. The specific sample sizes for this end point were difficult to
ascertain from the paper as only total number of test organisms across all lest treatments was
provided.

As indicated in the summary of the resulls abo\ e. toxicity \ alues through the experiment
decline with exposure duration. EPA took all of the author reported toxicity values between 10
(which was considered to be the start of Ihe chronic exposure) and 200 days of exposure into
account. Independently-calculated ECio values could not he determined given the level of data
that were presented in the paper Author-reported toxicity \ allies after 10 days of exposure were
aNOEC of 0.1 mu I. and a I.()!ฆ(' of I 0 mg/L. The LOEC was associated with a 79% decrease
in nymph sui \ i\ al compared to the control at this time. This NOEC and LOEC resulted in a
MATC of n .1162 mu I. Author-reported toxicity values after 150 days of exposure were a
NOEC of n "I mg/L and a I.OlvC of <>. I mg/L. The LOEC was associated with a 19% decrease
in nymph sur\ i\ al compared to the control at this time. This NOEC and LOEC resulted in a
MATC of 0.031(->2 mu I. I .astlv, the authors also reported an NOEC of 0.01 mg/L for survival
and an LOEC of < 0.01 mg/L for metamorphosis success after 320 days of exposure. Both of
these toxicity values fell outside the 200-day exposure duration and were not considered for use
in the freshwater chronic criterion calculation since control survival at this point was low (40%)
and considered unacceptable for quantitative use. Additionally, there was insufficient data

C-20


-------
provided in the paper to evaluate the reported results for the endpoints at 320 days of exposure.
Therefore, these toxicity values were considered as supporting information (see Section 4.1.1
below) and only the toxicity values from 10 to 200 days of exposure range were considered
further for the criterion derivation.

The 150-day MATC of 0.03162 mg/L was similar to the author-reported 10-day and 20-
day survival and growth MATCs of 0.0687 and 0.0454 mg/L for chironomid (MacDonald et al.
2004), which was the only other emergent insect toxicity study in the PI OS chronic dataset (see
Section 4.1.1) and the test organisms at these exposure durations would likely he in similar life
stages (later development and about to undergo metamorphosis). And these later toxicity values
were therefore more comparable than the 10-day MATC of 0.3 162 mg/L, which was focused on
the effects of PFOS on a much earlier instar of odonatc (which has a much longer development
time and life span) in relation to the 20-day M ATC of <) 11454 mu L for chironomid. These results
indicated that PI OS effects to aquatic insects was likely similar (see Section 4.1.1 for more
details); however additional data are needed to fully understand the effects of PFOS. The MATC
for nymph sui\ i\ al at I 5<)-da\ reported aho\ e was used quantitatively to derive the chronic water
column criterion \ nine Additionally. LP A ran additional analyses with some of the other toxicity
values for /.. cyathigerum to understand the influence of this study on the overall chronic
criterion (see Section 4 2.2)

C.2.3.1 Bots ci al. (2010) Concentration Response Curve - Enallagma (damselfly)

Publication: (Bots et al. 2010)

Species: Damselfly, Enallagma cyathigerum

Genus: Enallagma

EPA-Calculated EC10: Not calculable, concentration-response data not available

C-21


-------
C.2.4 Fourth Most Sensitive Freshwater Genus for Chronic Toxicity: Danio (zebrafish)

Wang et al. (2011) evaluated the full life-cycle effects of PFOS (> 96% purity) on Danio

rerio via a static-renewal study that reported nominal exposure concentrations. This test

evaluated the effects of PFOS on a parental (F0) generation and included breeding trials to assess

the effects of PFOS on an offspring (Fl) generation exposed via maternal transfer. PFOS stock

solutions were prepared in 100% dimethyl sulfoxide (DMSO). Adult zebrafish (wild-type strain

AB) were raised and kept at standard laboratory conditions ol"2N (' with a l4-hr:10-hr light:dark

cycle in a recirculation system according to standard zebrafish culture protocols Water supplied

to the system was filtered by reverse osmosis (pFl 7.0-7.5), and Instant Ocean sail was added to

the water to raise the conductivity to a range of 450 to I .<>(>(> uS'cm (system water). Zebrafish

embryos were obtained from spawning adults in tanks overnight with a sex ratio of 1:1. Embryos

were collected within one hour after spawning and rinsed in embryo medium. High-quality 8-hpf

embryos were divided into four treatment groups: DMSO \ chicle control (0.01% v/v), and PFOS

concentrations of 0 1105. 0 050. and 0 25<) nig L. Embryos were first exposed to PFOS in a petri

dish (100 embryos/treatment) lor li\ e days w ithout media change, and all embryos hatched and

survi\ ed in this stage After li\ e days, the fish were transferred into 2 L tanks for the period of 5-

dpf to .in dpi', and after that were raised in ^ I. tanks (30 fish/tank) until the end of the

experiment. I 5') dpi" Fish were kept in a static system, and 50% water was renewed with freshly

prepared solutions e\ cry fi\ e days. Each tank was checked for morbid fish on a daily basis, and

water quality was monitored on a weekly basis. Feeding was initiated at day five. Between five

and 14 dpf, fish were fed three times daily with zebrafish larval diet (Aquatic Habitats), and after

14 dpf they were fed twice daily with freshly hatched live Artemia. The experiment was repeated

three times with embryos derived from different parental stocks. At the end of exposure period

(150 dpf or five months), all fish were checked for their sex. However, the method used for

C-22


-------
determining sex, as either external morphology or genetic testing, was not stated in the paper.
EPA assumed external morphology was used and concluded that the effects on sex ratio may not
be reliable since determining sex through external morphology in zebrafish is difficult. A
subsample of 10 male and 10 female fish from each batch were also measured for standard body
length and wet weight. Condition factor (K) was tabulated to determine their overall fitness, and
sperm motility in male F0 fish was also determined after chronic PI-OS exposure. The most
sensitive endpoint was F0 parental male sperm density with a chronic \ nine of <0.005 mg/L
PFOS. However, as sperm density was not typically considered an apical endpoint and the
reported effects of PFOS on sperm density did not translate to other reproductive effects (i.e.,
fertilization), this endpoint was not considered further in the derivation of the PFOS freshwater
chronic criterion. Instead, the most sensiii\ e apical endpoinl lor the I 'd generation was
considered to be male growth (length and weight) with an author reported MATC of 0.01581
mg/L PFOS. Howe\ er. I -PA was unaMe to fit a concentration-response curve with significant
model parameters lor these endpoinls. and therefore, were unable to independently verify the
reported toxicity \alne for the l'<> generation

Bleeding trials were also carried out to produce F1 offspring. Six different crosses were
employed between FO females and males to incorporate both the exposure of the same treatment
groups throughout and crosses between the control and highest treatment group. Specifically, for
the components with consistent treatment groups throughout the experiment, females were paired
with males in the same treatment group (DMSO control or PFOS-exposed concentrations of
0.005, 0.050, and 0.250 mg/L). For the crosses between the control and the highest treatment
group, some females from the 0.250 mg/L PFOS treatment group were paired with males from
the DMSO controls, and some females from the controls were paired with males from the 0.250

C-23


-------
mg/L PFOS treatment group. For each of these crosses, eight randomly selected female fish were
paired with four male fish in two separate spawning tanks with four females and two males per
tank. Spawning was induced every other day for five days, and embryos were used for
monitoring their developmental progress. All eggs from each spawn were evaluated for
fertilization success. Percent fertilization was expressed as the number of fertilized eggs divided
by total number of eggs. Fifty fertilized embryos from each spaw 11 were further monitored for
continuous development. Percent hatch was calculated at 72 hpf. Larvae were also assessed for
their morphological appearance. Percent survival was monitored until 8 dpi". Snr\ i\ ing larvae at
5 dpf with normal morphology were further subjected to behavior assessment (lai \ al swimming
speeds were recorded when they responded to a 70-minule dark to light, 10-minute for each
period, transition stimulation). Following llic receipt of treatment level data from the study
authors, EPA independently calculated an ECio value of <) n|65 (0.01267 - 0.02033) mg/L for F1
survival. While this I X' i- has some uncertainty given the wide spacing (lOx) of the treatment
concentrations, this toxicity \ al lie was supported In others in the PFOS toxicity literature (see
Section 4.4.2.1.1 and Appendix (i) This study and the ECio value forFl survival was considered
quantitati\ ely in the deri\ ation of the aquatic lile criteria, despite the use of renewal exposure
and nominal test concentrations.

Guoel ill. (2019) e\ aluated the chronic effects of perfluorooctane sulfonate (PFOS
solution of-40% in water purchased from Sigma-Aldrich) to AB strain zebrafish (Danio rerio)
males in a 21-day static-renewal, unmeasured study. Official test guidelines were not cited by the
authors. Approximately 3.5-month-old male adult zebrafish were purchased from Taiyuan fish
hatcheries in Shanxi Province, PR China. Prior to exposure, fish were acclimated for 15 days in a
flow-through dechlorinated tap water system (<1% mortality during the holding period) that had

C-24


-------
the following water qualities: pH: 7.0-7.4, temperature: 28 ฑ 1ฐC and a 14-hr: 10-hr light to dark
cycle photoperiod. The fish were fed a commercially available adult zebrafish compound feed
during both acclimation and exposure. Nominal concentrations of PFOS dissolved in
dechlorinated tap water were reported to be 0 (control), 0.02, 0.04 and 0.08 mg/L, and a total of
660 fish were divided equally among the four concentration groups. Three replicates were
present for each concentration group with each containing 55 fish Water quality throughout the
experiment was maintained with standards listed above, as well as the Ibllowing conditions:
dissolved oxygen of 5 - 6 mg/L and total hardness of 20.0 mg/L reported as CaCO i. Exposure
media was completely changed every three days, and aquaria were completely cleaned during
testing. On days 7, 14 and 21, 50 fish from each group were sacrificed, with 30 fish measured for
length and body weight, while the other 2<) dissected on ice to evaluate PFOS concentrations in
the liver. The test fish had a mean weight of 0.19 ฑ 0.03 g and a mean length of 2.5 ฑ 0.3 cm at
test initiation. On day se\ en the fish lengths ranged Ironi 2 cm to <3 cm for all groups, and
weights were >0.3 lo iMu lor the control and 0.02 mg/L exposure. However, the 0.3 g fish
weight for the <~>4 mu I. and <> <>N mg I. exposures were significantly different from the control.
At days 14 and 21. fish length of only the highest concentration (0.08 mg/L PFOS) was
significantly different from the control, and the same fish weight effect levels were observed at
14 and 21 days as those reported at seven days. Therefore, weight was the most sensitive
endpoint at 21 days, with a NTOEC and LOEC of 0.02 and 0.04 mg/L PFOS, respectively. No
LCso value was reported due to lack of mortality (no mortality occurred for 21 days). However,
an independently-calculated EC10 could not reliably be estimated for mean body weight as the
data were sparse, was inconsistent with the author-reported toxicity values, and the confidence
bands were wide. Therefore, EPA's independently-calculated EC10 based on mean body length

C-25


-------
(in cm) at 21 days is 0.06274 (0.06229 - 0.06318) mg/L PFOS and was used quantitatively to
derive the draft chronic water column criterion for freshwater.

C.2.4.1 Wang et al. (2011) Concentration Response Carve - Danio (zebrafish)

Publication: (Wang et al. 2011)

Species: Zebrafish, Danio rerio
Genus: Danio

EPA-Calculated ECio: 0.01650 (95% C.I. 0.01267 - 0.02033) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

1.4238

0.1030

13.8260

< 2.2 e"16

e

80.1484

4.9349

16.2410

< 2.2 e"16

Concentration-Response Model Fit:

Wang et al. 2011

Danio rerio
Weibull type 1,2 para

PFOS ( ug'L )

C-26


-------
C.2.4.2 Gno el al. (2019) Concentration Response Curve - Danio (zebrafish)

Publication: Guo et al. (2019)

Species: Zebrafish, Danio rerio
Genus: Danio

EPA-Calculated ECio: 0.06274 (95% C.I. 0.06229 - 0.06318) mg/L
Concentration-Response Model Estimates:

Parameter

Estimate

Std.Error

t-stat

p-value

b

0.5686

4.4315 e4

1283.1200

0.0005

d

3.0630

7.7727 e5

39406.9800

1.616 e5

e

3.2830

9.8428 e3

333.55

0.0019

Concentration-Response Model Fit:

Danio rerio
Weibull type 1, 3 para

"le-05	le^M	le^03	le^02	le^OI

PFOS (mg L )

C.2.5 Fifth Most Sensitive Freshwater Genus for Chronic Toxicity: Daphnia (cladoceran)
Logeshwaran et al. (2021) conducted acute and chronic toxicity tests with the

cladoceran, Daphnia cannula, and PFOS-K (perfluorooctancesulfonate potassium salt, > 98%

purity, purchased from Sigma-Aldrich Australia). In-house cultures of daphnids were maintained

in 2 L glass bottles with 30% natural spring water in deionized water, 21ฐC and a 16-hr: 8-hr

light:dark photoperiod. The chronic test protocol followed OECD guidelines (2012). A PFOS

stock solution (20 mg/mL) was prepared in dimethylformamide and diluted with deionized water

to achieve a concentration of 200 mg/L PFOS. Cladoceran culture medium was used to prepare

C-27


-------
the PFOS stock and test solutions. One daphnid (6-12 hours old) was transferred to 100 mL
polypropylene containers containing 50 mL of nominal test solutions (0, 0.001, 0.01, 0.1, 1.0 and
10 mg/L PFOS). Each test treatment was replicated ten times with test solutions renewed and
daphnids fed daily. At test termination (21 days) test endpoints included survival, days to first
brood, average offspring in each brood and total live offspring. At the higher test concentrations
(1 and 10 mg/L) reproduction was completely inhibited. No mortality occurred in the controls
and lowest test concentration. However, reproduction was inhibited al the lowest test
concentration. The author-reported 21-day NOEC and LOEC, based on a\ crime offspring in each
brood and total live offspring, was <0.001 and 0.001 mg/L PFOS, respectively. Additionally, the
author-reported 21-day NOEC and LOEC based on the days lo first brood was 0.001 and 0.01
mg/L, respectively. EPA could not independently calculate 21 -day F.C io values for any of the
endpoints given the level of data provided in the paper by the study authors. And while the
endpoints of mean offspring per each brood and total living offspring appear to be more sensitive
than the days to lirsl brood, lliey result in less than LOECs of 0.001 mg/L and are not consistent
with other chronic toxicity \ allies for this species Therefore, the author-reported MATC of
0.003 I (->2 mu I. lor the days to first brood was quantitatively used to derive the chronic
freshwater criterion for PFOS

DroUsir nntl lvrueซor (2000i) reported the results of a life-cycle, 21-day renewal,
measured test of PI'OS (potassium salt, CAS # 2795-39-3, 90.49% purity) with Daphnia magna.
The GLP test was conducted at Wildlife International, Ltd. in Easton, MD in February, 1999.
The test followed OECD 211 (1997), U.S. EPA OPPTS Number 850.1300 (1996), and ASTM
Standard E 1193-87. D. magna used for the test were less than 24 hours old at test initiation.
Dilution water was 0.45 |im filtered and UV sterilized well water [total hardness: 124 (120-128)

C-28


-------
mg/L as CaC03; alkalinity: 169 (164-172) mg/L as CaC03; pH: 8.2 (8.0-8.3); TOC: <1.0 mg/L;
and conductivity: 329 (315-340) |amhos/cm], Photoperiod was 16-hr:8-hr, light:dark with a 30
minute transition period. Light was provided at an intensity of 329-383 lux. A primary stock
solution was prepared in dilution water at 46 mg/L. It was mixed until all test substance was
dissolved prior to use. After mixing, the primary stock was proportionally diluted with dilution
water to prepare the five additional test concentrations. Exposure \ essels were 250 mL plastic
beakers containing 200 mL of test solution. The test employed 10 repliants of one daphnid each
in six measured test concentrations plus a negative control. Nominal concentrations were 0
(negative control), 1.4, 2.9, 5.7, 11, 23, and 46 mg I. Mean measured concentrations were
<0.458 mg/L (the LOQ), 1.5, 2.9, 5.6, 12, 24, and 48 mg/L, respectively. Analyses of test
solutions were performed at Wildlife International Ltd. using I MM.(' MS The mean percent
recovery of matrix fortifications analyzed concurrently during sample analysis was 104%.
Measured values of new samples ranged from to 121" u of nominal. Measured values from the
old solutions ranged from Mo I < >8" (. of nominal. PFOS was stable throughout the renewal
periods. Dissolved oxygen in new and old test concentrations ranged from 8.3-8.9 mg/L in the
negative controls and 8 <) mg I. at the NOEC of 12 mg/L. Similarly, pH ranged from 8.1-8.4
and 8 2-8.5. respectively, and test temperature from 19.4-20.1ฐC (negative control and at the
NOEC). Sui \ i\ al in the 1 5. 2 l), 5.6, 12, and 24 mg/L treatment groups was 100, 100, 100, 90,
and 0%, respec1i\ ely After 48 hours, survival of the second generation in the negative control
was 95%. The 21-day NOEC (survival, growth, and reproduction) was 12 mg/L. The 21-day
LOEC was 24 mg/L and the calculated MATC is 16.97 mg/L. No second-generation D. magna
survived the 24 mg/L treatment. The independently-calculated ECio based on survival was 11.19

C-29


-------
(10.50 - 11.89) mg/L and was used quantitatively to derive the draft chronic water column
criterion.

Boudreau (2002) also conducted a chronic life-cycle 21-day renewal, unmeasured test of
PFOS (potassium salt, CAS # 2795-39-3, 95% purity) with Daphnia magna as part of a Master's
thesis at the University of Guelph, Ontario, Canada. The results were subsequently published in
the open literature (Boudreau et al. 2003a). The test followed ASTM I- I 193-97 (1997). IX
magna used for testing were less than 24 hours old at test initiation. I), magna were obtained
from abrood stock (Dm99-23) at ESG International (Guelph, ON, Canada). Dilution water was
clean well water. Hardness was softened by addition of distilled deionized water to achieve a
range of 200-225 mg/L of CaC03. Photoperiod was 16-hr:8-hr (light:dark) under cool-white
fluorescent light between 380 and 480 lux I .ahoratorv-grade distilled water was used for all
solutions with maximum concentrations deri\ ed from stock solutions no greater than 450 mg/L.
Test vessels consisted of 225 nil. polypropylene disposable containers containing 120 mL of test
solution. All toxicity testing in\ ol\ ed lour replicates of three daphnids each in five nominal test
concentrations plus a negati\ e control Nominal concentrations were 0 (negative control), 6, 13,
25, 5'). and I'm mg I. The test was conducted in environmental chambers at 21 ฑ1ฐC. Authors
noted that temperature and pi I were measured at beginning and end of study, but the information
was not reported Survival of daphnids in the negative control was 100%. The 21-day NOEC
(survival and reproduction) was 25 mg/L. The 21-day LOEC was 50 mg/L and the calculated
MATC is 35.36 mg/L. The independently-calculated ECio based on survival was 16.35 (7.377 -
25.33) mg/L and was used quantitatively to derive the draft chronic water column criterion.

In addition to the acute toxicity tests described above, Ji et al. (2008) conducted chronic
life-cycle tests of the effects of PFOS (acid form, CAS # 1763-23-1, purity unreported) on

C-30


-------
Daphnia magna. Tests were done under renewal conditions over a 21-day period. The test
followed OECD 211 (1998). D. magna used for testing were obtained from brood stock cultured
at the Environmental Toxicology Laboratory at Seoul National University (in South Korea).
Organisms were less than 24 hours old at test initiation. Dilution water was moderately-hard
reconstituted water (total hardness typically 80-100 mg/L as CaCCb). Experiments were
conducted in glass jars of unspecified size and fill volume. Phoioperiod was assumed to be 16-
hr:8-hr (light:dark) as was used for daphnid culture. Preparation of lest solutions was not
described. The test involved 10 replicates of one daphnid each in five nominal lest
concentrations plus a negative control. Nominal concentrations were 0 (negati\e control),
0.3125, 0.625, 1.25, 2.5, and 5 mg/L. Test temperature was 21 ฑ 1ฐC. Authors noted water
quality parameters (pH, temperature, conckicli\ ily. and dissol\ ed oxygen) were measured after
changing the medium, but the information was not reported Sm \ ival of daphnids in the negative
control was 100".. The author reported/), magna 21-day \OI-C for the reproductive endpoint of
number of young per snr\ i\ a I adult was 1.25 mg/L. The author reported 21-day LOEC for the
same endpoint was 2 5 nig I.. The calculated \I.\TC was 1.768 mg/L. In the independent
verification of the toxicity \ allies. I.IW recalculated the reproductive endpoint noted to be the
number of young per starting adult (instead of surviving adult). This recalculated reproductive
endpoint look the full effects of PFOS into account as it was representative of the full life cycle.
The calculated E('iฆ ฆ lor I >. magna was 0.7885 (0.7043 - 0.8726) mg/L. The results from this
study were acceptable for quantitative use despite the nominal test concentrations, as the study
design was sound, and the concentration-response curve was compelling for the tested species.
The independently-calculated ECio of 0.7885 mg/L was used in the aquatic life criteria
derivation for I), magna.

C-31


-------
Li (2010) conducted a chronic life-cycle 21-day test on the effects of PFOS (potassium
salt, >98% purity) on Daphnia magna. The test followed OECD 211 (1998). D magna used for
the test were maintained in the laboratory for more than one year and were less than 24 hours old
at test initiation. Dilution water was distilled water with ASTM medium (0.12 g/L CaS04.2H20,
0.12 g/L MgSC>4, 0.192 g/L NaHCCte, and 0.008 g/L KC1 - calculated total hardness 169 mg/L as
CaCCte). Photoperiod was 16-hr:8-hr, light:dark at an unreported light intensity. A primary stock
solution was prepared in dilution water and did not exceed 400 mg/L. L\posuie vessels were 50
mL polypropylene culture tubes with 50 mL fill volume. The test involved 1" replicates of one
daphnid each in five nominal test concentrations plus a negative control. Nominal concentrations
were 0 (negative control), 0.5, 1, 5, 10, and 20 mg/L. Test temperature was maintained at 20
ฑ1ฐC. Water quality parameters measured in test solutions ^eie not reported. Survival of
daphnids in the negative control was 96.7".. The / K magna 21 -day NOEC (reproduction - no.
young per female) was I nig I. The 21-day 1 .()!ฆ(' was 5 nig L and the calculated MATC was
2.236 (2.8642 - 2 l)73K) nig I. The independently-calculated toxicity value (ECio) based on total
neonates per female was 2 ^ I ^ mu I. and was used quantitatively to derive the draft chronic
water column criterion

Ysing ct nl. (2014) e\ aluated the chronic 21-day renewal, measured test of PFOS
(potassium salt, C AS # 27^5-.ป-3, 99% purity) with Daphnia magna. The test followed ASTM
1993c90-l 191 (liw.l) / K magna used for the test were donated by the Chinese Research
Academy of Environmental Sciences, and less than 24 hours old at test initiation. Dilution water
was dechlorinated tap water (pH, 7.0 ฑ0.5; dissolved oxygen, 7.0 ฑ0.5 mg/L; total organic
carbon, 0.02 mg/L; and total hardness, 190.0 ฑ0.1 mg/L as CaCCte). Photoperiod was 12-hr: 12-hr
(light:dark) at an unreported light intensity. A primary stock solution was prepared by dissolving

C-32


-------
PFOS in deionized water and cosolvent DMSO. The primary stock was proportionally diluted
with dilution water to prepare the test concentrations. Exposure vessels were 200 mL beakers of
unreported material type containing 100 mL of test solution. The test employed 10 replicates of
one daphnid each in six test concentrations (measured in low and high treatments) plus a
negative and solvent control. Nominal concentrations were 0 (negative and solvent controls),
2.00, 2.60, 3.38, 4.39, 5.71 and 7.43 mg/L. Mean measured concentrations before and after
renewal respectively were 1.74 and 1.98 mg/L (lowest concentration) and (ฆ> 78 and 7.54 mg/L
(highest concentration). Analyses of test solutions were performed using IIPI.(' MS and negative
electrospray ionization. The concentration of PI OS was calculated from standard curves (linear
in the concentration range of 1-800 ng/mL), and the average extraction efficiency was in the
range of 70-83%. The concentrations and chromatographic peak areas exhibited a significant
positive correlation (r=0.9987, p<0.01), and the water sampie-spiked recovery was 105%. Test
temperature was maintained at 22 ~2ฐC. The DO and pi I were reported as having been
measured, but results are not reported Negative and solvent control survival was 100%. The/).
magna 21-day LCi- lor reproduction was reported to be 2.26 mg/L from the study authors. The
study author reported \ nine was used <.|iianlitatively to derive the draft chronic water column
criterion

Lu el ;il. (2015) conducted a chronic life-cycle 21-day renewal, unmeasured test of PFOS
(purity 98%) with / ki/>/mia magna. The test followed OECD 211 (2012). I). magna used for the
test were originally obtained from the Chinese Center for Disease Control and Prevention
(Beijing, China) and cultured in the laboratory according to the International Organization for
Standardization (ISO, 1996). Daphnids were less than 24 hours old at test initiation. Dilution
water was the same used for daphnid culture and was reconstituted according to OECD (2004)

C-33


-------
with a total hardness of 250 mg/L as CaCCte, as calculated based on the recipe provided, and pH
ranging from 7.7 to 8.4. Photoperiod was 16-hr:8-hr (light:dark) at an unreported light intensity.
The test solution was prepared immediately prior to use by diluting the stock solution with
culture medium. Exposure vessels were 100 mL glass beakers containing 45 mL of test solution.
The test employed 20 replicates of one daphnid each in six nominal test concentrations plus a
negative control. Nominal concentrations were 0 (negative control). <> < >08, 0.04, 0.2, 1, and 5
mg/L. Exposure water quality was checked daily and maintained at 2" I ('. pU of 7.2 ฑ 0.3,
and dissolved oxygen of 5.3 mg/L. Negative control survival was 100%. The author reported I),
magna 21-day NOEC (no. offspring per brood per female) was 0.008 mg/L and the 2 I -day
LOEC was 0.04 mg/L. The calculated MATC was 0.017l) mu I. and the independently-
calculated EC 10 was 0.001712 (0.000998 n o<)2426) mg/L lor the same endpoint. Other
endpoints, including growth and other reproductive endpoints. could not be independently-
calculated by EPA. The independently-calculated ECio from this study was acceptable for
quantitative use and was used to deri\ e the PFOS chronic water criterion.

Liang et al. (2017) conducted a chronic life-cycle 21-day renewal, unmeasured test of
PFOS (potassium salt. CAS 27l)5->>-3. >98% purity) with Daphnia magna. The test organisms
were originally obtained from State Key Laboratory of Environmental Aquatic Chemistry (Eco-
Environmental Sciences of Chinese Academy of Sciences, Beijing) and cultured in the
laboratory according to Re\ el et al. (2015). Daphnids were less than 24 hours old at test
initiation. Dilution water was artificial medium "M4 (Elendt)" at 20ฐC and pH 7. Photoperiod
was 16-hr:8-hr (light:dark) at an unreported light intensity. The test solution was prepared
immediately prior to use by diluting the stock solution with M4 medium. Exposure vessels were
80 mL glass beakers containing an unspecified volume of test solution. The test employed 10

C-34


-------
replicates of one daphnid each in six nominal test concentrations plus a negative control.

Nominal concentrations were 0 (negative control), 1, 2, 4, 8, and 16 mg/L. No mention was made
of water quality being checked during the exposure. Negative control survival was 100%. The D.
magna 21-day NOEC (days to 1st brood, intrinsic rate of natural increase, r) was 4 mg/L. The
21-day LOEC was 8 mg/L and the calculated MATC was 5.657 mg/L. The independently-
calculated ECio based on survival was 3.596 (2.1207 - 5.0704) mu I. and was used quantitatively
to derive the draft chronic water column criterion.

Yang et al. (2019) evaluated the chronic effects of perfluorooctane sulfonate, potassium
salt (PFOS-K, CAS# 2795-39-3, 98% purity, purchased from Sigma-Aldrich in Si I .ouis, MO)
on Daphnia magna via a 21-day unmeasured, static-renewal test that evaluated growth and
reproductive effects. D. magna cultures were obtained from the Inslilule of Hydrobiology of
Chinese Academy of Science in Wuhan, China Organisms were cultured in Daphnia Culture
Medium according lo the parameters laid out in OECD (inideline 202 and all testing followed
OECD Guideline 21 I Cultures were led green algae daily and were acclimated for two to three
weeks before testing The 21 -day chronic study had nominal concentrations of 0 (control),
0.0001)1)124. i) 000420 mol/L (or 0 (control), 0.6674, 1.012,
1.512. and 2 261 mg/L given the molecular weight of the form of PFOS used in the study, CAS #
2795-39-3. of 53S 22 g/mol) Lach neonate (12-24 hours old) was placed in a 100 mL glass
beaker, in which there were 10 replicates, each filled with 80 mL of test solution maintained at
20 ฑ1ฐC and a 16-hr:8-hr light:dark photoperiod with a light intensity maintained at 1000 - 1500
lux. D. magna were fed green algae and test solutions were renewed every 72 hours. Test
organisms were counted daily, with any young also removed. The author-reported NOEC and
LOEC for reproduction (measured as mean offspring proportion relative to control at 21 days)

C-35


-------
was <0.6674 and 0.6674 mg/L PFOS, respectively. The author-reported NOEC and LOEC for
growth (measured as length) was 0.6674 and 1.012 mg/L PFOS (MATC = 0.8218 mg/L). The
independently-calculated ECio values for reproduction and growth are 0.3773 and 0.9093 mg/L,
respectively. However, the reproduction ECio of 0.3773 mg/L was determined to be less
statistically robust as the independently calculated toxicity values were control normalized and
could not be weighted given the level of data provided by the study authors in the paper.
Therefore, the independently-calculated ECio for growth of 0.9093 (<) 7423 I 076) mg/L was
used quantitatively to derive the draft chronic water column criterion freshwater

Seyoum et al. (2020) evaluated the chronic effects of perfluorooctane sulfonic acid
(PFOS, CAS# 1763-23-1, > 99%, purchased from Sigma) oil Daphnia magna neonates via a 21-
day unmeasured, static-renewal study. The study authors did not report following any specific
protocol. D. magna ephippia were purchased from MicroliioTests Inc. (Belgium) and were
activated by rinsing in tap water l-phippia were hatched In incubating at 20-22 ฐC for 72 to 90
hours in standard freshwater under a continuous light intensity (6,000 lux). Newly hatched
neonates (<24-hr old) were led a suspension of Spirulina micro-algae two hours before testing.
Nominal concentrations of <) (control). I. I n and 25 |iM (or 0 (control), 0.5001, 5.001, and 12.50
mg/I. ui\ en the molecular weight of the form of PFOS used in the study, CAS # 176-23-1, of
500.13 g/mol) were prepared hv mixing the respective amounts of PFOS in dimethyl sulfoxide
(DMSO). Ten <24-hr old neonates, exposed in triplicate, were placed into 250 mL crystallization
dishes with 100 mL of test solution. A mean temperature of 23ฐC, dissolved oxygen of 8 to 9
mg/L, total hardness of 250 mg/L as CaC03, pH of 7.5 ฑ0.25 and salinity of 0.02% were
reported in the exposure water. I), magna were fed a mixture of Spirulina microalgae and yeast
(Saccharomyces cerevisiae) daily during the test, and 50% of the test solution was changed every

C-36


-------
other day. Neonates were counted daily and removed. At day 21, neonate counts were reported to
be highest in the control with >40 to < 60 neonates, and >20 to <40 neonates were reported at the
0.5001 and 5.001 mg/L (or 1 and 10 |iM) concentrations, respectively. Neonate counts for the
12.50 mg/L (or 25 |iM) concentration were not reported. A reproductive NOEC of 0.5001 mg/L
and a LOEC of 5.001 mg/L were reported by the study authors. This LOEC of 5.001 mg/L was
associated with a 42.95% decrease in reproduction (measured as the mean number of daphnids at
21 days) compared to control. An independently-calculated ECio \alue could not be determined
as EPA was unable to fit a model with significant parameters Instead, the author-reported
MATC of 1.581 mg/L PFOS was used quantitati\ ely to deri\ e the draft chronic water column
criterion for freshwater.

C.2.6 Sixth Most Sensitive Freshwater (ienus for Chronic Toxicity Salmo (salmon)

Atlantic salmon, Salmo salar, emlnyos were e\ aluated by Spsichmo and Arukwe (2012)

via a 56-day unmeasured exposure to PFOS (lW\> purity). Eggs were obtained from Lundamo

Hatcheries, Norway (Aquagen) and transported to the Norwegian University of Science and

Technology Centre of Fisheries and Aquacullurc in Trondheim, Norway. The eggs were kept in

plastic tanks (25 I.) at 5-7 C with filtered, re-circulating and aerated water. Approximately one-

third of the w ater volume w as changed once per week. The eggs and larvae were exposed to

PFOS (100 iiu I.) for 49 days representing the developmental period from 404 to 679-degree

days. PFOS was dissoK ed in methanol (carrier solvent: 0.01%) and control group was exposed

to the carrier solvent only. Flatching occurred at 20 calendar days after start of exposure, at an

effective developmental age of 504-degree days, after which riverbed environment was

simulated by tank bed gravel and continuous water flow. Fish sampling was performed at 21, 35,

49 and 56 calendar days after exposure, or at respective developmental age of 549, 597, 679 and

721 degree days. The exposure was terminated at 679-degree days, and 712-degree days

C-37


-------
represents the end of a one-week exposure-free recovery period. Thus, day 49 sampling was
performed 24 hours after terminating the exposure and no exposure related differences in
hatching rate were observed. The 49-day growth NOEC and LOEC were 0.10 and >0.10 mg/L
PFOS, respectively. These data are deemed quantitative and considered acceptable to derive the
draft chronic water column criterion for freshwater.

C.2.6.1	Spachmo andArukwe (2012) Concentration Response Curve - Salmo (salmon)

Publication: (Spachmo and Arukwe 2012)

Species: Atlantic salmon {Salmo salaf)

Genus: Salmo

EPA-Calculated LCso: Not calculable, concentration-response data not a\ ailahle

C.2.7 Seventh Most Sensitive Freshwater Genus for Chronic Toxicity: Pimephales (minnow)
3MCompany (2000) includes the results of an in\ csliuution of the chronic effects of

PFOS-K (perfluorooctancesulfonate potassium sail. CAS ; 2795-3l)-3. unknown purity) on the

fathead minnow, Pimephales promelas, from a test conducted in 1978. The test methods

followed those proposed lor euu and fry stages of freshwater fish and EG&G Bionomics. Eggs

(48 hpf) from EG&G Bionomics were exposed until 30 dph under a flow-through regime with

measured PI OS concentrations PI'OS-K was dissolved in acetone to make the stock solution

(19.4 mu I.) and diluted with well water to make five test concentrations and two controls (well

and solvent control) The sol\ cut control contained the same amount of acetone (43 |iL/L) as in

the highest test concentration Average measured PFOS concentrations were <0.006 (solvent

control), <0.006 (well water control), 0.12, 0.28, 0.45, 1.0, and 1.9 mg/L PFOS; measured on test

day 3, 10, 17, 24 and 32. At test initiation sixty eggs each were assigned to each of the 14 egg

cups and assigned to one of the 14 aquaria (two aquaria per test treatment). Dead eggs were

counted and removed daily until hatching was complete (test day 3), afterwards 40 fry from each

egg cup were transferred to their respective aquarium and exposures continued for an additional

C-38


-------
30 days. Fry were fed brine shrimp three times daily throughout the exposure period.
Observations on behavior and appearance were made daily and fry counts were made weekly.
Test endpoints included hatch, survival, and growth (length and weight). Throughout the
exposure period average water quality parameters varied very little between treatments (average
D.O. of 8.6 mg/L, 22-26 ฐC, pH 6.6-7.3). PFOS-K did not have any significant effects on
hatchability of eggs or growth of fry. However, survival of fry in I lie I 9 mg/L treatment was
42% less than the controls, with fish exhibiting stressful behavior (e.g . erratic swimming). The
author reported 33-day NOEC and LOEC, based on survival, was 1.0 and 1.9 mu L PFOS,
respectively, with aMATC of 1.378 mg/L. The independently-calculated 33-day I-Cio value was
0.4408 (0.3076 - 0.5738) mg/L and was considered acceptable for quantitative use in the
derivation of the chronic freshwater criterion lor PI'OS

The chronic effects of PFOS on the fathead minnow. Pimcphales promelas, have been
reported by several researchers Drollsir sind Krueger (2000j), associated with Wildlife
International, conducted a (il.P 47-day llow-through measured early life-stage toxicity test with
<24-hour old P. promdas embryos A primary stock solution was prepared by dissolving PFOS
(90.49% purity) in dilution water at a concentration of 88.4 mg a.i./L, then proportionally diluted
with dilution water to prepare live secondary stock solutions at concentrations of 44.2, 22.1,
11.0, 5.52 and 2 7(-> mg a.i. I. Stock solutions were prepared every three to four days during the
test. The five stocks were injected into the diluter mixing chambers (at a rate of 6.0 mL/minute)
where they were mixed with dilution water (at a rate of 116 mL/minute) to achieve the desired
test concentrations. The water used for culturing and testing was freshwater obtained from a well
approximately 40 meters deep located on the Wildlife International Ltd. site. The well water was
characterized as moderately-hard water. The well water was passed through a sand filter to

C-39


-------
remove particles greater than approximately 25 |iin and then pumped into a 37,800-L storage
tank where the water was aerated with spray nozzles. Prior to delivery to the diluter system, the
water again was filtered (0.45 (j,m), then passed through a UV sterilizer to remove
microorganisms and particles. Fathead minnow embryos used in this test originated from
cultures maintained by Wildlife International Ltd., Easton, MD. The embryos were removed
from spawning substrates and examined under the dissecting microscope to select healthy
specimens at approximately the same stage of development. Embryos collected for use in the test
were from six individual spawns. Embryos were exposed to a geometric scries (if six test
concentrations and a negative (dilution water) control under flow-through conditions at 24.5ฐC,
pH of 8.2, total hardness of 140 mg/L as CaCCte and a pholoperiod of 16 hours light and eight
hours dark. Four replicate test chambers (lH. glass aquaria) were maintained in each treatment
and the control group. Each test chamber contained one incubation cup with 20 embryos,
resulting in a total of S<) embryos per treatment The exposure period included a five-day embryo
hatching period, and a 42-da\ post-hatch juvenile growth period. Nominal test concentrations
were 0.14, n 29. <> 57. I 1. 2 3 and 4 (•> mu I. a i Mean measured test concentrations (0, 0.15,
0.30, i) (•><). I 2. 2 4 and 4 (•> mu I.) were determined from samples of test water collected from
each treatment and the control group at the beginning of the test, on day four, at weekly intervals
during the test, unci at test termination. To start the test, embryos less than 24 hours old were
collected from cultures and groups of one and two individuals were impartially distributed
among incubation cups until each cup contained 20 embryos. One cup was then placed in each
treatment and control test chamber. Twice during the next twenty-four hours and daily thereafter,
all dead embryos were counted and removed from the cups to avoid contaminating viable
embryos. All eggs that remained were considered viable. Dead embryos continued to be removed

C-40


-------
daily. After hatching, the larvae were counted and released into the test chambers, where
exposure continued until test termination. Observations of mortality and other clinical signs were
made daily during the test. Time to hatch, hatching success, growth, and survival were monitored
in each treatment and control group. The most sensitive 47-day chronic value (MATC) of 0.4243
mg/L PFOS was based on post-hatch survival as reported by the study authors. The
independently-calculated ECio was 0.4732 (0.3308 - 0.6156) mu I. hascd on survival and was
used quantitatively to derive the draft chronic water column criterion

Ankley et al. (2005) also exposed Pimephalespvomelas to PFOS (potassium salt, > 98%
pure) under flow-through measured conditions for 21 days Stock solutions were prepared by
dissolving crystals in Lake Superior control water with stining (mean measured test conditions:
25ฐC, pH of 7.3, total hardness of 46 mg, I. as CaCO;. alkalinity of 4<) mg/L as CaC03 and
dissolved oxygen of 6.2 mg/L). Two stock solutions of approximately 9.7 and 97 mg/L were
used to span the desired range of target concentrations in test water. Final test concentrations
were generated by appropriate dilution of the PFOS stocks with Lake Superior water and were
supplied to the test tanks at a How rate of approximately 45 mL/min. Sexually mature fathead
minnow s (six to se\ en months old) obtained from the on-site culture facility were used for the
toxicilv test I light pairs of fish (one male and one female) were exposed at each treatment level,
0 (control), <> <>3. <> I. 3. and 1.0 mg PFOS/L. Assays were conducted using glass aquaria
containing 10 L of test solution, with two pairs of fish separated by perforated nylon screening in
each tank. Reproductive viability of the fish used for the test was documented during a 27-day
acclimation phase in the same tanks in which the tests were conducted. The number of eggs
spawned by each pair was evaluated daily by inspecting the underside of a polyvinyl chloride
spawning tile placed on the bottom of the test chambers. Egg fertility was assessed using an

C-41


-------
optical microscope. The animals were held at 25 ฑ1ฐC under 16-hr: 8-hr light:dark photoperiod
and fed frozen brine shrimp to satiation twice daily. Conditions during the 21-day reproduction
phase of the PFOS exposure were the same as during the acclimation phase. To evaluate possible
early developmental toxicity of PFOS, 50 to 75 eggs from single viable spawns were collected
during the final 7-day of the reproduction phase of the test. A subset of eggs was reserved for the
determination of PFOS concentrations. Embryos were held in 3<~><) nil. Pyrex beakers in the same
aquaria as the parental fish. Embryos hatched within four to five days and I hereafter were fed
live brine shrimp twice daily. After 12 days, fry were randomly sampled for PI OS analysis and
to reduce the number of animals per chamber to 3< > Remai ni ng fry were maintained in a larger
chamber (1 L plastic container) within the original tank. I)e\ eloping fish were inspected daily to
assess survival. After 24 days, they were anesthetized and weighed A subset of the fry was
collected for PFOS measurements, while others were preser\ ed in Bouin's fixative for
histological analyses The authors reported a 21 -day EO.. (fecundity) of 0.23 mg/L PFOS, and a
chronic value of 0.47lM mu I. PI-OS lor percent hatch (21-day), probability of survival and larval
weight endpoints (21-day (I'D) 24-day (l; I)) The independently-calculated ECio was 0.05101
(0.04')S () ()(•> 13) mu I. bused on fecundity and was used quantitatively to derive the draft
chronic water column criterion

Suski ol ;il. (2021) reported the chronic toxicity of PFOS-K (perfluorooctancesulfonate
potassium salt, CAS 27l)5-3l)-3, > 98%, purchased from Sigma-Aldrich) on the fathead
minnow, Pimephalesprumelas. Adult (5-month old) fathead minnows were purchased from a
commercial supplier (Aquatic Biosystems) and were sexually mature when the test was initiated.
Fish were fed twice a day and held in dechlorinated tap water at test conditions (mean
conditions: 24.96ฐC, D.O. of 7.68 mg/L, pH 7.9 and conductivity of 347.3 |iS/cm). Stock

C-42


-------
solutions of PFOS (150 mg/L) were made without a solvent and prepared weekly with stock
solutions shaken at 80 rpm for 24 hours to ensure mixing. Test solutions were made by diluting
the stock with dechlorinated tap water and shaking the solutions for 10 minutes prior to water
exchanges. Half of the total volume (10 L) in each exposure 5-gallon polycarbonate tank was
renewed three times per week. Measured PFOS concentrations were 0.14 (control), 44, 88, 140
and 231 |ig/LPFOS (or 0.00014 (control), 0.044, 0.088, 0.K and <) 231 mg/L PFOS). Each test
treatment was replicated six times for each treatment and consisted of i\\ o females and one male
per tank with exposures lasting 42 days. Tanks were expected daily for eggs and all eggs
collected were assumed to be per single female regardless of the number of females per tank. On
the last week of testing, eggs were carried through hatching in their respective test treatments,
and 20 larval fish per concentration were exposed for an additional 21 days to investigate
developmental effects. One liter polypropylene beakers were used lor F1 generation exposure
with solutions renewed daily Sur\ i\al of adult fathead minnows in the control and two lowest
test concentrations was K<)"(. at test termination. Survival of male fish in the highest test
treatment was significantly less than male control fish, and while female survival was also less
compared to control fish. the effects were not significant. The mean number of spawning events
per female was also reduced in the two high test treatments, but the effect was only significant in
the 140 |ig/l. (i) 14 mg/L) treatment. Larval survival intheFl generation was significantly
reduced in the highest test treatment. The most sensitive endpoint from the study was a
significant decrease in the mean mass of individuals in the larval F1 generation with reported
values of 3.76, 3.53, 3.09, 2.64 and 2.00 mg for the test treatments of control, 0.044, 0.088, 0.14,
and 0.231 mg/L PFOS, respectively. The author-reported NOEC and LOEC, based on growth in
the F1 generation, were 0.044 (6% reduction in growth compared to controls) and 0.088 mg/L

C-43


-------
PFOS (associated with an 18% reduction in growth), respectively, with a MATC of 0.06223
mg/L. The independently-calculated ECio value was 0.0549 (0.0396 - 0.0701) mg/L and was
considered acceptable for quantitative use in the derivation of the chronic freshwater criterion for
PFOS.

C.2.8 Eighth Most Sensitive Freshwater Genus for Chronic Toxicity: Proccimbarus (crayfish)
Funkhouser (2014) conducted a chronic 28-day renewal icsl ol'PFOS (potassium salt,

>98% purity) with a crayfish species, Procambarus fallax (f. vn^maHs) The study was

conducted as part of a Master's thesis at Texas Tech I Iniversity, Lubbock. T.\ ,lu\ enile P. fallax

(4-weeks old, 0.056 g) used for the test were originally purchased from a prh tile col lector. The

crayfish reproduced for several generations before being used for experiments. Based on an

average reproductive age of 141-255 days, an inlerclutch period of 50-85 days, and a brooding

time of 22-42 days, the author estimated the experimental animals to he F4-F6 (Seitz et al. 2005).

Dilution water was moderately hard reconstituted laboratory water (3.0 g CaSC>4, 3.0 g MgSC>4,

0.2 g KC1, and 4 ^ u \al ICO; added to 5<) L deionized water). Photoperiod was 14-hrs:10-hrs

(light:dark) at an unreported light intensity PFOS was dissolved in dilution water to prepare the

test concentrations Lxposuic \ essels were 1 L polypropylene containers containing 500 mL of

test solution The test employed eight replicates of one crayfish each in five test concentrations

plus a neuati\ e control. Nominal concentrations were 0 (negative control), 0.2, 0.5, 1.3, 3.2, 8

and 20 mg/L. Exposure concentrations were reportedly measured, but concentrations were not

provided. Analyses of test solutions were performed using LC-MS/MS. Standards were used as

part of the analytical method, but details were not reported. The reporting limit was 0.010 mg/L.

Experiments were conducted in an incubator set at 25 ฑ1ฐC and covered with plastic opaque

sheeting to limit evaporation. No other water quality parameters were reported as having been

measured in test solutions. Negative control survival was 85% after 28 days. The 28-day LC20

C-44


-------
was reported as 0.167 mg/L. The independently-calculated ECio could not be calculated at this
time given the level of data that were presented in the paper. The study author-reported LC20
value was used quantitatively to derive the draft chronic water column criterion.

C.2.9 Ninth Most Sensitive Freshwater Genus for Chronic Toxicity: Moina (cladoceran)
In addition to the acute toxicity tests described above, Ji et al. (2008) conducted a

chronic life-cycle test of the effects of PFOS (acid form, CAS I 703-23-1, purity unreported) on

Moina macrocopa. The test was performed under renewal conditions o\ or a 7-day period. The

M. macrocopa test followed a protocol developed and reported by Sutherland and krueger

(2001) was similar to OECD 211 (1998), but with slight modification. M macrocopa used for

testing were obtained from brood stock cultured at the I-n\ iron mental Toxicology Laboratory at

Seoul National University (in South Korea) Test organisms were less than 24 hours old at test

initiation. Dilution water was moderately hard reconstituted water (total hardness typically 80-

100 mg/L as CaCOO F.\peri 111 en is were conducted in glass jars of unspecified size and fill

volume. Photoperiod was assumed 10-hr K-hr (light.dark) as was used for daphnid culture.

Preparation of test solutions was not described The test involved 10 replicates of one daphnid

each in ll\ e nominal test concentrations plus a negative control. Nominal concentrations were 0

(negati\ e control), n 3 125, n (->25, I 25. 2 5, and 5 mg/L. Test temperature was 25 ฑ 1ฐC. Authors

noted that water quality parameters (pH, temperature, conductivity, and dissolved oxygen) were

measured after changing the medium, but the information was not reported. Survival of daphnids

in the negative control was 100%. The author reported M macrocopa 7-day LOEC for the

reproductive endpoint of number of young per surviving adult was 0.3125 mg/L. The author

reported 7-day NOEC and MATC was <0.3125 mg/L for the same reproductive endpoint. In the

independent verification of the toxicity values, EPA recalculated the reproductive endpoint to be

the number of young per starting adult (instead of surviving adult). This recalculated

C-45


-------
reproductive endpoint took the full effects of PFOS into account as it was representative of the
full life cycle. The independently-calculated ECio for M macrocopa was 0.1789 (0.041 - 0.399)
mg/L. The results from this study were acceptable for quantitative use. The independently-
calculated ECios of 0.1789 mg/L was used in the aquatic life criteria derivation forM
macrocopa.

C.2.10 Tenth Most Sensitive Freshwater Genus for Chronic Toxicity Brachionus (rotifer)

Zhang et al. (2013) conducted a chronic life-cycle renewal test of PFOS (potassium salt,

CAS # 2795-39-3, >98% purity) with Brachionus calyciflorus. The test duration was five days in

a full-life cycle test (primary emphasis) and 28 days in a multi-generation population growth test

(secondary emphasis - only two exposure concentrations plus a control). Test organisms were

less than two hours old at test initiation. All animals were parlhenogenetically-produced

offspring of one individual from a single resting egg collected from a natural lake in Houhai Park

(Beijing, China). The rotifers were cultured in an artificial inorganic medium at 20ฐC (16-hr:8-

hr, light:dark; 3,000 lux) lor more than six months before toxicity testing to acclimate to the

experimental conditions Culture medium was an artificial inorganic medium and all toxicity

tests were carried out in the same culture medium and under the same conditions as during

culture (i e . pFT, temperature, illumination). Solvent-free stock solutions of PFOS (1,000 mg/L)

were prepared In dissolving the solid in deionized water via sonication. After mixing, the

primary stock was proportionally mixed with dilution water to prepare the test concentrations.

Exposures were carried out in 24-well cell culture plates (assumed plastic) containing 2 mL of

test solution per cell. The test employed four measured test concentrations plus a negative

control. Each treatment consisted of one replicate plate of 15 rotifers each in individual cells.

Treatments were repeated six times. Nominal concentrations were 0 (negative control), 0.25, 0.5,

1.0, and 2.0 mg/L. PFOS concentrations were not measured in the rotifer exposures, but rather,

C-46


-------
in a side experiment using HPLC/MS. The side experiment showed that the concentration of
PFOS measured every eight hours over a 24-hour period in rotifer medium with green algae
incurs minimal change in the concentration range 0.25 to 2.0 mg/L. 100% survival was observed
at 24 hours in the negative control in the corresponding acute test but was not provided for the
life-cycle test. The B. calyciflorus 5-day LOEC (net reproductive rate and intrinsic rate of natural
increase) was 0.25 mg/L. The NOEC and MATC were <0.25 mu I. The study author reported
value was used quantitatively to derive the draft chronic water column criterion.

C.2.11 Eleventh Most Sensitive Freshwater Genus for Chronic Toxicity: Xiyhonhorns

(swordfish)

The toxicity of PFOS (potassium salt, > Wn purity) lo the swordtail fish. Xi/'hophorus
helleri, was evaluated by Han and Fang (2010) A PFOS slock solution (250 mg/L) was
prepared by dissolving crystals in dechloi i nalccl lap water ( from the same water source as that
used in fish keeping). Six- to seven-month old adull swordtails were purchased from a local fish
farm with no water pollution. The fish were separated by sex into different aquaria. Both the
males and females were acclimated for eight weeks under semi-static conditions in charcoal
filtered, aerated lap water at 27 I ('with a 14-hr I <)-hr (light:dark) photoperiod. The water in
each aquarium was completely renewed e\ cry 4K hours. The fish were fed once daily in the
morning with fluke food and once daily at dusk with frozen blood worms. Adult male fish were
then randomly distributed into 30 L tanks containing 20 L dechlorinated tap water or a
corresponding PFOS solution. Swordtail fish were exposed to 0 (control), 0.1, 0.5 or 2.5 mg/L
PFOS for three weeks and then transferred into clean water for one-week recovery. Every day,
half of the water in each tank was replaced with fresh water, and the fish were exposed to the
appropriate concentrations daily. Exposure conditions were the same as those during the
acclimation period. Each aquarium housed 10 swordtails. Three aquaria were used for each

C-47


-------
exposure concentration and for the controls, resulting in three full biological replicates for each
exposure group. Body, liver and testis weights were determined at 7, 14, 21 and 28 days after
ice-bath anaesthetization. The livers were weighed immediately, then frozen in liquid nitrogen
and stored at -80ฐC for RNA extraction. The hepatosomatic index (HSI) and gonadal somatic
index (GSI) values were also calculated. Nonpregnant adult female fish were housed under the
same exposure conditions as the males for the six-week exposure period. At the same time, to
ensure impregnation of the females, nine adult females were paired with three adult males and
kept in each aquarium for one week, after which the males were taken out There were also three
biological replicates for each exposure group. One pregnant female per aquarium was isolated
and housed until giving birth. Larvae were maintained in clean water for up to 14 days after birth
to calculate their survival rate. At the end of llie exposure period, the survival rate, HSI and GSI
values of all groups were determined. The total number of puerperal females and females with
eggs or embryos in each group was recorded to determine their corresponding ratios. More than
100 adult swordtails (with a male lemaleratio of about l:3)were housed together to obtain at
least 240 juveniles (2<ป-.><> days old) All of the fry were then randomly separated into two
exposure groups (<) and <> I mu I.) and kept under the same housing conditions as the males.

Each tank contained 40 fry There were also three biological replicates in each group. After a 90-
day exposure period, the FISI. GSI, and condition factor (CF) values and the sex ratio of each
group were calculated l\\ se\ category. Body length from the snout to the end of the caudal fin
and sword length from the distal end of the middle rays of the caudal fin to the tip of the sword
were measured for each young male. After an extended period of stable breeding, part of the
juveniles became young females and some of them were with eggs, embryos or puerperal. So,
just like adult females, the total number of puerperal females and females with eggs or embryos

C-48


-------
in each group were recorded as a single entity to determine their corresponding ratios. The 4-
week (adult male), 6-week (adult female) and 90-day (juvenile female and male) survival chronic
values were >2.5, 1.118 and >0.1 mg/L PFOS, respectively. The study-author reported survival
chronic value for offspring of females exposed for six weeks was 0.2236 mg/L PFOS, and the
90-day growth and percent females with eggs chronic value was <0.1 mg/L PFOS. The
independently-calculated ECio for female survival was 0.5997 (<) 233(-> - n 9658) mg/L, which
was acceptable for quantitative use.

C.2.12 Twelfth Most Sensitive Freshwater Genus for Chronic Toxicity: Xcnonns (frog)

Lou et al. (2013) evaluated the chronic toxicity of PFOS to the African claw ed frog,

Xenopus laevis. PFOS (98% purity) stock solutions (S mu ml.) were prepared by dissolving in

DMSO every four weeks and stored at 4 (' Slock solutions were diluted by charcoal-filtered tap

water to prepare test water. DMSO concent ml ions were <> <")<") 1% (\ \) in all tanks including the

solvent control group. The same charcoal-filtered tap water (pH 6.5-7.0, dissolved oxygen >5

mg/L, and total water hardness, as CaCOi, of approximately 150 mg/L) was used to raise X.

laevis frogs and tadpoles. Adult female and male X. laevis (3 years old, obtained from Nasco,

USA ) were raised separately in glass tanks at 22 ฑ 2ฐC with a 12-hr: 12-hr light:dark cycle and

fed w iih chopped pork li\ er (commercial amphibian diet three times a week). A pair of X. laevis

was injected l\\ human chorionic gonadotropin to induce breeding. Fertilized eggs were

incubated in the same dechlorinated tap water at 22 ฑ 2ฐC for six days (and were fed live

Artemia starting on the 5lh day). On the fifth day postfertilization, tadpoles atNF stage 46/47

were exposed to PFOS (nominal: 0.0001, 0.001, 0.100 and 1.0 mg/L; measured: 0, 0.00009,

0.001, 0.1117, 0.7160 mg/L) until two months post-metamorphosis. Each exposure group and

control group consisted of three replicated tanks. Each tank with 18 L water was assigned

randomly 25 tadpoles. The tadpoles were fed with live Artemia three times daily. After

C-49


-------
metamorphosis, the juvenile frogs were fed with live Artemia daily and chopped pork liver every
other day. The test water (22 ฑ 2ฐC) was completely replaced every other day. Fluorescent
lighting provided a photoperiod of 12 hours and a light intensity ranging from 600 to 1,000 lux at
the water surface. During the exposure, the animals were observed for mortality and growth daily
and dead tadpoles were removed. At the end of exposure, the survival rate of the frogs in each
tank was recorded. After anaesthetization, the frogs were weighted and dissected. The liver tissue
of each frog was weighted, and hepatosomatic index (HS1) was calculated The sex or intersex of
each frog was determined by examining the gross gonadal morphology with a stereo microscope.
The survival, weight and sex ratio/intersex chronic value were all > 1 mg/L PFOS The study-
author reported value was used quantitatively to derive the draft chronic water column criterion.

Fort et al. (2019) evaluated the chronic effects of peril uorooctane sulfonic acid (PFOS,
>98% purity, CAS # 1763-23-1, lot # BCBII2K34Y from Sigma-Aldrich in St. Louis, MO) on
clawed frogs {Xcnopns iropicahs. formerly Si/nrana tropicalis) in a 150-day post-metamorphosis
flow-through, measured study Stock solutions were prepared by dissolving PFOS into filtered,
dechlorinated tap water in IS I. glass carboys, u liich were then pumped into the master mixing
cell of the continuous flow cli I Liter Adult frogs were obtained from Xenopus 1 and fed salmon
starter pellets daily for 30 days during acclimation prior to breeding. Temperature during
acclimation was maintained at 26 ฑ0.5ฐC. Researchers followed the breeding guidance of Fort et
al. (2002), and added human chorionic gonadotropin the day before breeding began. Three pairs
of frogs were isolated and allowed to breed, but only a single clutch with a >70% spawn rate was
utilized for the experiment. Normal appearing dejellied embryos (Nieuwkoop and Faber Stage
10) were randomly selected, and 20 were placed in each of four aquaria, each 4-L in size, for a
total of 80 embryos per concentration. The frogs were subjected to a 12-hr: 12-hr light-dark

C-50


-------
photoperiod with a light intensity of 600 ฑ 50 lux, and the pH was maintained naturally at 7.5 ฑ
0.3. The diluter system achieved a complete volume change every 6.5 hours, and diluter
performance, flow rates, temperature, dissolved oxygen and light intensity were measured daily.
Test organisms were exposed to mean measured concentrations of <0.03 (control), 0.05, 0.13,
0.31, 0.59 and 1.05 mg/L PFOS until metamorphosis, and liquid chromatography mass-
spectrometry was used to verify differences in PFOS concentrations At metamorphosis (NF
Stage 66), weight and snout-vent lengths were measured. Frogs were kept an additional 150 days
past metamorphosis to determine weights, lengths, and sex differences amongst the organisms.
Mortality data showed a NOEC value >1.10 mg I. u hi I e the pre-metamorphosis portion of the
study showed a NOEC of 0.59 mg/L and a LOEC of 1.05 nig/ L for both snout-vent length and
weight. This LOEC of 1.05 mg/L was associated with 5% (snonl-\ enl length) and 14% (weight)
decrease compared to controls, respectively A significant increase in the median metamorphosis
time was obsen ed in the I .<>5 mg I. PFOS treatment relati\ e to the control. The post-
metamorphosis I .()!ฆ(' was reported as I 05 mg/L. No LCso value was reported in that only 5.2
percent mortality was obsei \ ed in the highest exposure concentration (1.05 mg/L) at test
termination Independently-calculated ECios could not be calculated as EPA was unable to fit a
model with significant parameters Instead, the author-reported MATC of 0.7871 mg/L PFOS
based on grow th (measured as mean body weight at metamorphosis) was used quantitatively to
derive the draft chronic water column criterion for freshwater.

C.2.13 Thirteenth Most Sensitive Freshwater Genus for Chronic Toxicity: Lithobates (frog)

The chronic flow-through measured toxicity of PFOS (potassium salt, 98% purity) to the

northern leopard frog, Lithobates pipiens (formerly, Ranapipiens), was investigated by Ankley

et al. (2004) Two PFOS stock solutions (708 and 21.7 mg/L) were prepared by dissolving solid

PFOS with one liter of Lake Superior water in a glass carboy for 24 hours and then brought to a

C-51


-------
volume of 18 L for the final stock solutions. Contents were stirred at room temperature (~20ฐC)
for 24 hours prior to being used. Solutions were pumped from the carboys to the glass aquaria
through Teflonฎ tubing using fluid metering pumps equipped with stainless-steel rotary
dispensers. Target concentrations were achieved by diluting the high and low stock solutions
with an appropriate volume of the Lake Superior (control) water. The PFOS stock solutions were
renewed every seven days. Fertilized eggs were collected from (irand Lake (St. Louis County,
MN), near a sandy shoreline with no development. Tests were initiated with stage 8/9 embryos;
animals were gently separated with a plastic spatula from the egg mass, inspected under a
microscope for viability (evidence of cell division), and randomly allocated to treatment groups.
Exposures were conducted in glass aquaria in 10 L of water, which was continually renewed at a
flow rate of about 50 mL/minute (72 L/davJ Duplicate tanks at target (nominal) PFOS
concentrations of 0.03, 0.1, 0.3, 1, 3, and 1<> mu I. and lour replicate control aquaria were used.
Embryos (n=120) were placed in each aquarium, in addition, two of the control tanks and the
duplicate tanks at <>. I and I mu PI-OS I. received an extra 80 organisms (total of 200) at test
initiation to pro\ ide animals lor determination of PFOS concentrations during the early part of
the assay. Although hiomass \ aried between the tanks with 120 versus 200 tadpoles, in both
situations total loading to the system w as more than two orders of magnitude lower than
guidance recommended for a test at this flow rate. Water temperature was maintained at 20 ฑ
0.5ฐC, and the photoperiod (provided by fluorescent lights) was a constant 16-hr:8-hr light:dark
cycle. On hatching (at approximately six days), animals were fed a mixture of live brine shrimp,
ground trout chow, and Tetrafin ad libitum two times daily. Dead organisms were removed daily
and inspected for gross abnormalities. On test day 6, 10 newly-hatched (<24 hours) animals were
randomly removed from each tank, preserved in 10% neutral buffered formalin, and

C-52


-------
subsequently examined for developmental anomalies. Groups of animals were randomly selected
from each treatment (excluding the 10 mg/L group, which had been terminated because of high
mortality) on test days 35 (10 tadpoles/tank) and 54 (three tadpoles/tank). The animals were
weighed, and developmental stage was recorded, before being processed for PFOS tissue
analysis. The first tadpoles to undergo complete metamorphosis (defined as emergence of the
forelimbs) were observed on test day 60. Metamorphs were remo\ ed from the test tanks,
sacrificed with an overdose of MS-222, weighed, measured (total and snoni-\ ent length), and
assessed for gross abnormalities. Metamorphosis of the tadpoles continued o\ er the next 51 days,
until the test was terminated, when remaining tadpoles were counted, staged, and weighed. A
subset of tadpoles from the control and 3 mg PFOS/L treatments were processed for histological
analysis of the thyroid gland when they were sampled at forelimh emergence. The most sensitive
apical chronicvalue was the 112-day growth M.VI'C of 1.732 mu L PFOS, followed by the 5-
week LCso of 6 21 mu I. PI-OS These data are considered quantitative even though the control
mortality was >2<)"() at lest termination fNote' Excessive mortality of amphibian larvae should be
expected within the lull duration of this experiment given the life history strategy employed by
amphibians Therefore, the ohser\ ed control survival for this study was considered within the
acceptable range for this species and the toxicity data should be limited to the first 10 weeks of
the experiment) The study author-reported value (112-d growth MATC of 1.732 mg/L) was
used quantitative!\ to deri\ e the draft chronic water column criterion.

Hoover et al. (2017) also evaluated the chronic toxicity of PFOS (>98% purity) to
Lithobatespipiens. Test solutions were renewed every four days and exposure concentrations
were measured prior to and after each water change. Stock solutions consisted of 1 g of chemical
dissolved in 2 L of Milli-Q water, then vacuum-filtered before storage in polycarbonate bottles.

C-53


-------
Eight northern leopard frog egg masses were collected during early spring from a temporary
pond at the Purdue Wildlife Area in West Lafayette, IN, and randomly assigned to outdoor ~100
L wading pools. After hatching, larvae were checked daily for mortality and fed Purina Rabbit
Chow ad libitum. Treatments consisting of control and exposure to PFOS at three concentrations
(nominally 0.010, 0.100, and 1.0 mg/L) were placed in two replicates on adjacent shelves within
an environmental chamber. Experimental units consisted of 15 T. pUisiic aquaria filled with 7.5 L
of filtered, UV-irradiated well water. Tadpoles (n=35 per aquarium) w civ randomly assigned to
the experimental units. Prior to addition to aquaria, a subset of animals was e\ami ned to confirm
development at Gosner stage 26, when hind limb buds start to develop. Tadpoles with visible
irregularities in morphology, coloration, or behavior were excluded. Animals were maintained at
20 ฑ 2ฐC with a 12-hr: 12-hr light:dark photoperiod for 10 days lo acclimate to indoor conditions
and were fed a Tetramin slurry ad libitum. \\ ater changes (were conducted every four
days. Tadpoles were exposed lor 4<) days and were monitored daily for mortality and
abnormalities. A water sample ( ~5 ml.) was taken immediately prior to and after each water
change to monitor concentration of test chemicals I-very 10 days, six animals were randomly
collected from each aquarium The animals were euthanized, measured (total length at 10 days,
snoul-\ cut length otherwise), and staged (Gosner) prior to storage at -20ฐC for chemical
analyses. After 4') days, the depuration phase was initiated by removing animals, cleaning each
aquarium with a methanol-soaked sponge, and rinsing to remove adsorbed compound. Aquaria
were refilled with clean water; animals were returned to the same aquarium and monitored as
described above. Water changes were carried out every four days with fresh water, and a water
sample was taken prior to each water change. Two tadpoles were sampled every 10 days for an
additional 30 days. The 40-day chronic value was 0.0316 mg/L PFOS based on Gosner stage

C-54


-------
reached at test termination. This study was deemed quantitative, even though the exposure was
renewal rather than the required flow-through design for chronic ALC development. Also, PFOS
was detected in the control organisms. While the concentrations were much lower than any of the
PFOS treatment groups (3 orders of magnitude lower), it indicated that some potential
contamination may have occurred in the controls. The study author-reported value (a
developmental-based MATC of 0.0316 mg/L) was used quanlitali\ civ to derive the draft chronic
water column criterion.

C.2.14 Fourteenth Most Sensitive Freshwater Genus for Chronic Toxicity, rhvsclla (snail)
Funkhouser (2014) conducted a 44-day renewal test of PFOS (potassium salt, >98%

purity) with Physella heterostrophapomilia as part of a Master's thesis at Texas Tech

University, Lubbock, TX. Egg masses from I no />. pomilia adults were collected from Canyon

Lake 6, Lubbock Lakes System, Lubbock. T.\. in May 2<)|3 and used lor testing. Dilution water

was moderately hard reconstituted laboratory water (3.0 u CaSO i. 3 n g MgS04, 0.2 g KC1, and

4.9gNaHC03 added to 5<) I. deionixed water). Pholoperiod was 12-hr: 12-hr, light:dark at an

unreported light intensity PI-OS was dissoKed in dilution water to prepare the test

concentrations l-xposure \ essels were 25<) mL polypropylene containers containing 200 mL of

test solution The test employed two replicates composed of four egg masses each with an

average of 3 7 25 euus/egg mass at start, then truncated to just four snails per replicate once snails

hatched. The test consisted of seven test concentrations plus a negative control. Nominal

concentrations were 0 (negative control), 10, 20, 40, 50, 70, 80, and 90 mg/L. Exposure

concentrations were reportedly measured, but concentrations were not provided. Analyses of test

solutions were performed using LC-MS/MS. Standards were used as part of the analytical

method, but details were not reported. The reporting limit was 0.010 mg/L. Experiments were

conducted in incubators set to 25ฐC, which did not vary more than 1ฐC during the course of the

C-55


-------
studies. No other water quality parameters were reported as having been measured in test
solutions. Negative control survival was not reported specifically for the test, but was reported to
be 85-100% across all experiments. The 44-day life-cycle MATC was 14.14 mg/L (from the
study author-reported NOEC and LOEC, 10 and 20 mg/L, respectively) for mean number of eggs
per egg mass. The independently-calculated ECio for the same endpoint was 8.831 (6.170 -
10.88) mg/L. The independent statistical analysis was conducled using data that was estimated
(using Web plot digitizer) from the figures presented in the paper. This chronic value was
acceptable for quantitative use and was used to deri\ e the criteria

C.3 Comparison of Quantitative Data usetl to Derive Freshwater Criteria

C.3.1 Aquatic Insects

While comparing the effects of PI- OS across studies presented a number of challenges,

especially with differences in test species, methodologies, exposure durations, and observed
endpoints, in general there appeared to be a number of similarities and few differences between
the two aquatic insect toxicity studies used quantitatively to derive the PFOS chronic freshwater
criterion. In both IJois el al (2<)|<)) and MacDonald et al. (2004) effects of chronic exposures to
PFOS on sur\ i\ al and emergence were observed in the damselfly (Enallagma cyathigerum) and
the midge (( hironomus ililmiis). respecli\ ely. Additionally, both studies measured effects of
PFOS on grow th I lowever. si nee the studies focused on very different life stages for this
endpoint (on growth in emerged adult damselfly from Bots et al. 2010; and on growth in larval
midge from MacDonald et al. 2004), the toxicity data for growth could not be compared.

In the early phases of both experiments, the effects of PFOS on aquatic insects did not
appear to be similar. Bots et al. (2010) reported a 10-day NOEC of 0.1 mg/L, a LOEC of 1.0
mg/L, and a MATC of 0.3162 mg/L for survival. The LOEC was associated with a 79% decrease
in survival compared to the control group. In contrast, MacDonald et al. (2004) reported a 10-day

C-56


-------
NOEC of 0.0491 mg/L, a LOEC of 0.0962 mg/L, and a MATC of 0.0687 mg/L for survival. The
midge LOEC was associated with an 8.2% decrease in survival compared to the control group.
During the early phases and exposure durations of both experiments, it appeared that the effects
of PFOS were not similar between the two species and that the midge was more the sensitive
than the damselfly. However, after this initial phase, the effects of PFOS on the damselfly and
the midge became more similar.

In the later phases of the tests, Bots et al. (2010) reported a 1>-chi\ NOliC of 0.01 mg/L,
aLOEC of 0.1 mg/L, and aMATC of 0.0316 mg/L for survival. This 150-day I.OIvC was
associated with a 19% decrease in nymph survival compared to the control. Similarly.
MacDonald et al. (2004) reported a 20-day NOEC of 0.021 7 mg/L, a LOEC of 0.0949 mg/L, and
a MATC of 0.0454 mg/L for survival. The I.OI-C was associated with a 61% decrease in larvae
survival compared to the control. EPA's independently-calculated ECio was 0.0171 mg/L using
estimated, treatment le\ el summary data presented in the paper (using Web plot digitizer). This
independently-calculated I X"i- and the author reported 20-day MATC of 0.0454 mg/L for
chironomid survival was similar to the author-reported 150-day survival MATC of 0.0316 mg/L
for the damselfly. The greatest sensiti\ ity of the test organisms based on the survival endpoint of
each rcspecli\ e test likely occurred in a similar life stage (later development and about to
undergo metamorphosis). And therefore, were more comparable than any of the other survival
toxicity values from these studies (i.e., the 10-day values for damselfly and the 10-day values for
midge), which were focused on the effects of PFOS on much less comparable instars, especially
given that odonates have a much longer development and life span compared to midge. These
results indicated that PFOS exposure to aquatic insects in later life stages were likely similar.

C-57


-------
Likewise, both studies observed similar effects of PFOS on emergence. However, this
endpoint was less certain compared to the survival endpoints given the level of data provided in
both of the papers. Bots et al. (2010) observed that nymphs exposed to the lowest PFOS
treatment group metamorphosed less (75.5%) than those in the control (92.5%) and reported a
LOEC of <0.01 mg/L for metamorphic success. This less than LOEC was associated with a
18.4%) decline. Metamorphic success could not be measured in the remaining PFOS treatment
groups because only nymphs in the control and the lowest PFOS treatment group (or 0.01 mg/L)
survived until metamorphosis. In comparison, MacDonald et al. (2004) obser\ ed a reduction in
total emergence in all of the PFOS treatment groups compared to the control. The sludy authors
reported aNOEC of <0.0023 mg/L, aLOEC of 0.0217, a MATC of 0.0071 mg/L and anECio of
0.0893 mg/L. The reported LOEC was associated with a 41.(•>"<> decline in midge emergence
compared to the control. EPA's independently-calculated I X'iฆ ฆ was 0.0102 mg/L using
estimated, treatment le\ el summary data presented in the paper (using Web plot digitizer). This
independently-calculated I X"i- and the author reported LOEC of 0.0217 mg/L for midge were
similar to the I .()!ฆ(' of <> <> I mu I. reported for damselfly by Bots et al. (2010), and indicates
that like the later stage. sur\ i\ al endpoints summarized above, PFOS may have a similar level of
effect on insect metamorphosis and emergence.

These apparent similarities in the chronic effects of PFOS to aquatic insects support the
toxicity values LIW used lor chronic criteria derivation from the two studies, as well as the
ranking of these two species (as first and third most sensitive) to derive the chronic freshwater
criterion. However, additional replicate level data would have been helpful to fully understand
the observed effects of PFOS individually and to compare across the full PFOS toxicity dataset.

C-58


-------
Appendix D Acceptable Estuarine/Marine Chronic PFOS Toxicity Studies

D.l Summary Table of Acceptable Quantitative Estuarine/Marine Chronic PFOS Toxicity Studies

Species (

Method'

Tesl
Dui'iilioii

( Ik-iii ic;il /
Piiriu

pll

Temp.

<ฐC)

S;ilinil\


C'hroilic Yiiluc
liidpoinl

Author
Koporlcd
Chronic

\ illllC
(iiiu/l.)

IP A
( iilculiilcd
C h ron ic
Yiiluc
(ni}ป/l.)

liiiiil
C h ron ic
Yiiluc
(iiiu/l.

Species
Mc;in
Chronic
\ ;iluc

(inii/l.)

Kcl'crcncc

\sian green mussel

(60-65 mm),

Perna viridis

R, M

7 d

PFOS-K

98%



25

25

i:ciu

(growill condition
index)

0.03190

1)0033

0.0033

0.0033

Liu et al.
(2013)



Copepod (nauplii),

Tigriopus japonicus

R, U

20 d

PFOS
Unreported

-

25

32

\1 VK

(do\ olopnionlal stage)

0.7071

-

0.7071

0.7071

Han et al.
(2015)



Mysid (< 24 hr),

Americamysis bahia

F, M

35 d

PFOS-K
90.49%

8.2-
8.4

25



\1 VK

(roproduclion. growili)

0.3708

-

0.3708

0.3708

Drottar and

Krueger

(2000h)

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D-dissnl\ ed. I )icl diel;ii\. \l I -maternal transfer
b Values in bold used in SMCV calculation.

D-l


-------
D.2 Detailed PFOS Acute Toxicity Study Summaries and Corresponding

Concentration-Response Curves (when calculated for the most sensitive
genera)

The purpose of this section was to present detailed study summaries for tests that were
considered quantitatively acceptable for criteria derivation, with summaries grouped and ordered
by genus sensitivity. Data for chronic PFOS toxicity were available for two saltwater
invertebrate species, representing two genera and two families. The data a\ ailable for saltwater
fish fulfilled only two of the eight MDRs, therefore EPA could not de\ clop chronic
estuarine/marine criteria following the 1985 Guideline methods.

D.2.1 Most Sensitive Estuarine/Marine Genus: I'erna (mussel)

Liu et al. (2013) evaluated the chronic effects of peril uorooctanesulfonate, potassium salt

(PFOS-K, CAS# 2795-39-3, 98% purity, purchased from Sigma-Aldrich) on green mussels,

Perna viridis) via a 7-day measured, static-renewal study The mussels were obtained from a

local farm in Singapore, and subsequently acclimated to laboratory conditions for seven days

before testing Mussels were kept at a salinity of 25 ppt (artificial seawater) and a temperature of

25ฐC. Forty mussels (MMo mm length) per 5<)-l. polypropylene tank, with duplicate, were

exposed to measured PTOS concentrations of 0 (control), 0.00012, 0.0011, 0.0096, 0.106 and

0.96S mg I. Mussels were led a commercial marine micro-alga purchased from Reed

Mariculture on renewal days, which occurred every two days, two hours before the solution

renewal. PFOS concentrations were verified through water and muscle tissue samples via liquid

chromatography-tandem mass spectrometry. Weights and lengths were determined on days 0 and

7. A NOEC of 0.0096 mg/L and a LOEC of 0.106 mg/L was determined for the growth condition

index. No LCso value was reported. EPA's independently calculated ECio is 0.0033 (0.00330 -

0.00332) mg/L. This ECio was used quantitatively to derive the draft chronic water column

criterion for freshwater.

D-2


-------
D.2.2 Second Most Sensitive Estuarine/Marine Genus: Americamysis (mysid)

Drottar and Krueger (2000h) reported the results of a life-cycle, 35-day flow-through,

measured test of PFOS-K (potassium salt, 90.49% purity) with Americamysis bahia (formerly

Mysidopsis bahia). This Good Laboratory Practice (GLP) test was conducted at the Wildlife

International, Ltd. toxicology facility in Easton, MD in June, 1999. The test followed U.S. EPA

OPPTS 850.1350, and ASTM Standard E 1191-90 test guidelines. Mysids used for the test were

neonates less than 24 hours old at test initiation. The dilution water was liIk-red natural seawater

collected at Indian River Inlet, DE diluted to a salinity of approximately 20 ppl with well water

[pH: 8.3 (8.2-8.4); TOC: >5.8 mg/L; temperature 25 2 (']. Photoperiod was lo S hours,

light:dark with a 30-minute transition period. Light as pro\ ided at an intensity of 023-815 lux.

A primary stock solution was prepared in dilution water at Kl) 5 muL. Tt was mixed until all of

the test substance was dissolved prior to use. AIk-r mixing. Ihe piiniai\ stock was proportionally

diluted with dilution water to prepare the five additional lest concentrations. Exposure vessels

were glass beakers with nylon mesh screens on each side placed in 9 L glass aquaria with 5 L of

test solution. After mysids reached sexual maturity, they were placed in pairs in glass petri dishes

to obsei \ e reproduction The test employed lour replicates of fifteen mysids each in six

measured test concentrations plus a neuati\ e control. Nominal concentrations were 0 (negative

control), 0 <>X(\ () I 7. 0.34, n W, 1.4, and 2.7 mg/L. Mean measured concentrations were <

0.0458 mg/L (the LOO), n 1157, 0.12, 0.25, 0.55, 1.3, and 2.6 mg/L, respectively. Analyses of test

solutions were performed at Wildlife International Ltd. using HPLC/MS. Measured values

ranged from 66 to 96% of nominal. Mortality from test initiation to pairing (day 20) in the 0.057,

0.12, 0.25, 0.55, 1.3, and 2.6 mg/L treatment groups was 8, 25, 18, 17, 32 and 100%),

respectively, and mean control mortality was 22%. From pairing until test termination (day 20 to

day 35) survival was greater than 90% in the control and all but the 1.3 mg/L treatment, which

D-3


-------
had 57% survival during that period. The 35-day NOEC (reproduction and growth) was 0.25
mg/L, and the corresponding 35-day LOEC was 0.55 mg/L. The EPA-calculated MATC was
0.3708 mg/L. The ECio could not be calculated at this time given the level of data that was
presented in the paper. The chronic value was considered acceptable for quantitative use despite
the control survival of 78% because it was only slightly below the So0 (. survival threshold, and
because there were no other deficiencies in the study design.

D.2.3 Third Most Sensitive Estuarine/Marine Genus: Tiizriopus (conenod)

A 20-day renewal, unmeasured full life-cycle test with PFOS (analytical grade) was

conducted on the copepod, Tigriopus japonicus (non-North American species) In I l:m et al.

(2015). Copepods were cultured and maintained in 0.2 urn tillered artificial seawaler adjusted to

32 psu salinity and 25ฐC under a 12-hour pholopcriod. 1. japomciis were fed with green algae,

Tetraselmis suecica. PFOS (100 mg/L in MeOI I) was concentrated In evaporation and re-

dissolved in DMSO lo oblain a maximum slock concentration (1,000 mg/L). The PFOS stock

was diluted with artificial seawaler lo obtain four nominal test concentrations (0, 0.25, 0.5 and 1

mg/L PFOS) The final concentration of DMSO in seawater was 0.001% (v/v) or less for each

treatment Ten new l\ -hatched nauplii ( 12 hour post hatch) were allocated to each well of a 12-

well tissue culture plate with 4 nil. of lest solution. There were three replicates per each

treatment. Organisms were led algae during testing and 50% of test media was replaced daily.

Over the next 20 days, the development of the copepod's growth from nauplii to copepodite and

from nauplii to adults was determined daily based on morphological characteristics. Results were

presented as the number of days needed to reach the normal development stages. The highest test

concentration (1 mg/L PFOS) significantly increased the amount of time it took the copepods to

reach the development stage. Additionally, the authors assessed the reproduction of the copepods

by counting the nauplii produced by eight ovigerous females for 10 days in each well exposed to

D-4


-------
PFOS. However, it was unclear if this was a subsampling of the organisms used in the 20 day
developmental test or if an independent assay with adult females. Results are presented
graphically as daily nauplii production/individual. There was a statistically significant decrease
in production (daily nauplii production/individual) in the 0.25, 0.5 and 1.0 mg/L PFOS
concentrations compared to the control. It was decreased by approximately 50% in the highest
concentration (1 mg/L). While this endpoint was more sensiti\ c that the growth endpoint, the
publication is unclear about the method used for the reproduction test endpoint and whether it
was an independently-conducted 10-day test or a subsample of reproducing aclnlis were observed
from the 20-day test. EPA sought but did not receive responses to clarifying questions posed to
the authors. Additionally, the authors were asked if control survival for the test was above 80%
and if the authors could provide the data, liased on the information presented in the paper
without additional information and data pro\ it led In the authors to clarify adherence to EPA data
quality objectives and independent calculation and \ erilicalion of point estimates, the
developmental stage is considered lor quanlilali\ e use and the reproductive endpoint for
qualitative use The use of the ieproducli\ e endpoint could be changed based on input on
clarifying questions from the study authors The 2'i-day MATC (based on time to reach
development stage) was 0.7t)71 mg/l. and currently recommended by EPA as acceptable for
quantita1i \ e use

D-5


-------
Appendix E Acceptable Freshwater Plant PFOS Toxicity Studies

E.l Summary Table of Acceptable Quantitative Freshwater Plant PFOS Toxicity Studies

Species

Mi-Muxr

Tesl

Dui'iilioii

( hcniiciil
/ Piirilv

Pll

Temp.

<ฐo

iriw-i

Reported
I'.ITecl
C oncenI r;ilion
diiii/l.)

Reference

Diatom,

Navicula pelliculosa

S,M

96 hr

PFOS-K

86.9%

7.5-8.9

24

EC50

(area under growth curve)

252

(Sutherland and
Krueger 2001)



Green alga (1.5 x 104 cells/ml),

Chlorella vulgaris

s,u

96 hr

PFOS-K

95%

-

2'

IC50

(cell density)

81.6

Boudreau et al.
(2003a)



Green alga,

Raphidocelis subcapitata
(formerly, Selenastrum
capricornutum)

s,u

96 hr

PFOS-K
24-28ฐ/.,

-

23

EC50

(specific growth rate)

49.28b

3MCompany (2000)

Green alga,

Raphidocelis subcapitata

s,u

4 d

PFOS-K
Unknown

-

2 '

i :c5o

(cell count)

77.19

3MCompany (2000)

Green alga,

Raphidocelis subcapitata

s,u

7 d

PFOS-K
Unknown

-

2 '

EC50

(cell count)

76.68

3MCompany (2000)

Green alga,

Raphidocelis subcapitata

s,u

lu d

PFOS-K
Unknown

-

2 '

EC50

(cell count)

83.92

3MCompany (2000)

Green alga,

Raphidocelis subcapitata

s,u

14 d

PFOS-K
U nknown

-



EC50

(cell count)

76.78

3MCompany (2000)

Green alga,

Raphidocelis subcapitata (formerly
Pseudokirchneriella subcapitata)

S. \1

l><> hr

PFOS-K

90.49%

"4-S4

24

EC50

(cell density)

71

Drottar and Krueger
(2000g)

Green alga (1.5 x 104 cells/ml),

Raphidocelis subcapitata (

S. 1

9(> III'

PFOS-K
95%

-

23

IC50

(cell density)

48.2

Boudreau et al.
(2003a)



Green alga,

Scenedesmus quadricauda

S. \1

96 hr

PFOS-K

99%

7

22

EC50

(growth inhibition rate)

89.34

Yang et al. (2014)



Duckweed,

Lemna gibba

S, M

"d

PFOS-K

86.9%

7.5

25

IC10

(frond number)

18.06

Desjardins et al.
(2001)



Water milfoil (5 cm apical shoots),

Myriophyllum sibiricum

S,M

14 d

PFOS-K
Unreported

-

-

EC10

(wet weight)

0.7

Hanson et al. (2005)

Water milfoil (5 cm apical shoots),

Myriophyllum sibiricum

S,M

28 d

PFOS-K
Unreported

-

-

EC10

(wet weight)

0.19

Hanson et al. (2005)

E-l


-------
Speeies

Method"

Test
Dunilion

( hcmiciil
/ Piiriu

pll

Temp.

(ฐC)

iiivci

Reported

r.nwi

( nneenlnilion
(inii/1.)

Reference'

Water milfoil (5 cm apical shoots),
Myriophyllum sibiricum

S,M

42 d

PFOS-K
Unreported



-

EC 10

(wot weight)

0.6

Hanson et al. (2005)



Water milfoil (5 cm apical shoots),

Myriophyllum spicatum

S,M

14 d

PFOS-K
Unreported

-

-

j :c in

(plant length)

4.8

Hanson et al. (2005)

Water milfoil (5 cm apical shoots),

Myriophyllum spicatum

S,M

28 d

PFOS-K
Unreported

-

-

EC in

(dry weight)

3.3

Hanson et al. (2005)

Water milfoil (5 cm apical shoots),

Myriophyllum spicatum

S,M

42 d

PFOS-K
Unreported

-

-

EC10

(wet weight)

3.5

(Hanson et al. 2005)

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D=dissol\ cd. \k not reported

b The independently-calculated EC50 value was 176.0 mg/L as the test substance, or 49.28 nig/L based on (lie percentage ol' PFOS-K (active ingredient 28%) in the test substance.

E-2


-------
E.2 Summary of Plant PFOS Toxicity Studies Considered in the Aquatic
Life Criterion Derivation

E.2.1 Diatom. Navicula pelliculosa

Sutherland and Krueger (2001) conducted a 96-hour static acute algal growth

inhibition test on PFOS (potassium salt, 86.9% purity) with the freshwater diatom, Navicula

pelliculosa. The GLP test was conducted at the Wildlife International. Ltd. in Easton, Maryland

in February-March, 2000. The test followed USEPA OPPTS K5<) 54<)i) (I S.EPA 1996) and

ASTM 1218-90E (ASTM 1990). The freshwater diatom was provided iVoni in-house cultures

that had been actively growing in the culture medium for at least two weeks. The test media was

prepared by adding the stock nutrient solution to purified well water according to ASTM 1218

and adjusting pH to 7.5. Seven measured concentrations (0, 62 3. 83.2, 111, 150, 206, 266, 335

mg/L PFOS) were tested from one negath e control and six nominal concentrations: 61.5, 81.3,

110, 147, 198, 264 and 347 mg/L based on PSOI'-K purity Solutions were stirred for

approximately 24 hours before testing Lxposures were conducted in 250 mL plastic Erlenmeyer

flasks containing |no ml. solution and plugged with foam stoppers. Each flask contained lxlO4

cells/mT. and each test concentration had three replicates. Flasks were incubated in

environmental chambers at 24 2 (' under constant illumination (4,300 lux) and shaken

continuously at 100 rpm. pi I in the test solutions ranged from 7.5-8.9 over the exposure period.

Samples were collected daily to determine cell density and to calculate area under the curve and

growth rates. The cell density of the control replicates increased by greater than two orders of

magnitude during the test. The 96-hour EC so, based on area under the growth curve, was 252

mg/L PFOS (NOEC<62.3 mg/L). The plant value was acceptable for quantitative use.

E-3


-------
E.2.2 Green alga. Chlorella vulgaris

Boudreau et al. (2002) performed a 96-hour static acute algal growth inhibition test on

PFOS (potassium salt, 95% purity) with Chlorella vulgaris as part of a Master's thesis at the

University of Guelph, Ontario, Canada. The same information was subsequently published in the

open literature as Boudreau et al. (2003a). The acute algal growth inhibition tests followed

protocols found in ASTM E 1218-97a (ASTM 1999b) and Geis el ill (2000). All treatment

concentrations were based on the PFOS anion (without T<) and solutions were prepared in

laboratory-grade distilled water. C. vulgaris (UTCC 266 strain) were obtained as slants from the

University of Toronto Culture Collection (UTCC; Toronto, Canada). Toxicity testing consisted

of initial range-finder tests (0, 28, 56, 113, 225, and 45<) mg I.) followed by at least two

definitive tests (0, 12.5, 25, 50, 100, 200. and 4<)i) mg/L.) Tests were conducted in 20 mL

solution in 60 x 15 mm polyethylene disposable petri dishes Each dish contained 1.5xl04

cells/mL and each test concentration had four replicates Tests were continuously illuminated

with cool-white fluorescent light between 3.S00 and 4,200 lux and incubated at 23ฑ1ฐC. Each

dish was manually shaken twice a day dining testing. Toxicity test endpoints included cell

density and chlorophyll a content The most sensitive endpoint, cell density, had a reported

NOEC of S 2 nig/L and an IC*.. SI (•> mg I.. and was quantitatively acceptable for use.

E.2.3 Green alua. Kaphidocelis subcapitata

3M( 'ompany (2000) provides the results of a 96-hour toxicity test completed in 1991

with the green alga, Raphidocelis subcapitata (formerly Selenastrum capricornutum), and PFOS-

K (perfluorooctancesulfonate potassium salt, CAS # 2795-39-3) in a formulated mixture with

diethylene glycol butyl ether and water (mixed product FM-3820, with 24-28% PFOS-K). Based

on this purity author made calculations to adjust test concentrations using 28% active ingredient,

but the presence of diethylene glycol could also contribute to toxicity. The toxicity test followed

E-4


-------
OECD test guidelines with five test concentrations and control in a static unmeasured exposure.
A stock culture of the alga was obtained from the Culture Collection of Algae at the University
of Texas at Austin. Alga were transferred to 250 mL flasks with an initial density of lxlO4
cells/mL and 100 mL of test solution. There were three replicates for each of the five nominal
test concentrations (62.5, 125, 250, 500 and 1,000 mg/L) and control. Synthetic nutrient medium
was used as the dilution media for all test treatments. Alga were grow n at 23 ฐC and continuously
shaken. The author reported EC50, based on average specific growth rule, was 255 mg/L as the
test substance, or 71 mg/L based on the percentage of PFOS-K (active ingredient 28%) in the test
substance. The independently-calculated ECso value was 176 n mg/L as the test substance, or
49.28 mg/L based on the percentage of PFOS-K (active ingredient 28%) in the test substance and
is acceptable for quantitative use.

3MCompany (2000) provides the results of lour separate toxicity tests completed in
1981 with the green alga. Ruplmlocelis subcapiiani (formerly Selenastrum capricornutum), and
PFOS-K (perfluorooclancesullbnale potassium salt, CAS # 2795-39-3, unknown purity). The
toxicity tests followed a protocol modi lied from IJ.S.EPA-600/9-78-018 (1978), ASTM-E-35.23
(1981). OI X'I) 2<)| (l^7lJ). and ASTM STP #667. There were four separate exposure regimes: 1)
four day exposure + 10 day recovery period; 2) seven day exposure + seven day recovery period;
3) 10 day exposure four day recovery period; and 4) 14 day continuous exposure. A bacteria-
free culture of the alga was obtained from the U.S.EPA (Corvallis, OR) and stored in the dark
until testing. Seven-day old stock cultures with an initial density of lxlO4 cells/mL were placed
in 250 mL flasks with 50 mL of test solution. There were three replicates for each of the six
nominal test concentrations (26, 40, 61, 93, 145 and 225 mg/L) and control. Nutrient medium
was used as the dilution media for all test treatments and were not renewed during the exposure.

E-5


-------
Alga were grown at 23ฐC and continuously shaken at 100 rpm. The author-reported EC so, based
on cell counts, was 82, 99, 98, and 95 mg/L, for the 4-, 7-, 10- and 14-day exposures,
respectively. However, it should be noted that the authors do not specify if the ECsos were
determined after the exposure period or the post observation period. The independently-
calculated ECso values were 77.19, 76.68, 83.92, 76.78 mg/L and are acceptable for quantitative
use.

Drottar and Krueger (2000g) conducted a 96-hour static acute algal growth inhibition
test on PFOS (potassium salt, 90.49% purity) with the freshwater alga, Rap/m/occ/is subcapitata
(formerly Selenastrum capricornutum). The (il.P test was conducted at the Wildlife
International, Ltd. in Easton, Maryland in April, 1999. The lest followed USEPA OPPTS
850.5400 (U.S.EPA 1996), OECD 201 (OECD NX4). and ASTM 1218-90E (ASTM 1990)
methodologies. The green alga was originally obtained from the culture collection at University
of Texas at Austin (or another supplier) and maintained at Wildlife International Ltd. for a
minimum of two weeks in culture medium Algae used in tests were in exponential growth
phase. The test media was prepared In adding the stock nutrient solution to purified well water
according to ASTM l2ISand adjusting pH to 7.5. Seven measured concentrations (<0.115, 5.5,
11, 2 1. 44. K(\ 11K) mg/L PI-OS) were tested from a negative control and six nominal
concentrations 5 7. II. 23. 4(\ 91, 183 mg/L based on PFOS-K purity. Test concentrations were
measured at test initiation, at 72 hours, and at test termination by HPLC-MS with a mean 99.1%
recovery. Solutions were stirred for approximately 24 hours before testing. Exposures were
conducted in 250 mL polycarbonate flasks containing 100 mL solution and plugged with foam
stoppers. Each flask contained lxlO4 cells/mL and each test concentration had three replicates.
Flasks were incubated in environmental chambers at 24ฑ2ฐC under constant illumination (4,300

E-6


-------
lux) and shaken continuously at -100 rpm. The pH in test solutions ranged from 7.4-8.4 over the
exposure period. Samples were collected daily to determine cell density and to calculate area
under the curve and growth rates. The 96-hour EC so, based on cell density and area under the
growth curve, was 71 mg/L PFOS (NOEC=44 mg/L). The plant value was acceptable for
quantitative use.

Boudreau et al. (2002) performed a 96-hour static acule algal growth inhibition test on
PFOS (potassium salt, 95% purity) with Raphidocelis subcapitata (formerly Pseudokirchneriella
subcapitata). The study was part of a Master's thesis at the University of Guclph. Ontario,
Canada and subsequently published in the open literature as Boudreau et al. (2003a). The acute
algal growth inhibition tests followed protocols found in ASTM E 1218-97a (ASTM 1999b) and
Geis et al. (2000). All treatment concentralions were based on the PFOS anion (without K) and
solutions were prepared in laboratory-grade distilled water R. subcapitata (UTCC 37 strain)
were obtained as skints from the I ni\ersil\ of Toronto Culture Collection (UTCC; Toronto,
Canada). Toxicity testing consisted of initial range-finder tests (0, 28, 56, 113, 225, and 450
mg/L) followed by at least two deliniti\ e tests (<). 12.5, 25, 50, 100, 200, and 400 mg/L). Tests
were conducted in 2d ml. solutions in (ฆป<> \ I 5 mm polyethylene disposable petri dishes. Each
dish contained I 5\T04 cells mL and each test concentration had four replicates. Tests were
continuously illuminated with cool-white fluorescent light between 3,800 and 4,200 lux and
incubated at 23: I (' l .ach dish was manually shook twice a day during testing. Toxicity test
endpoints included cell density and chlorophyll a content. The reported NOEC and ICso based on
most sensitive endpoint, cell density, were 5.3 mg/L and 48.2 mg/L. The plant values from the
study were acceptable for quantitative use.

E-7


-------
E.2.4 Green alga. Scenedesmus quadricauda

Yang et al. (2014) conducted a 96-hour static, measured test on the growth effects of

PFOS (potassium salt, CAS # 2795-39-3, 99% purity) with the green alga, Scenedesmus

quadricauda. Algae were obtained from in-house cultures from the Chinese Research Academy

of Environmental Sciences. The algae used for testing were inoculated at a cell density equal to

2.0xl04 cells/mL in 50 mL beakers. PFOS was dissolved in deionized water and DMSO (amount

not provided) and then diluted with M4 medium. Alga were exposed lo <> (solvent control),

50.00, 65.00, 84.50, 109.85, 142.81 and 185.65 mg/L PFOS (each treatment was replicated three

times). While the text implied the exposures were sialic, the supplemental information provided

the measured test concentrations in the highest and lowest test treatments both before and after

renewal. Measured concentrations ranged from 42.56 mg I. (before renewal) to 49.78 mg/L

(after renewal) in the lowest treatment, and from 165 M (before renewal) to 183.90 mg/L (after

renewal) in the highest treatment. The experiments were conducted at; 22 ฑ 2ฐC with a 12 hour

light:dark cycle. The initial pH of the test solution was 7.0 ฑ 0.5, total hardness was 190ฑ0.1

mg/L as CaC03, and total organic carbon was 0.02 mg/L. Algae concentrations in the beakers

were measured daily with a microscope The 96-hour ECso (based on growth inhibition) was

89.34 mu I. and was acceptable lor quantitative use.

E.2.5 Duckweed. Lemnauibba

Desjardins of al. (2001) performed a static, measured 7-day growth inhibition study on

the duckweed Lemna ^ibba withPFOS-K (perfluorooctanesulfonate potassium salt, 86.9% purity

from 3M Corporation). The test protocol from USEPA, OPPTS Number 850.4400 was followed.

Duckweed was cultured and tested at Wildlife International Ltd. in 20X AAP medium and were

actively growing for at least two weeks prior to testing. The pH of the medium was adjusted to

pH 7.5 with HC1 and filtered to sterilize before use. Test chambers were covered 250 mL plastic

E-8


-------
beakers with 100 mL of culture medium or test concentration and held at 25ฐC under continuous
warm-white lighting with a target intensity of 5,000 lux. Fronds of duckweed were exposed to
six test concentrations and a control with three replicates for each treatment. PFOS
concentrations in the test medium were measured on day 0, 3, 5 and 7 with mean reported
concentrations of <4.39 (method limit of quantitation, control), 7.74, 15.1, 31.9, 62.5, 147 and
230 mg/L PFOS active ingredient. Growth was defined as an increase in the total number of
fronds in each replicate and measured by direct count on day 3, 5 and 7 Frond numbers on day
seven in the 147 and 230 mg/L test treatments were inhibited by 65 and 81" 0 as compared to the
control. The reported 5-day ICio based on frond number was 3<~> 7 mg/L PFOS. The
independently-calculated 5-day ICio value was 18.06 mg/L and is acceptable for quantitative use.

E.2.6 Watermilfoil. Myriophyllum sp.

Hanson et al. (2005) conducted a 42-day toxicity study of PI OS (potassium salt, purity

not provided) with the submerged waterni ill oils. \ lyviophyllum spicatum and M sibiricum. The
study was conducted in I Z.noo I. outdoor microcosms at the University of Guelph Microcosm
Facility located in ()ntaiio. Canada luich microcosm was below ground and was flush with the
surface Plastic trays H11 eel with sediment (1:1:1 mixture of sand, loam and organic matter,
mostly manure) were placed in the bottom of each microcosm. The total carbon content of the
sediment ^as I (•> 3".. Ten apical shoots, 5 cm in length, from in-house cultures using the same
sediment were transferred to each microcosm, with three separate microcosms used for each
treatment (0, 0.3, 10 and 30 mg/L). Endpoints of toxicity that were monitored on days 1, 14, 28
and 42 of the study included growth in plant length, root number, root length, longest root, node
number, wet mass, dry mass and chlorophyll a and b content. PFOS treatments were dissolved in
the same water (well water) used to supply the microcosms. Measured concentrations in the
microcosms were reported in a companion publication (Boudreau et al. 2003b). Results from the

E-9


-------
companion paper showed that measured concentrations remained similar to nominal
concentrations throughout the entire exposure period and did not change appreciably over the
course of the study. Water quality (i.e., pH, temperature, D.O., hardness, and alkalinity) and light
levels were measured regularly, but were not reported. M. sibiricum was more sensitive to PFOS
concentrations thanM spicatum. The 42-day ECio (based on wet weight) was 0.6 mg/L forM
sibiricum and 3.5 mg/L forM spicatum. The plant values were acceptable for quantitative use.

E-10


-------
Appendix F Acceptable Estuarine/Marine Plant PFOS Toxicity Studies

No data at this time.

F-l


-------
Appendix G Other Freshwater PFOS Toxicity Studies

G.l Summary Table of Acceptable Qualitative Freshwater PFOS Toxicity Studies

Species (life's!;iiicI

Mclhori'1

Tesl

l)iir;ilitiii

( Ik-iii ic;il /
I'llril\

I'"

Temp.

(ฐC)

r.nwi

Chronic
l.imils
(NOI'.C-

i.or.C)

(111^/1.)

Reported

i.nvci

(one.

Delieieneies

Reference

Cyanobacteria,

Anabaena sp.

S,M

24 hr

PFOS
98%



-

LC50

(bioluminescence)

-

I(> y>

Duralion loo shorl lor
a planl losi. non-apical
endpoint

kudea-l'alnmaies

et al. (2012;

Cyanobacteria,

Anabaena sp.

S,U

24 hr

PFOS-K

98%

7.8

28

rc5o

(bioluniincscence)

-

83.51

Duration too short for
a plant test, non-apical
endpoint

Rodea-Palomares
et al. (2015)



Green alga

(7.0 x 105 cells/ml),

Chlorella vulgaris

S,M

96 hr

PFOS-K

98%

-

-

i.oi:t

(chlorophyll a)

-

40

Missing exposure
details

Xu et al. (2017)



Green alga,

Raphidocelis subcapitata

S,M

72 hr

PFOS-K

98%

-

2 1 -24

i:( 50

(growth)

-

35

Duration too short for
a plant test, missing
some exposure details

Rosal et al. (2010)

Green alga,

Raphidocelis subcapitata

S,U

72 hr

H< >S-k

-

22

EC50

(growth inhibition)

-

35

Duration too short for
a plant test

Boltes et al.
(2012)



Green alga,
Scenedesmus obliquus

S,U

72 hr

H OS
L IIIVpol led

" 5

::

IC50

(growth rate reduction)

-

78.02e

Duration too short for
a plant test

Liu et al. (2008)

Green alga,
Scenedesmus obliquus

S,U

hr

H< )S

" 5

::

NOEC

(growth)

-

40

Duration too short for
a plant test

Liu et al. (2009)



Duckweed,

Lemna gibba

S, I

"d

H< )S-K
<>5"„

-

25

IC50

(wet weight)

-

31.1

Culture water not
characterized, missing
some exposure details

Boudreau et al.
(2003a)



Aquatic microcosm
(mixed invertebrate and
aquatic plant community)

S,M

35-42 d

PFOS-K
86".,

8.3-
8.6

15.9-
20.5

MATC

(zooplankton community
abundance)

3.0-10

5.478

Mixed species
exposure, static
chronic exposure

Boudreau (2002);
Boudreau et al.
(2003b)

Aquatic microcosm
(mixed invertebrate and
aquatic plant community)

S,M

35 d

PFOS-K
U lire ported

8.3

18

MATC

(zooplankton abundance;

Cyclops diaptomus
abundance)

1.0-10

3.162

Mixed species
exposure

Sanderson et al.
(2002)



G-l


-------
Species (lilcslii^c)

Method'

Tesl
Diimlion

( Ik-iii ic;il /
I'llril\

pll

Temp.

(ฐC)

r.nwi

Chronic
Limits

<\oi:( -
I.OKC)
(inii/l.)

Reported
r.ffccl
(one.
(inii/1.)

Deficiencies

Reference

Tubificid worm
(0.03g, 0.8cm),
Limnodrilus hoffmeisteri

S,M

96 hr

PFOS-K

99%

7

22

LC50

-

120.97

Atypical source of
organisms

Yang et al. (2014)

Tubificid worm,

Limnodrilus hoffmeisteri

S,U

24 hr

PFOS

>98%

5.0

23

I.C50

-

45.26

Duration too short for
an acute test, missing
some exposure details

Liu et al. (2016)

Tubificid worm,

Limnodrilus hoffmeisteri

S,U

24 hr

PFOS

>98%

6.0

23

LC50

-

4(> 23

Duration too short for
an acute test, missing
some exposure details

Liu et al. (2016)

Tubificid worm,

Limnodrilus hoffmeisteri

S,U

24 hr

PFOS

>98%

7.0

23

l.( 5(1

-

60.70

Duration too short for
an acute test, missing
some exposure details

Liu et al. (2016)

Tubificid worm,

Limnodrilus hoffmeisteri

S,U

24 hr

PFOS

>98%

80



l.( 5(1

-

64.48

Duration too short for
an acute test, missing
some exposure details

Liu et al. (2016)

Tubificid worm,

Limnodrilus hoffmeisteri

S,U

24 hr

PFOS

>98%

9.0



T.C50

-

65.74

Duration too short for
an acute test, missing
some exposure details

Liu et al. (2016)

Tubificid worm (3-4 cm),
Limnodrilus hoffmeisteri

R, U

48 hr

PFOS-K
98",,

6.2

::

l.( 50

-

23.81

Duration too short for
an acute test, missing
some exposure details

Qu et al. (2016)

Tubificid worm (3-4 cm),
Limnodrilus hoffmeisteri

R, U

48 hr

I'lOS-K

" (i

22

LC50

-

35.89

Duration too short for
an acute test, missing
some exposure details

Qu et al. (2016)

Tubificid worm (3-4 cm),
Limnodrilus hoffmeisteri

R, U

48 hr

\>\( )S-K

S.(i

22

LC50

-

39.80

Duration too short for
an acute test, missing
some exposure details

Qu et al. (2016)



Planarian (10-12 mm),
Dugesia japonica

R, U

lu d

H< )S-K

-

:u

LOEC

(regeneration: decreased
appearance of auricles)

<0.5-0.5

0.5

Duration too long for
an acute test and too
short for a chronic test

Yuanetal. (2014)



Freshwater mussel
(6 cm),

Unio ravoisieri

R, L

l><> hr

PFOS-K
>9S",i

8

18

LC50

-

65.9

Test species fed from
the natural freshwater
used

Amraoui et al.
(2018)



Mud snail (4.0 g, 2.0 cm)

Cipangopaludina
cathayensis

S,M

96 hr

I'lOS-K

99%

7

22

LC50

-

247.14

Source of organisms
may be problematic

Yang et al. (2014)



Snail (adult),

Lymnaea stagnalis

S,M

96 hr

PFOS-K

95%

-

20

LC50

-

196

Test species fed

Olson (2017)

G-2


-------
Species (lilcslii^c)

Method-'

Tesl
Diimlion

( Ik-iii ic;il /
Piiriu

pll

Temp.

(ฐC)

r.nwi

Chronic
l.imils

<\oi:( -
I.OKC)
(inii/l.)

Reported
r.lTecl
(one.
(inii/l.)

Deficiencies

Reference

Snail

(0-3 week, juvenile),

Lymnaea stagnalis

S, M

96 hr

PFOS-K

95%



20

LC50

-

150

Test species fed

Olson (2017)

Snail

(0-3 week, juvenile),

Lymnaea stagnalis

R,M

21 d

PFOS-K

95%

-

20

NOEC

(survival, feeding rate,
mass change, length
change, carbohydrate
concentration)

5o- 5< >

50

Duration too short for
a chronic test

Olson (2017)

Snail

(3-6 week, juvenile),

Lymnaea stagnalis

R, M

21 d

PFOS-K

95%

-

20

\T \TC

(mass change, length
change)

25-50

35.35

Duration too short for
a chronic test

Olson (2017)

Snail

(6-9 week, juvenile),

Lymnaea stagnalis

R, M

21 d

PFOS-K

95%

-

:u

noi:c

(survival, mass change.

length change,
carbohydrate and prolein
concentration)

50->50

>50

Duration too short for
a chronic test

Olson (2017)

Snail

(9-12 week, juvenile),

Lymnaea stagnalis

R, M

21 d

PFOS-K
<>5"„

-

20

\oi:c

(survival, feeding rate,
mass change, length
change, protein
concentration)

50->50

>50

Duration too short for
a chronic test

Olson (2017)

Snail (adult),

Lymnaea stagnalis

R, M

21 d

H< >S-k
<>5",,

-

20

MATC

(survival)

3.0-6

4.243

Duration too short for
a chronic test

Olson (2017)



Snail (5 mm),

Physella heterostropha
pomilia

(formerly, Physa pomilia)

S, M

5(i hr

H< )S-K

w;,

-

25

NOEC-LOEC

(avoidance)

< 30-30

-

Duration too short for
an acute test; atypical
endpoint

Funkhouser
(2014)

Snail (adult, 4 mo.),

Physella heterostropha
pomilia

(formerly, Physa pomilia)

R, \1

14 d

PFOS-K
>9S",i

-

25

LC50

-

94.99

Duration too long for
an acute test and too
short for a chronic test

Funkhouser
(2014)



Rotifer

(< 2 hr old neonates),

Brachionus calyciflorus

R,Ub

4 d

H< )S-K

98%

-

20

MATC

(intrinsic rate of
population increase and
resting egg production)

0.125-0.25

0.1768

Atypical

concentration-response
pattern

(Zhang et al.
2014)



G-3


-------
Species (lil'csliiiic)

Method'

Tcsl
Diimlion

( hcmic;il /
I'llril\

pll

Temp.

(ฐC)

r.nwi

Chronic
l.imils

i\oi:( -
I.OKC)
(niii/l.)

Reported

i.nvci

(one.
(niii/l.)

Deficiencies

Reference

Cladoceran (<24 hr),
Daphnia magna

R, U

25 d

PFOS-K

99%

7.8

20

MATC

(reproduction F0
generation)

(i () 1-0.1

0.03162

No consistent

concentration-response

relationship

Jeong et al. (2016)

Cladoceran (<24 hr),
Daphnia magna

S,U

48 hr

PFOS-Li
24.5%

8.6

20.1-
21.0

EC50

(death/immobility)

-

51.45

Inability to verify
author-reported LC50

3MCompany
(2000)

Cladoceran (0-12 hr),

Daphnia magna

R, U

28 d

PFOS-K
Unknown

7.6

22

MATC

(reproduction)

7.0-18 II

1 1 12

Inability to calculate
an EC 10 and
comments by authors

3MCompany
(2000)



Crayfish

(3 wk juvenile, 0.048 g),

Procambarus fallax f.
virginalis

R, M

38 d

PFOS-K

>98%

-

25

\1 \T(

(survival growth)

0.2->0.2

>0.2

Only two organisms
per exposure
concentration; invasive
species

Funkhouser
(2014)

Crayfish

(juvenile, 2 wk, 0.041 g),

Procambarus fallax f.
virginalis

R, M

7 d

PFOS-K
>9S"„

-

25

l.( 50

-

39.71

Duration too long for
an acute test and too
short for a chronic test,
only six organisms per
exposure

concentration, test
species fed; invasive
species

Funkhouser
(2014)



Oriental river prawn
(0.30 g, 4.0 cm),
Macrobrachium
nipponense

S,M

96 hr

I'lOS-K

99%

"

::

LC50

-

19.77

Source of organisms
may be problematic

Yang et al. (2014)



Yellow fever mosquito
(1st instar),

Aedes aegypti

S,U

4X hr

H'< )S
Unre purled

-

25

LC50

-

1.18

Duration too short for
an acute test, missing
some exposure details

Olson (2017)

Yellow fever mosquito
(1st instar),

Aedes aegypti

R, U

42 d

PF( )S
I 'nrepniied

-

25

MATC

(average time to
emergence)

0.05-0.125

0.079

Missing some
exposure details

Olson (2017)



Blue damselfly (larva, F2
instar stage),

Enallagma cyathigerum

R, U

4 mo

i'oiiluorooctanesulfomc
acid

tetraethylammonium

98%

-

21

MATC

(general activity, burst
swimming, foraging
success)

0.01-0.1

0.0316

Sporadic solution
renewal, behavioral
endpoints

(Van Gossum et
al. 2009)



G-4


-------
Species (lil'cs(;i;ic)

Mi'lhori1

Tesl
Diinilinn

( Ik-iii iciil /
PuriU

pll

Temp.

(ฐC)

HITi'ii

('limn ic
Limits

<\oi:( -
I.OKC)

(iiiii/l.)

Reported
I.ITcCl
('one.

(inii/1.)

Deficiencies

Reference

Mi due Hi "5 u. 1 2 am.

(Ivronomus plumosus

S. M

0(. lu-

H< )S-k

00",,

"

::

l.( 5(1

-

is: i:

Source of organisms
may be problemalic

Yaimclal (2<>I4)



Midge

(multi-generational),

Chironomus riparius

S,M

ll)

generations

(-20-28 d ea.)

PFOS
Unspecified

oo

l< 00

20

NOEC

(emergence,
reproduction, sex ratio)

-

0.0035

Only one exposure
concentration, static
chronic test

Stefani et al.
(2014)

Midge

(multi-generational),

Chironomus riparius

S,M

10

generations

(-20-28 d ea.)

PFOS
Unspecified

oo

l< 00

20

LOEC

(increased mutation rale)

-

(i 0035

Only one exposure
concentration, static
chronic test

Stefani et al.
(2014)

Midge (1st instar larva),
Chironomus riparius

S,M

-36 dd

(1st of 10
generations)

PFOS
Unreported

7.5-
8.2

20 1

loi:c

(F1 developmental lime,
adult weight, exuvia
length)

-

0.004

Only one exposure
concentration, static
chronic test,
significant responses
not observed in every
generation

Marziali et al.
(2019)



European eel
(juvenile, 138.3 g),

Anguilla anguilla

R, M

28 d

PFOS-K
OS",,

-

20

noi:c

(survival, growth)

(1.011-
>0.011

0.011

Not true ELS test (28
days beginning with
juvenile)

Roland et al.
(2014)

European eel
(juvenile, 138.3 g),

Anguilla anguilla

R, M

28 d

H< )S-k
OS",,

-

20

LOEC

(proteomic growth)

<0.00081-
0.00081

0.00081

Not true ELS test (28
days beginning with
juvenile), atypical
endpoint

Roland et al.
(2014)



Rainbow trout (immature,
16.4 cm, 22.7 g),

Oncorhynchus mykiss

Microcosm

12 d

H '< )S
so",,

o:

Mi-
ll. 5

NOEC

(mortality)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

Rainbow trout (immature,
16.4 cm, 22.7 g),

Oncorhynchus mykiss

Microcosm

i: d

PF( )S
89".,

o:

Mi-
ll. 5

LOEC

(decrease LSI and
condition index (K) in
females)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

Rainbow trout (female,
mature, 34.8 cm, 511.1 g),

Oncorhynchus mykiss

s,u

14 il

PF( )S
SO",,

-

12

NOEC

(mortality)

-

1

Atypical exposure, not
a true ELS test

(Oakes et al.
2005)

Rainbow trout (female,
mature, 34.8 cm, 511.1 g),

Oncorhynchus mykiss

s,u

14 d

HOS

80%

-

12

LOEC

(decrease LSI)

-

1

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

Rainbow trout (11 mo),
Oncorhynchus mykiss

Diet, U

15 d

PFOS-K
Unknown

-

12

NOEC

(growth - weight)

-

250
mg/kg bw
per day

Dietary exposure

Benninghoff et al.
(2011)

G-5


-------
Species (lilcslii^c)

Method-'

Tesl
l)iir;ilitiii

( Ik-iii ic;il /
I'llril\

pll

Temp.

(ฐC)

r.nwi

Chronic
l.imils

<\oi:( -
I.OKC)

(iiiii/l.)

Reported
r.lTecl
(one.
(inii/l.)

Deficiencies

Reference

Rainbow trout (fry, 15
week),

Oncorhynchus mykiss

Diet, U

8 mo

PFOS-K
Unknown



12

LOEC

(survival, tumor
incidence)

-

: 5 Mi— ku
h\\ per

day

Dietary exposure,
mixture exposure

(P,ennindk>riel

al. 2ul2)



Goldfish

(6.91 g, 6.01 cm),

Carassius curatus

R, U

48 hr

PFOS-K

>99%

-

18

NOEC-LOEC

(swimming behavior:
motion distance and % of
actionless time)

2 (I-S

-

Atypical endpoint and
source of organisms,
duration too short for
an acute test

Xia et al. (2013)

Goldfish (6.0 g, 7.0 cm),
Carassius curatus

S,M

96 hr

PFOS-K

99%

7

22

1.(

-

81.18

Source of organisms
may be problematic

Yang et al. (2014)

Goldfish

(juvenile, 27.85 g),

Carassius curatus

S,M

96 hr

PFOS

>98%

7.25

23

Anlioxidanl en/\ ine
acli\ n\

-

5.oor

Atypical endpoint, no
point estimate

Feng et al. (2015)



Common carp (juvenile,
3.72g, 5.18 cm),
Cyprinus carpio

R, Uฐ

14 d

PFOS
>9X"„

-

-

\oi:c

(liver prolein)

1->1

1

Duration too short for
a chronic test, atypical
endpoint

Hagenaars et al.
(2008)

Common carp (juvenile,
3.72g, 5.18 cm),
Cyprinus carpio

R, Uฐ

14 d

H< )S

-

-

MATC

(liver glycogen)

0.5-1

0.7071

Duration too short for
a chronic test, atypical
endpoint

Hagenaars et al.
(2008)

Common carp (juvenile,
3.72g, 5.18 cm),
Cyprinus carpio

R, Uฐ

14 d

Perfluorooctanesulfonic

H< )S

-

-

NOEC

(liver lipid)

1->1

1

Duration too short for
a chronic test, atypical
endpoint

Hagenaars et al.
(2008)

Common carp (juvenile,
3.72g, 5.18 cm),
Cyprinus carpio

R, Uฐ

14 d

H< )S

-

-

LOEC

(relative condition factor)

<0.1-0.1

0.1

Duration too short for
a chronic test

Hagenaars et al.
(2008)

Common carp (juvenile,
3.72g, 5.18 cm),
Cyprinus carpio

R, I

14 d

PF( )S
>9S",,

-

-

MATC

(HSI)

0.1-0.5

0.2236

Duration too short for
a chronic test, atypical
endpoint

Hagenaars et al.
(2008)

Common carp (juvenile,
~12 cm; ~20 g),
Cyprinus carpio

F, M

96 hr

PF( )S
100.3".,

6.9

23

LOEC

(DNA damage)

-

5.395

Atypical endpoint

(Kim et al. 2010)



Zebrafish

(female fry, 14 dpf),
Danio rerio

R, U

70 d

PI'OS-K

>99%

-

27

EC10

(male weight)

0.01-0.05

0.001990

Missing some
exposure details

Du et al. (2009)

G-6


-------
Species (lilcsliiue)

Method'1

Tesl

l)iir;ilitiii

( hcmic;il /
Piiriu

I'"

Temp.

(ฐC)

r.nwi

Chronic
l.imils

<\oi:( -
I.OKC)
(inii/l.)

Reported
r.lTecl
(one.
(inii/l.)

Deficiencies

Reference

Zebrafish

(embryo - blastula stage),
Danio rerio

R, U

Fert. up to
15 dpf

PFOS-K

99%



-

MATC

(body length and average
weight)

u :uu-
0 400

o :s:s

Duralion loo sliorl lor
a chronic test

Shi el al 			

Zebrafish
(embryo, 6 hpf),

Danio rerio

S,M

114 hr

PFOS

>96%

7.0-
7.5

28

I.C50

-

2.20

Duration too long for
an acute test

Huang et al.
(2010)

Zebrafish
(embryo, 6 hpf),

Danio rerio

S,M

114 hr

PFOS

>96%

7.0-
7.5

28

EC50

(niaHbrnialion)

-

1.12

Duration too long for
an acute test, atypical
endpoint

Huang et al.
(2010)

Zebrafish (embryo),

Danio rerio

R,M

21 d

PFOS Isomers

-

26

i.oi:c

(reduce fecundity)

<0.5-0.5

0.5

Only one exposure
concentration, control
issues

Sharpe et al.
(2010)

Zebrafish (embryo),

Danio rerio

R, M

48 hr

PFOS Isomers

-

:<•

LC50

(range of 3 tests)

-

7.7-38.9

Duration too short for
an acute test, results
are not reproducible

Sharpe et al.
(2010)

Zebrafish
(embryo, 4 hpf),

Danio rerio

S,U

96 hr

PFOS-K

>99%

-

28.5

\oi:c-loi:c

(increased ROS
formation)

n.2-0.4

-

Atypical endpoint,
missing exposure
details

Shi and Zhou
(2010)

Zebrafish (embryo),

Danio rerio

R, U

96 hr

H< )S-k

-

26

LC50

-

54.36e

Problems with
reported data to be
used for LC50 analysis

(Ding et al. 2012)
2013)

Zebrafish

(F2 embryo, 0 hpf),

Danio rerio

F, M

300-330 d

H'< )S-K

S 25-
X "5

26

MATC

(F2 180 d survival)

0.1-0.3

0.1732

Poor concentration-
response, test design
complications

Keiter et al.
(2012)

Zebrafish
(embryo, 6-8 hpf),

Danio rerio

R, U

6 d

H'< )S
I iiiepni'ial

-

2(.

AC50

(toxicity score: includes
survival, hatchability, and
malformation index)

-

16.44e

Duration too long for
an acute test and too
short for a chronic test,
atypical endpoint

(Padilla et al.
2012)

Zebrafish (embryo),

Danio rerio

S,U

In

PF( )S
98".,

8.3

28.5

LC50

-

68

Duration too short for
an acute test, missing
some exposure details

Zheng et al.
(2012)

Zebrafish (embryo),

Danio rerio

S,U

-2 hr

PF( )S
W„

8.3

28.5

EC50

(malformation)

-

37

Duration too short for
an acute test, atypical
endpoint, missing
exposure details

(Zheng et al.
2012)

Zebrafish

(embryo, FO generation),

Danio rerio

R, U

120 dpf

\>\ ( )S

90%

6.8-
7.6

28

LOEC

Increase mortality and
malformations in the F1
generation

<0.250-
0.250

0.250e

Only one exposure
concentration

Chen et al. (2013)

G-7


-------
Species (lilcslii^e)

Mclhori'1

Tesl
Diimlion

( Ik-iii ic;il /
I'llril\

pll

Temp.

(ฐC)

r.nwi

('limn ic
l.imils

<\oi:( -
I.OKC)

(iiiii/l.)

Reported

i.nvci

(one.
(inii/1.)

Deficiencies

Reference

Zebrafish (embryo, 4hpf),

Danio rerio

R, U

120 hr

PFOS

>98%



28

NOEC-LOIX

(suppression of
steroidogenic enzyme
synthesis)

u 1-0.2

-

Duralion loo long lor
an acute test, atypical
endpoint, missing
exposure details

Duetal. (2013)

Zebrafish

(embryo - 4 cell stage),
Danio rerio

S,U

Fert. To
144 hpf

PFOS
Unreported

7.2-
7.6

26

EC50

(lellial and sublethal
effects)

-

1 5

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Ulhaq et al.
(2013)

Zebrafish

(embryo - 4 cell stage),
Danio rerio

S,U

Fert. To
144 hpf

PFOS
Unreported

7.2-
7.6

26

l.( 5(1

-

10

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Ulhaq et al.
(2013)

Zebrafish (embryo),

Danio rerio

S,U

48 hr

PFOS

>96%

-

-

Malfoi'inalKiii
(lOii"..)

-

8.002e

Duration too short for
an acute test, atypical
endpoint, no point
estimate

(Chen et al.
2014a)

Zebrafish (embryo),

Danio rerio

R, U

6 d

PFOS-K

98%

7.5

28.5

I.C50

-

6.25

Duration too long for
an acute test and too
short for a chronic test

Hagenaars et al.
2014

Zebrafish (embryo),

Danio rerio

R, U

6 d

PFOS-K
<>X"„

" 5

28.5

EC 50

(uninflated swim bladder)

-

2.29

Duration too long for
an acute test and too
short for a chronic test,
atypical endpoint

(Hagenaars et al.
2014)

Zebrafish
(embryo, 2 hpf),

Danio rerio

S,U

6d

H< )S-K
<>X"„

"4

28

NOEC-LOEC

(behavior: spontaneous
swimming activity)

0.1-1.0

-

Duration too long for
an acute test and too
short for a chronic test,
atypical endpoint, only
two exposure
concentrations

Spulber et al.
(2014)

Zebrafish (embryo, 6
hpf),

Danio rerio

S,U

1 14 hr

I'lOS-K

1 IlkllOWII

-

:x

LOEC

(mortality)

3.307-
33.07

33.07e

Duration too long for
an acute test and too
short for a chronic test

Truong et al.
(2014)

Zebrafish (embryo, 6
hpf),

Danio rerio

S, L

1 14 hr

PF( )S
Unknow n

-

28

LOEC

(mortality)

0.32-3.2

3.2e

Duration too long for
an acute test and too
short for a chronic test

Truong et al.
(2014)

Zebrafish
(embryo, 8 hpf),

Danio rerio

R, U

42 dpi

PF( )S

^•96%

7.0-
7.5

28

LOEC

(increased condition
index)

-

0.25

Only one exposure
concentration

Chen et al. (2016)

Zebrafish
(embryo, 8 hpf),

Danio rerio

R, U

150 dpf

PIOS

>96%

7.0-
7.5

28

LOEC

(increased estradiol in

male/females and
testosterone in males)

-

0.25

Only one exposure
concentration

Chen et al. (2016)

G-8


-------
Species (lilcsliiuc)

Mclhori'1

Tesl

l)iir;ilitiii

( Ik-iii ic;il /
Piiriu

pll

Temp.

(ฐC)

r.nwi

Chronic
Limits

<\oi:( -
I.OKC)

(iiiii/l.)

Reported
r.ffccl
(one.
(inii/l.)

Deficiencies

Reference

Zebrafish (larva, 120 hpf),

Danio rerio

S,M

24 hr

PFOS

>98%

7.0-
7.5

28

NOEC

(various metabolites)

-

9.700

Duration loo short for
an acute test, atypical
endpoint

( I liianu el al

2016) Huang et al.
2016

Zebrafish (embryo),

Danio rerio

R, U

6 d

PFOS
Unknown

-

28

LOEC

(liver size and gene
expression)

(i (KKI5-
(i (i( i( )5

0.0005

Duration too long for
an acute test and too
short for a chronic test,
atypical endpoint

(Tse et al. 2016)

Zebrafish
(embryo, 8 hpf),

Danio rerio

R, U

180 d

PFOS

>96%

7.0-
7.5

27

MATC

(altered sex ratio: female
dominance, l'l oil spring
survival)

0.05-0.25

() 1 118e

Non-apical endpoint

Cui et al. (2017)

Zebrafish (3 hpf),

Danio rerio

S,U

5 d + 9 d
observation

PFOS
Unreported

7.2-
7.7

:<-:x

\1 VIC

(growth - total body
length)

0.02-0.2

0.06325e

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Jantzen et al.
(2017)

Zebrafish (3 hpf),

Danio rerio

S,U

5 d + 9 d
observation

PFOS
Unreported

7.2-
7.7

2(ป-28

LOEC

(interoccular distance)

0.02-
0.02

0.02e

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Jantzen et al.
(2017)

Zebrafish (3 hpf),

Danio rerio

S,U

5 d + 9 d
observation

PF< )S
Unreported

7.2-

26-28

MATC

(yolk sac area)

0.02-0.2

0.06325e

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Jantzen et al.
(2017)

Zebrafish (3 hpf),

Danio rerio

S,U

5 d + 9 d
obsci'\ aiion

PI OS
Unreported

" :-

26-28

LOEC-

(swimming activity -
crossing frequency)

< 0.02-
0.02

0.02e

Duration too long for
an acute test and too
short for a chronic test,
static chronic exposure

Jantzen et al.
(2017)

Zebrafish (embryo),

Danio rerio

S,M

4X hr

H'< )S

1 IlkllOW II

-

2"

LC50

-

107.6

Duration too short for
an acute test

(Rainieri et al.
2017)

Zebrafish
(embryo, 3 hpf),

Danio rerio

R, U

"d

H'< )S
I iireporied

-

:s 5

MATC

(islet morphological
anomalies)

8.0-16.0e

11.3 le

Duration too long for
an acute test and too
short for a chronic test

(Sant et al. 2017)

Zebrafish (sperm),

Danio rerio

S,U

:u see

PFOS-K
>9S",i

8

25

NOEC-LOEC

(sperm motility)

0.09-0.9

-

Duration too short for
an acute test, atypical
endpoint

Xia and Niu
(2017)

Zebrafish (sperm/egg),
Danio rerio

S,U

2 111111

H< )S-K

8

25

NOEC-LOEC

(fertilization success)

0.09-0.9

-

Duration too short for
an acute test, atypical
endpoint

(Xia and Niu
2016)Xia and Niu
(2017)

Zebrafish (embryo, 3
hpf),

Danio rerio

S,U

5 d

PFOS
Unknown

7.2-
7.7

27

LOEC

(gene expression of
Leptin A mRNA)

<0.01-0.01

0.01e

Duration too long for
an acute test and too
short for a chronic test,
atypical endpoint

(Annunziato
2018)

G-9


-------
Species (lil'csliiiic)

MiMhori'1

Tcsl
Diimlion

( Ik-iii ic;il /
PuriU

pll

Temp.

(ฐC)

r.nwi

('limn ic
l.imils

<\oi:( -
I.OKC)

(iiiii/l.)

Koporiod

i.nvci

(one.
(inii/1.)

Dcl'icic'iicii's

Reference

Zebrafish
(embryo, 1-2 hpf),

Danio rerio

R, U

96 hr

PFOS

>99%



25

NOEC-LOIX

(growth: body length)

<0.050-
0 050

-

Atypical endpoinl,
missing exposure
details

(Dang et al. 2018)

Zebrafish (embryo, 2
hpf),

Danio rerio

R, U

72 hr

PFOS
Unknown

-

28.5

l.OF.C

(malformations)

(i 5-1 (i

1.0e

Duration too short for
an acute test

Ortiz-Villanueca
et al. (2018)

Zebrafish (embryo, 2
hpf),

Danio rerio

R, U

72 hr

PFOS
Unknown

-

28.5

LOEC

(survival)

5.0-10

10

Duration too short for
an acute test

Ortiz-Villanueca
et al. (2018)

Zebrafish
(embryo, 1 hpf),

Danio rerio

R, U

96 hr

PFOS
Unreported

76

28 5

\OLC-I.Oi:(

(pericardial area)

8-16e

-

Atypical endpoint,
missing exposure
details

Sant et al. (2018)

Zebrafish (female, 4 mo),
Danio rerio

R, U

21 d

PFOS
Unknown

7 ii-

7.5

:x

xoi:(

(growlli - length and
weight)

0.2->0.2

0.2

Inability to
independently verify
effect values, partial
life cycle test

Bao et al. (2019)

Zebrafish (embryo,
maximum of 4 hpf),

Danio rerio

R, M

96 hr

PFOS

1 IlkllOW II

-

26

NOEC

(hatching success,
embryo mortality,
deformation)

0.0007-
>0.0007

0.0007

Greater than low value

(Cormier et al.
2019)

Zebrafish (embryo, 2
hpf),

Danio rerio

R, U

72 hr

H< >S-k
W"„

-

28

LOEC

(growth - total body
length)

2.691-
5.832

5.382e

Duration too short for
an acute test

Martinez et al.
(2019)

Zebrafish (embryo, 2
hpf),

Danio rerio

R, M

1 IS hr

PFOS-k

-

:x

EC50

(mortality,
malformations)

-

2.045e

Duration too long for
an acute test and too
short for a chronic test,
mixed test endpoints

Vogs et al. (2019)

Zebrafish (embryo, 6
hpf),

Danio rerio

S,U

(>(> hr

H OS
97".,

-

:s

NOEC

(survival)

25->25

25e

Duration too short for
an acute test

(Dasgupta et al.
2020)

Zebrafish (embryo, 6
hpf),

Danio rerio

R, M

90 III'

PFOS-k
>9X"„

-

28.5

LOEC

(malformations,
locomotive behavior)

<20-20

20

Only one exposure
concentration; atypical
endpoint

(Huang et al.
2021)

Zebrafish (embryo, <1
hpf),

Danio rerio

R, U

96 hr

H'< )S
I iiknown

-

28.5

LOEC

(increase lauric C12:0
and myristic C14:0 fatty
acids)

<8.002-
8.002

8.002e

Atypical endpoint

Sant et al. (2021)

Zebrafish (dechorionated
embryo, 1 dpf),

Danio rerio

R, U

30 d

PFOS
Unknown

-

28.5

NOEC

(growth - length)

16->16

16e

Growth effects not the
focus of study rather
other non-apical
endpoints

(Sant et al. 2021)

G-10


-------
Species (lil'csliiiic)

Mclhocl-1

Tcsl
Diimlinn

( Ik-iii ic;il /
PuriU

pll

Temp.

(ฐC)

I.ITccl

('limn ic
l.imils

<\oi:( -
I.OKC)

(iiiu/l.)

Rcpork'd

i.nvci

(one.
(iiiu/l.)

IkTicicnck's

Reference

/vhialish (adiili. 'ซ) dpi ).

Danio rerio

k. 1

lu d

H< )S-K

" :i

:so

i.oi:c

(gone expression)

<0.5-0.5

0.5

Atypical endpoint

(Zhuetal. 2021)



Spottail shiner (female,
mature, 8.9 cm, 6.7 g),

Notropis hudsonius

Microcosm

14 d

PFOS
89%

-

-

NOEC

(mortality)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

Spottail shiner (female,
mature, 8.9 cm, 6.7 g),

Notropis hudsonius

Microcosm

14 d

PFOS
89%

-

-

LOEC

(increase TBARS in
liver ovary and l'AO
aoliviiv in liver)

-



Atypical exposure, not
a true ELS test

Oakes et al.
(2005)



Fathead minnow
(mature, 6.1 cm, 2.0 g),

Pimephales promelas

Microcosm

28 d

PFOS
89%

9:

it. (.-

:: s

LCM

-

3.5

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)



Topmouth gudgeon
(juvenile female, 0.81 g,
4.03 cm),

Pseudorasbora parva

R, U

96 hr

PFOS-K

-

15

\olc-i.oi:c

(spontaneous swim
behavior: swim distance)

0.5-2

-

Atypical endpoint and
source of organisms

(Xia et al. 2014)

Topmouth gudgeon
(4.0 g, 4.0 cm),
Pseudorasbora parva

S,M

96 hr

H< )S-k
W'„

ฆ

22

I.C50

-

67.74

Source of organisms
may be problematic

Yang et al. (2014)

Topmouth gudgeon
(4.0 g, 4.0 cm),
Pseudorasbora parva

R,M

"(id

H< )S-K
W',,

ฆ

22

EC10

(survival)

-

2.12

Not a true ELS test
(started with older life
stage), renewal chronic
exposure, source of
organisms may be
problematic

Yang et al. (2014)



Creek chub

(mature, 11.8 cm, 16.3 g),

Semotilus atromaculatus

Microcosm

14 d

PF( )S
S')"„

-

-

NOEC

(mortality)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

Creek chub

(mature, 11.8 cm, 16.3 g),

Semotilus atromaculatus

Microcosm

14 d

I'I'OS
x
-------
Species (lil'csliiiic)

Mclhnri'1

Tcsl
Diimlinn

( Ik-iii ic;il /
PuriU

nil

Temp.

(ฐC)

r.nwi

('limn ic
l.imils

<\oi:( -
I.OKC)
(mป/l.)

Kcpnrlcri

i.nvci

( nnc.
(inu/l.l

IkTicicnck's

Reference

Quinbo (juvenile, 2.77 g,
5.62 cm),

Spinibarbus sinensis

R, U

30 d

PFOS-K

>99%

OO

vo r-

18

MATC

(% mobile, % highly
mobile, swim distance,
swim speed, freq. highly
mobile, % social, resting
metabolic rate)

(1 ^2-u SO

0.506

Test was not replicated

(Xiaetal. 2015b)

Quinbo (juvenile, 2.77 g,
5.62 cm),

Spinibarbus sinensis

R, U

30 d

PFOS-K

>99%

00

18

MATC

(decrease maximum
linear acceleration)

0.32-u No

u 5(16

Atypical endpoint

(Xiaetal. 2015c);
Xia et al. 2015d

Quinbo (juvenile, 2.77 g,
5.62 cm),

Spinibarbus sinensis

R, U

30 d

PFOS-K

>99%

00

28

\1 \T(

(decrease maximum
linear acceleration)

0.32-0.80

0.506

Atypical endpoint

Xia et al. (2015d)



White sucker (mature,
22.7 cm, 114.5 g),

Catostomus commersonii

Microcos
m

14 d

PFOS
89%

-

-

noi :c

(mortality)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)

White sucker (mature,
22.7 cm, 114.5 g),

Catostomus commersonii

Microcos
m

14 d

PFOS

ST.,

-

-

LOEC

(decrease LSI in females)

-

3

Atypical exposure, not
a true ELS test

Oakes et al.
(2005)



Bluegill

(28.6 mm, 0.60 g),

Lepomis macrochirus

S,U

96 hr

H'< )S 1)1: \ sail

1 IlkllOW II

s

s .

-

LC50

-

31

Only one replicate per
treatment

3MCompany
(2000)



Medaka (adult, male),

Oryzias latipes

R, U

14 d

H '< )S
I iii'qxn'la.1

-

25

NOEC

(adult survival, GSI%,
HSI%, condition factor)

1->1

1

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka (adult, female),

Oryzias latipes

R, U

14 d

PF< )S
I Jnrepmial

-

25

NOEC

(adult survival, condition
factor)

1->1

1

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka (adult, female),

Oryzias latipes

R, U

14 d

PF( )S
I 'nrcpni'ioJ

-

25

LOEC

(GSI%)

<0.01-0.01

0.01

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka (adult, female),

Oryzias latipes

R, U

14 d

PFOS
I iiiqinrlcd

-

25

MATC

(HSI%)

0.1-1

0.3162

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka (F1 generation,
<12 hr, embryo),

Oryzias latipes

R, U

7-14 d
(assumed)

PFOS
Unreported

-

25

MATC

(% hatchability, time to
hatch)

0.1-1

0.3162

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

G-12


-------
Species (lilcslii^c)

Method'1

Tesl
Dnmlion

( Ik-iii ic;il /
PuriU

pll

Temp.

(ฐC)

r.iTcci

Chronic
l.imils

<\oi:( -
I.OKC)
(inii/l.)

Reported

i.nvci

(one.
(inii/l.)

Deficiencies

Reference

Mcdaka (I I ueiiercilinii.
12 In; cinhiwM,

Oryzias latipes

R, I

2X d pusl-
lialcli

(assumed)

H '< )S

Unreported



25

\1 \T(

(swim up success)

0 1 -1

0.3162

Duration loo long lor
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka(Fl generation,
<12 hr, embryo),

Oryzias latipes

R, U

100 d post-
hatch

PFOS
Unreported

-

25

EC 10

(grow ill - length)

(1 (11 -(1 (11

0.0013

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka(Fl generation,
<12 hr, embryo),

Oryzias latipes

R, U

28 d post-
hatch

PFOS
Unreported

-

25

LOEC

(larval survival)

<0.01-0(11

(i.01

Duration too long for
an acute test and too
short for a chronic test

Ji et al. (2008)

Medaka (adult, 16 week,
0.38g)

Oryzias latipes

R, U

21 d

PFOS

>98%

-

25

loi:c

(lecundily)

<1.0-1.0

1

Only one exposure
concentration

(Kang et al. 2019)



African clawed frog
(embryos),

Xenopus laevis

R, M

96 hr

PFOS-K

86.9%

7.3

24

i:c50

(teratogenesis)

-

12.1

Atypical acute
endpoint

Palmer and
Krueger (2001)

African clawed frog
(embryos),

Xenopus laevis

R, M

96 hr

PFOS-K
S(. o hr

PFOS-K
86.,>"„

" :_

24

LOEC

(growth)

-

7.97

Atypical acute
endpoint

Palmer and
Krueger (2001)

African clawed frog
(embryos),

Xenopus laevis

R, M

9o hr

PFOS-K
S6.0.1

0.1

Control issues

(Cheng et al.
2011)

G-13


-------
Species (life's!;iiicI

Method'1

Tesl
Diimlion

( Ik-iii ic;il /
PuriU

I'"

Temp.

(ฐC)

r.iToci

('limn ic
l.imils

<\oi:( -
I.OKC)

(iiiu/l.)

Koporiod

i.nvci

(one.
(inii/1.)

Dcl'iciciicii's

Reference

Mi ieaii clawed ling

(embryo, NF 10),
Xenopus laevis

R, M

96 hr

PFuS

>99%



24

LC50

-

>96

Non-definitive value

(San-Segundo et
al. 2016)



Asiatic toad

(tadpole, 1.8 cm, 0.048 g),

Bufo gargarizans

R, M

96 hr

PFOS-K

99%

7

22

I.C50

-

4s :i

Source of organisms
may be problematic

Yang et al. (2014)

Asiatic toad

(tadpole, 1.8 cm, 0.048 g),

Bufo gargarizans

R, M

30 d

PFOS-K

99%

7

22

1 ( in

(survival)

-

2.00

Renewal chronic
exposure, not a true
ELS test, source of
organisms may be
problematic

Yang et al. (2014)



Northern leopard frog
(Gosner stage 25),
Lithobates pipiens

S, M

116 d

PFOS
Unknown

7.41-

8.54

1 ^ 1-

2') S

\OLX

(survival and growlli)

0.0128-
>0.0128

0.0128

Outdoor mesocosm

(Foguth et al.
2020)

Northern leopard frog
(Gosner stage 26.5, 0.109
gX

Lithobates pipiens

R, U

10 d

PFOS-K

7.9

22

noi:c

(developmcnl, growth,
survival)

0.1->0.1

0.1

Duration too long for
an acute test and too
short for a chronic test

(Brown et al.
2021)

Northern leopard frog
(larva, Gosner stage 25),
Lithobates pipiens

S, M

30 d

H'< )S

" S

:<-2

LOEC

(developmental stage)

<0.00006-
0.00006

0.00006

Duration too long for
an acute test and too
short for a chronic test

(Flynnetal. 2021)

a S=static, R=renewal, F=flow-through, U=unmeasured. M measured. I Inial. I) dissnl\ al. I)iet=dietary, MT=maternal transfer

b Chemical concentrations made in a side-test rcprcscnlali\ e of exposure and \ en lied stability of concentrations of PFOS in the range of concentrations tested under similar

conditions. Daily renewal of test solutions.

0 Water concentrations were not measured, bul PFOS concern nil ions were measured in the liver.

d 36 days corresponds to the first of leu ueneralions, the one u illi ihe niosi cousisieui negative response. The value 36 days is 1/10 of the duration of this year-long 10-generation
study.

e Reported in moles converted to gram based mi a molecular weidil of 500.13 g/mol (PFOS); 538.22 g/mol (PFOS-K); 629.4 g/mol (PFOS-TEA).

G-14


-------
G.2 Summary of Acute PFOS Toxicity Studies Used Qualitatively in the
Freshwater Aquatic Life Criterion Derivation

G.2.1 Freshwater Invertebrates
G. 2.1.1 Worms (flat and annelids)

Yang et al. (2014) conducted a 96-hour measured, renewal acute test of PFOS

(potassium salt, CAS # 2795-39-3, 99% purity) with the annelid worm, Limnodrilus hoffmeisteri.

The test followed ASTM E729 (1993). L. hoffmeisteri (0.03 g, 0.S cm) used for the test were

obtained from Beijing City Big Forest Flower Market and allowed lo acclimate for seven days

before testing. Dilution water was dechlorinated tap water (pH, 7.0ฑ0.5; dissol\ cd oxygen,

7.0ฑ0.5 mg/L; total organic carbon, 0.02 mg/L; and hardness. I90.0ฑ0.1 mg/L as CaCCte).

Photoperiod was 12-hr: 12-hr (light:dark) at an unreported intensity. A primary stock solution

was prepared by dissolving PFOS in deionized water and cosol\ enl DMSO and proportionally

diluted with dilution water to prepare the test concentrations Exposure vessels were 90 cm petri

dishes containing in ml. of test solution. The test employed three replicates of 10 worms each in

six test concentrations (measured in low and high treatments only) plus a negative and solvent

control. Nominal concentrations were <) (neuati\e and solvent controls), 60.00, 72.00, 86.40,

103. ON. 12442 and I4l> 3<) mill I. The authors provided arithmetic mean measured

concentrations before and after renewal: 51.42 and 58.94 mg/L (lowest concentration) and

127.73 and I 5<) S4 mg/L (highest concentration). Analyses of test solutions were performed

using high performance liquid chromatography with mass spectrometric detection (HPLC/MS)

and negative electrospray ionization. The concentration of PFOS was calculated from standard

curves (linear in the concentration range of 1-800 ng/mL), and the average extraction efficiency

was in the range of 70-83%. The concentrations and chromatographic peak areas exhibited a

significant positive correlation (r=0.9987, p<0.01), and the water sample-spiked recovery was

G-15


-------
105%. The temperature, D.O., and pH were reported as having been measured every day during
the acute test, but results are not reported. Negative control survival was 100%. Solvent control
survival was 96%. The 96-hour LCso was 120.97 mg/L (C.I. 103.97-140.76). The acute value
was acceptable for qualitative use because the source of organisms was atypical.

Liu et al. (2016) conducted several 24-hour static, unmeasured acute tests on PFOS
(>98% purity) withLimnodrilus hoffmeisteri. The test organisms were obtained from an
aquarium market in Jinan, Shandong, China and acclimatized in a large aquarium containing
river sediment (7.0-10.0 cm thickness) and aerated water for seven days before use. Only worms
of uniform size were used for testing. Dilution water was carbon-filtered, dechlorinated tap water
(pH 7.25ฑ0.25, conductivity 340.67ฑ16.4 |iS/cm, total hardness 135.5ฑ9.3 mg CaCCte/L,
alkalinity 40.7ฑ5.2 mg CaCOs/L, Na = 1 I 2 n 2 mg'I.. K = 2 34 <) 07 mg/L, Mg = 7.74ฑ0.02
mg/L, Ca = 41.07ฑ0.82 mg/L, CI = 28.3ฑ1 2 mu I.) No details were provided regarding
photoperiod or light intensity A primary stock solution was prepared by dissolving PFOS in
DMSO. The stock solution w as further diluted with dilution water to obtain different
concentrations l-xposure \ essel size and material type was not reported, but fill volume was 20
mL. Prior to joint toxicity and Moacaimulation of Zn and PFOS experiments, the 24-hour
toxicity of PI OS to L. hoJjiiK-isteri was determined at pH 5.0, 6.0, 7.0, 8.0, and 9.0 in
preliminary experiments. The final pH values for the preliminary tests were adjusted to the
desired values within a range of ฑ 0.1. The tests employed three replicates of 10 worms each and
an unspecified number of test concentrations plus a negative and presumably a solvent control.
The authors reported nominal test concentrations in their supplemental Table SI. The test
temperature was maintained at 23ฑ1ฐC. No other water quality parameters were reported as
having been measured for test solutions. Survival of control animals (and presumably solvent)

G-16


-------
were also not reported. The 24-hour LCso values at pH 5.0, 6.0, 7.0, 8.0, and 9.0 were reported to
be 45.26, 46.23, 60.70, 64.48, and 65.74 mg/L, suggesting a slight trend toward increasing LC50
with increasing pH for L. hoffmeisteri. The acute values from the study were acceptable for
qualitative use because of the short test duration.

Qu et al. (2016) similarly conducted a set of 48-hour renewal, unmeasured tests on PFOS
(potassium salt, 98% purity) with Limnodrilus hoffmeisteri at different pH levels. Test organisms
were obtained from an aquarium market in Jinan, Shandong, China and acclimatized in a large
aquarium containing river sediment (7.0-10.0 cm thickness) and aerated water lor 10 days before
use. Only worms of uniform size (body length: 3 <)-4 <) cm) were used for testing. Dilution water
was carbon-filtered, dechlorinated tap water (pH, 7.70 n- 0.15; conductivity, 340.6 ฑ 16.4 |iS/cm;
total hardness, 135.5 ฑ 9.3 mg CaCCte/L; alkalinity. 4<~> 7 ฃ 5 2 mu CaCCb/L; Na, 11.2 ฑ 0.2
mg/L; K, 2.34 ฑ 0.07 mg/L; Ca, 41.07 ฑ 0.S2 mu I.. Mu. 7 74 <") 02 mg/L; CI, 28.3 ฑ 1.2 mg/L).
No details were pro\ ided regarding photoperiod or light intensity. A primary stock solution was
prepared by dissol\ inu PI-OS in DMSO (<0.1%). The stock solution was further diluted with
carbon-filtered and aerated tap water to obtain different concentrations. Exposure vessels were
beakers of a size and material type not reported. Fill volume was 20 mL. The three different tests
conducted at pi I (\2, 7.0, and S (J employed three replicates of 10 worms each in three test
concentrations plus a solvent control. Nominal concentrations were 0 (solvent control), 5, 10,
and 20 mg/L. The test temperature was maintained at 22 ฑ 1ฐC. No other water quality
parameters were reported as having been measured for test solutions. Survival of solvent control
animals were also not reported. The 48-hour LCso values were 23.81 ฑ 1.14, 35.89 ฑ 0.49 and
39.80 ฑ 1.15 mg/L at pH 6.2, 7.0 and 8.0. Similar to the preliminary findings from Liu et al.
(2016), the PFOS 48-hour LCso for L. hoffmeisteri increased with increasing pH, but the change

G-17


-------
was not large. The acute values from the study were acceptable for qualitative use because of the
short test duration.

G.2.1.2 Mollusks

(Amraoui et al. 2018) conducted a 96-hour renewal, unmeasured toxicity test of PFOS
(potassium salt, CAS # 2795-39-3, >98% purity) with the freshwater mussel, Unio ravoisieri (a
non-North American species). Test organisms (6 cm) were wild-caught from Sejenane river, a
tributary of Ichkeul Lake (Bizerte, northern Tunisia). The mussels were acclimated under
laboratory conditions for 14 days before use. Dilution water was natural aerated freshwater from
an unspecified and uncharacterized source. Use of this natural freshwater for the lesl reportedly
provided enough food to prevent starvation during the lesl hul not so much as to cause any
discrepancy in observed effects resulting from the interaction of PFOS and food. Photoperiod
was 16-hr:8-hr (light:dark) at an unreported light intensity A stock solution was prepared by
directly dissolving the powder in 1 <"><">% I)\|S() Stock solution was proportionally diluted with
system water resulting in a final L)\ISO concentration of 0.01% (v/v). The test employed three
replicates of five mussels each in li\ e lest concentrations plus a solvent control. Nominal test
concentrations were <> (sol\ enl control), 10, 25, 50, 75, and 100 mg/L PFOS. Exposure vessels
were 3 I. tanks of unreported material, dimensions and fill volume. The test temperature was
controlled at IS (' and water pi I 8 in an acclimated room. No other water quality parameters
were reported as ha\ inu been measured in test solutions. Survival of solvent control animals
appeared to be 100% (as depicted in a figure). The 96-hour LCso was reported as 65.9 mg/L. The
acute value was acceptable for qualitative use because the organisms were exposed using an
uncharacterized natural freshwater source as dilution water.

Yang et al. (2014) conducted a 96-hour acute test of PFOS (potassium salt, CAS # 2795-
39-3, 99% purity) with a non-North American snail species, Cipangopaludina cathayensis. The

G-18


-------
test followed ASTM E729 (1993). The test organisms (4.0 g, 2.0 cm) were purchased from the
Beijing Dahongmen Jingshen Seafood Market and were held seven days prior to testing. Dilution
water was dechlorinated tap water (pH, 7.0 ฑ 0.5; dissolved oxygen, 7.0 ฑ 0.5 mg/L; total
organic carbon, 0.02 mg/L; and hardness, 190.0 ฑ0.1 mg/L as CaCCte). Photoperiod was 12-
hr: 12-hr (light:dark) at an unreported light intensity. A primary stock solution was prepared by
dissolving PFOS in deionized water and cosolvent DMSO. The primary stock was proportionally
diluted with dilution water to prepare the test concentrations. Exposure \ essels were 200 mL
beakers of unreported material type containing 100 mL of test solution. The lesl employed three
replicates of 10 snails each in six test concentrations (measured in low and high treatments) plus
a negative and solvent control. Nominal concentrations were 0 (negative and solvent controls),
100, 130, 169, 219.7, 285.61 and 371.29 mu I. Mean measured concentrations before and after
renewal were 86.50 and 99.85 mg/L (lowest conceniralion) and 328.84 and 368.24 mg/L (highest
concentration). Analyses of lesl solutions were performed using HPLC/MS and negative
electrospray ionization The conceniralion of PFOS was calculated from standard curves (linear
in the conceniralion range of l-S<)<) nu ml.), and lhe average extraction efficiency was in the
range of 7<)-S3".. The concentrations and chromatographic peak areas exhibited a significant
positi\ e correlation (r=0.9^S7. p<0.<)|), and the water sample-spiked recovery was 105%. The
temperalure. DO. and pFl were reported as having been measured every day during the acute
test, but results are nol reported. Negative and solvent control survival was 100%. The 96-hour
LCso was 247.14 mg/L (C.I. 188.81-323.48). The acute value was acceptable for qualitative use
because the source of organisms was atypical.

Olson (2017) conducted 96-hour static acute tests on PFOS (potassium salt, CAS # 2795-
39-3, 95% purity) with adult and juvenile (0-3 week old) Lymnaea stagnalis as part of a Ph.D.

G-19


-------
thesis at the Texas Tech University, Lubbock, TX. The test followed methodology established in
Ducrot et al. (2010). Snails were fed during the acute test. The test organisms were randomly
selected from the appropriate age group from a snail culture maintained at Texas Tech
University. Dilution water was reconstituted laboratory (13.38 g CaSC>4, 5.6 g MgSC>4, 0.25 g
KC1, and 2.95 g NaHCCte added to 50 L deionized water). Photoperiod and light intensity were
not reported. Stock solutions were prepared by dissolving between n 2< >00 and 0.2010 g of the
chemical powder in 1 L of lab water and placing in HDPE bottles on a shaker overnight. Stock
solution was then diluted as necessary to obtain the final exposure concentrations l-xposure
vessels were 1 L polyethylene beakers containing 1 L of test solution. The test employed two
replicates of five to 10 snails each in six test concentrations (measured in low and high
treatments) plus a negative and solvent control Nominal concentrations in the test with adults
included five individuals in each replicate and were <> (neuati\ e control), 15, 30, 60, 125, 200,
and 250 mg/L. Nominal concentrations in the test with <)-3 week old snails included 10
individuals in each replicate and were 0. 12 5. 25, 50, 100, 150, and 200 mg/L. Exposure
concentrations were reportedly measured initially and after three days for verification, but were
not reported Analyses of test solutions were performed HPLC/MS. Standards were used as part
of the analytical method, but details were not reported. The reporting limit was 0.010 mg/L.
Experiments were conducted in incubators set to 20ฐC, which did not vary more than 1ฐC during
the course of the studies No other water quality parameters were reported as having been
measured in test solutions. Negative control survival was >90%. The 96-hour LCso reported for
adult snails was 196 mg/L. The 96-hour LCso reported for juveniles was 150 mg/L. The acute
values were acceptable for qualitative use because snails were fed during acute toxicity testing.

G-20


-------
G. 2.1.3 Zooplankton (rotifers andplanktonic crustaceans)

3MCompany (2000) provides the results of a 48-hour static, unmeasured acute toxicity

test completed with the cladoceran, Daphnia magna, and PFOS-Li (perfluorooctancesulfonate

lithium salt, CAS # 29457-72-5). A stock solution was made with carbon-filtered well water at a

test sample concentration of 1,000 mg/L and where the test sample was reported as a mixture of

PFOS-Li (24.5%) in water (75.5%). Daphnids obtained from the I 'SI-PA (Duluth, MN) were

used for testing and were less than 24 hours old at test initiation Exposure \ essel were 100 mL

glass beakers containing 50 mL of solution and five daphnids per beaker, Each lest treatment was

replicated four times with nominal test concentrations (control, 100, 180, 320. 5(-><) and 1,000

mg/L test sample). Throughout the experiment the D.O mimed from 7.0-7.8 mg/L, pH 8.6 and a

test temperature of 20.1 - 21.0ฐC. No immobility or mortality occurred in the control treatment

and 100%) was observed in treatments > 32<) nig I. The author reported that the test sample

containing 24.5% PFOS-Li exhibited a 48-hour ECso (mortality and immobilization) of 210

mg/L, which equates to 5 I 45 mu I. as PFOS-Li. EPA was unable to independently calculate a

48-hour ECso value based 011 the le\ el data provided in the paper by the study authors.

Additionally, the statistical analysis and methods used by the study authors could not be

evaluated by EPA gi\ en that the details provided in the paper. Specifically, the EC50 was

calculated using u hat appears to be proprietary statistical software in which few details are

provided by the study authors Therefore, the author-reported ECso value was used qualitatively

to derive the acute freshwater criterion for PFOS. However, this toxicity value of 51.45 mg/L is

consistent with others for this species that were used quantitatively to derive the PFOS criterion.

G.2.1.4 Benthic Crustaceans

Yang et al. (2014) performed a 96-hour acute test of PFOS (potassium salt, CAS # 2795-

39-3, 99% purity) with the freshwater prawn species, Macrobrachium nipponense (a non-North

G-21


-------
American species). The test followed ASTM E729 (1993). M nipponense (0.30 g, 4.0 cm) used
for the test were purchased from the Beijing Dahongmen Jingshen Seafood Market and were
held seven days prior to testing. Dilution water was dechlorinated tap water (pH, 7.0 ฑ 0.5;
dissolved oxygen, 7.0 ฑ 0.5 mg/L; total organic carbon, 0.02 mg/L; and hardness, 190.0 ฑ0.1
mg/L as CaCCte). Photoperiod was 12-hr: 12-hr (light:dark) at an unreported light intensity. A
primary stock solution was prepared by dissolving PFOS in dcionized water and cosolvent
DMSO. The primary stock was proportionally diluted with dilution water lo prepare the test
concentrations. Exposure vessels were 2 L beakers of unreported material type containing 1.5 L
of test solution. The test employed three replicates of 1 prawn each in six test concentrations
(measured in low and high treatments) plus a negative and solvent control. Nominal
concentrations were 0 (negative and solvent controls). 6 00, in So. |9 44, 34.99, 62.99 and
113.37 mg/L. Mean measured concentrations before and after renewal were 4.88 and 5.95 and
mg/L (lowest concentration) and ^7 S5 and 10l) 22 nig/I. (highest concentration). Analyses of
test solutions were performed using IIPI.(VMS and negative electrospray ionization. The
concentration of I'TOS was calculated lYom standard curves (linear in the concentration range of
l-80<) nu ml.), and the a\eraue extraction efficiency was in the range of 70-83%. The
concentrations and chromatographic peak areas exhibited a significant positive correlation
(r=0.9987. p n <)|). and the water sample-spiked recovery was 105%. The temperature, D.O.,
and pH were reported as ha\ ing been measured every day during the acute test, but results were
not reported. Negative control survival was 100%. Solvent control survival was 96%. The 96-
hour LCso was 19.77 mg/L (C.I. 12.42-31.48). The acute value was acceptable for qualitative use
because the source of organisms was atypical.

G-22


-------
G. 2.1.5 Aquatic Insects

Olson (2017) conducted a 48-hour static acute test on PFOS with first instar of the

mosquito Aedes aegypti as part of a Ph.D. thesis at the Texas Tech University, Lubbock, TX.

The colony was originally donated by Texas A&M and maintained in the laboratory at Texas

Tech University since summer 2013. Dilution water was moderately hard reconstituted water (3

g CaSC>4, 3 g MgSC>4, 0.2 g KC1, and 4.8 g NaHCCte added to 50 I. deionized water).

Photoperiod was 12-14 hours light and 10-12 hours dark. Light intensity was not reported. Stock

solutions were prepared by dissolving soluble amounts of powdered chemical in dilution water.

Diluted stock concentrations were equal to the maximum test concentration. The slock was

mixed on a shaker table at 125 rpm for at least 18 hours before being added to exposure

containers and proportionally diluted. Exposure \ essels ^ere 5<) mL HDPE plastic beakers

containing an unspecified amount of test solution The lest employed 10 mosquito larvae each in

six test concentrations plus a negative control. Replication was not reported. Nominal

concentrations in the test were 0 (negative control), 0.050, 0.125, 0.250, 0.500, 1.000 and 2.000

mg/L. Experiments were conducted in incubators set to 25ฐC and covered with plexiglass to limit

evaporation No other water quality parameters were reported as having been measured in test

solutions Negati\e control sur\i\al was ^5"o. The 96-hour LCso was 1.18 mg/L. The acute

value was acceptable for qualitative use due to its short duration because the publication was

missing some exposure details, including the purity of PFOS, it was an unmeasured test.

Additionally, the author-reported LCso and concentration-response curve could not be assessed

by EPA on a statistical basis since model parameters and sufficient treatment level data were not

provided to independently calculate toxicity values (e.g., LCso). And finally, the species is an

invasive, pest species to the U.S.

G-23


-------
Yang et al. (2014) performed a 96-hour acute test on PFOS (potassium salt, CAS # 2795-
39-3, 99% purity) with the midge, Chironomusplumosus. The test followed ASTM E729 (1993).
C. plumosus (0.05 g, 12 cm) used for the test were purchased from the Beijing City Big Forest
Flower Market and were held seven days prior to testing. Dilution water was dechlorinated tap
water (pH, 7.0ฑ0.5; dissolved oxygen, 7.0ฑ0.5 mg/L; total organic carbon, 0.02 mg/L; and
hardness, 190.0ฑ0.1 mg/L as CaCCte). Photoperiod was 12-hr 12-hr (light:dark) at an unreported
light intensity. A primary stock solution was prepared by dissolving PFOS in dcionized water
and cosolvent DMSO. The primary stock was proportionally diluted with dilution water to
prepare the test concentrations. Exposure vessels were 90 cm petri dishes containing 10 mL of
test solution. The test employed three replicates of 10 midges each in six test concentrations
(measured in low and high treatments) plus a negative and sol\ cut control. Nominal
concentrations were 0 (negative and solvent controls), loo on. 130.00, 169.00, 219.70, 285.61
and 371.29 mg/l. Mean measured concentrations before and after renewal were 92.24 and 99.46
mg/L (lowest concentration) and 34<) 45 and 369.29 mg/L (highest concentration). Analyses of
test solutions were performed using I MM.(' MS and negative electrospray ionization. The
concentration of PI-OS was calculated from standard curves (linear in the concentration range of
1-80O nu ml.), and the average extraction efficiency was in the range of 70-83%. The
concentrations unci chromatographic peak areas exhibited a significant positive correlation
(r=0.9987, p<0.0l). and the water sample-spiked recovery was 105%. The temperature, DO, and
pH were reported as having been measured every day during the acute test, but results were not
reported. Negative and solvent control survival was 96%. The 96-hour LCso was 182.12 mg/L
(C.I. 158.71-209.00). The acute value was acceptable for qualitative use because the source of
organisms was atypical.

G-24


-------
G.2.2 Freshwater Fish
G. 2.2.1 Carassius auratus

Three acute PFOS studies were conducted with Carassius auratus, but all are classified

as qualitative as indicated below. Yang et al. (2014) exposed C. auratus to the potassium salt of
PFOS (CAS # 2795-39-3, 99% purity) for 96 hours using static, measured conditions (the
authors note that the experiments followed ASTM 1993b88-729 (ASTM 1993). The goldfish
(6.0-7.0 g) were purchased from the Beijing Chaoyang Spring How or Market, which was
considered an atypical source. The organisms were allowed to acclimate lor se\ en days before
testing, and the test was conducted at 22ฑ2ฐC with a light:dark cycle of 12-hr 12-hr. with 10 fish
per replicate and three replicates per concentration. Beakers used for exposure were assumed
glass, but was not specified by study authors PFOS was dissolved in deionized water and carrier
solvent DMSO to obtain a 7 mg/mL stock solution, and then diluted w ith dechlorinated tap water
to yield nominal exposure concentrations of 2'). 32. 51 2. SI l)2. 13 1.07 and 209.72 mg/L PFOS.
Water quality parameters reported were pi I 7 <) n 5. dissoh ed oxygen=7.0 ฑ 0.5 mg/L, total
organic carbon= n <>2 mu I. and hardness ll)<) n n I mg/L as CaC03. The supplemental data
provided for the study includes a com pari son of measured PFOS concentrations before and after
solution renewal in the low and high acute and chronic test concentrations. PFOS concentrations
in the test water did not fluctuate by more than 15% during experiments. The 96-hour LCso
reported for the study of 81 IS mg/L was deemed qualitative due to the atypical fish source and
unknown composition of test beakers and statistical method used for calculating the LCso.

The effect of PFOS (CAS # 1763-23-1, >98% pure) on oxidative stress enzyme responses
of Carassius auratus juveniles (27.85 g) was evaluated by Feng et al. (2015). The 96-hour static
measured exposure was conducted at a temperature of 23ฐC, pH of 7.25, dissolved oxygen of 6.5
mg/L and total hardness of 174.3 mg/L as CaC03. The fish were purchased from a local aquatic

G-25


-------
breeding base and acclimated in dechlorinated tap water for at least for 10 days, with the total
mortality near zero. After acclimatization, five fish were randomly selected and placed in each
glass tank (two replicates for treatments and five control replicates) containing 20 L of test
solutions (nominal concentrations of 1 or 10 |imol/L PFOS) or 20 L of dechlorinated tap water.
PFOS was dissolved in DMSO to prepare a 100 mmol/L stock solution. The tanks were
continuously aerated, and water was refreshed to minimize the contamination from metabolic
wastes. Antioxidant enzyme activity (CAT, SOD and GPx) and lipid peroxidation were
adversely impacted at 10 |imol/L (5.001 mg/L) PFOS at test termination, but these data were
considered qualitative because of the atypical endpoints reported and only two exposure
concentrations were evaluated.

The swimming behavior of Cantssms alliums exposed to PFOS was reported by Xia et
al. (2013). Juvenile goldfish (6.91 g, 6.01 cm) were exposed to the potassium salt of PFOS
(>99% pure) in glass aquaria under renewal unmeasured conditions for 48 hours at 18ฐC and
photoperiod of I 5-hrs lMus. light dark The fish were obtained from a local market in
Chongqing. China and acclimated al the test temperature for three weeks in dechlorinated tap
water The PI OS stock solution (<) S g ml.) was dissolved in DMSO, with the final exposure
solutions diluted with dechlorinated tap water. The fish were divided into six groups (n=8) and
were exposed to <> (water only). 0 (DMSO vehicle control), 0.5, 2, 8 or 32 mg/L of PFOS (one
replicate per concentration) The concentration of DMSO in the water did not exceed 0.004%
(v/v). The contaminants were administered by replacing 50% of the water with water
contaminated with the appropriate concentration of PFOS each day. After exposure to PFOS for
48 hours, the swimming performance behavior (spontaneous activity and fast-start performance)
of each fish was examined. The 48-hour NOEC and LOEC for swimming behavior (motion

G-26


-------
distance and % of actionless time) was 2 and 8 mg/L, respectively. However, the effects levels
were classified as qualitative due to the non-apical endpoints reported, duration of the exposure,
only eight fish per exposure concentration, and atypical source of test organisms (local market).

G. 2.2.2 Cyprinus carpio

Kim et al. (2010) evaluated the effects of PFOS (100.3% purity) to biomarker responses

exhibited by Cyprinus carpio exposed for 96 hours under flow-ill rough measured conditions.

PFOS stock solutions were prepared in N,Ndimethylformamide ( 100 mg I.) and diluted with

carbon-filtered and dechlorinated tap water to give nominal concentrations of n <>50, 0.500, 5.000

and 50.00 mg/L. Dechlorinated tap water was used as a control. The exposure concentrations of

PFOS ranged from 90 to 124% of the nominal concentrations (or 0, 0.044, 0.620, 5.395, 48.242

mg/L), and where appropriate, the average of I lie PFOS measured concentrations was used to

calculate endpoints when not within ฑ 20?.. of the nonii nal concentrations. The carp were

obtained from the Chungchcongnam-do Experimental Station for Inland Waters Development

and held in 2,000 I. tanks with flowing dechlorinated tap water at 23 ฑ 2ฐC, which was also used

in the study (pH, 6 'ฆK alkalinity. 2S o nig I. as CaCOi; total hardness, 47.8 mg/L as CaC03). Ten

juvenile lisli ( 12 cm. 2d g) were held in each exposure tank (assume one replicate per

concentration) under a lO-hrK-hr (light dark) photoperiod, with water temperature maintained at

23 ฑ 1ฐC. At test termination, the fish were removed from the tanks and evaluated for

biochemical and genetic responses. DNA single-strand breaks was determined to be the most

sensitive endpoint, with a %-hour LOEC of 5.395 mg/L. This study was deemed qualitative due

to the non-apical endpoints reported.

G. 2.2.3 Danio rerio

Ding et al. (2012; 2013) evaluated the acute effects of PFOS-K (perfluorooctane

sulfonate, potassium salt, CAS # 2795-39-3, 98% purity, purchased from Sigma-Aldrich) to

G-27


-------
Danio rerio embryos via a 96-hour static-renewal unmeasured exposure. Adult AB strain
zebrafish were cultured in aerated and biologically-filtered reconstituted freshwater at 26 ฑ 1ฐC.
The day before test initiation, male and female zebrafish, at a ratio of 1:1, were placed in
spawning tanks before the onset of darkness. Mating, spawning and fertilization took place
within 30 minutes after light onset in the morning. Eggs were collected from spawn traps and
washed with clean OECD water. Unfertilized or abnormal eggs u ere removed under a
stereomicroscope. PFOS was dissolved in reconstituted water to achie\ e the desired target test
concentrations; no solvents were used. Six exposure concentrations were lesled with three
replicates each. Graphically, these concentrations are shown as logio(mol/L) concentrations.
These were converted to mg/L (15.74, 35.23, 125.3, 142 2. Ho 9, and 207.5 mg/L, respectively)
given the molecular weight of the form of PI'OS-K used in the study. CAS # 2795-39-3, of
538.22 g/mol. Twenty fertilized eggs per exposure concentration were divided into a 24-well
plate with one embryo per well, containing 2 nil. lest solution (three replicate plates per
concentration x 2d embryos per concentration for a total of 60 embryos per concentration). The
remaining four wells were tilled with control water and one embryo each (four control embryos
per plate \ three plates per concentration for a total of 12 control embryos). Test solutions were
half renewed e\ cry 24 hours An embryo was considered dead when one of four end points (i.e.,
coagulation of the embryo, non-detachment of the tail, non-formation of somites and non-
detection of the heartbeat) was observed. Mortality was monitored and documented at 72- and
96-hour post fertilization (hpf). The author-reported 96-hour LCso was 0.101 mM (or 54.36
mg/L) PFOS. The presentation of the data is problematic, however, as the authors did not appear
to report the control response either graphically or in the text. Specifically, for PFOS, there are
six pairs of symbols (72 hour - plus sign, and 96 hour - x). Most of these symbols overlap. To

G-28


-------
the left of these symbols is what appears to be a (minus sign. Initially, it was thought that
this was a poorly reproduced (plus sign) and digitized it as the 72 hour control. However, on
further inspection, it was noticed that if you increase the size of the figure greatly (800%), the
"symbol" is actually tilted slightly, and appears to be a discontinuous extension of the fitted
PFOS toxicity curve. EPA reached out to the study authors on July 13, 2021, to request the data
from the paper, but have not heard back as of April 12, 2022. Gi\ en the level of data presented in
the paper and the unreported data from the control treatment, HIJA changed the overall use
classification of this paper from quantitative to qualitative.

In a follow-up study to one conducted earlier by the investigators, Hagenaars et al.
(2014), again exposed Danio rerio to the potassium salt of PFOS (CAS # 2795-39-3, purity
>98%), but for six days via renewal unmeasured methodology The objective of the study was to
determine the exposure windows during early /ehralish de\ elopment that are sensitive to PFOS
exposure and result in impaired swim bladder inflation in order to specify the mechanisms by
which this effect might he caused Since PFOS was fully soluble in water in the tested
concentrations (maximum solubility of oso mg I. at 24-25ฐC), no solvents were used. Adult
wildtype zebra fish (breeders) were maintained and tested in reconstituted fresh water (Instant
Ocean i< Sea Salt) at 28 ฑ 0 2 (' and a 14-hr: 10-hr, light:dark cycle. Male and female fish were
separated in breeding tanks with a perforated bottom using a divider. The divider was removed
when the lights turned on in the morning. Within 45 minutes after the lights turned on, spawning
and fertilization took place and eggs were collected and then transferred to clean reconstituted
fresh water with the same composition as in the breeding tanks. Seven different time windows of
exposure (1-48, 1-72, 1-120, 1-144, 48-144, 72-144 and 120-144 hour post fertilization, hpf)
were tested based on the different developmental stages of the swim bladder. These seven time

G-29


-------
windows were tested at four concentrations (0.70, 1.14 3.07 and 4.28 mg/L) at a water
temperature of 28.5ฐC, pH of 7.5, and 14-hr: 10-hr (light:dark) photoperiod. Forty-two normally
shaped fertilized eggs per exposure concentration were divided over a 48-well plate (sterile
tissue culture plates) and each egg was placed individually in 1 mL of the test solution. The
remaining six wells per plate were filled with reconstituted fresh water and used as internal
negative control embryos. A plate was considered valid if no more than one internal negative
control embryo showed lethal effects. Apart from the negative control embryos, a separate plate
with control embryos in clean water was used in each test and this plate was used as the control
for statistical comparison with exposed embryos A lest was considered valid i I" >" u of the
controls successfully hatched and showed neither sub-lethal nor lethal effects. At six days post
fertilization, effects on survival, hatching, swim bladder inflation and size, larval length and
swimming performance were assessed. The reported o-day I .('*ฆฆ and ECso (uninflated swim
bladder) were 6.25 and 2 2l) mu I. PFOS, respecti\ ely, and were considered qualitative due to
test duration.

Huang et al. (2016) later in\ estimated the effect of PFOS (CAS # 1763-23-1, >98% pure)
on metabolic responses ol'Ikmio ivno (measured metabolites (208 in total) included amino
acids, biogenic amines, fatty acids, bile acids, sugars and lipids). Ethanol (0.05-0.1%) was used
as the carrier sol\ ent for PFOS. with results reported as measured values. Adult AB/Tubingen
zebrafish were reared and maintained in a Zebtec aquatic system with re-circulating water (pH
7.0-7.5) and in standard laboratory conditions of 28ฐC and a 14-hr: 10-hr (light:dark cycle). Five
biological replicates comprised a treatment group and each replicate consisted of 80 zebrafish at
either the embryonic (24-48 hpf) or larval (96-120 hpf) stage. All exposures were carried out
statically for 24 hours. No apparent toxic effects were observed during and after the exposure

G-30


-------
period in either the control (embryo media [HE3] and carrier control [CC]) or treatment groups
at either life stage. Thus, the 24-hour LOEC for the different metabolites was >9.7 mg/L PFOS.
The test was classified as qualitative due to test duration and non-apical endpoints.

Effects of PFOS on reactive oxidative stress (ROS) biomarkers of Danio rerio was
investigated by Shi and Zhou (2010). The authors subjected 4 hpf zebrafish embryos to the
potassium salt of PFOS (>99% purity) for 96 hours under static mini ensured conditions. The
stock solution (50,000 mg/L) was prepared by dissolving the crystals in I MM .('-grade DMSO.
The wild-type (AB strain) zebrafish were maintained at 28 ฑ 0.5ฐC in a 14 hour light, 10 hour
dark cycle in a continuous flow-through system in charcoal-filtered tap water. Fertilized eggs
were obtained from natural mating of these adult zebrafish. Zebrafish eggs were collected within
four hours of spawning from several breeding tanks, pooled, washed, and then randomly
transferred into a glass beaker (four replicates) containing 5 DC) ml, of exposure solution (0, 0.2,
0.4 and 1.0 mg/l.) of PI OS ( 3 <)() eggs per beaker) Both the control and the treated embryos
received 0.03% (\ \ ) DMSO The beakers were kept in a humidified incubator at 28.5 ฑ 0.5ฐC
under controlled lighting conditions (14-hr I o-lir light:dark cycle). The reported MATC
(increased ROS formation) was n 2SS2 mg/L PFOS (NOEC = 0.2 mg/L and LOEC = 0.4 mg/L),
but the study was deemed <.|iialitati\e due to the non-apical endpoint and lack of dilution water
information

Padilla ol al. (2012) exposed Danio rerio embryos (6-8 hpf) to PFOS for six days
employing renewal unmeasured procedures. The goal of this study was to describe the broad
application of a zebrafish screening model to 309 chemicals, including PFOS. A stock solution
of PFOS was prepared in 100% DMSO at a concentration of 20 mM, with exposure solutions of
0, 0.001, 0.004, 0.012, 0.030, 0.110, 0.320, 1.00, 2.96, 8.80, 26.6, and 80.0 |iMPFOS or 0.0005,

G-31


-------
0.002, 0.006, 0.015, 0.055, 0.16, 0.5, 1.48, 4.4, 13.3 and 40 mg/L (DMSO at 0.4% v/v) prepared
in 10% Hanks' solution. Wild type adult zebrafish obtained from Aquatic Research Organisms
(Hampton, NH) were held at 28ฐC with a 14-hr: 10-hr light:dark cycle. Adult breeding fish (2-3
females per male; density=15-20 adults per tank) were kept in one of several 9 L flow-through
colony tanks. Typically, adults from two to three colony tanks were mated on the same day. Two
hours after light onset the adults were returned to the colony lank All embryos were gathered
from each breeder tank, pooled, and placed in a 28ฐC water bath for two hours, followed by two
washes with 0.06% bleach (v/v) in 10% Hanks' Balanced Salt Solution for fi\ e minutes in order
to remove any residual bacteria or fungi. Zebrafish embryos were exposed in 96-well plates. On
day 0, approximately 6-8 hours after fertilization, zebrafish embryos were placed one embryo per
well in Millipore Multiscreen Nylon mesh plates and exposed to nominal concentrations of the
chemicals. All embryos and larvae were kepi in a 2(-> <> I (' incubator with a 14-hrs:10-hrs,
light:dark cycle I jnlnyos were exposed lo the chemicals lor ll\ e days post fertilization (dpf)
(i.e., 120 hpf) with daily dosing (i e . complete solution change with chemical renewal every 24
hours), followed In a wash-oul in I lanks" huffcr lor one day prior to the lethality, hatching, and
malformation assessments performed on (•> dpf. The half-maximal toxicity score concentration
(AC:") was lo 44 mg/L. The toxicity score incorporates survival, hatchability, and malformation
(maximum score of 40), and the PFOS score of 28.0 indicates observed inhibition of hatchability
and malformation Ik-cause of the limited details presented for any specific chemical, the
unconventional exposure duration, and the unconventional endpoint, the study was classified as
qualitative.

A 72-hour exposure of Danio rerio embryos to PFOS (98% purity) was conducted by
Zheng et al. (2012) following OECD (1996) methodology. No solvent was used for PFOS

G-32


-------
because of its high water solubility (500 mg/L). Exposure solutions were diluted from the stock
solutions with embryonic water. Adult wild-type zebrafish were obtained from Model Animal
Research Center of Nanjing University and kept in a semiautomatic rearing system (tap water),
with five females and ten males in each 10 L tank at 28 ฑ 1ฐC. Water was exchanged at a rate of
1/3 daily and the lighting was 14-hr: 10-hr (light:dark) photoperiod at 1000 lux. Spawning and
fertilization took place within 30 minutes after the lights were turned on in the morning.

Embryos were transferred to exposure solutions (embryo water) immediately after fertilization
and examined under a stereomicroscope. Damaged or unfertilized embryos were discarded.
Zebrafish embryos were exposed in 24-well cell culture plates (material not identified) with 2
mL solution per well (pH of 8.3 ฑ 0.2, dissolved oxygen concentration of 6.07 ฑ 0.24 mg/L at the
beginning and end of experiments). Twenty normally shaped fertilized embryos were assigned to
each treatment (0, 6.25, 12.5, 25, 50, 100, 200 mg/L.) or control group. All concentrations were
repeated in triple at different days with different batches of eggs. Embryos were cultured in an
incubator at 28.5 (' after exposure The reported 72-hour LCso and ECso (malformations) were
68 and 37 mg/T. PFOS. rcspecli\ ely I lowe\ er. the data were considered qualitative because the
duration is too short lor an acute exposure

l)u el ;il. (2013) in\ estigated the effect of PFOS (>98% purity) on the survival,
malformation and suppression of steroidogenic enzyme synthesis of Danio rerio embryos
exposed via renewal unmeasured conditions for 120 hours. PFOS stock solutions were prepared
in DMSO at a concentration of 0.1 M and stored at -20ฐC. They were diluted to desired
concentrations in culture medium immediately before use, and the final concentration of DMSO
in the culture medium did not exceed 0.1% (v/v). Wild-type adult male and female zebrafish,
obtained from the Model Animal Center of Nanjing University, were maintained on a 14-hr: 10-

G-33


-------
hr light:dark cycle at 28ฐC under semi-static conditions with charcoal-filtered water. Spawning
was induced in the morning when the light was turned on. Fertilized eggs were collected 30
minutes later and examined under the microscope. Only those that had developed normally were
selected. Embryos were incubated with embryo medium in Petri dishes for subsequent
experiments. Zebrafish embryos at 4 hpf were exposed (three replicates, 40 fish per replicate) to
0.100, 0.200 and 0.500 mg/L PFOS and 0.001% DMSO (control) at 2KT. with daily renewal of
the embryo medium. Embryo survival and stage of embryonic development were recorded daily
until test termination (120 hpf). The NOEC and LOEC for suppression of steroidogenic enzyme
synthesis was reported as 0.100 and 0.200 mg/L, respecti\ elv Since only non-apical endpoints
were reported, these data are classified as qualitative In addition, the authors noted that no
effects of mortality or malformation were oltsei \ ed. thereby resulting in a LOEC of >0.500
mg/L. However, this was a low effect concentration compared to other acute values and therefore
of little value in dei i\ ing criteria

(Chen el ;il. 20l4h) e\ aluated the effects of PFOS (>96% purity) on malformation of
Danio rerio statically exposed lor 4K hours following procedures described by Westerfield
(1993) PI OS was dissoKed in l<)<)'\> dimethyl sulfoxide (DMSO) to prepare PFOS stock
solutions (32 niM), followed In serial dilution with embryo medium (EM) to prepare the
exposure solutions (DMSO of 
-------
was turned on. Embryos were collected within 0.5 hours of spawning, rinsed in EM and staged
under a dissecting microscope to select those with normal morphology. Embryos/larvae were
waterborne exposed to nominal PFOS concentrations of 8, 16, 32 |iM (4.001, 8.002, and 16.00
mg/L) in 6-well plates (20 embryos per well with 5 mL solution) from 0 to 48 hpf or 48-96 hpf.
At the end of each exposure period, the embryos or larvae were rinsed three times with EM and
transferred to 96-well plates (1 embryo per well with 200 |iL solnlion) for continuous
development until 120 hpf, where the incidence of various malformation was scored. Complete
malformation (100%) was observed at 8.0 mg/L PFOS, but since the duration was only 48 hours,
the study was deemed qualitative.

Danio rerio behavioral alterations in response to acule lJFOS exposure (potassium salt,
CAS # 2795-39-3, >98% purity) was investigated hy Spulbcr ol ;il. (2014). PFOS was dissolved
in DMSO (1 mg/mL), and further diluted in DM SO and rearing water to yield exposure
concentrations of either <> I or I mg/L (0.1% DMSO). Control embryos were exposed to 0.1%
DMSO in E3 water Wiklt\ pe \li xebrallsh embryos were obtained from the zebrafish core
facility at Karolinska Institute IJiceding groups of adult fish (three males and two females) were
housed together o\ernight in in I. spawning tanks containing environmental enrichment
(commercially a\ ailable aquaria made of non-toxic plastic). Thirty minutes after turning the light
on, the fertilized eggs were collected and stored at 28.5ฐC until further processing. The eggs
were washed twice with fresh L3 water (pH 7.4) at room temperature and under constant
illumination (approximately 500 lux). The exposure to PFOS was initiated about 2 hpf. The
zebrafish larvae were then plated and maintained individually in 48-well plates (cylindrical
wells, 10 mm inner diameter) in 750 ju.1 E3 water at 28ฐC in a 14-hr: 10-hr light:dark cycle (300
lux intensity, daylight-matching spectrum white light) until behavioral testing at six days post

G-35


-------
fertilization (dpf). The exposure followed a static, non-replacement regime. All treatments were
present in each plate, and the larvae were distributed such that an equal proportion from each
group would be placed in wells at the periphery of the plate. The mortality, successful hatching,
and the occurrence of embryonal malformations was assessed at the 24 (developmental
failure/coagulation, light-induced coiling movements), 48 (developmental failure/coagulation,
blood circulation, pericardial oedema, pigmentation), 72, and 12') hpf (hatching, eye
development, swimming bladder inflation, morphological abnormalities such as scoliosis or bent
spine), as well as after completing the behavioral experiments (6 dpf). The lai \ ae displaying
morphological abnormalities at 6 dpf were excluded from analyses. The behavioral \OEC and
LOEC (spontaneous swimming activity) were 0.1 and 1 .0 mg/L PFOS, respectively. The data
were considered qualitative since only two exposures, duration and non-apical endpoints.

The acute toxic effects of PFOS (potassium salt. lW\. purity) on sperm vitality,
kinematics and fertilization success in Danio rcrio was in\ estigated by Xia and Niu (2017).
PFOS was initially dissoK ed in DM SO. and the stock solution (0.5 g/mL) was stored at 4ฐC until
preparation of the final exposure solutions in dilution water (Hank's Balance Salt Solution).

Adult zebrallsh ( \Ii strain) were maintained according to standard culture protocols (Westerfield
1995) The male and female lish were housed in separate aquariums for four weeks prior to the
experiment The rearing water was dechlorinated tap water, maintained at 25ฑ1ฐC, dissolved
oxygen level >6 mu I.. the pH ranged from 7.0 to 7.8, and the rearing system was maintained
under a 14-hr: 10-hr light:dark cycle. After the acclimation period, male and female zebrafish of
uniform size (0.41 g and 0.49 g, respectively) were selected as the experimental fish. Sperm
vitality and kinematics were determined with a CASA system. To avoid the possible effects of
inter-male variation in sperm quality, sperm from the same fish were activated by mixing 1 [xL

G-36


-------
sperm suspension with 9 [iL different activation solutions containing a range of PFOS
concentrations (0, 0.1, 1 and 10 mg/L). Specifically, the concentrations of PFOS in the treatment
groups were 0, 0.09, 0.9 and 9 mg/L. The concentration of DMSO in the activation solutions did
not exceed 0.002% (v/v). The viabilities and kinematics of sperm exposed to the different
treatments were assessed at 20, 40 60 and 80 seconds after activation at room temperature (25ฐC)
and pH of 8. The percentage of motile sperm, the curvilinear velocity: the straight-line velocity,
the angular path velocity and the mean angular displacement (MAD) of spermatozoa were
calculated for each group from three recordings of at least 50 sperm (30 frames s) A total of 24
male fish (n=24) were used for each PFOS treatment group F.ggs (approximately from
individual females were collected by gentle abdominal massage, mixed with 100 [xL sperm
suspension (sperm to egg ratio 3000:1), and immediately exposed to 5<~> mL solutions in 9 cm
diameter petri dishes containing a range of PI-OS concentrations (0, 0.09, 0.9 and 9 mg/L). After
two minutes, the euus were washed three times and then transferred to uncontaminated water
without PFOS. The fertilized euus at the uastrulation stage were examined using a
stereomicroscope. and the fertilization rate was then determined six hours after exposure to
spermatozoa All of the manipulations were performed at 25ฐC. A total of 24 male and 24 female
fish (n 24) were used for each PF()S treatment group. The NOEC and LOEC for sperm motility
and fertilization success were <> 09 and 0.9 mg/L PFOS, respectively. These data were deemed
qualitative due to the non-apical endpoints evaluated and duration of the exposure.

Annunziato (2018) evaluated the acute effects of perfluorooctane sulfonate (PFOS) on
zebrafish (Danio rerio) via a 5-day static, unmeasured study. AB strain zebrafish were sourced
from Zebrafish International Resource Center and followed Rutgers University Animal Care and
Facilities Committee guidelines protocol 08-025. Fish were maintained at a pH of 7.2-7.7,

G-37


-------
temperature of 27 ฑ 1ฐC and a 14-hr: 10-hr light:dark cycle, and were fed twice daily a diet of
artemia in the mornings and aquatox/tetramin flake mix in the evenings. The stock solution was
prepared as 2,000 |iM in the same water used to incubate the eggs. Twenty-five embryos (3 hpf)
were exposed in 20 mL glass vials at a concentration of 0 (control), 0.02, 0.2 and 2.0 |iM for five
days. There were four or five replicate vials per concentration. At five days, larvae were snap
frozen in liquid nitrogen, and RNA was isolated. Authors reported a I 17-hr LOEC for gene
expression of Leptin A mRNA of 0.02 |iM, or 0.01 mg/L PFOS based on a molecular weight of
500.13 g/mol. The atypical test endpoint and test duration makes the study acceptable for
qualitative use only.

Dang et al. (2018) evaluated the acute effects of lJFOS (CAS # 2795-39-3, >99% purity)
to Danio rerio via a 96-hour renewal unmeasured exposure. Slock solutions were made in
DMSO, but the concentration in exposure solutions was not identified. Sixteen-week old adult
zebrafish (wild type. \li strain) were maintained in a flow-through system as previously
described (Dang et al . 2<> I 5. I .in et al . 2009). Sexual mature zebrafish were cultured at 28 ฑ
0.5ฐC with a 12-hr 12-hr light dark cycle and were fed thrice daily newly hatched Artemia
nauplii Fertilized eggs were rapidly collected and counted from natural crosses after lights on in
the morning and then were examined under a stereomicroscope. Normally developed embryos
were randomly distributed in l->0 mm culture dishes containing 30 mL rearing water (60 mg/L
instant ocean salt in aerated distilled water) for subsequent experiments at 1-2 hours post
fertilization (hpf). The experiment included two parts. First, 450 embryos from six pairs of fish
were collected and cultured in three dishes, and each dish contained 150 embryos. The
embryos/larvae were sampled at 2, 24, 48, 72, 96, and 120 hpf to examine the mRNA expression
profiles of GH/IGF axis genes; and larvae were sampled at 72, 96, and 120 hpf to measure body

G-38


-------
length. The body length of 20 larvae from each replicate at post-hatch along the body axis from
the anterior-most part of the head to the tip of the tail was measured with digital images using the
Image Pro Plus software. Twenty embryos/larvae were pooled to produce one replicate for the
subsequent quantification of mRNA expressions of genes involved in GH/IGF axis and each
concentration contained three replicates. Second, based on the results of the first part of the
experiment, PFOS (and other chemicals) were chosen to study the responses of GH/IGF axis of
zebrafish embryos/larvae at 96 hpf. A total of 1,200 embryos and four concent rations were used
for toxicity testing each chemical. Each concentration contained three replicates, and each
replicate contained 100 embryos. Stock solutions of PI OS were prepared in DMSO and each
group contained the same concentration of DMSO or rearing water. Nominal exposure
concentrations were 0.1, 1 and 10 |iM PFOS (or <> (>5, 0.50, and 5 <) mg/L). During the exposure
period, dead larvae were removed from the culture dishes and exposure solutions were renewed
at 48 hpf. Endpoints of embryo toxicity test at 96 hpf included survival rate, hatching rate,
malformation incidence and body length Heart rates were recorded at 72 hpf. For the
measurement of snr\ i\ al. hatching and malformation incidences, all the 100 eggs were used, and
for the calculations of heart rates and body length, twenty larvae from each dish, a total of sixty
larvae, were used. The most scnsiti\ e endpoint was length, which was atypical for an acute
exposure. The lK->-hourNOI-C and LOEC for growth were <0.05 and 0.05 mg/L, respectively.
However, these data were considered qualitative since it was atypical for an acute assessment. In
addition, there was no effect on survival, so the 96-hour LCso was >5.0 mg/L PFOS.

The effects of PFOS on oxidative stress exhibited by Danio rerio 1 hpf embryos (wild
type) was evaluated by Sant et al. (2018). Stock solutions of 160, 320, and 640 mM for embryo
exposures were prepared by dissolving PFOS into DMSO, and stored at room temperature in

G-39


-------
glass bottles inside of light-prohibitive, airtight containers until use. Adult fish populations were
maintained in an automated Aquaneering zebrafish system at 28.5ฐC and following at 14-hr: 10-
hr light:dark cycle daily. Breeding populations were housed at an appropriate density with a 2:1
female-to-male ratio. Embryos for experiments were collected with 1 hpf from homozygous
genotyped tanks, washed thoroughly, and confirmed for fertilization prior to experimental
proceedings. Pools of 10-15 mid-blastula stage embryos of each genotype were separately
collected into fresh polystyrene petri dishes containing 20 mL of 0.3x Danieau's media (pH of
7.6 and 28.5ฐC). Polystyrene plates were used because they have lower matrix
retention/adherence rates for PFOS than other materials such as glass. Stock solutions of PFOS
or DMSO were added to the dishes at 0.01% v/v, resulting in exposures to 0 (DMSO control),
16, 32, or 64 |iM PFOS (or 0, 8.002, 16.0<). 32 n I mg'I.Y All exposure media was refreshed
daily. At 24 hpf, all embryos were manually dcchoiionated using watchmakers' forceps and
chorion debris >\as remo\ ed from dishes. All experiments were repeated 3-4 times. Embryos
were imaged at hpf lbr embryonic morphology using a FBS10 Fluorescence Biological
Microscope Embryos screened lbr morphological deformities were imaged at 5x magnification
using transmitted light microscopy The pericardial area NOEC and LOEC values were 8 and 16
mg/I. PI-OS. respectively, and were considered qualitative due to the non-apical endpoint.

Daiiio rcno was e\aluated by Huang et al. (2010) in a 114-hour static measured
exposure to PFOS (CAS # 1763-23-1, > 96% pure). PFOS stock solutions (32 mg/L) were
prepared in DMSO, followed by serial dilution with the embryo medium (DMSO of 0.1%). The
control also received 0.1% DMSO. Wild-type (AB strain) zebrafish founding stocks were
obtained from the Environmental Health Sciences Center at Oregon State University. All fish
were raised and kept at standard laboratory conditions of 28ฐC on a 10-hr: 14-hr light:dark

G-40


-------
photoperiod in a recirculation system. Zebrafish embryos were obtained from spawning adults in
tanks overnight with the sex ratio of 1:1. Embryos were collected within 0.5 hours of spawning
and rinsed in embryo medium (EM). Fertilized and normal embryos were staged under a
stereomicroscope. To determine the LCso, zebrafish embryos were exposed to 0, 0.25, 0.5, 1.0,
2.0, 4.0, and 8.0 mg/L PFOS from six to 120 hour post-fertilization (hpf). Embryos were kept in
sterile 96-well plates, with one embryo per well, containing 200 ul. treatment or control
solutions (EM). For each exposure condition, five replicates each with 32 embryos were
performed at 28ฑ0.5ฐC in a light-controlled incubator. For ECso determination, lower exposure
concentrations of 0, 0.005, 0.05, 0.5, 2.0, and 4.0 mu I. were used. The end points of toxicity
included bent spine, malformed tail, pericardial edema, yolk sac edema, uninflated swim bladder,
failed hatching, single eye, and opaque head (apparent necrosis) The I 14-hour LCso was 2.2
mg/L PFOS, and the malformation ECso was I 12 mu I. PI OS These data were considered
qualitative due to exposure duration, which was longer (at I 14 hours) than fish acute toxicity test
guidelines (OCSPP X5<) 1 < >75) with exposures of 96 hours. Both of these toxicity values were
within a factor of two of the I AY of 2 222 muL and indicated that this genus might be more
sensiti\ e to acute exposures of PI OS than the quantitative data for the genus indicate. However,
it is unlikely that these toxicity values would substantially change the FAV, as this genus was not
among the four most sensiti\ e genera (Danio GMAV of 29.31 mg/L ranks eighth), and therefore
would have little impact on the acute freshwater criterion.

A sub-chronic static unmeasured test was utilized by Ulhaq et al. (2013) to determine the
toxicity of PFOS to Danio rerio. PFOS stock solutions were freshly prepared in reconstituted
water in concentrations below the limit for water solubility. Adult zebrafish (AB strain) were
held in charcoal-filtered tap water. Breeding groups including three males and two females were

G-41


-------
placed in 10 L glass aquaria equipped with spawning nets separating the parental fish from the
eggs. Half an hour after onset of lights the eggs were collected, rinsed for removal of debris, and
then only normally developed fertilized eggs at least in the four-cell stage were selected using a
stereomicroscope. Within 15 minutes after collection, the zebrafish eggs were exposed to a series
of concentrations of the test substance dissolved in reconstituted water (exposure medium).
Fertilized eggs (4-cell stage) were randomly distributed individuall\ into flat bottom, 48-well
polystyrene plates along with 750 |iL of the exposure medium. PFOS was tested at six
consecutive concentrations differing by a factor of 3.3 based on logarithmic scale lilting. For
each test, four 48-well plates were used, with a total of 24 embryos per concentration as well as
24 in the water control group. Each treatment group was equally distributed to each of the four
well plates (i.e., six embryos/concentration plate giving a total of 168 embryos). The plates were
covered with parafilm and the embryos were exposed to the chemical until 144-hour post
fertilization (hpf) l-'ish laboratory conditions throughout the study were kept at pH 7.2-7.6, a
water temperature of 2o I (' and a light cycle of 14 hours. Observations of mortality and
sublethal endpoints were made after 24. 4S. 12<~) and 144 hpf using a stereomicroscope. Sublethal
endpoints such as presence of edema, malformations, not-hatched eggs, lack of circulation and
reduced pigmentation were also observed. Fleart rate was recorded at 48 hpf and hatching time
was determined using time-lapse photography. The 144-hour LCso was >10 mg/L PFOS and the
ECso (lethal and sublethal effects, including spinal curvatures, oedemas, uninflated swimbladder,
and side-lying) was 1.5 mg/L PFOS. Both were considered qualitative data because of test
duration, which was longer (at 114 hours) than fish acute toxicity test guidelines (OCSPP
850.1075) with exposures of 96 hours. Furthermore, the LC50 for mortality was a less certain
greater than value. However, these toxicity values were within a factor of two of the FAY of

G-42


-------
2.222 mg/L and indicated that this genus might be more sensitive to acute exposures of PFOS
than the quantitative data for the genus. However, it is unlikely that these toxicity values would
substantially change the FAV, as this genus was not among the four most sensitive genera
(Danio GMAV of 29.31 mg/L ranks eighth), and therefore would have little impact on the acute
freshwater criterion.

Sharpe et al. 2010 examined the bioaccumulation and toxicity of PFOS isomers on
Danio rerio through three different tests, a 96-hour renewal toxicity test on adults, a 48-hour
renewal toxicity test on embryos, and a chronic exposure test that evaluated maternal transfer
and fecundity of PFOS isomers. The 48-hour tests are described in this present section, as these
results were used qualitatively. The 96-hour test was used quantitatively to derive the acute water
column criterion (see Appendix A) and the chronic toxicity tests were used qualitatively and are
summarized in Section 4.4.2.1.4. Zebrafish were purchased from a pet store local to the
University of Alberta and were reared at university facilities for six to ten months. Conditioned
zebrafish water obtained from the liiolouical Sciences Zebrafish Facility at the University of
Alberta was used to acclimate the fish in 7<) I. glass aquaria where they were fed powdered trout
chow (I 'nifeed) daily, occasionally supplemented with live brine shrimp. An automated reverse
osmosis system was used 1o maintain conditioned zebrafish water, used for acclimation and
testing, at a total hardness of around 160 mg/L and a calcium carbonate hardness at 20 mg/L.

Test concentrations were diluted from a 25 mg/mL stock solution in a methanol (MeOH) solvent
for dosing in all experiments.

The 48-hour embryo toxicity test followed the same OECD guideline. This experiment
was also performed in triplicate. Embryos for the experiment were collected one hour post-
spawn and their fertilization was verified with a compound microscope. Forty fertilized embryos

G-43


-------
were randomly selected and placed into 24-well cell plates for 72 hours at the following
concentrations; control (0 mg/L PFOS), solvent control (0.65% MeOH v/v), and 1.1, 2.4, 5.3,
11.7, 25.8 and 56.8 mg/L measured PFOS concentrations. The testing water and dosages were
renewed daily and mortality was measured after 48 hours to calculate a 48-hour LCso. In addition
to mortality observations, 24-hour checks for the presence of detached tails and somite
formations were performed. At 48 hours, observers looked for l lie presence of a heartbeat and
eyespot, and at 72 hours, the number of hatched embryos were conn led

The author-reported LCso at 48 hours for the embryo toxicity ranged from 7.7 - 38.9
mg/L PFOS; however, the test results were considered to be inconsistent given the w ide range of
the LC50S. Additionally, the study authors note that developmental deformities (i.e., delayed
eyespot formation, lack of a heartbeat at 4K-hours. yolk sac deformities, and a lack of
cephalization) were observed in two of the three assays. I low ever, given that the test was not of a
standard duration lor criteria deri\ ation per EPA's test quality guidelines and the inconsistency
in the author-reported toxicity \ allies, the 4S-liour embryo study was considered for qualitative
use. However, these toxicity \ allies were similar to the SMAV of 24.44 mg/L and support the
toxicity \ al ne used for this genus in the deri\ ation of the acute freshwater criterion.

Uninicri el al. (2017) evaluated the acute effects of PFOS (perfluorooctane sulfonate
purchased from Siuma-Aldrich) on zebrafish (Danio rerio) in a 48-hr static measured study. A 2
mg/mL stock solution was prepared by dissolving PFOS in methanol and stored at <4ฐC until
use. Wild type fish were obtained from AZTI Zebrafish Facility and maintained at 27ฐC, a 12-
hr: 12-hr light:dark cycle, and were fed twice daily with commercial feed. Embryos were held in
culture water for seventy-two hours before testing. Twenty-five hatched embryos were exposed
in 10 mL of test solution in glass petri dishes 6 cm in diameter at 27ฐC under a 12-hr: 12-hr light-

G-44


-------
dark photoperiod for 48 hours. Triplicate exposures ranged from 10 to 500 mg/L PFOS with a
maximum of 0.45% DMSO in any exposure. Samples of each exposure solution were taken at
the beginning and at the end of the test to determine PFOS concentrations. The reported 48-hr
LCso value was 107.6 mg/L PFOS is acceptable for qualitative use because of the short test
duration.

Ortiz-Villanueva et al. (2018) evaluated the acute effects of pcrfluorooctane sulfonate
(PFOS, purchased from Sigma-Aldrich) on wild-type zebrafish (Danio rcrio) in a 2-day
unmeasured, static-renewal study. Adult fish were maintained in reverse-osmosis water mixed
with 90 |ig/mL Instant Ocean salt and 0.58 mM CaS04 2FhO and were fed twice daily with
Tetramin flakes. A stock solution was prepared by mixi nu PI OS with dimethyl sulfoxide on the
day of the experiment, with exposure concentrations of 0 (control). 0 2. 0.5, 1.0, 2.0, 5.0, 10.0,
15.0, 20.0 and 200 |iM PFOS (maximum of 0.2% DMSO). Six males and three females were
placed in a 4 L spawning lank with a mesh bottom At two hp I", fertilized embryos were
collected, rinsed and maintained in o-well multi-plates (10 embryos per well) at 28.5ฐC under a
12-hrs:12-hrslight dark pholoperiod until 48 hpf. Test solutions were added, and fish were
obser\ ed lor an additional 72 hours (I 2d hpf) Each treatment was replicated five times for a
total of fifty embryos per exposure concentration. Study authors reported following protocols
DAMM 7669 and 7(^4. The reported PFOS LOEC was 2.0 |iM (1.0 mg/L) for malformations,
and 20 |iM (10 mu I.) for sur\ ival (based on a molecular weight of 500.13 g/mol PFOS). The
short test duration made the study acceptable for qualitative use only.

(Martinez et al. 2019b) evaluated the acute effects of perfluorooctane sulfonate
potassium salt (PFOS-K, >98% purity, CAS No. 2795-39-3, obtained from Sigma-Aldrich) on
zebrafish (Danio rerio) in a 3-day unmeasured, static-renewal study. The stock solution was

G-45


-------
prepared by dissolving PFOS salt in dimethyl sulfoxide and stored at -20ฐC. Adult, wild type
zebrafish, twelve to eighteen months old, were kept at 28ฐC under a 12-hr: 12-hr light:dark
photoperiod and fed dry flakes twice daily in accordance with protocols DAMM 7669 and 7964.
Eggs were collected and rinsed at two hours post-fertilization (hpf) and put in 6-well multi-plates
with ten fertilized eggs per 3.0 mL of test solution with five or six replicates per test
concentration. Reverse-osmosis purified water was combined with l)<) ug/mL Instant Ocean and
100 |ig/mL CaS04 2H2O to create culture medium and test concentrations of <) (control, 0.2%
DMSO), 0.2, 0.5, 1.0,2.0, 5.0, 10.0, 15.0, 20.0 and 200 |iM PFOS. After fi\ e dpi", surviving fish
were sacrificed and measured for body length, which yielded a LOEC value of I <> <> iiM, or
5.382 mg/L PFOS calculated using the molecular weight of 538.22 g/mol for PFOS-K. The study
is acceptable for qualitative use because of the short test duration

Vogs et al. (2019) evaluated the acute effects of peril uorooctane sulfonic acid potassium
salt (PFOS-K, 5 ox"., purity. CAS No. 2795-3^-3. 7 Oxl <)"" mu L solubility at 25ฐC) on zebrafish
(Danio rerio) embryos in a II 8-hour measured, static-renewal study. AB strain fish used in this
study were pro\ iclecl In the /ehrallsh Core facility at Comparative Medicine, Karolinska
Institute Three male and three female adults were grouped, and embryos were collected in E3
medium directly after spawning. Study authors reported following OECD TG236. A stock
solution was prepared by dissolving PFOS into dimethyl sulfoxide to achieve initial measured
concentrations of <> <>4. n <)K and 0.76 |iM in E3 medium. Thirty embryos (2 hpf) were placed in
30 mL exposure medium in 50 mL glass petri dishes maintained at 28ฑ1ฐC under dark
conditions throughout the exposure (until 120 hpf). When the medium was renewed was not
provided. A 118-hr ECso value of 3.8 |iM PFOS (or 2.045 mg/L based on molecular weight of
538.22 g/mol PFOS-K), was reported for mortality, non-inflated swim bladders, pericardial and

G-46


-------
yolk sac edemas, and scoliosis. The atypical test duration makes the study for acceptable for
qualitative use only.

Dasgupta et al. (2020) evaluated the acute effects of perfluorooctane sulfonic acid
(PFOS, CAS No. 1763-23-1, 97% purity, purchased from Synquest Laboratories) on zebrafish
(Danio rerio) via a 66-hour unmeasured, static study. A stock solution was prepared with either
DMSO or NaOH and stored in 5 mL glass vials and kept at room temperature. The test solution
(50 mM PFOS) was prepared by spiking stock solution into water deri\ ed from the recirculating
water system used to maintain and breed adult, wildtype (5D) zebrafish. Light embryos (6 hpf)
were incubated and exposed to 10 mL of either a control or 5<~> tiM PFOS until 72 hpf at 28ฐC
under a 14-hr: 10-hr light:dark photoperiod. At test termination there was no significant effect of
survival or development on zebrafish embryos The 66-hour \OI-(" of 50 |iM PFOS (or 25 mg/L
based on molecular weight of 500.13 g/mol PI-OS), based on sur\ i\al, is acceptable for
qualitative use only due to the short exposure period

Truong ol al. (2014) e\ aluated the acute effects of potassium perfluorooctanesulfonate
(PFOS-K, CAS 27l)5-3l)-3) and perlluorooclane sulfonic acid (PFOS, CAS #. 1763-23-1) on
zebrafish (Pernio rcrio) in a I 14-hour unmeasured, static study. A stock solution was prepared in
100% dimethyl sulfoxide at a concentration of 20 mM and stored at -20ฐC until 30 minutes prior
to embryo exposures Exposure concentrations were 0 (control), 0.00614, 0.0614, 0.614, 6.144
and 61.44 |iM PI 'OS-K and O.O064, 0.064, 0.64, 6.4 and 64 |iM PFOS. 5D wild-type zebrafish
were obtained from the Sinnhuber Aquatic Research Laboratory at Oregon State University in
Corvallis, Oregon. Fish were maintained at 28ฐC under a 14-hr: 10-hr light:dark photoperiod in a
synthetic fish culture water consisting of reverse osmosis water supplemented with a
commercially available salt (Instant Ocean). Embryos were collected from breeding tanks using

G-47


-------
a spawning funnel and dechorionated using pronase. Dechorionated embryos (6 hpf) were placed
one embryo per well in a 96-well plate (32 replicates per concentration) filled with 90 |iL
embryo medium and 10 |iL of stock solution (final DMSO concentration of 0.64%). The well
plates were sealed and covered in foil until embryos reached 120 hpf. The reported 114-hr
mortality LOEC was 61.44 |iM PFOS-K, or 33.07 mg/L, based on a molecular weight of 538.22
g/mol. The 114-hr LOEC, based on mortality was 6.4 |iM PFOS. or 3 2 mg/L, based on a
molecular weight of 500.13 mg/L. Both values are qualitative only because of the atypical test
duration.

Cormier et al. (2019) evaluated the acute effects of (1. 1.2.2,3,3,4,4,5,5.(•>.(•>.7.7.8,8,8-
heptadecafluorooctane-1-sulfonic acid (PFOS, purity lW\>. CAS No. 2785-37-3, purchased
from Sigma-Aldrich in St. Louis, MO) oil /.eh m fish (Danio rcrio) \ ia a 96-hour measured, static-
renewal study. Three universities participated in this study, and although zebrafish husbandry
conditions varied between the laboratories, all were within OI-CD TG 236 guidelines. Note: only
the Orebro Unh ersity tested PI-OS l-'ish were fed ad libitum with TetraMin and newly hatched
brine shrimp nanplii two times daily /ehralish embryos were collected and tested according to
OECI) T(i 23(\ and embryo exposure started at a maximum of 4 hpf. Triplicate 100 mL glass
vials co\ ered with lids were tilled with 20 mL test solution with five embryos per container.
Aqueous exposures of PFOS were conducted at concentrations of 0 (control), 0.1% DMSO
(negative control) and 7<)<) nu L maintained at 26 ฑ 1ฐC and a 10-hr: 14-hr dark/light rhythm. The
authors reported a 96 hr NOEC value of 700 ng/L PFOS (or 0.0007 mg/L) for hatching success,
embryo mortality, and developmental deformations. Because the value represents a greater than
low value (see 2013 Ammonia rule; U.S. EPA 2013) the study is only used qualitatively in the
acute criterion.

G-48


-------
G. 2.2.4 Pseudorasbora parva

Juvenile female topmouth gudgeons, Pseudorasbora parva, were exposed to the

potassium salt of PFOS (>99% purity) for 96 hours by Xia et al. (2014). PFOS was initially

dissolved in DMSO, and the stock solution (0.8 g/mL) was kept at 4ฐC until prepared for the

final exposure solutions in water. Live fingerlings of P. parva were obtained from a local market

in Chongqing, China (note: P. parva is not a North American species) Individuals of uniform

size (0.81 g body weight and 4.03 cm body length) were selected and acclimatized in a 120 L

recirculating water tank system at Chongqing Normal I Iniversity for at least two weeks prior to

the experiment. Fish were maintained in water at a constant temperature of 15 I (' under a

photoperiod of 14-hr: 10-hr light:dark. The dissolved oxygen lc\ el was kept above 7 mg/L. The

rearing water was dechlorinated and filtered through activated carbon. Fish were fed daily with

commercial tubifex, and were used when no mortality was observed in the acclimation

population. Waterborne exposures were conducted in a renewal exposure experimental

apparatus, which consisted of se\ eral glass aquariums with a capacity of approximately 22 L of

water. Prior to exposure, a total of So lopmouth gudgeon were randomly divided into five groups

(n=l(->). and were gently transferred to the aquariums. Fish were maintained at same conditions as

described above for one week to eliminate stress. After the habituation period, the fish were

exposed to <> (DMSO vehicle control), 0.5, 2, 8 and 32 mg/L of PFOS (nominal concentrations)

for 96 hours. The concentration of DMSO in the water did not exceed 0.004% (v/v). The

toxicants were administered by replacing 50% of the water with water contaminated with the

appropriate concentration of PFOS every day. Food was not provided during the course of

exposure and subsequent testing. After termination of PFOS exposure, the metabolic rate and

swimming performances were examined. For each treatment, eight fish (n=8) were used for

RMR determination, and another eight fish (n=8) were used for swimming measurements. The

G-49


-------
96-hour NOEC and LOEC for spontaneous swim behavior (swim distance) were reported as 0.5
and 2 mg/L PFOS, respectively, and were qualitative due to the non-apical endpoint and atypical
source of test organisms (local market). However, since all 16 fish of those originally initiated
were evaluated for swimming, it was assumed that no mortality occurred which gave a 96-hour
LCso of >32 mg/L PFOS.

Yang et al. (2014) also evaluated the toxicity of PFOS (potassium salt, CAS # 2795-39-
3, 99% purity) to Pseudorasboraparva via 96-hour renewal measured exposures (the authors
noted that the experiments followed ASTM standards and USEPA procedures lor deriving water
quality criteria). The topmouth gudgeon (4.0 g, 4 " cm) were purchased from the lieijing
Chaoyang Spring Flower Market, which was considered ail atypical source. The organisms were
allowed to acclimate for seven days before testing. and the test was conducted at 22ฑ2ฐC with a
light:dark cycle of 12-hr: 12-hr, with 10 fish per rcplicale and three replicates per concentration.
Beakers used for exposure were assumed glass, but glass was not specified by study authors.
PFOS was dissoK ed in deionixed water and carrier solvent DMSO to obtain a 7 mg/mL stock
solution, and then diluted with dechlorinated tap water to yield nominal exposure concentrations
of 30. 45. (->7 5. 11) | 25. 151 SS and 227 81 mg/L PFOS. Water quality parameters reported were
pH=7 i) i).5. dissolved ox\ uen=7 n <),5 mg/L, total organic carbon=0.02 mg/L and
hardness I > < > < > I mg/]. as CaC03. The supplemental data provided for the study includes a
comparison of measured PI OS concentrations before and after solution renewal in the low and
high acute and chronic test concentrations. PFOS concentrations in the test water did not
fluctuate by more than 15% during experiments. The 96-hour LCso reported for the study of
67.74 mg/L PFOS was deemed qualitative due to the atypical fish source and unknown
composition of test beakers and statistical method used for calculating the LCso.

G-50


-------
G. 2.2.5 Lepomis macrochirus

3MCompany (2000) evaluated the acute effects of perfluorooctane sulfonate, DEA salt

(CAS # 70225-14-8), also known as PFOS DEA salt FC-99, or 3M Sample No. 2, on bluegill

sunfish (Lepomis macrochirus) via a 4-day static, unmeasured test. Bluegill (average length 28.6

mm, average weight 0.60 g) were sourced from Osage Catfisheries, Inc. in Osage Beach, MO.

Fish were exposed in 40 L glass aquaria containing 30 L of test solution at a loading rate of 0.2 g

fish/L. Dilution water was laboratory well water with hardness 255 mu I. as CaC03, alkalinity

368 mg/L as CaC03, pH 7.8 and conductivity 50 |jmhos/cm. A primary slock solution prepared

in deionized water at a concentration of 150 mg/L was diluted with dilution water to achieve six

nominal concentrations (18, 37, 75, 160, 320, and 65<> mu I. PI OA) plus a negative control. The

exposure consisted of a single replicate with in fish lor each treatment. Dissolved oxygen

measured in the control and lowest treatment (IS mu I.) remained above 5.8 mg/L during the

exposure, and pH ranged from 8 2 to 8.3 The author-reported 96-h LC50 was 31 mg/L but due

to there being just a single replicate often fish in each treatment, the acute value is only being

considered qualitati\ ely

G.2.3 Amphibians

(}. 2./ Xenopus Utc\•is

Ssin-Segundo et al. (2016) evaluated the acute toxicity of PFOS (potassium salt, CAS #

2795-3-3, > o puritv) to Xenopus laevis embryos via 96-hour renewal measured exposures

using the FETAX assay PI OS was dissolved in boiling distilled water to obtain a stock solution

of 300 mg/L, and testing solutions were prepared in FETAX solution. X laevis embryos were

obtained from the broodstock at the National Institute for Agricultural and Food Research and

Technology in Madrid, Spain. Frog housing and husbandry procedures were carried out as

described by Martini et al. (2010). X laevis mating and breeding were induced by intra-

G-51


-------
lymphatic injections of human chorionic gonadotropin (hCG). Adults were subsequently
transferred together to a spawning tank that was half filled with FETAX solution. Eggs were
collected the next day and dejellied in a 2% w/v L-cysteine solution, adjusted to pH 8.1.

Normally developing embryos were selected under a dissecting microscope in NF stage 10 (early
gastrula). Embryos were exposed in water to nominal solutions of PFOS (0, 0.5, 6, 12, 24, 48 and
96 mg/L) at 24ฑ1ฐC. Thirty total embryos for (ten per each of replicates) each treatment group
were distributed into 90-mm polystyrene Petri dishes, each containing 25 nil. of test solution.
The experiment was conducted in triplicate and control embryos were maintained in FETAX
solution. Embryo mortality and morphological abnormalities were recorded at 24. 4S. 72 and 96
hours of exposure with the dissecting microscope. Coagulation of the embryo or lack of
embryonic heartbeat was taken as criteria to define death and to estimate mortality. Embryonic
deformities were also identified using the morphological descriptions provided by Nieuwkoop
and Faber (1994), and liantle et al (1998) At the end of the assay (96 hours), surviving embryos
were anesthetized and were subsequently photographed to measure their total length. The study
author-reported lH->-hour sui \ i\ al I .()!ฆ(' was 96 mg/L PFOS. The study author-reported value
was used qualitati\ ely to deri\ e the draft acute water column criterion as it was a greater than (>)
high toxicity \ alue compared to the other acute toxicity data for PFOS. This LOEC of > 96 mg/L
was substantially higher (43 2 times) than the FAV of 6.011 mg/L and indicated that this genus
may be less sensi ti \ e to acute exposures of PFOS compared to the results reported in the
quantitative study by Palmer and Krueger (2001) with LCso values of 15.53, 18.04, and 14.60
mg/L. Therefore, the acute freshwater criterion of 3.0 mg/L will likely be protective of this
genus. Moreover, it is unlikely that these toxicity values would substantially change the FAV, as

G-52


-------
this genus would be ranked as the tenth most sensitive genus in the acute PFOS dataset and
another genus (Lampsilis) with a similar GMAV of 16.5 mg/L would be ranked fifth.

G. 2.3.2 Bufo gargarizans

Yang et al. (2014) evaluated the acute toxicity of PFOS (potassium salt, CAS # 2795-3-

3, 99% purity) to the Asiatic toad, Bufo gargarizans via 96-hour renewal measured exposures

(the authors note that the experiments followed ASTM standards and I .S. EPA procedures for

deriving water quality criteria). The tadpoles (0.048 g, 1.8 cm) were purchased from the Beijing

Olympic Park, which was considered an atypical source. The organisms were allowed to

acclimate for seven days before testing, and the lest was conducted at 22ฑ2ฐC with a light:dark

cycle of 12-hr:12-hr, with 10 toads per replicate and three replicates per concentration. Beakers

used for exposure were assumed glass, hul glass was not specified by study authors. PFOS was

dissolved in deionized water and carrier sol \ enl DM SO to obtain a 7 ing/mL stock solution, and

then diluted with dechlorinated lap w tiler lo yield nominal exposure concentrations of 20, 30, 45,

67.5, 101.25 and 151 SS mu I. PI-OS \Vtiler quality parameters reported were pH=7.0 ฑ 0.5,

dissolved oxygen 7 i) i) 5 mu I., lolal organic carbon=0.02 mg/L and hardness= 190.0 ฑ0.1

mg/L as CaCO; The supplemental data pro\ided for the study included a comparison of

measured PI OS concentrations before and after solution renewal in the low and high acute and

chronic test concentrations PI OS concentrations in the test water did not fluctuate by more than

15% during experiments The 96-hour LCso reported for the study of 48.21 mg/L PFOS was

deemed qualitative due to the atypical test organism source and unknown composition of test

beakers and statistical method used for calculating the LCso.

G-53


-------
G.3 Summary of Chronic PFOS Toxicity Studies Used Qualitatively in the
Freshwater Aquatic Life Criterion Derivation

G.3.1 Freshwater Invertebrates
G. 3.1.1 Worms (flat and annelids)

Yuan et al. (2014) conducted a 10-day renewal, unmeasured chronic test on PFOS

(potassium salt, >99% purity) with the planarian, Dugesia japonica (a non-North American
species). The test organisms were originally collected from a fountain in Quan HetouBoshan,
China, and cultivated in the laboratory for an unspecified lime period before use. The planarians
had a body length of 10-12 mm at test initiation. Dilution water was aerated lap water. No details
were provided regarding photoperiod or light intensity A primary stock solution was prepared
by dissolving the salt in DMSO. The control and exposed planarians received 0.005% DMSO
(v/v). Exposure vessels were beakers of unreported material type and dimensions and 50 mL fill
volume. The test employed three replicates of I <> planarians each in li\ e test concentrations: 0
(solvent control), 0.5, 1,5, 8 and 10mg/l. PI-OS The test temperature was reported as 20ฐC. No
other water quality parameters were reported as having been measured in test solutions. Survival
of solvent control animals was not reported The 10-day LOEC based on regeneration and
decreased appearance of auricles was 0.5 mg/L (NOEC and MATC, <0.5 mg/L). The chronic
value was acceptable for qualitath e use because of the short test duration.

G.3.1.2 Mollnsks

Hazelton el al. (2012) and Hazelton (2013) conducted a test of the effects of PFOS
(acid form, >98% purity) on glochidia of Lampsilis siliquoidea. The test exposed brooding
glochidia (in marsupia) for 36 d. Brooding female fatmucket were collected from Perche Creek,
MO. Dilution water was dechlorinated tap water. Mean hardness (47.5 ฑ 9.2 mg CaC03/L) and
alkalinity (34.8 ฑ4.1 mg CaC03/L) were measured by titration twice weekly (n=8) prior to water
changes. Replicates used for water quality measurements were changed daily to allow

G-54


-------
measurements from all four replicates every four days. For all treatments, water temperature
ranged from 14.6 to 16.1ฐC, dissolved oxygen ranged from 6.1 to 7.3 mg/L, and pH ranged from
7.6 to 8.5, but did not differ across treatments. Photoperiod and light intensity were not reported.
No details were provided regarding primary stock solution and test solution preparation.
Experiments were conducted in 3.8 L glass jars of unspecified fill volume. The test employed a
single replicate of four brooding females each in two measured icsl concentrations plus a
negative control. Nominal concentrations were 0 (negative control), 0 mil and 0.100 mg/L.
Mean measured concentrations were 0.00211 (negative control), 0.00452 and <> oe5 mg/L.
Analyses of test solutions were performed at the U.S. EPA National Exposure Research
Laboratory in Research Triangle Park, NC using HPLO MS. Two standard curves were used to
quantify PFOS water concentrations during the experiment: low range (<>.00005, 0.00025,
0.0005, 0.00075, 0.001, 0.0025, 0.005 mg I.) and high range (ซ) 001, 0.005, 0.010, 0.025, 0.050,
0.100, 0.150 mg I.) Two replicate samples were measured at each standard concentration.
Accuracy (reco\ cry) of PI OS in the low-range standard curve ranged from 89.5 to 123% (n=7)
and for the high-range standard air\ e accuracy w as 85.3 to 123% (n=7). Exposures were
maintained at a target temperature of I 5 (' and varied from 14.6 to 16.1ฐC during the exposure.
Mussels in the negative control was l)0%. The test resulted in an LOEC of 0.00452 mg/L based
on reduced \ iahilily of free glochidia (NOEC and MATC <0.00452 mg/L). The chronic value
was acceptable for <.|iialitati\ e use because there were only two test concentrations.

Olson (2017) conducted a series of chronic tests on the effects of PFOS (potassium salt,
CAS # 2795-39-3, 95% purity) with juvenile, pre-adult, and adult Lymnaea stagnalis as part of a
Ph.D. thesis at the Texas Tech University, Lubbock, TX. Chronic toxicity of PFOS to L.
stagnalis was observed under renewal (every 3.5 days) conditions over a 21-day exposure

G-55


-------
period. The test followed methodology established in Ducrot et al. (2010). Experimental animals
were divided into four pre-adult groups of ages equal to the length of the experiment (i.e., 0-3
weeks, 3-6 weeks, 6-9 weeks, 9-12 weeks). The test series also included at 21-day test with adult
L. stagnalis. Dilution water was reconstituted laboratory (13.38 g CaSC>4, 5.6 g MgSC>4, 0.25 g
KC1, and 2.95 g NaHCCte added to 50 L deionized water). Photoperiod and light intensity were
not reported. Stock solutions were prepared by dissolving between n 2< >00 and 0.2010 g of the
chemical powder in 1 L of lab water and placing the solutions in HDP I- hollies on a shaker
overnight. Stock solutions were then diluted as necessary to obtain the final exposure
concentrations. Exposure vessels were 1 L polyethylene beakers containing 1 I. of lest solution.
Number of replicates and number of snails in each replicale were not reported. Snails of various
age classes were exposed to each of six test concentrations (measured in low and high
treatments) plus a negative control. Nominal concentrations in the test were 0 (negative control),
1.5, 3, 6, 12.5, 25 and 5<) mu I.. l-.\posure concentrations were reportedly measured initially and
after three days lor \ erillcation. Inn concentrations were not reported. Analyses of test solutions
were performed using I MM.(' MS Standards were used as part of the analytical method, but
details were not reported The reporting limit was 0.010 mg/L. Experiments were conducted in
incubators set to 20VC, which did not vary more than 1ฐC during the course of the studies. No
other water quality parameters were reported as having been measured in test solutions. Negative
control survival w as )" <> The 21-day effect concentrations were as follows: 0-3 week old
juveniles - the NOEC for survival, feeding rate, mass change, length change, and carbohydrate
concentration was 50 mg/L (LOEC and MATC >50 mg/L); 3-6 week old juveniles - the MATC
for mass and length change was 35.35 mg/L (NOEC and LOEC, 25 and 50 mg/L respectively);
6-9 and 9-12 week old juveniles - the NOEC for survival, feeding rate, mass change, length

G-56


-------
change, and carbohydrate concentration was 50 mg/L (LOEC and MATC >50 mg/L); adult - the
MATC for survival was 4.243 mg/L (NOEC and LOEC, 3.0 and 6.0 mg/L, respectively).
Through this study design, all life stages were included in a 21-d (short-term) chronic toxicity
test, but not as a continuation so that the same organisms were exposed to PFOS across all life
stages as in a full-life cycle, or for a duration of continuous chronic exposure required to satisfy
EPA requirements for an acceptable partial or early-life stage lest Thus, the chronic values from
this study were acceptable for qualitative use. The qualitative values confirm the relative
insensitivity of this and other snail species (i.e., I'/iysclla hcicrosiropha poniilia: SVICV = 8.831)
to PFOS in the chronic criteria dataset.

G. 3.1.3 Zooplankton (rotifers andplanktonic crustaceans)

Zhang et al. (2014) reported the results of a chronic lilc-cycle test of PFOS (potassium

salt, CAS # 2795-39-3, >98% purity) with Hrac/iionns ca/ycifloras. The full life-cycle test used

renewal conditions for approximately four days />. calyciHorns used for the test were less than

two hours old attest initiation. All animals were parlhenogenetically-produced offspring of one

individual from a single resting euu collected from a natural lake in Houhai Park (Beijing,

China) The rolilers were cultured in an artificial inorganic medium at20ฐC (16-hr: 8-hr,

light dark. 3<)00 lux) for more than six months before toxicity testing to acclimate to the

experimental conditions. Culture medium was an artificial inorganic medium and all toxicity

tests were carried out in the same culture medium and under the same conditions as during

culture (i.e., pH, temperature, illumination). Solvent-free stock solutions of PFOS (1,000 mg/L)

were prepared by dissolving the solid in deionized water via sonication. After mixing, the

primary stock was proportionally diluted with dilution water to prepare the test concentrations.

Exposure vessels and size were not reported for the four-day reproductive assay, but were likely

in 6-well cell culture plates (assumed plastic) each containing at total of 10 mL of test solution.

G-57


-------
The test employed eight test concentrations plus a negative control. Each treatment consisted of
six replicates of 10 rotifers each in individual cells. Nominal concentrations were 0 (negative
control), 0.125, 0.25, 0.50, 1.0, 2.0, 4.0, 8.0, and 16.0 mg/L. PFOS concentrations were not
measured in the rotifer exposures, but rather, in a side experiment using HPLC/MS. The side
experiment showed that the concentration of PFOS measured every eight hours over a 24-hour
period in rotifer medium with green algae incurs minimal change in the concentration range 0.25
to 2.0 mg/L. Negative control survival is not provided for the life-cycle test The B. calyciflorus
4-day NOEC (intrinsic rate of population increase (R) and resting egg production) was 0.125
mg/L. The 4-day LOEC was 0.25 mg/L. The calculated MATC was 0.1768 mu I. The results
from this study were acceptable for qualitative use because of the atypical concentration-
response pattern.

3MCompany (2000) provides the results of a 2S-da\chronic toxicity test completed in
1984 with the cladoceran. Pa/>/mia magna, and Pl OS-k (periluorooctancesulfonate potassium
salt, CAS # 2795->Kv unknou n purity) The chronic test followed proposed standard practice
for life cycle tests with /kip/mm magna and OI-CD (1981). In-house culture of daphnids were
tested in unchloiinaled. caihon filtered well water under a 16:8 hr light:dark photoperiod, pH 7.6,
total hardness of 256 mg/L as CaCCb, dissolved oxygen > 70% and average temperature of 22
ฐC. A stock solution of PFOS was made in deionized water and diluted with test water to achieve
five nominal test concentrations (0.26, 1.0, 26., 7.0 and 18.0 mg/L). The chronic test was split
into two components based on the test endpoint. For mortality, five daphnids (12 ฑ 12 hour) were
added to 250 mL beakers with 200 mL of test solution and three replicates for each test
treatment. For reproductive endpoints, one daphnid was added to each beaker with seven
replicates for each test treatment. The specific number of replicates and organisms per replicates

G-58


-------
for the control was not well defined by the author, other than noting that 20 beakers in total were
used for the controls. However, the specific number of organisms and replicates is further
complicated by conflicting values for the number of replicates reported in the results section of
the publication which do not match the stated text. Over the 28-day exposure period, test
solutions were renewed and daphnids were fed three times per week. Authors reported that
reproduction was a more sensitive endpoint when compared to the control with an 18% reduction
in the cumulative number of live young per adult, a 55% reduction in a\ eraue number of live
young per brood and a 62% reduction in number of broods per adult at the highest test
concentration (18.0 mg/L). The author reported 28-day NOF.C and LOEC, based on all three
reproductive endpoints, was 7.0 and 18.0 mg/L PFOS, respectively, with a MATC of 11.22
mg/L. EPA was unable to independently calculate an F.C 10 \ alue bused on the level data provided
in the paper by the study authors. Additionally, the study authors noted that additional PFOS
studies are needed on this species to understand the effects at additional concentrations. Thus,
EPA assumed that the more recent tests resulted in better understanding of the effects of PFOS
for this species Therefore, the author-reported MATC of 11.22 mg/L PFOS was used
qualitati\ ely to deri\ e the draft chronic freshwater criterion for PFOS.

.Icon" ol sil. (2016) also conducted a chronic life-cycle 25-day renewal, unmeasured test
of PFOS (potassium salt, purity 99%) with Daphnia magna. The multi-generational exposure
study was conducted in two sets with five continuous exposure generations and three
discontinuous exposure generations. The initial generation (F0) was shared for continuous and
discontinuous exposure sets. In the continuous exposure set, the exposure conditions were not
changed during five generations (F0-F4). D. magna used for the test were originally obtained
from the Korea Institute of Toxicology and cultured in the laboratory according to EPA-821-R-

G-59


-------
02-012 (USEPA 2002; an acute toxicity testing protocol). Daphnids were less than 24 hours old
at test initiation. Dilution water was the same used for daphnid culture and was USEPA (2002)
hard reconstituted water with a hardness of 170 ฑ 5 mg/L as CaCCte, alkalinity of 110 ฑ 5 mg/L
as CaCCte, and pH 7.8. Photoperiod was 16-hr:8-hr (light:dark) at an unreported light intensity. A
primary stock solution of PFOS was said to be prepared in methanol (0.1% maximum level) at
10,000 mg/L, which seemed high. The primary stock was proportionally diluted with dilution
water to achieve the test concentrations. Exposure vessels were 30 ml. polypropylene beakers
containing 20 mL of test solution. The test employed 20 replicates of one daphnid each in five
nominal test concentrations plus a solvent control. Nominal concentrations were <> (solvent
control), 0.0001, 0.010, 0.100, 1.000, and 10.00 mg/L. The reported test temperature was 20ฐC.
No other water quality parameters were reported as having been measured in test solutions.
Solvent control survival was not reported. The most sensitive apical endpoint was reproduction
(number of offspring). Inn it was not observed in all generations. In the first generation, the
offspring number significantly decreased at concentrations above 0.100 mg/L. As the generation
number increased, the degree of the ad\ crsc effect tended to diminish, implying adaptation. The
impairment of offspring reproduction was greatest during re-exposure among all of the exposed
generations The D. magna 25-day NOEC (reproduction - F0 generation) was 0.010 mg/L. The
25-day LOI-C was<) lOOmg I. The calculated MATC was 0.03162 mg/L, and independently-
calculated ECio was ii ()(>41 mg/L. Independent statistical analyses were conducted using data
that were estimated (using Web plot digitizer) from the figures presented in the paper. However,
this independently-calculated toxicity was not considered reliable as this ECio was much lower
than the author reported NOEC and any other reproductive toxicity value for this species
(ranging between 0.001712 and 16.35 mg/L; see C.2). In addition to not being able to calculate

G-60


-------
an ECio for this study, the test concentrations were widely spaced (with each treatment group
increasing by one order of magnitude and ranging between 0.0001 and 10 mg/L), the data
presented in the paper was control normalized, and there was no consistent concentration-
response relationship. Therefore, the study author reported value was used qualitatively to derive
the draft chronic water column criterion. The toxicity value was within a factor of two of the
FCV of 0.008398 mg/L.

G. 3.1.4 Benthic Crustaceans
G. 3.1.5 Aquatic Insects

Olson (2017) conducted a chronic, approximately 42-day renewal lest of PI OS with first

instar of mosquito Aedes aegypti as part of a Ph.D. thesis at the Texas Tech University, Lubbock,

TX. The colony were donated by Texas A&M and had been maintained in the laboratory at

Texas Tech University since summer 2013 Dilution water was moderately hard reconstituted

water prepared according to USF.PA (2002; 3 g CaSCk 3 u \lgSO4, 0.2 g KC1, and 4.8 g

NaHCCte added to 5<) I. deionized water). Photoperiod was 12-14 hours light and 10-12 hours

dark. Light intensity was not reported Stock solutions were prepared by dissolving soluble

amounts of powdered chemical in dilution water. Diluted stock concentrations were equal to the

maximum test concentration The stock was mixed on a shaker table at 125 rpm for at least 18

hours before being added to exposure containers and proportionally diluted. Exposure vessels

were 50 mL 11 DPI- plastic beakers containing an unspecified amount of test solution. The test

employed 10 mosquito larvae each in six test concentrations plus a negative control. The number

of replicates was not reported. Nominal concentrations in the test were 0 (negative control),

0.050, 0.125, 0.250, 0.500, 1.000, and 2.000 mg/L. Experiments were conducted in incubators

set to 25ฐC and covered with plexiglass to limit evaporation. No other water quality parameters

were reported as having been measured in test solutions. Negative control survival was > 95%.

G-61


-------
The A aegypti 42-day NOEC (average time to emergence) was 0.05 mg/L. The 42-day LOEC
was 0.125 mg/L. The calculated MATC was 0.079 mg/L, and an ECio could not be
independently-calculated. The results from this study were considered acceptable for qualitative
use to derive the freshwater acute criteria. Additional data had been requested from the study
authors to help with independent verification of the toxicity values.

Van Gossum et al. (2009) conducted a chronic, approximately 4-month renewal test of
PFOS (tetraethylammonium salt, 98% purity) with damselfly, Enallagmu cyuiliigerum. The test
organisms were larvae that had reached the F2 instar stage. Dilution water was dcchlorinated tap
water. Photoperiod was 16-hr:8-hr light:dark. Light intensity was not reported. A primary stock
solution was prepared and proportionally diluted with dilution water to prepare the test
concentrations. Exposure vessels were plastic containers (15 cm \ I n cm \ 11 cm) with a 2 cm
depth of test solution. The test employed ll>-2<) lai \ ae each in two test concentrations plus a
negative control. Nominal concentrations were 0 (negati\ e control), 0.01, 0.1, 1, and 10 mg/L.
All larvae were housed (and presumably tested) in temperature-controlled rooms at 21 ฑ 1.3ฐC.
No other water quality parameters were reported as having been measured in test solutions.
Negati\ e control mortality was said to be much lower than the 100% mortality that occurred at 1
and I" mu I.. but was not reported. The 4-month NOEC (behavioral - including general activity,
swimming performance, foraging success) was 0.010 mg/L. The 4-month LOEC was 0.100
mg/L. The calculated M ATC was 0.03163 mg/L. The chronic value was acceptable for
qualitative use because non-apical endpoints.

Stefani et al. (2014) conducted a chronic (10 generation) test of PFOS (form and purity
not reported) with midge, Chironomus riparius. The 10 generations (each approximately 20 to
28 days) were tested under static conditions. The 10 generations (each approximately 20 to 28

G-62


-------
days) were tested under static conditions. The test followed OECD (2004a); OECD (2004b);
OECD (2010), and a specific protocol for multigenerational assays using C. riparius developed
and published by Nowak et al. (2008); Nowak et al. (2006); Nowak et al. (2007a); Nowak et al.
(2007b); Nowak et al. (2009) and Vogt et al. (2007a); Vogt et al. (2010); Vogt et al. (2007b).
The specific protocol for multigenerational assays was designed to highlight neutral evolutionary
responses caused by exposure to contaminants. A native population collected in the Lambro
River (Milan, Lombardy, Italy) was used as a starting population for the test (riparius used to
initiate the test were LI (first instar) larvae. Dilution water was reconstituted u alcr according to
U.S. EPA (2000b) - hardness not specified, pH 7.8-8.2. Photoperiod was 16-hr S-lir. Iight:dark
with an intensity 500 - 1,000 lux. Treatments with two replicates each (i.e., two cages with five
vessels each per treatment and 60 larvae per \ essel. or approximately 300 larvae per treatment)
were tested, by spiking 15 L of test water with I 5<) ill. of methanolic solution at 1 g/L of PFOS
to achieve a nominal concentration of 0.01 mg/L. Exposure \ essels were glass tanks (19 cm x 19
cm x 18 cm) containing and un sped lied amount of test solution and 1 cm thick layer of
formulated sediment (75ฐ.. of the \olume constituted by 250-300 |im grain size aquarium quartz
sand and 25".. of the \ olume In (\i-25<) mil grain size natural sediment collected in an
unimpacteil ii\er, sieved anil sterilized). The measured exposure concentration diminished
significantly o\ er the course of the exposure (mean concentration at beginning of experiment:
0.0089ฑ0.21 mg/l. anil concentration at end of experiment: 0.0016ฑ0.2 mg/L), meaning later
generations were exposed to less PFOS than earlier generations. The reported time-weighted
measured concentration was 0.0035 mg/L. Mass-balance evaluations at the end of a generation
showed that most of the PFOS was detected in the sediment (36% of the added amount). Since
45% of the PFOS added to the test vessel was lost (the authors hypothesize because of air

G-63


-------
stripping), only roughly 17% of the PFOS added to the test vessel ended up in the water. Test
temperature was controlled at 20 ฑ 1ฐC, and dissolved oxygen remained above 66% saturation.
No other water quality parameters were reported as having been measured in the test solutions.
Controls were considered acceptable because they fulfilled the validity criteria for mortality
according to OECD guideline 218 (OECD 2004a). In the control group, most vessels in all
generations reached the emergence of at least 70% individuals The NOI -C and LOEC based on
reproduction and emergence were 0.0035 and > 0.0035 mg/L (as time-weigh led average) as there
were no effects on emergence, reproduction, or sex ratio at this concentration The results from
this study are acceptable for qualitative use because of the use of only a single test concentration,
the lack of observed effects in the one exposure concentration that was considered a greater than
low value, and lack of details pertaining to the characteristics of the sediment used in the
exposure, particularly considered the difference between measured concentrations over the
exposure duration This particular study provided few details pertaining to the effects of chronic
PFOS exposure on midue Therefore, it was determined that the ECio of 0.05896 mg/L from
MacDonald et al (2<)i)4) that was used quantitatively in the chronic criterion derivation was
more robust than the toxicity \ allies reported in Stefani et al. (2014) and the chronic freshwater
criterion of <) <)<)X4 mg/L would likely be protective of this genus.

In a companion paper to Stefani et al. (2014), Marziali et al. (2019) similarly conducted
a chronic (10 generation) test of PFOS (form and purity not reported) with midge, Chironomus
riparius. The test was done under static conditions for 10 generations, each approximately 36
days (or 1/10 of this year-long, 10 generation test). The test followed OECD 218 and 233
(OECD 2004a, 2010), with slight variations. C. riparius used for testing were from in-house
cultures originating from a native population collected in the Lambro River (Milan, Lombardy,

G-64


-------
Italy). C. riparius used to initiate the test were first instar larvae. Dilution water was
reconstituted water according to U.S. EPA/600/R-711 99/064 (U.S. EPA 2000b) - hardness not
specified, pH 7.8-8.2. Photoperiod was 16-hr:8-hr (light:dark). Light intensity was not reported.
A single treatment of 0.01 mg/L (nominal) and solvent control with 10 replicates of 60 larvae
each were tested. PFOS was dissolved in pure methanol (> 99%) in order to achieve stock
solutions at 1 g/L of PFOS. Each stock solution was then diluted in reconstituted water in order
to achieve the nominal concentration of 0.01 mg/L. Exposure vessels were glass tanks (19 cm x
19 cm x 18 cm) containing 1 L of test solution and 1 cm of formulated sediment (75% of the
volume aquarium quartz sand and 25% of sterilized natural sediment). The measured exposure
concentration diminished significantly over the course ol" the exposure, meaning later generations
were exposed to less PFOS than earlier generations. The reported time-weighted measured
concentration was 0.004 mg/L. PFOS was observed to sorb sediment. To check the potential
presence of PFAS in the fish food used in the test, PFOS concentrations were determined in the
larvae of the control after 3<~> days exposure Measured concentrations of PFOS in larvae tissue
were always below detection limits (2 nu u wet weight). Water temperature, dissolved oxygen
and pi I were measured e\ cry three to li\ e days in two to three replicates per treatment. Test
temperature was controlled at 20.1 n 7 (', and dissolved oxygen remained equal to or above
66% saturation pi I stayed w ithin the range of 7.8-8.2. Each generation test was considered valid
if emergence in the control was > 0% in at least six replicates (i.e., vessels) of the 10 included.
Emergence in the control groups by generation was as follows: 88 (primary emphasis for criteria
development), 71, 53, 61.6, 78.6, 91.9, 62, 53.5, 79.1, 75.5. Thus, generations 1, 2, 5, 6, 9, and
10 met control survival acceptability. The LOEC based on F1 developmental time and F1 adult
weight was < 0.004 mg/L (time-weighted average). The were no effects on F1 exuvia length at

G-65


-------
this concentration. The results from this study were acceptable for qualitative use because of the
use of only a single test concentration, the lack of consistent observed effects in both the control
and the treatment groups across the generations, the LOEC was considered a less than low value
that is not very robust since there was only one treatment concentration, and lack of details
pertaining to the characteristics of the sediment used in the exposure, particularly considered the
difference between measured concentrations over the exposure duration. This particular study
provided few details pertaining to the effects of chronic PFOS exposure 011 midge. Therefore, it
was determined that the EC10 of 0.05896 mg/Lfrom MacDonald et al. (2004) thai was used
quantitatively in the chronic criterion derivation was more robust than the toxicity \ alues
reported in Marziali et al. (2019) and the chronic freshwater criterion of 0.0084 mg/L would
likely be protective of this genus.

G. 3.1.6 Microcosm Data Summaries

Mixed species exposures are typically not used quanlilati\ ely in EPA aquatic life criteria

documents, but two well-designed microcosm experiments provided results that are supportive of

the preliminary SMCYs lor \ loma iiiacrocopa and lkiphnia magna representative of other

cladocerans and cope pods

Sanderson et al. (2002) e\ aluated the chronic effects of PFOS on the invertebrate

community in a 35-day indoor microcosm study. Indoor 30 L transparent polyvinyl chloride

aquariums were ll 11 ed with sediment and water from natural ponds at the University of Roskilde,

Copenhagen, Denmark. Microcosms were allowed to stabilize for four weeks before PFOS-K

(potassium salt, CAS # 2795-39-3, purity unreported) additions were made. Each measured

treatment (nominal: 0, 1, 10 or 30 mg/L) had five replicate aquariums. Measured concentrations

remained relatively stable over the exposure period; the 1, 10 and 30 mg/L nominal

concentrations were 1.33, 12.3 and 33.9 mg/L at test initiation, but decreased slightly to 1.08,

G-66


-------
11.7 and 29.8 mg/L at day 35. PFOS was added only once to each aquarium as a sub-surface
injection. Test endpoints focused on the zooplankton community (overall abundance and number
of taxa) and abundance of individual species (Cyclops diaptomus, Daphnia magna, Cyclops
canthocamptus staphylinus, and total Rotifer sp). Aquariums were under constant aeration and
maintained an oxygen concentration of 6 mg/L over the entire study. Microcosms had a 12-hour
light period (2,852 lumens), and were maintained at 18ฐC. The pi I a\ eraged 8.3 over the
exposure period (range 8.28-8.37). Zooplankton community abundance and structure was
affected at 10 and 30 mg/L. Effects on rotifers were variable (i.e., some species decreased and
others increased). The authors noted that due to variability between replicates, it was not possible
to determine with statistical confidence whether or not treatment related effects were present at 1
mg/L. The most sensitive species was the copepod. C. diaptomns. w hich was eliminated in the
10 and 30 mg/L by day 14. The MATC for the study. 3 I (->2 mu I., based on zooplankton
abundance was acceptable lor qualitative use because the \ alue was from a non-definitive
microcosm test.

Boudrenu ol ;il. (2003h) examined the ecoloxicological impact associated with PFOS
exposure across multiple le\ els of biological organization using 35- to 42-day exposures in
outdoor microcosms. The studies were conducted as part of a Master's thesis at the University of
Guelph, Ontario. Canada (Boudreau et al. 2002) and later published in the open literature. Four
nominal concentrations of PFOS (potassium salt, purity of 86%) were tested: 0.3, 3, 10, or 30
mg/L. The persistence of PFOS in the outdoor microcosms was evaluated over 285 days. The
measured concentrations, as calculated by the time-weighted average (TWA) over 285 days,
changed by <6.0% from nominal concentrations (the TWA for the nominal 3 mg/L exposure was
2.8 mg/L, for the nominal 10 mg/L exposure, 9.8 mg/L, and for 30 mg/L, 30.1 mg/L). For this

G-67


-------
reason, nominal values were used in all analyses. Fifteen 12,000 L microcosms were used for
this analysis, with three replicates for each treatment and controls. Water for the microcosms was
provided from a spring-fed irrigation pond and was circulated among microcosms for 10 days
prior to study initiation at a rate equivalent to 11,000 L per day. This circulation period enabled
establishment and equilibrium of the indigenous animals and plants in the microcosms. The
microcosms were open to allow for aerial colonization of flying insects and the lining was
allowed to accumulate periphyton and algae. This design feature increased \ ariability in the
organisms and invertebrate and primary producer community. One day prior lo treatment,
circulation from the irrigation pond was terminated producing isolated systems and water and
zooplankton samples were collected as a pre-treatment reference. Water quality varied little
across all treatments: temperature varied from I 5 l)-2<~> 5ฐC, DO from 7 2-8.8 mg/L, pH from
8.3-8.6, and hardness from 294-300 mg/L as CaCO; /ooplanklon populations were significantly
affected at 10 and 3d mu I.. and a community-lc\ el no ohscr\ ed effect concentration
(NOECcommnnit-) of 3 n mu I. was determined over 35 days. The most sensitive taxonomic
groups, Cladocera and Cope|->oda. were \ irtually eliminated at 30 mg/L by seven days. The
zooplankton communities from both " 3 and 3 mg/L treatments showed little change from the
control o\ er time Based on these results, the MATC for zooplankton community abundance was
5.478 mg/l. and was acceptable for qualitative use because the value was from a non-definitive
microcosm test.

G.3.2 Freshwater Fish
G. 3.2.1 Anguilla anguilla

Roland et al. (2014) evaluated the chronic effects of the potassium salt of

perfluorooctane sulfonate (PFOS, CAS # 2795-39-3, 98% purity) to juveniles of the European

eel, Anguilla anguilla. A set of 162 juvenile female eels (138.3 g) were purchased from a Dutch

G-68


-------
eel farm (Zon-Aquafarming, Helmond, The Netherlands) and were randomly distributed at a
fixed number into 18 tanks filled with 70 L tap water. Fish were acclimated to laboratory
conditions in aerated water at 20ฑ2ฐC under a 12-hr: 12 hr light:dark photoperiod for two months
before the experiment. Eels were exposed to nominal PFOS concentrations of 0, 0.001 and 0.010
mg/L (mean measured concentrations of 0.00001, 0.00081, 0.011 niul.. respectively) at 20ฐC
and 12-hr: 12-hr light:dark photoperiod during 28 days while the control lish were kept in clean
water. Each treatment included six replicate tanks with nine fish per tank One third of the water
was renewed every 72 hours. Water samples were taken at days 0, 7, 14 and 2S after the
beginning of the exposure for PFOS measurements (absolute PFOS recoveries in the
concentration range of 0.00001 and 0.0001 mg/L were in the range of 50-90%). Animals were
fed daily during exposure and no mortality was recorded during the experiment. The 28-day
survival NOEC and LOEC were 0.011 and <> <> I I mu I. PI-OS. respectively, but were
considered qualitati\ e data since the test was not an acceptable early life-stage (ELS) test (used
juveniles not embryos) and . I. anifiii/la is not aNorth American species. In addition, the
proteomic change \OI-C and I.()!ฆ(' were <> rinnsi and 0.00081 mg/L PFOS, respectively, also
qualitati\e (non-apical endpoint)

G.3. 2.2 (h il orhynchus m \ k iss

Benninghoff et al. (2011) evaluated the chronic effects of perfluorooctane sulfonic acid

potassium salt (PFOS-k. C AS 4 2795-39-3, purchased from Sigma Aldrich in St. Louis, MO) on

rainbow trout (Oncorhynchus mykiss) juveniles in a 15-day static, unmeasured study. Mount

Shasta strain rainbow trout were hatched and reared in the Sinnhuber Aquatic Research

Laboratory at Oregon State University in Corvallis, OR. Fish were maintained at 12ฐC and a 12-

hr: 12-hr light:dark cycle in a 375 L flow-through tank filled with carbon filtered water. Two

weeks before testing, fish were fed a semipurified casein-based diet with menhaden oil at a rate

G-69


-------
of 2% body weight. A stock solution was prepared by dissolving PFOS in dimethyl sulfoxide
(DMSO) that was then added to oil in the fish diet. Eleven-month-old fish weighing
approximately 70 g were placed in treatment groups, six fish per group, and were fed either 0.1,
1 or 5 mg/kg body weight/day PFOS laced food five times per week for 15 days: correlating to a
diet concentration of 5, 50 or 250 ppm PFOS per day. Four replicates were included for each
dietary treatment, along with a negative control group, and a \ chick- control group (treated
dietarily with 0.5 ppm DMSO). A positive control group (treated dietarily with 5 ppm estradiol)
was also included with an additional twelve fish. Fish were sacrificed and weighed on day 15.
The authors reported a NOEC of 250 ppm PFOS lesi diet for growth (weight). The lack of
description of the dilution water and the test methodology (dietary exposure) makes the study
acceptable for qualitative use only.

Benninghoff et al. (2012) also evaluated the chronic effects of perfluorooctane sulfonic
acid potassium salt (PI'OS-K. purchased from I'luka Chemical Corp in St. Louis, MO) on
rainbow trout (Oncorhynchiis myktss) in an K-monlh unmeasured study. Mount Shasta strain
rainbow trout were hatched and reared in the Sinnluiber Aquatic Research Laboratory at Oregon
State I 'diversity in Cor\ allis. OR I'ish were raised in a 375 L tank filled with carbon filtered tap
water and maintained at 12VC and a 12-hr: 12-hr light:dark cycle. Fry (15 week post spawn) were
exposed to a cancer causing agent for 30 minutes, then fed a semi purified casein-based diet for
one month. Fish ^crc led experimental diets containing 100 ppm PFOS (approximately 25
mg/kg body weight/day) five days per week for a period of six months. Fish were sacrificed at
test termination (12.5 months post spawn) and examined for tumor presence. Both survival and
tumor incidence LOEC were both observed in the 100 ppm PFOS test diet (2.5 mg PFOS/kg

G-70


-------
body weight/day). The lack of description of the dilution water, mixture based exposure, and the
test methodology (dietary exposure) makes the study acceptable for qualitative use only.

G. 3.2.3 Cyprinus carpio

Hagenaars et al. (2008) evaluated the effects of PFOS (98% purity) to gene expression,

condition factors and energy storage endpoints exhibited by Cyprinus carpio exposed for 14 days

under renewal unmeasured conditions. Juvenile carp (3.72 g and 5 IS cm) were acclimatized for

three weeks in plastic 20 L aquaria prior to treatment. The water used din ing acclimatization and

treatment was filtered and aerated (source not provided). Every 48 hours, the water was totally

renewed with nominal PFOS concentrations. The fish were exposed to a 14-hr 1 <>-hr light:dark

cycle and fed 2% of body weight. After acclimatization. fish were exposed to PFOS at nominal

concentrations of 0.1, 0.5 and 1 mg/L during two weeks while the control carp were kept in clean

water during exposure. For each exposure concentration as well as for the controls, three aquaria

were used resulting in three full biological replicates for each exposure condition. Twelve carp

were housed in each aquarium Relative condition factor (RCF) was determined in all fish at day

0, 7 and 14. After 14 days, fish were sacrificed by decapitation. The liver was immediately

remo\ ed. weighed for hepatosomalic index (HSI) calculation, frozen in liquid nitrogen, and

stored at -So C. From each aquarium, six livers were randomly selected for use in the microarray

analysis, lour to determine the energy reserves and two to measure PFOS concentrations. The

two most sensitive endpoints were reduced hepatosomatic index (HSI) and liver glycogen, with

14-day chronic values (MATCs) of 0.2236 and 0.7071 mg/L PFOS, respectively. This study was

classified as qualitative due to non-apical endpoints, nominal water concentrations and lack of

exposure detail.

G-71


-------
G. 3.2.4 Danio rerio

A sub-chronic static unmeasured test was utilized by Ulhaq et al. (2013) to determine the

toxicity of PFOS to Danio rerio. PFOS stock solutions were freshly prepared in reconstituted

water in concentrations below the limit for water solubility. Adult zebrafish (AB strain) were

held in charcoal-filtered tap water. Breeding groups including three males and two females were

placed in 10 L glass aquaria equipped with spawning nets separating the parental fish from the

eggs. Half an hour after onset of lights the eggs were collected, rinsed lor removal of debris, and

then only normally developed fertilized eggs at least in the four-cell stage were selected using a

stereomicroscope. The zebrafish eggs, within 15 minutes after collection, were exposed to a

series of concentrations of the test substance dissolved in reconstituted water (exposure

medium). Fertilized eggs (4-cell stage) ^ere randomly distributed individually into flat bottom,

48-well polystyrene plates along with 75<) ill. of the exposure medium PFOS was tested at six

consecutive concentrations differing by a factor of 3 3 based on logarithmic scale fitting. For

each test, four 48-well plates were used, with a total of 24 embryos per concentration as well as

24 in the water control group luich treatment group was equally distributed to each of the four

well plates (i e . six embryos concentration/plate, for a total of 168 embryos). The plates were

covered with parafilm and the embryos were exposed to the chemical until 144 hours post

fertilization (hpf) l-xposure conditions throughout the study were kept at pH 7.2-7.6, a water

temperature of 2o I (' and a light cycle of 14 hours. Observations of mortality and sublethal

endpoints were made after 24, 48, 120 and 144 hpf using a stereomicroscope. Sublethal

endpoints such as presence of edema, malformations, not-hatched eggs, lack of circulation and

reduced pigmentation were also observed. Heart rate was recorded at 48 hpf and hatching time

was determined using time-lapse photography. The 144-hour LCso was >10 mg/L PFOS and the

G-72


-------
ECso (lethal and sublethal effects) was 1.5 mg/L PFOS (both were qualitative data because of test
duration and unmeasured static test).

The effects of PFOS on pancreatic organogenesis in Danio rerio 3 hpf embryos (wild
type) were evaluated by Sant et al. (2017). PFOS stock solutions (160-640 mM) for embryo
exposures were prepared by dissolving PFOS into DMSO and stored at room temperature in
glass bottles inside of light-prohibitive containers until use. Transgenic zebrafish of the
Tg(ins:GFP) and Tg(ptfla:GFP) strains were each obtained as a heterozygous population from
the University of Massachusetts Medical School and hied in house to homozygosity. The Tg(ins-
GFP) strain expresses green fluorescence in the insulin-producing beta cells, allowing for
visualization of pancreatic islets. The Tg(ptfla:GFP) strain expresses green fluorescence in the
exocrine pancreas tissues, and also in the retina and parts of the brain Adult fish were housed in
an Aquaneering zebrafish system maintained al 2K 5 (' and a 14-hr: 10-hr light:dark cycle.
Breeding populations were housed in tanks containing roughly 15 males and 30 females.

Embryos were collected from breeding tanks <)-1 hour post fertilization (hpf), washed, and
housed with no more than 25 other embryos in glass 100 mm petri dishes containing 0.3x
Danieau's medium (pi I 7 (•>) throughout the experiments. At 3 hpf, embryos staged at the
midblastula transition were exposed to PFOS solutions with a total of 0.01% DMSO v/v in a
total of 20 ml of <> 3x Danieau's medium. Final concentrations of PFOS were 0 (DMSO control),
16, 32, or 64 |iM (or N.<)<)2. 16.00, and 32.01 mg/L), and were refreshed daily to mimic
subchronic developmental exposures. All embryos were manually dechorionated using
watchmaker's forceps at 24 hpf and debris removed from dishes prior to refreshing exposures.
Experiments were replicated 3-4 times on groups of 8-12 embryos per concentration. The authors

G-73


-------
reported a 7-day chronic value (islet morphological anomalies) of 11.31 mg/L PFOS. The
chronic value was classified as qualitative due to duration and an unmeasured chronic exposure.

Du et al. (2009) investigated the effect of PFOS (> 99% purity) on the survival, growth
and hepatotoxicity of Danio rerio female fry exposed via renewal unmeasured conditions for 70
days. The PFOS stock solution (50,000 mg/L) was prepared in HPLC-grade DMSO and stored at
4ฐC. Adult zebrafish (AB strain) were maintained in charcoal-filtered. recirculating, aerated tap
water with a 12-hr:12-hr light:dark cycle and a temperature of 27ฑ0.5 (' fertilized eggs were
collected and the fry were maintained until 14 days post-fertilization (dpf) for subsequent
experiments. Zebrafish fry were randomly distributed into 2<~> I. glass tanks for control and
exposure groups. There were three replicates for each group, with each group containing about
50 fry in each tank. The fry were exposed in a renewal system and the water was half-renewed
every other day. Both the control and exposure groups received DMSO (0.002%, v/v), with
nominal PFOS concentrations of <) 01, 0.05 and 0.25 mg I. The exposure regime included a
PFOS exposure period (7<) d) and reco\ erv in clean water (30 d). Fish were sampled after 40 and
70 days of exposure lo determine lengths and weights, histological examination of the testis and
liver. Ii\ er V'l'Ci gene expression and w hole body T3 measurement. The gonad weights of the
females were recorded after 7<) days of exposure and after 30 days of recovery. After 70 days of
exposure, the remaining fish were transferred to a 20 L glass tank and reared in dechlorinated
municipal tap water to allow for 30 days of recovery, while a subset of the exposed female fish
was placed into clean water and paired with unexposed male fish. The hatching rates,
malformation and survival in the F1 embryo-larvae were assessed. The 70-day MATC for
increased malformation and decreased survival of F1 fish was reported as 0.0224 mg/L PFOS.
This MATC was associated with a mean percent of malformation of 0, 37.5, and 100% in the

G-74


-------
0.01, 0.05, and 0.25 mg/L treatment groups, respectively. However, an independently-calculated
ECio could not be determined as the treatment level data needed for this analysis appear to have
been lost (personal communication with Bingsheng Zhou, corresponding study author). Instead,
only the independently-calculated ECio for male growth as weight of 0.001990 mg/L could be
determined. This ECio was substantially lower than the author reported MATC of 0.0224 mg/L
and the MATC of 0.0158 mg/L observed for the same endpoinl in Wang et al. (2011). Therefore,
the independently-calculated ECio of 0.001990 mg/L was not used in the deri\ ation of the
freshwater chronic criterion since the curve fitting for this study was questionable given that the
independently-calculated ECio for male weight does not appear to be reasonable, was
substantially lower than the MATC from this and other studies, and was not consistent with the
observed toxicity of PFOS for this species Additionally, the endpoint of male growth was
difficult to tie to a population effect compared to the survival endpoint that was used to derive
the freshwater chronic criterion I lo\\e\ er. the author-reported MATC of 0.0224 mg/L for
increased F1 malformation and decreased sur\i\al was similar to the independently-calculated
ECio of 0.01650 mu I. lor I ' I sur\ i\ al from \Yanu et al (2011), which was used quantitatively in
the freshwater chronic criterion

Cui el ;il. (2017) in\ estigated the toxic effects of PFOS (> 96% purity) to Danio rerio in
a near full lile-cvvie (unmeasured) static renewal test. The PFOS stock solution was prepared by
dissolving PFOS in DM SO Wild-type zebrafish (AB strain) were raised under standard
laboratory condition of 28ฐC (water temperature) with a 10-hr: 14-hr light:dark photoperiod in a
recirculating system according to standard zebrafish breeding protocol. Water supplied to the
system was filtered by reverse osmosis (pH 7.0-7.5), and Instant Ocean salt was added to
maintain the conductivity to 450-1,000 mS/cm. The adult fish were fed twice daily with live

G-75


-------
Artemia and dry flake diet. Zebrafish embryos were obtained from adults in tanks with a sex
ratio of 1:1, and spawning was induced in the morning when the light was turned on. Embryos
were collected within one hour after spawning and rinsed in embryo medium. The fertilized
embryos were staged using a stereomicroscope according to the standard method. High-quality
embryos at 8-hour post fertilization (hpf) were divided into four groups: vehicle control (0.01%
DMSO, v/v), and PFOS at 0.02, 0.1, and 0.5 mM (or 0, 0.01, 05. and 25 mg/L given the
molecular weight of the form of PFOS used in this study, CAS # 1703-23-1. of 500.13 g/mol).
Embryos were first exposed to PFOS in a petri dish (100/group) for five days w ilhout media
change, and all embryos hatched and survived in this stage After five days, fish were transferred
into 2 L tanks until 30 dpf. After 30 dpf, fish were raised in 10 L tanks (30/tank) until the end of
experiment. Throughout the whole exposure period. 5<~>ฐo water was renewed with freshly
prepared solutions every five days. Each tank was checked lor fish morbidity on a daily basis
and water quality was monitored on a weekly basis. Feeding was initiated at 5 dpf. Between five
and 14 dpf, fish were led three times daily with standard larval diet, and after 14 dpf, they were
fed twice daily with freshly hatched li\ e . \ricmia Equal amounts of feed were given among
different groups each day All experiments performed in this study were repeated three times
with embryos derived from three different parental stocks. At the end of chronic exposure (180
dpf), fish from each group (n 30) were checked for their sex, body weight, and length
(measured from snout to the fork point of caudal fin). Condition factor (K) was tabulated to
determine overall fitness. Breeding trials were also carried out to produce F1 offspring (F0
females were paired with F0 males from the same treatment group). Malformation and survival
rate of both generations were evaluated and the study authors indicate that F1 offspring derived
from the parental fish exposed to 0.25 mg/L were observed to have severe deformities (including

G-76


-------
uninflated swim bladders, bent spine, pericardial edema, yolk sac edema, and necrosis) and low
survival rates. A MATC of 0.1118 mg/L was calculated from the authors reported values for
effects on altered sex ratio (female dominance) and low F1 offspring survival. However, an
independently-calculated toxicity value could not be calculated with the data provided in the
paper. The author-reported MATC was used qualitatively since EPA was unable to
independently verify the reported toxicity value with the data r>ro\ iclccl in the paper, since only
the growth data were provided, and because there was insufficient information. including the
treatment level data, on the reproductive effects observed in the study. I'ni lIkt. the author-
reported MATC of 0.1118 mg/L for decreased F1 offspring survival was higher than the
independently-calculated ECio of 0.0165 mg/L for F1 survival from Wang et al. (2011) (which
was used quantitatively in the freshwater chronic criterion) and the MATC of 0.0224 mg/L for
increased F1 malformation and decreased sur\ i\ al lYom l)u el al (2009) (which was used
qualitatively as supporting information for this species), indicating that the author-reported
toxicity value for this study may not he representative of the effects of PFOS on zebrafish.

Effects of PI-OS on disruption of the hypothalamus-pituitary-thyroid axis in Danio rerio,
was in\ cstiualed In Shi el ;il. (2009) The authors subjected blastula stage zebrafish embryos to
the potassium salt of PFOS ( l>9% purity) for 15 dpf under renewal unmeasured conditions. The
stock solution was prepared hy dissolving the crystals in HPLC-grade DMSO. Adult zebrafish
(AB strain) embryos were collected at 2 hpf and embryos that had developed normally and
reached the blastula stage were selected for subsequent experiments. Approximately 400 normal
embryos were randomly distributed into glass beakers containing 500 mL of PFOS exposure
solution (0, 0.10, 0.20, and 0.40 mg/L) with three replicates for each exposure concentration.
During the experimental period, 50% of the exposure solution was renewed daily. The control

G-77


-------
and exposure embryos received 0.003% (V/V) DMSO. The larvae were randomly sampled until
15 dpf and immediately frozen in liquid nitrogen and stored at -80ฐC for further gene expression
and thyroid hormone assays. The body length of the larvae was also measured. The 15-day
chronic value, MATC, was 0.2828 mg/L PFOS and is considered qualitative as this study was a
rapid early-life stage test focused on the developmental toxicity of PFOS starting with embryos
that had developed normally to the blastula stage and lasting 1h rough 15 dpf with limited
procedural detail provided in the paper. No effects on survival were ohsei \ cd (with a NOEC >
0.40 mg/L). The study author-reported NOEC, LOEC, and MATC for growth as both total body
length and weight were 0.20, 0.40, and 0.2828 mg I., respectively. This 15-day growth MATC of
0.2828 mg/L was roughly one order of magnitude higher than the FCV of 0.008398 mg/L and
suggested that this genus may be less sensili\ e to chronic exposures of PFOS than the
quantitative study by Wang et al. (2011) ith an l-Ci- of <) <)|65 mg/L. However, the toxicity
value from Shi et al (2<)<)l>) was lor a much shorter exposure duration (15 days in a rapid early-
life stage test compared to I 5<) days in a full life-cycle test) and a less sensitive endpoint (growth
compared to reproduction) Therefore, the chronic freshwater criterion of 0.0084 mg/L would
likely be protective of this genus

/ kmio rcrio embryos were also in\ estigated by Keiter et al. (2012) in a long-term flow-
through measured study with PFOS (potassium salt, CAS # 2795-39-3, >98% purity). PFOS was
delivered to the test \ essels without use of a carrier solvent. A first stock solution of PFOS (300
mg/L) was prepared by dissolving 1.5 g PFOS in 5 L of deionized water with overnight magnetic
stirring. The solution after the first dilution step, hereafter named second stock solution (0.016,
2.6 and 7.8 mg/L), was freshly prepared four times a week by diluting the first stock solution
(300 mg/L) with deionized water. Nominal concentrations of PFOS in the test vessels were

G-78


-------
0.0006, 0.1, 0.3 mg/L. The test was initiated with fertilized zebrafish eggs obtained from non-
exposed adults reared in the laboratory of Aquatic Ecology and Toxicology, Heidelberg
University, Germany. Throughout the study, the fish were maintained in a light-isolated room
with an artificial 14-hr: 10-hr light:dark period. Adult fish were fed twice daily with freshly
hatched Artemia nauplii complemented with TetraMin flake food. Larvae were initially fed twice
daily with liquid starter food followed by Sera micron powder food and freshly hatched Artemia
nauplii. Tap water and deionized water were mixed until conductivity ((ฆ><)< 1-750 |iS), hardness
(276ฑ17.8 mg/L as CaCCte) and pH (8.0-8.2) were stably balanced. The water mix was supplied
from an aerated reservoir and used to culture all embryos and fish. The final test water was
routinely characterized for pH (8.25-8.75) and total hardness (167-356 mg/L). Temperature
(26.0ฑ1.0ฐC) and dissolved oxygen (6.45-1n ^7 niu'I.) was checked weekly.

During the course of the study, fish were conliiuiously exposed to PFOS with each
treatment group replicated twice holding a starling number of 80 fish per replicate (160
individuals per treatment) At 2-4 hours post fertilization (hpf), eggs were transferred to glass
dishes and exposed to the different treatments at 26ฑ1ฐC under semi-static conditions (complete
renewal of solutions after 24 hours) until 48 hpf, when they were transferred to respective test
vessel Whole-glass tanks, adjusted lor a 10 L working volume were utilized as test vessels. A
flow-through system w ith a three-fold water exchange per day was applied throughout the study
in order to provide adequate supply of fresh test solution. External aeration by pressurized air
was installed for each test vessel. Test solutions were daily refilled into light-isolated 10 L glass
bottles located above the test vessels. Each test solution was constantly held in motion by
magnetic stirring. Peristaltic pumps were used for a continuous delivery of test solution (50
mL/hour) from each glass bottle to paired test vessels serving as replicates A and B for each

G-79


-------
treatment group. For each test vessel, a water flow rate of 1.25 L/hour was adjusted by means of
rotameters.

Each replicate of the F1 and F2 generations was sub-sampled at 30 and 90 dpf. At 30 dpf,
the number of fish in each replicate test vessel was reduced by 35 individuals for measurements
of length and weight. At 90 dpf, each replicate was further reduced by 25 individuals for
measurement of length, weight and vitellogenin (Vtg), as well as lor histological evaluation of
liver, thyroid and gonads. For each replicate test vessel, a total of 10 males and 10 females were
retained for reproduction experiments and breeding of the F2 and F3 generations After
termination of the breeding experiments (approximately at 180 dpf), remaining adults were
sampled following the exact procedure as described above for sub-sampling at 90 dpf. Post-
hatching survival was documented for the I 3 generation at 14 dpf. w hen the experiment was
terminated without subsequent sampling. Bleeding experiments for evaluation of fecundity and
fertilization rate were performed with I' I and 12 adults starting at approximately four months of
age. Breeding trials lor each treatment group were repeated six to seven times (Fl) and nine to
ten times (F2) with a minimum of one week of recovery in between to avoid stress related bias.
At the day before spaw ning. Ii\ e indi\ iduals of either sex from each test vessel were randomly
selected and transferred to breeding tanks prior to the onset of darkness. The spawning facility
was constructed of six breeding tanks which were held together under un-exposed semi-static
conditions with constant air supply (7.37-8.10 mg/L) and heating (25.0ฑ1.0ฐC). The bottom of
the breeding tanks was covered by a stainless steel grid (mesh size 1.25 mm) to allow the eggs to
pass through into separate spawning trays and thus to avoid cannibalism by parental fish. About
20-30 minutes after the onset of light, spawning trays were removed and the eggs were collected
and any further debris was removed. Eggs were counted and visually inspected under a stereo

G-80


-------
microscope and transferred to petri dishes containing freshly prepared artificial water according
to ISO (1996) (maximum 100 eggs/200 mL water). The eggs were incubated at 26.0ฑ1.0ฐC
overnight, after which coagulated and fertilized eggs were counted.

The chronic value (MATC F2 180-day survival) presented for this study was 0.1732
mg/L PFOS, and was considered qualitative as this study had a poor concentration-response
relationship and test design complications. These test design complications included a pseudo-
replication issue with this study that needs to be considered. There were lour PFOS treatment
groups with 160 fish per treatment, separated over two replicate tanks (80 lisli per lank). A
decrease in fish density was observed shortly after swim-up in one of the two replicates of the
PFOS 0.1 mg/L treatment where the survival rate was 5"... u hich may indicate a problem with
the test organisms. The NOEC and LOEC for I ' I and F2 180-day male lengths and weights, and
F2 180-day female weights, was 0.0006 and 0.1 mg/L. The LOEC for the remaining growth
endpoints (lengths and weights across all generations) was < 0.0006 mg/L. The lowest LOEC
values were less than the I CV of t) oos.i98 mg/L, suggesting this genus may be more sensitive to
chronic exposures of PI-OS than indicated In the quantitative study by Wang et al. (2011) with
an EC" i ฆ ฆ of <> <)!(ฆ> 5 mu I. llo\\c\ei\ this study had poor concentration-response relationships,
across holh endpoints and generations, and had test design complications, including a pseudo-
replication issue Therefore, the ECio from Wang et al. (2011) that used quantitatively in the
chronic criterion dei i\ ation w as determined to be more robust than the toxicity values reported in
Keiter et al. (2012), and the chronic freshwater criterion of 0.0084 mg/L is expected to be
protective of this genus.

Chen et al. (2016) evaluated the estrogenic effects of PFOS (> 96% purity) to Danio
rerio via renewal unmeasured tests. The stock solutions of PFOS (2.5 g/L) were prepared by

G-81


-------
directly dissolving PFOS in 100% DMSO and storage at 4ฐC. The working solution of PFOS
(0.250 mg/L) was prepared by a series of dilutions of the 2.5 g/L stock solution with system
water, resulting in a final DMSO concentration of 0.01% (treatments and control). Adult
zebrafish (wild type AB strain) were raised and kept at standard laboratory conditions of 28ฑ1ฐC
with a 14-hr: 10-hr light:dark cycle in a recirculation system according to standard zebrafish
culture protocols. Water supplied to the system was filtered by re\ eise osmosis (pH 7.0-7.5), and
Instant Ocean salt was added to the water to raise the conductivity to 45<) lo I <)00 mS/cm (system
water). The adult fish were fed twice daily with live Anemia and dry flake food Zebrafish
embryos were obtained from spawning adults (sex ratio of I • I) in tanks overnight I jnbryos were
collected within one hour after spawning and rinsed in embryo medium (EM). Fertilized and
normal embryos were staged under a stereomicroscope The embryos were divided into two
treatment groups: DMSO vehicle control (<> <)|"n \ \ ) and PI OS (0.250 mg/L). Embryos at 8 hpf
were first exposed lo PI-OS in a IVtri-dish (10<) cmlnyos 5<) niL EM) till 5 dpf, and all embryos
hatched and sun i\ ed lo this slaue At 5 dpf, the fish were transferred into 2 L tanks for the
period of 5-30 dpi" The fish were led three limes with zebrafish larval diet between five and 14
dpf and after 14 dpi", they were led twice daily with freshly hatched live Artemia. The fish were
raised in tanks (30 fish per tank) after 30 dpf until the end of the experiment. Half the
solution volume was renewed every five days. There were three replicates from embryos derived
from different parental lineages. To cover the entire sex differentiation period, fish were sampled
at 21, 35, and 42 dpf for measurements of general growth, sex hormone levels and related gene
expression. For general growth measurements, 80 fish were used. The body length was measured
on individual fish while the body weight averages were obtained by pooling samples of 10 fish
per sample. For the measurements of estradiol (E2) and testosterone (T), 60, 40, and 20 fish were

G-82


-------
collected on 21, 35, and 42 dpf, respectively. Whole fish tissue homogenates were used for the
hormone measures due to difficulty in obtaining sufficient serum volume from juvenile fish.
Similarly, whole fish tissue homogenates of 40 pooled juvenile fish from each of these three
sampling time periods were used for RNA extraction. All experiments were repeated three times.
After 150-day PFOS exposures, fish were anesthetized and sexed, body length, body weight and
gonad weights were measured. Sperm samples were collected from adult males for quality
analysis, and serum samples from adult fish (15 fish per sex with blood sample pooled from five
fish serving as one replicate) were collected for sex hormone measurements. The 42 dpf
(increased condition index) and 150 dpf (increased estradiol in male/females and testosterone in
males) LOECs were both 0.250 mg/L PFOS, but these data are classified as qualitative because
there was only one exposure concentration The I.()l-C of 0.25<) mu'l. was one order of
magnitude higher than the FCV of 0.0083^S mu I. and indicated that this genus may be less
sensitive to chronic exposures of PFOS than the quanti1ali\e study by Wang et al. (2011) with
ECio of 0.0165 mu I. lor I ' I sui \ i\ al I lowever, it was difficult to compare these two papers
since different endpoints were assessed and the present study only included one exposure
concentration. v\ liich pro\ ides limited understanding of the magnitude of effects compared to the
control Therefore, the chronic freshwater criterion of 0.0084 mg/L would likely be protective of
this genus

Jantzen ol al. (2U17) evaluated the effects of PFOS on the morphometric, behavioral and
gene expression in Danio rerio exposed via 5-day static unmeasured exposures (OECD Method
212). The AB strain of zebrafish (Zebrafish International Resource Center, Eugene, OR) were
used for all experiments. Breeding stocks were bred and housed in recirculating systems under a
14-hr: lOhr light:dark cycle. System water was obtained by carbon/sand filtration of municipal

G-83


-------
tap water and water quality was maintained at a pH between 7.2 and 7.7, and water temperature
between 26 and 28ฐC. Zebrafish embryos were exposed at 3 hpf to PFOS at concentrations
reported by the authors of 0, 0.02, 0.2 or 2.0 [xM (or 0, 0.02, 0.2, and 2.0 mg/L using a molecular
weight of 500.13 |ig/|imol for PFOS) for 120 hours (4 replicates, 24-35 fish per replicate). All
compounds were dissolved in water. After this time, fish were transferred to non-treated system
water and fed two times daily with Zeigler Larval AP50. Therefore. I he only exposure was
through the water from 3 hpf to 120 hpf (5 days), which corresponds lo embryonic to yolk sac
larval exposure. At 120 hpf, morphometric measurements were recorded and gene expression
analyzed. The OECD protocol was to extend the study beyond the exposure timepoints which
allowed for removing any chemical exposure from 120 hpf to 14 dpf. Morphometric
measurements were also taken at 7 dpf and 14 dpi" At 14 dpi", uene expression data and swim
activity endpoints were collected. Each treatment compound and corresponding control group
was set up as individual experiments, and the sample size was dependent on number of embryos
produced from the stock breeding sets Mo experiment had mortality greater than 20% of the
starting sample size The li\ e day (plus nine days for observation) NOEC, LOEC, and MATC for
growth as total body length were <> <>2. n 2. and 0.06325 mg/L, respectively. The LOEC was
associated with a 3.79% decrease in growth compared to control. This study was considered for
qualitati\e use due to the short exposure duration of five days and since it was an early-life stage
test focused on the de\ elopmental toxicity of PFOS. The MATC of 0.06325 mg/L was similar to
the FCV of 0.008398 mg/L and supports the quantitative study by Wang et al. (2011) with ECio
of 0.0165 mg/L. Therefore, the chronic freshwater criterion of 0.0084 mg/L would likely be
protective of this genus.

G-84


-------
Sharpe et al. (2010) examined the toxicity and bioaccumulation of PFOS isomers on
Danio rerio through three different tests, a 96-hour renewal toxicity test on adults, a 48-hour
renewal toxicity test on embryos, and a chronic exposure test that evaluated maternal transfer
and fecundity of PFOS isomers. The chronic toxicity tests are described in this present section,
as they were used qualitatively (also see Section 4.4.2.1.4). The 48-hour tests were used
qualitatively (see G.2.2.3) and the 96-hour tests were used quaniilaii\ ely and are summarized
above in A.2.10. Zebrafish were purchased from a pet store local to the I ni\ ersity of Alberta and
were reared at university facilities for six to ten months. Conditioned zebrafish water obtained
from the Biological Sciences Zebrafish Facility at the University of Alberta. Fish were
acclimated to and kept in 70 L glass aquaria where they were led powdered trout chow (Unifeed)
daily, occasionally supplemented with li\ e brine shrimp An automated reverse osmosis system
was used to maintain conditioned zebrafish water, used lor acclimation and testing, at a total
hardness of around I (ฆ>< > mu I. and a calcium carbonate hardness at 20 mg/L. Test concentrations
were diluted from a 25 mu ml. stock solution in a methanol (MeOH) solvent for dosing in all
experiments.

A 14-day experiment was conducted to examine PFOS accumulation and changes in
isomer profiles in response to maternal transfer. Over a period of 14 days, eight tanks (two
controls: 0 mu I., two solvent controls: 0.01% MeOH v/v, and four treatments: 2 mg/L PFOS)
received daily water changes with daily dose renewals. After 14 days, fish were transferred to
breeding cages in conditioned zebrafish (control) water to spawn. The next day, following
spawning, adults were sacrificed and stored as -80ฐC, and eggs were collected and stored at -
80ฐC for further analysis.

G-85


-------
A 21-day experiment was conducted to test the potential of PFOS to reduce fecundity.
Over a period of 21 days, five tanks (1 control: 0 mg/L, 1 solvent control: 0.01% MeOH v/v, and
three treatments: 0.5 mg/L) received daily water changes with daily dose renewals. Fish were
spawned prior to exposure (day 0), and on days 14 and 21 of the exposure period, where adult
fish were removed from their tanks and held in breeding cages (in conditioned zebrafish water)
to spawn. Fish that spawned on day 14 were returned to their experimental tanks after spawning,
and fish that spawned on day 21 were sacrificed after spawning. Eggs from all tanks were
counted from day 0, 14, and 21, and eggs spawned from fish held in treatment tanks were
compared to eggs spawned from fish held in control tanks Results for this experiment are
reported qualitatively because data from one control tank were lost due to unusual fish
aggression.

Results for the fecundity study showed at 34" <> reduction in fecundity relative to control
in fish exposed to <> 5 mu I. PI OS for 14 days and a 47% reduction in fish exposed 21 days. In
the maternal transfer experiment, approximately 10% (wt) of the burden for PFOS was
transferred from adult fish to their embryos Resulting egg concentrations were 116 ฑ 13.3 mg/L
PFOS. which were significantly higher than whole body PFOS concentrations of 72.1 ฑ 7.6
mg/I. Similar isomer patterns were observed in adults and eggs, suggesting isomer fractionation
is more likely to occur from either biomagnification or isomer-specific accumulation instead of
maternal transfer. Results from the fecundity study indicates an author-reported NOEC of < 0.5
mg/L and LOEC of 0.5 mg/L which was associated with a 34% reduction in fecundity relative to
control in fish exposed for 14 days. This study was considered for qualitative use as a result of
this test consisting of one experimental concentration (of 0.5 mg/L) and because one of the two
control replicates was lost, which the study authors note was due to unusual aggression among

G-86


-------
the test organisms. The author-reported LOEC of 0.5 mg/L was one order of magnitude higher
than the FCV of 0.008398 mg/L and indicated that this genus may be less sensitive to chronic
exposures of PFOS than the quantitative study by Wang et al. (2011) with ECio of 0.0165 mg/L
for F1 survival. However, it was difficult to compare these two papers since the present study
only included one exposure concentration, which provides limited understanding of the
magnitude of effects compared to the control. Therefore, the chronic freshwater criterion of
0.0084 mg/L would likely be protective of this genus.

Chen et al. (2013) examined the behavioral effects of zebrafish resulting from prolonged
chronic exposure to PFOS. Adult Danio rerio (US-AB strain) used for spawning the lest
organisms were maintained following standard protocols. A stock solution of 0.5 mM (250.07
mg/L) PFOS was dissolved in 100% dimethyl sulfoxide Fertilized embryos were collected
within one hour after spawning, rinsed with embryo medium, then staged and inspected with a
Nikon stereomicroscope Testing concentrations included lour treatments: a vehicle control
(0.01% dimethyl sulfoxide) and three') 5 liVI (0.2501 mg/L) PFOS treatments that were
maintained for different periods of time, to assess the effects of PFOS exposure during different
zebrafish life stages All groups began with S hpf embryos. Of the three groups exposed to
PFOS. one group was exposed from 1 -20 dpf, the second from 21-120 dpf, and the third from 1-
120 dpf.

Testing began with 100 embryos placed into Petri dishes (one Petri dish per group) filled
with 150 mL of treatment water. Each petri dish received a daily media renewal. On day 5,
hatched embryos were transferred to 3.75 L stainless steel tanks (one tank per group) for 15 days,
with each tank receiving a water renewal every three days. On day 21, 30 fish from each group
were transferred to three replicate tanks (10 fish per tanks), for a total of 12 tanks (three tanks per

G-87


-------
group). Solutions and water were renewed every three days. During renewal, the tanks were
monitored for morbid fish and water quality. Temperature was maintained at 28ฑ0.5 ฐC, pH was
measured at 6.8-7.6, conductivity ranged from 450-1,000 |iS/cm, and ammonia ranged from 0-2
mg/L. A 14-hr: 10-hr light:dark cycle was maintained throughout the study period. Before 60 dpf,
nitrite concentrations were measured at 0.073ฑ0.12 mg/L, and after 61 dpf, nitrite concentrations
ranged from 1.02ฑ0.65 mg/L. Feeding methods varied as the fish aged Beginning at 5 dpf,
animals were fed zebrafish larval diets three times per day. From 15 dpf lo dpf, zebrafish
were fed live brine shrimp three times a day, which was reduced to two times a day at 97 dpf
until the end of the test.

Following the 120-day exposure period, swimming behavior was measured in F0 adults
as follows: 12 fish from each treatment group were placed in their own individual 1.75 L tank
filled with 1.5 L of water. The test organisms were lasted during the behavioral experiments and
given two hours to acclimate before the trials began The position of each fish was recorded
every 0.2 seconds to calculate distance moved, which were then averaged for each 30 second
interval during the test l-\pei iments lasted 3<> minutes, and the final 16 minutes were analyzed.
This period included 12 minutes of mo\ ement; analysis described above, plus four minutes of
startle response analysis. To induce a startle simulation, an electromagnetic solenoid was
attached to the bottom of each tank and programmed to tap each tank at the same time.

At approximately three months of age, F0 adults from the same treatment group were
bred. Embryos (Fl) hatched from these adults were monitored for developmental progression,
and hatched larvae were monitored for 8 dpf for malformation and mortality. Embryos were
monitored in 6-well plates with 5 mL of water per well. This experiment was repeated five times

G-88


-------
using 20 embryos per replicate per group. Whole body tissue concentrations were analyzed from
40 pooled embryos as one sample which was replicated four times per group.

Second generation (Fl) zebrafish were maintained in water free from PFOS. Larvae aged
4 dpf that were free from malformations were subjected to an assessment involving light and
dark cycles for a period of 50 minutes. Six fish per treatment group, including the control, were
placed in 24-well plates inside a ZebraLab behavior monitoring station Larvae were allowed to
acclimate for 20 minutes before stimulus testing began. For fifty minutes, fish were exposed to
alternating cycles of light and dark periods. Each light and dark period was ten minutes each. At
the end of the experiment, the average activity of IS fish per each treatment group was
calculated.

Movement speeds in adult male and female I'd fish exposed to PFOS from 1-120 dpf
were significantly higher (P<0.05) than control fish PI OS whole body tissue residues in the Fl
larvae from the 21 -12<> dpf group and the 1-120 dpf group were not statistically different from
each other. Ftowe\er. concentrations in both groups were statistically significantly higher
(P<0.001) than the control group and I -2d dpf group. Additionally, Fl larvae hatched from F0
adults from the 21 -1 2d dpf and I -1 2d dpf groups showed much higher rates of mortality and
malformation than the control and the 1-20 dpf groups. Finally, during the Fl larvae light/dark
test, it was found that the basal swim rate of the 1-20 dpf group and the 21-120 dpf group was
statistically significantly higher (P<0.05) than that of the control in both light and dark
conditions, while the 1-120 dpf group showed a statistically significantly lower (P<0.001) basal
swim rate during the light cycles and higher (P=0.028) basal swim rate during the dark cycles
when compared with the control group. The author-reported NOEC was < 0.250 mg/L and the
LOEC was 0.250 mg/L for mortality and malformation. This study was considered for qualitative

G-89


-------
use because the test consisted of one experimental concentration (of 0.250 mg/L). The author-
reported LOEC of 0.250 mg/L was one order of magnitude higher than the FCV of 0.008398
mg/L and indicated that this genus may be less sensitive to chronic exposures of PFOS than the
quantitative study by Wang et al. (2011) with ECio of 0.0165 mg/L for F1 survival. However, it
was difficult to compare these two papers since the present study only included one exposure
concentration, which provides limited understanding of the magnitude of effects compared to the
control and different endpoints were assessed. Therefore, the chronic freshwater criterion of
0.0084 mg/L would likely be protective of this genus.

Tse et al. (2016) evaluated the chronic effects of perfluorooctane sulphonic acid (PFOS,
purchased from Sigma-Aldrich) on zebrafish (Danio reno) in a 6-day unmeasured, static-
renewal study. AB wild-type zebrafish were obtained from /eMailsh International Resource
Center and were kept in a stand-alone system at 2S (' under a 14-hr: 10-hr light-dark
photoperiod. The study was appro\ ed and followed guidance as given by the Animal
Experimental Committee at I long Kong Baptist I Iniversity. The test consisted of a control and
PFOS treatment The PI-OS treatment was prepared by dissolving PFOS into dimethyl sulfoxide
and diluting the stock solution with egg medium (E3) to a concentration of 0.5 |ig/L. Wild-type
AB embryos at the 1-4 cell stage were exposed to PFOS in 2 mL of E3 medium in a 6-well plate
for six days with daily renewals. Test temperature was maintained at 28ฐC. On day six,
organisms were sacrificed and measured for liver enlargement and taken for genetic analysis. All
PFOS-exposed fish showed increased liver size in comparison with the control, and all showed
significant differences in gene expression of LPL, IGFBP1, PITPNA, DDX5, LECT2, MMP9,
APOE, AGT, APOA1, GALE and GMDS. The 6-day LOEC for all non-apical, sublethal

G-90


-------
endpoints included in the study was 0.5 |ig/L PFOS (or 0.00050 mg/L). The lack of apical
endpoints and atypical duration makes the study acceptable for qualitative use.

Bao et al. (2019) evaluated the chronic effects of perfluorooctane sulfonate (PFOS
purchased from Dr. Ehrenstorfer in Augsburg, Germany) on juvenile zebrafish (Danio rerio), via
a 21-day unmeasured, static-renewal study. Wild-type AB strain zebrafish were also obtained
from Dr. Ehrenstorfer for the study. PFOS was dissolved with dimethyl sulfoxide (DMSO) to
obtain nominal test concentrations of 0 (solvent control), 0.002, 0.02 and <> 2 mg/L. Female
zebrafish approximately four months old were acclimated to laboratory conditions for two weeks
in a recirculating water system under a 12-hr: 12-hr lighfdark photoperiod, temperature of 28 ฑ 1
ฐC, pH of 7.0 - 7.5 and conductivity of 500 to 800 |iS/cm. Eighty zebrafish were divided into two
subgroups for analysis, with test initiation of one group at 8 no am and the other group at 7:00
pm. Within those subgroups, the 40 fish were di\ ided equally among the concentrations, and
each fish was considered a single hiological replicate. Six fish were used for growth and
fecundity testing, u hile the other lour were sacrificed for genetic analysis. Fish were randomly
placed into four glass aquaria measuring 1S \ 13 x 15 cm3 and half of the exposure water was
replaced e\ ery day Ik-lore sampling, each female was matched with a male zebrafish, and
fertilized eggs were collected at either 8:30 am or 7:30 pm according to subset type. An endpoint
for fecundity was not reported by the authors, and there were no significant differences between
the exposure concentrations in terms of growth length or weight (NOEC > 0.2 mg/L PFOS). No
mortality was observed. Independently-calculated ECios could not be calculated as EPA was
unable to fit a model with significant parameters. Therefore, given EPA was unable to
independently calculate toxicity values based on the level data provided in the paper by the study
authors, the test duration was a partial-life cycle test as opposed to the preferred life-cycle test

G-91


-------
for which there were studies on this species (Wang et al. 2011), and the author-reported toxicity
values resulted in a NOEC > 0.2 mg/L, this study was used qualitatively to derive the draft
chronic water column criterion.

Huang et al. (2021) used PFOS as a positive control in a series of short-term, 4-day tests
as part of a comparative developmental toxicity assessment against sodium /;-perfluorous
nonenoxybenzene sulfonate (OBS). Normally developing zebrallsli embryos (6 hours post
fertilization, hpf) from adult wild-type (AB strain) breeding stock were used lor the experiments.
A stock solution of PFOS-K (>98% purity, CAS # 2795-39-3) purchased from Siuma-Aldrich
(Oakville, ON, Canada) was prepared at 2 g/L in DM SO and diluted with culture medium to
achieve the desire concentration of 20 mg/L. The final DMSO concentration in the treatment and
control groups was 0.1% (v/v). In the de\ dopmcnlal toxicity assay, xebrafish embryos were
randomly allocated to 6-well plates at a density of 3') per well containing 5 mL of exposure
solution. Each treatment contained three replicates with 3d embryos per replicate. The embryos
were maintained at 2K 5 (' with 14-hr 10-hr light:dark photoperiod for four days. The exposure
solutions were renewed daily, and dead enilnyos w/ere gently removed. Coagulation of the
embryo, failure of somite de\ elo|">menl, lack of heartbeat and non-detachment of the tail from the
yolk sac were identified as dead embryos. Several types of malformations were looked for in the
embryos. Hatching rate was determined as the percentage of hatched embryos at 48 hpf and 72
hpf. In the bioconcentration. physiological and biochemical assay, zebrafish embryos were
randomly exposed to a 1 L glass beaker filled with 400 mL of exposure solution that had 300
individuals for four days. Each treatment contained three replicates with 300 embryos per
replicate. At the end of exposure, visually normal larvae were harvested and divided into several
subsets for PFAS determination, locomotor behavior assessment, dopamine concentration

G-92


-------
measurement, transcriptome and qRT-PCR analysis, Western blot analysis, and immunostaining
of cilia. The remaining larvae served as backup samples. Finally, for the angiogenesis assay,
embryos at 6 hpf were assigned to a Petri dish filled with 100 mL exposure solution that had 25
individuals. The embryos were exposed for four days to evaluate any effects on embryonic
vasculature. After exposure for four days, 36 zebrafish larvae were randomly selected from each
treatment group for locomotor behavior assay. Measured concentrations of PFOS were
comparable to nominal concentrations (range: 21.88-22.33 mg/L). After lour days of exposure to
20 mg/L PFOS, mean whole-body burden of PFAS (poly- and perfluoroalkyl substances) in
zebrafish larvae was 464.7 mg/kg wet weight. PFOS at this concentration induced a series of
malformations, mainly including pericardial edema, curved lail, spine curvature and shortened
body. A significant 44.4% reduction in hatching rales at 48 hpf was observed compared to the
negative control group but hatching rates al 72 hpf reached I oo",, in all treatment groups.
Examination of genes related to the cell cycle re\ ealed that PI OS treatment resulted in a
significant reduction in the niRY\ le\ els of ccnel, cdk2, cdk6 andpena genes in zebrafish
larvae, but did not affect the transcript expression of ccndl and myca. Locomotor behavior was
also reduced by exposure to 2d mu I. PFOS, which also significantly reduced dopamine content
in zehralish lar\ ae by at least I 5 times compared to the control group. The transcription of
several crucial genes invoh ed in neuronal development was also significantly reduced in the
PFOS treatment group (ii\ en the lack of apical responses in the study and single PFOS
treatment, this value is considered qualitative only.

The effects on zebrafish, Danio rerio, embryonic metabolism, pancreas development, and
adiposity due to developmental and sub-chronic PFOS exposures and their persistence into later
larval and juvenile periods were evaluated by Sant et al. (2021). Stock solutions were prepared

G-93


-------
by dissolving PFOS (Sigma-Aldrich, St. Louis, MO, Catalog # 77283) into DMSO, and stored at
room temperature in amber glass bottles inside of light-prohibitive containers until use. Wildtype
AB zebrafish were originally obtained from Boston Children's Hospital (Boston, MA). Embryos
<1 hour post fertilization (hpf) used for experimentation were collected from breeding tanks of
adults. Embryonic exposures to either 0 (0.01% v/v DMSO), 16, or 32 |iM PFOS for
biochemical analyses were performed on wild-type (AB) embryos Successful embryos (3 hpf)
were randomized and individually transferred to wells of 24-well polystyrene plates, and
exposure media was prepared daily in 0.3X Danieau's medium prior to this transfer. One mL of
the assigned exposure medium was added to each well containing one embryo. l-\posure media
was refreshed daily to model subchronic exposures throughout development, and embryos were
maintained in a dedicated incubator at 28 5 (' l-\periments were repeated three times each for
the biochemical assays and fatty acid analysis, each with 4-0 samples per exposure group each
containing 15 pooled embryos (f->( )-80 embryos per treat menu At four days post fertilization
(dpf), larvae were rinsed thoroughly and transferred to 1.5 mL polypropylene microcentrifuge
tubes. For protein, cholesterol, glucose, and triglyceride quantification, 15 larvae were pooled
per sample I or fully acid analysis. 2d lar\ ae were pooled per sample. Concentrations of lauric
(C12 D) and myrislic (C14:0) saturated fatty acids were increased by PFOS at 4 dpf, and PPAR
gene expression was reduced The 4 dpf LOEC based on these apical endpoints was 16 |iM
PFOS (8.002 mg/l. PI-OS based on a molecular weight of 500.13 g/mol) and is acceptable for
qualitative use only.

A 10-day subchronic unmeasured test was conducted by Zhu et al. (2021) to investigate
the effect of PFOS exposure on hepatocellular carcinoma (HCC) progression in transgenic
zebrafish Tg (fabplO:rtTA2s-M2; TRE2:EGFP-KRASG12V). PFOS-K (potassium salt, purity

G-94


-------
>98%) purchased from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan) was dissolved into
dimethyl sulfoxide (DMSO) to expose male krasV12 transgenic zebrafish to 500 |ig/L PFOS for
10 days under renewal conditions. Both treated and control groups received 0.01% (v/v) DMSO
previously shown not to cause a significant toxicological response in zebrafish. The
concentration of PFOS selected for use in the study was based on preliminary experimentation (4
day tests) conducted over a range of concentrations to elicit an increase in liver size and
fluorescence intensity. For the definitive experiment, adult transgenic zchralish (l->0 dpf) were
maintained in 5 L glass beakers with 3 L charcoal-filtered tap water al 28 C in llie dark. During
the exposure period, the water chemistry was recorded (dissolved oxygen: 8.06 =1 n <15 mg/L;
water temperature: 28.0 ฑ 0.5ฐC and pH: 7.21 ฑ 0.09). /ehiafish were fed with Artemia nauplii
two times every day. Treatments included three replicate beakers each containing eight zebrafish.
Exposure solutions were changed daily. A Tier l<> days of exposure, zebrafish, were euthanized,
weighed, and dissected /elualish livers were obser\ ed and photographed by fluorescence
microscopy. The li \ ers were then separated and weighed to calculate the hepatosomatic index
(HSI). Six livers from each treatment were randomly selected and fixed in 4% paraformaldehyde
solution lor histological examination, and three li\ers from each treatment were randomly
selected lor transcriptomics analysis No significant effect on HSI was observed in zebrafish
exposed to 0.5 mg I. PFOS. and livers showed normal hepatocyte nuclear structure compared to
control livers. Cytoplasmic \ acnolation was evident in the PFOS group compared with the
control, but PFOS had no significant effect on liver enlargement. Gene expression level of
cyp24al was significantly upregulated by 8.2 fold and expression of cyp27bl was significantly
downregulated by 4.8 fold after exposure of zebrafish to PFOS for 10 days. The results indicated
that the synthesis of calcitriol was reduced, and the degradation of calcitriol was enhanced,

G-95


-------
resulting in a decrease in intracellular concentration. Additionally, three of the top 16 enriched
pathways (p value < 0.05) were related to lipid metabolism in zebrafish liver after exposure to
0.5 mg/L PFOS for 10 days, including PPAR signaling pathway, adipocytokine signaling
pathway and fatty acid metabolism pathway. Given the lack of apical responses in the study and
single PFOS treatment, this value is considered qualitative only.

G. 3.2.5 Pimephales promelas

Oakes et al. (2005) exposed fathead minnows to PFOS in an outdoor microcosm

experiment. The University of Guelph Microcosm Facility is located at the Guelph Turfgrass

Institute (ON, Canada) and consists of 30 artificial ponds of approximately 12,00') I. The

microcosms were constructed below grade to a depth of I 2 meters using galvanized steel panels

lined with food-grade polyvinylchloridc l .ach microcosm had a diameter of 3.9 meters, was

filled with water to a depth of approximately one meter, and was flush with ground level. The

water supply for the microcosms was an irrigation pond supplied by a well located on site.

Sediment trays containing a 1.1.1 (\ \ \) mixture of sand, loam, and organic matter, as well as

potted macrophytes {Myriophyllnm spicamm) were added to each microcosm. Prior to being

treated with PI OS. water was circulated among all microcosms for two weeks at a flow rate of

approximately 12 m3/d, ensuring homogeneous water chemistry, zooplankton, and algae

assemblages PI OS (S9% purity) stock solutions were premixed in 40 L Rubbermaidฎ

containers before introduction to the microcosms by subsurface injection. Water samples from

each microcosm were obtained using a metal depth-integrating water-column sampler at one

hour and 1, 2, 4, 7, 14, 21, and 28 days after PFOS addition to calculate time-weighted mean

PFOS concentrations. All treatment concentrations were based on the PFOS anion (without K+).

Microcosms were treated in triplicate at nominal concentrations of 0.3, 3, 10, and 30 mg/L

PFOS. Three additional microcosms served as controls and did not receive any PFOS. Fathead

G-96


-------
minnow (6.1 cm, 2.0 g) were purchased from Silhanek Baitfish Farms (Bobcageon, ON, Canada)
and acclimated in the adjacent irrigation pond for 10 days prior to PFOS exposure (under a
natural photoperiod). Fish were held in two wooden frames with 5 mm aperture
polyvinylchloride mesh cages. Each microcosm held two cages, with each PFOS concentration
replicated in three microcosms. Cages were divided into four quadrants, and each quadrant
contained two female and one male fathead minnow for a total of 24 fish per microcosm. Fish
were initially sexed prior to exposure based on size and presence of secondary sex
characteristics. Sexes were subsequently confirmed at the conclusion of the exposure after the
fish were killed. A 15 cm piece of 10 cm polyvinyl chloride pipe cut in half lengthwise served as
a breeding substrate within each quadrant and was examined for egg deposition daily. Both egg
production and oviposition (spawning) frequency w ere recorded and used for the subsequent
calculation of egg and oviposition frequency per female, per microcosm, and cumulatively per
dose. At the conclusion of llie 2S-da\ exposure, measurements of total length, total weight,
gonad weight, and li\ er weight were taken, and uonadosomatic indices (GSI), liver-somatic
indices (LST). and condition factor (k) were calculated. Mean water-quality parameters
(collected mid-depth) sampled o\er the course of the exposure include dissolved oxygen (8.7
mg/I.). temperature (18.0VC). pH (9 2). and alkalinity (114.5 mg CaC03/L). The 28-day LCio
was 3.5 mg/I. PI OS (qualitati\ e) based on the time weighted average over the course of the
experiment. Even though the microcosm experiment contained sediment, algae, macrophyte and
zooplankton, the results of this study should be compared with other single species studies.

G. 3.2.6 Pseudorasbora parva

Yang et al. (2014) evaluated the toxicity of PFOS (potassium salt, CAS # 2795-39-3,

99% purity) to Pseudorasbora parva via 30-day renewal measured exposures (the authors note

that the experiments followed ASTM standards and USEPA procedures for deriving water

G-97


-------
quality criteria). The topmouth gudgeon (4.0 g, 4.0 cm) were purchased from the Beijing
Chaoyang Spring Flower Market, which was considered an atypical source. The organisms were
allowed to acclimate for seven days before testing, and the test was conducted at 22ฑ2ฐC with a
light:dark cycle of 12-hrs:12-hrs, and 10 fish per replicate with three replicates per concentration.
Beakers used for exposure were assumed glass, but glass was not specified by study authors.
PFOS was dissolved in deionized water and carrier solvent D\TS() lo obtain a 7 mg/mL stock
solution, and then diluted with dechlorinated tap water to yield nominal exposure concentrations
of 30, 45, 67.5, 101.25, 151.88 and 227.81 mg/L PFOS. Water quality parameters reported were
pH=7.0 ฑ 0.5, dissolved oxygen=7.0 ฑ 0.5 mg/L. total organic carbon=0.02 mg/L and total
hardness=190.0 ฑ0.1 mg/L CaCCte. The supplemental data provided for the study included a
comparison of measured PFOS concentrations before and after solution renewal in the low and
high acute and chronic test concentrations. PI OS concentrations in the test water did not
fluctuate by more than I 5".. during experiments The 30-day survival ECio reported for the study
of 2.12 mg/L PI'OS was deemed qualitati\ e due to the atypical fish source, not a valid ELS test
(started with older unspecified lile stage), and unknown composition of test beakers.

G.3.2. ~ S/>inibarbns sinensis

The toxicity of PI-OS (potassium salt, 99ฐ u purity) to the qingbo, Spinibarbus sinensis

was evaluated l\\ Xia et al. (20L5a) via a 30-day renewal unmeasured test. PFOS was initially

dissolved in DMSO. and the slock solution (0.5 g/mL) was kept at 4ฐC until preparation of the

final exposure solutions in water. Juveniles of uniform size (2.77 g, 5.62) were obtained from

local farmers in Chongqing Municipality, China. The fish were housed in a 120 L recirculating

water tank system at Chongqing Normal University for two weeks prior to the experiment. The

rearing water was dechlorinated and filtered through activated carbon. During this time, the

temperature of the water was maintained at 22 ฑ 1ฐC, the dissolved oxygen level was kept above

G-98


-------
7 mg/L, and the photoperiod was 15 hours light and nine hours dark. The fish were fed
commercial tubifex twice daily. A semi-static exposure experiment apparatus was used for
waterborne PFOS exposure. The apparatus consisted of ten glass aquariums each with a capacity
of approximately 22 L of water. Prior to exposure, the fish were gently transferred to the
aquariums. They were maintained at conditions similar to that described above for one week to
eliminate stress effects and were fed at the level of maintenance ration daily. After that, the water
temperature was increased or decreased by lฐC/day until it reached the prescribed temperature
(18 and 28ฐC). Once the water temperature reached the prescribed values, the toxicants were
administered. The fish were then exposed to a range of PFOS concentrations (0. <> .>2. 0.8, 2 and
5 mg/L) under two different temperatures (18 and 28ฐC) for 3D days. A total of 160 fish were
used in the experiment (one replicate per concentration. 16 fish per replicate). The concentration
of DMSO in the water did not exceed 0.001 ".. (\ \ ) I )nri nu the experimental period, 50% of the
exposure solution was renewed dailyฆ. After termination of PI-'OS exposure, the fish were fasted
for 48 hours, and then their beha\ ior. metabolic characteristics and aerobic swimming
performance were examined I'irst. indi\ idnal fish were moved into the behavior observation
system in which their spontaneous swimming behavior (SSB) and social interactions (SI) with
fish were assayed using the Noldus \ icleo tracking software. Second, the fish were individually
transferred to continuous-flow respirometer chambers for routine metabolic rate (RMR)
determination. Finally, individual fish were placed in a swim tunnel respirometer, and the
aerobic swimming performance and the oxygen consumption rate MO2 during swimming were
determined using the critical swimming speed method. Subsequently, the metabolic scope (MS),
factorial metabolic scope (F-MS) and energetic cost of transport (COT) were calculated. The fish
were kept in a fasted state over the course of the entire experimental tests. The 30-day chronic

G-99


-------
MATCs determined at 18ฐC (% mobile, % highly mobile, swim distance, swim speed, freq.
highly mobile, % social, resting) and at 28ฐC (critical swim speed) were both 0.5060 mg/L
PFOS. These data were considered qualitative because the test was not replicated, qingbo is a
non-North American species, the test was unmeasured, water quality details were not reported
(assumed that dilution water was the same as holding water, but this was not specified), and the
biomass loading density was higher than recommended for both icsl temperatures.

In a companion study, Xia et al. (2015b) again evaluated the 3< ป-cla\ toxicity of PFOS
(potassium salt, >99% purity) to Spinibarbus sinensis. PFOS was initially dissol\ ed in DMSO,
and the stock solution (0.5 g/mL) was stored at 4 (' until preparation of the final exposure
solutions in water. Juvenile fish of uniform size (2.69 g, 5.39 cm) were obtained from local
farmers in Chongqing Municipality, China Prior to the experiment, the fish were reared in a 120
L recirculating water tank system for two w eeks The rcari nu u ater was dechlorinated and
filtered through acti\ ated carbon Water temperature was maintained at 22 ฑ 1ฐC, water oxygen
content was kept al">o\ e 7 mu I.. pi I ranged from 6.8 to 7.5 and the rearing system was
maintained under a I 5 hours light and nine hours dark cycle. The fish were fed to satiation daily
with commercial /nbijcx s/>/> After the acclimation period, healthy fish of similar sizes were
selected for the study. Juvenile southern catfish, Silurus meridionalis were used to provide the
predator stressor in this study A semi-static exposure experiment apparatus consisted of 10 glass
aquaria, each with a capacity of approximately 22 L. Prior to exposure, the juvenile qingbo were
randomly selected and divided into 10 groups (n=16 for each group) and were gently transferred
to the aquariums. Fish were maintained at conditions similar to those described above for one
week to eliminate stress effects. For each group, the same amount of food (approximately 10%
of body weight) was provided daily during this period and during subsequent processing. After

G-100


-------
that, the water temperature was increased or decreased by lฐC/d until it reached a prescribed
temperature (18 or 28ฐC). The selected temperatures were chosen based on the habitat/ season
temperature. For qingbo, the habitat temperature is approximately 18ฐC in spring and autumn
and approximately 28ฐC in summer. Once the water temperature reached the prescribed values,
the toxicants were administered. Fish were exposed to a range of PFOS concentrations (0, 0.32,
0.8, 2 and 5 mg/L) under the two different temperatures (18 and 2X (') for four weeks. The
concentration of DMSO in the water did not exceed 0.001% (v/v). During the experimental
period, 50% of the exposure solution was renewed daily. After termination of PTOS exposure,
feeding was withheld for 48 hours and the antipredator behavior and fast-start performance of the
fish were examined successively. Again, the 30-day chronic value (decreased maximum linear
acceleration) was 0.5060 mg/L PFOS at each temperature and the data are classified as
qualitative (non-apical endpoints reported and unmeasured exposure). In addition, control
survival was not reported. ho\\e\ er. the study was still classilied as qualitative because the study
authors did not indicate any pioMems with exposure related mortality.

G.3.2.8 (hyuus /
-------
than eight eggs per breeding and bred more than five times per week were selected and randomly
separated into four groups. Nine mating pairs were assigned to each treatment group and the
control. The following PFOS concentrations were used for definitive tests: 0.01, 0.1, and 1 mg/L,
based on the preliminary range-finding results using adult medaka. The exposure duration for F0
fish was limited to 14 days, during which the fish were fed Artemia nauplii (< 24 hours after
hatching) ad libitum twice daily. The exposure medium was renewed at least three times per
week. Dead fish were removed as soon as possible. Eggs were counted e\ cry day, and the eggs
spawned on the seventh day were saved for the F1 generation exposure study On day 14, all
surviving fish were euthanized, and body length and weight were measured, from u liich the
condition factor (K) was calculated. The gonads and livers were also measured, and the
gonadosomatic index (GSI) and the hepatosomatic index (I ISI) were calculated. For the F1 fish
exposure study, fertilized eggs collected from I'd lisli exposed lo each concentration of PFOS and
the control were randomly separated into groups of 25 eggs each and then assigned to varying
concentrations of PI-OS (t). i) i) I. <). I. or I muL), with only one replicate per treatment. Because
eggs were compiled into a single replicate lor the hatching stage, results reported beyond
hatching (e\ en w hen lar\ ae ju\ eniles were separated into replicates) are based on pseudo-
replication During the egg stage for the F1 generation, investigators maintained all possible
combinations of l'<> \ F1 exposure concentrations for a given compound. Exposure was initiated
in 50 mL beakers less than 12 hours after fertilization. The developing embryos were observed
daily under a stereoscopic microscope, and dead embryos were removed. This procedure was
repeated until all living embryos had hatched. Hatching was defined as the disruption of the
chorion. Newly hatched larvae were then randomly transferred to 100 mL beakers and observed
daily for swim-up success and survival for an additional two weeks. Larvae were fed Artemia

G-102


-------
nauplii ad libitum twice daily. After 14 days, replicates with five fry each were randomly
selected from each treatment group and transferred to 1 L beakers for the 100-day post hatch
observation. All survivors were sacrificed 100 days after hatching, and body length and weight
were measured. The gonads and livers were weighed to determine GSI and HSI.

The F0 (parental generation) adult survival, condition factor and adult male GSI and HSI
14-day NOECs were all > 1 mg/L PFOS, whereas the 14-day adull female HSI and GSI MATC
and LOEC were 0.3162 and < 0.01 mg/L PFOS, respectively. For the I I (progeny generation),
the percent hatchability, time to hatch, and swim-up success, the MATCs were all <>.3 162 mg/L
PFOS, and the EC 10 for larval growth (as organism weight and length) and LOEC for survival
were 0.0013 mg/L and 0.01 mg/L PFOS, respectively. The reduction in organism weight at 0.01
mg/L was only 12%, however, and the concentration-response cur\ e for weight was shallow: 1.0
mg/L yielded a 29% reduction in weight. Many of these toxicity values, particularly those for
apical endpoints. suggested that this genus is no more sensiti\ e than the most sensitive genus that
was used to deri\ e the criterion, u hicli had a GMCV of 0.009676 mg/L. EPA notes, however,
that the LOEC for lar\ al growth was actually <>01 mg/L, because the effects were observed in
the lowest concentration tested Therefore, larval growth reported by Ji et al. (2008) may be a
relati\ ely sensi ti \ e endpoint compared to the freshwater chronic criterion of 0.0084 mg/L and
indicated that this genus might be among the more sensitive genera for chronic exposures of
PFOS. However, the apical endpoints were considered to be qualitatively acceptable in the
criterion derivation because there was a lack of replication during the egg stage of the F1
generation. Although Ji et al. (2008) reported F1 growth as a relatively sensitive endpoint, the
chronic criterion was expected to be protective of this genus based on quantitative data for other
fish species.

G-103


-------
Kang et al. (2019) evaluated the chronic effects of perfluorooctane sulfonic acid (PFOS,
>98% purity, CAS No. 1763-23-1 purchased from Sigma Aldrich, St. Louis, MO) on Japanese
medaka (Oryzias latipes) in a 21-day unmeasured, static-renewal study. A stock solution was
prepared by dissolving PFOS into dimethyl sulfoxide and stored at 4ฐC. The 1 mg/L working
solution was prepared by diluting the stock solution in fish culture water (carbon-filtered
dechlorinated tap water). Adult fish (16 ฑ 2 weeks, 0.38 ฑ 0.06 u wet weight) were obtained
from the fish culture facility at the Korea Institute of Technology in Jinju. (i\ eongnam, South
Korea. Fish were acclimated for seven days in carbon-filtered dechlorinaied lap water at 25ฐC
with a 14-hr: 10-hr light:dark photoperiod. Eight male and eight female fish were introduced to a
20 L glass tank filled with 15 L of test solution at concentrations of 0 (control), 0 (solvent
control) and 1 mg/L PFOS. Fish were fed brine shrimp and Tel rum in daily, and the test solutions
were renewed twice weekly. Authors reported follow inu the exposure protocol given by OECD
229, with conditions maintained the same as during the acclimation period. Eggs were harvested
and counted twice daily at 7. I4and 21 days. Significant reduction in fecundity was shown for
all time periods, and spawning became limited at day four. The 21-day fecundity LOEC was 1.0
mg/L PI-OS and is acceptable lor qualitative use.

G.3.3 Amphibians
G.3.3.1 Bnjo gargarizai /\

Yang ol al. (2014) e\ aluated the chronic toxicity of PFOS (potassium salt, CAS # 2795-

3-3, 99% purity) to the Asiatic toad, Bufo gargarizans via 30-day renewal measured exposures

(the authors note that the experiments followed ASTM standards and USEPA procedures for

deriving water quality criteria). The tadpoles (0.048 g, 1.8 cm) were purchased from the Beijing

Olympic Park, which was considered an atypical source. The organisms were allowed to

acclimate for seven days before testing, and the test was conducted at 22 ฑ 2ฐC with a light:dark

G-104


-------
cycle of 12-hr: 12-hr. There were 10 fish per replicate and three replicates per concentration.
Beakers used for exposure were assumed glass, but was not specified by study authors. PFOS
was dissolved in deionized water and carrier solvent DMSO to obtain a 7 mg/mL stock solution,
and then diluted with dechlorinated tap water to yield nominal exposure concentrations of 1.5,
1.95, 2.54, 3.30, 4.28 and 5.57 mg/L PFOS. Water quality parameters reported were pH=7.0 ฑ
0.5, dissolved oxygen=7.0 ฑ 0.5 mg/L, total organic carbon=0 02 mg I. and total hardness=190.0
ฑ0.1 mg/L as CaCCte. The supplemental data provided for the study included a comparison of
measured PFOS concentrations before and after solution renewal in the low and high acute and
chronic test concentrations. PFOS concentrations in the test water did not fluctuate In more than
15% during experiments. The 30-day ECio (survival) reported for the study of 2.00 mg/L PFOS
was deemed qualitative due to the atypical lest organism source, unknown composition of test
beakers and a non-North American species (the test with this species is currently considered for
qualitative use as details related to the source of the test organisms were limited (obtained from
Beijing Olympic Park) and there was no mention of any potential previous exposure to PFAS or
any other contaminant)

G.3..\2 / illiohalcspipicns

l-'ogulli ct al. (2020) e\ aluated the chronic effects of perfluorooctanesulfonate (PFOS) on

northern leopard frogs (Lithohatespipiens, formerly, Ranapipiens) via a 116-day measured,

outdoor mesocosm study Six partial egg masses were collected from an ephemeral wetland at

the Purdue Wildlife Area Eggs were grown outdoors until reaching Gosner stage (GS) 25 in

twelve 150 L cattle tanks. Empty tanks were filled with 100 L well water, 100 g leaf litter

(predominately oak), and 5 g rabbit chow. Mesocosms were also inoculated with periphyton and

phytoplankton (i.e., algal food resources for leopard frog larvae), as well as zooplankton from the

wetland where eggs were collected. After setting for twelve days, four control mesocosms and

G-105


-------
four treatment mesocosms were spiked with a 0 (control) and 12.8 ppb measured PFOS,
respectively. After seven days, 25 tadpoles at GS 25 were added to each mesocosm to initiate the
experiment. Water quality conditions during the test were reported as pH of 7.41 to 8.54 s.u., DO
of 2.1 to 9.4 mg/L and temperature of 13.1 to 29.8ฐC. After 30 days, eight tadpoles were
removed from each mesocosm, sacrificed and measured for snout-vent length and mass.
Organisms that reached GS 46 were sacrificed and measured for the same endpoints. The test
was terminated at day 116, and tadpoles that had not begun metamorphosis were sacrificed and
measured. Survival was calculated as survival to GS 46 and as total survi\ al. The reported
survival and growth (length and weight) NOEC were both 12 8 ppb PFOS (or 0.<)| 28 mg/L). The
study is acceptable for qualitative use only, because of the test design (outdoor mesocosm
exposure with algal and zooplankton communities present).

Brown et al. (2021) evaluated the chronic effects of PFOS-K (>98% purity, CAS # 2795-
39-3, purchased from Sigma-.\klrich) on northern leopard frogs (/Jthobatespipiens) in a 10-day
unmeasured, static-re new a I study Six egg masses were collected from ponds at the Purdue
Wildlife Area in West l.alayetle. Indiana Tadpoles were maintained in outdoor wading pools
and'were led rahhit chow ad libitum A 5<)i) mg I. stock solution was used to prepare PFOS
exposure solutions of 10 ppb and I'm ppb. Twenty tadpoles of a median Gosner stage (GS) of
26.5 and median weight of <> I <>9 g were exposed to PFOS in 15 L tubs filled with 7.5 L solution
made from filtered. nltra\ iolet-irradiated aged well water with a pH of 7.9, dissolved oxygen of
7.4 mg/L and specific conductivity of 579 |iS/cm at 22ฐC under a 14:10 light-dark photoperiod.
Tadpoles were given a one-day acclimation period prior to testing. During the exposure tadpoles
were fed rabbit chow ad libitum and the experimental units were checked daily for mortality. A
complete water change was done on day five and fresh chemical treatments were applied. After

G-106


-------
10 days, tadpoles were transferred to clean water in the same 15-L bins. Tadpoles were held in
clean water for seven days before parasite {Echinoparyphium) exposure to examine tadpole
susceptibility to parasitic infection after PFOS exposure. PFOS-exposed tadpoles were assigned
to individual 1-L cups filled with 0.5 L of clean water. Each tadpole was exposed to 50
trematode cercariae for six hours. The authors reported a NOEC value of 100 ppb (or 0.1 mg/L)
for development stage, snout-vent length, weight, survival and parasite susceptibility. The study
is acceptable for qualitative use only because of the short test duration

Flynn et al. (2021) evaluated the chronic effects of perfluorooctanc sulfonic acid (PFOS,
CAS# 1763-23-1, > 96% purity, purchased from Sigma-Aldrich) on Northern Leopard frogs,
Lithobatespipiens (formerly, Ranapipiens), via a 30-day sediment-spiked measured, static
mesocosm study. Frog egg masses were collected from an ephemeral pond at the Purdue
Wildlife area in West Lafayette, Indiana, l-uu masses were held in covered 190-L outdoor tubs
containing 80 L of well water Once hatched, the larvae were led ad libitum with Purina rabbit
chow. A control (lour replicates) and live replicates of measured exposure concentrations of
0.00006, 0.0<) I and <> <) I (ฆ> mu I. were set up as 1S0-L plastic wading pools filled with 75 L well
water w liich contained 25 iVous in each The stock solution was made by dissolving 0.5 g PFOS
into I I. of re\ erse osmosis Milli-Q water in polycarbonate bottles. Sediment was collected from
the upper 5 to N cm of a permanent pond in the same wildlife area. The sediment was air dried
for eight days, with in I ku of the dried homogenized sediment placed in each experimental unit.
Sediment was spiked with the assigned PFOS dose by adding the appropriate volume of stock
solution to 6 L of water, stirred for five minutes, and then allowed to equilibrate for seven days.
Once equilibrated, 75 L of water was added to the experimental chamber and allowed to sit for
an additional three days. The water was then inoculated with algae and zooplankton from local

G-107


-------
pond water and allowed to establish for five days, after which the Gosner stage 25 frog larvae
were added to each tank. Reported average water quality conditions include pH of 7.8 and
temperature of 26.2ฐC. The study authors reported a 30-day NOEC of 0.016 mg/L for weight,
snout-vent length and mortality and a 30-day LOEC of 0.00006 mg/L for developmental stage
(measured as Gosner stage). Independently-calculated ECios could not be calculated as EPA was
unable to fit a model with significant parameters. Therefore, gi\ en this was an outdoor
mesocosm with spiked sediment and the addition of algal and zooplanklon communities and
EPA was unable to independently calculate toxicity \ allies bused on the replicate level data
provided by the study authors, this study was used t|Lialitali\ el\ to derive the draft chronic water
column criterion.

G.4 Summary of Plant PFOS Toxicity Studies I'soil Qualitatively in the
Freshwater Aquatic Life Criterion

G.4.1 Cvanobacteria. Anabaena sp.

Rodea-Palomares et al. (2012) examined the toxicity of PFOS (98% purity) with the

bioluminescent cyanohaclerium.. \nabacna sp. (CPB4337 strain) following the OECD
Guidelines\o 23 (OIX'I) 2<)i)()) The inhibition of constitutive luminescence was examined
over a 24-hour test period. Very little detail was provided about the exposure details (i.e., test
media, test \ essel, cell density per replicate, water quality parameters). PFOS was dissolved in
the exposure media with no solvent and was measured in the highest test concentration and one
concentration near the reported ECso. The cyanobacteria were exposed to five to seven test
concentrations (specifics not provided) with replicate samples. Each test was repeated three
times. The reported ECso was 16.29 mg/L based on bioluminescence inhibition and was
considered to be acceptable for qualitative use, based on the short test duration and lack of
exposure details. However, the authors in a later publication tested the cyanobacterium again to

G-108


-------
PFOS and it may be assumed that some of the missing test information for this earlier
publication was the same.

Rodea-Palomares et al. (2015) conducted a similar 24-hour static, unmeasured test on
PFOS (potassium salt, 98% purity) with the bioluminescent cyanobacterium, Anabaena sp.
(CPB4337 strain). The test was performed in 1.5 mL of cyanobacterial growth media (AA/8+N,
Allen and Arnon 1955) in transparent microtiter plates. The pTT of the growth media was 7.8.
The plates were incubated at 28ฐC under continuous illumination on a rotary shaker. The
cyanobacteria in the log-growth phase were exposed to seven concentrations of PI OS-K which
ranged from 0-200 mg/L. The description of how the lest solutions were prepared was not
provided, but it does not appear that a solvent was used, Lach test was repeated three times. The
reported EC50 was 83.51 mg/L based on Moluminescence inhibition and was considered to be
acceptable for qualitative use, given the short test duration

G.4.2 Green alga. ( hlorella \ uluaris

Xu et al. (2017) similarly conducted a 90-hour static acute algal growth inhibition test on

PFOS (potassium he|">ladecalluoio-1 -octanesulfonate; 98% purity) with Chlorella vulgaris.

Toxicity was determined from the logarithmic growth phase of C. vulgaris inoculated into 100

mL conical llasks. At the beginning of the experiment, the density of the algae cells was

approximately 7 t)\ 105 cells/ niL in a total volume of 50 mL. The green alga was exposed to one

of five nominal concentrations (0, 40, 80, 120, 160, 200 mg/L PFOS) that were verified by

UHPLCMS/MS confirmation methods. Measured concentrations were not reported. Water

quality conditions were also not reported by the authors other than photoperiod which consisted

of a 12-hr: 12-hr light:dark cycle. A number of endpoints were measured and included reactive

oxygen species (ROS) production, catalase (CAT) activity, chlorophyll a and b content,

superoxide dismutase (SOD) activity, cell permeability, and malondialdehyde (MDA) content.

G-109


-------
The LOEC for decreased chlorophyll a content (40 mg/L) was considered qualitative for use
because the study was lacking experimental details regarding nutrient medium, water quality,
and exposure vessels.

G.4.3 Green alga. Raphidocelis subcapitata

Boltes et al. (2012) conducted a 72-hour static toxicity test on PFOS (potassium salt,

CAS # 2795-39-3, 98% purity) with Raphidocelis subcapitata (identified as Pseudokirchneriella
subcapitata in the test). The exposure included an unknown number of nominal PFOS
concentrations. However, the toxicity test followed the algal growth inhibition test described in
OECD TG 201 (OECD 2011). Algal beads of R subcapitata, dissolving matrix and growing
media were purchased from MicroBioTest Inc. (Belgium) l .ach test concentration had four
replicates, with an unknown density of algal cells in the log growth phase incubated in plastic
96-well plates containing a small amounl of test solution (2<">0 ill.) Details for how the test
solutions were prepared were not provided, but it appeared no sol\ ent was used. The only test
condition reported was temperature (22 ฑ 2ฐC) with a continuous source of illumination. Optical
density was recorded at 72 hours to calculate an ECso (based on growth inhibition) of 35 mg/L.
The limited exposure details and short exposure duration prevented the study from being used
quantitati\ ely

Rossil ot ;il. (2010) perforined a 72-hour static, measured growth inhibition test with
PFOS-K (98% purity) on the green alga, Raphidocelis subcapitata following OECD TG 201
Protocol (Note: the species name has changed since paper publication and is no longer
Pseudokirchneriella subcapitata). While limited details are provided about the exposure, the
authors stated they were following the OECD protocol. Green alga was cultured in 96-well
microplates with a total volume of 200 |iL. No solvents were used to make test solutions.
Specific test concentrations were not provided, but the authors noted that nominal and measured

G-110


-------
concentrations did not have significant deviations. The 72-hour growth inhibition ECso of 35.0
mg/L PFOS was acceptable for qualitative use only because of the short test duration.

G.4.4 Green alga. Scenedesmus obliquus

Liu et al. (2008, 2009) published two studies examining the toxicity of PFOS to the

green alga, Scenedesmus obliquus. In Liu et al. (2008), a 72-hour exposure was conducted to
evaluate the effects on PFOS (acid form, CAS # 1763-23-1, purity not reported) at the cellular
level, measured by flow cytometry. The test followed OECD (2<)iP) methodology with S.
obliquus that were obtained from the Freshwater Algae Culture Collection. Institute of
Hydrobiology, Chinese Academy of Sciences (Beijing). The number of test treatments was not
provided but based on figures in the study, test concentrations ranged from approximately 60-
350 |iMPFOS (or 30-175 mg/L PFOS when coin cried based on a molecular weight of 500.13
g/mol). The authors did not provide details regarding how the PFOS treatments were prepared.
Experiments were initiated by inoculating equal cell numbers of 1 x 104 cells/mL into flasks
containing a total volume of 20 ml. ol"algal cell suspension per flask (with three replicates per
each treatment) The algal test media was prepared according to OECD (2002) using deionized
water and analytically pure chemicals, adjusted to pH 7.5 ฑ 0.2. Algal cells were incubated at 22
ฑ 1(C under cool-w hite lights ((•ซ,<)<)<) lux) with a 10-hrs:14-hrs light:dark cycle. The 72-hour ICso
(growth inhibition) ol'PFOS was 77.8 mg/L and 99.9 mg/L based on fluorescence and optical
density at 650 nm. respecti\ ely. The plant values from the study were acceptable for qualitative
use because of the short exposure duration (less than 96 hours) and the missing exposure details.

Under similar conditions Liu et al. (2009) conducted a 72-hour toxicity test on PFOS
(>98% purity based on dry mass) with the alga, but with lower test concentrations (0, 10, 20, 30,
40 mg/L). PFOS alone exhibited no inhibition on the growth rate of Scenedesmus obliquus
within the concentration range of 10-40 mg/L. The PFOS concentration applied was in the range

G-lll


-------
where PFOS was previously found to not show growth inhibition but disturb the algal membrane
properties in S. obliquus (Liu et al. 2008). Cells in exponential phase of growth collected from
stock cultures were used for experiments. Experiments were initiated by inoculating equal cell
numbers of lxlO4 cells/mL into flasks, containing a total volume of 200 mL of algal cell
suspension per flask (with three replicates per each exposure treatment). The same growth media
was used as in the previous experiment under the same test conditions The PFOS stock solution
was prepared in methanol and the working solution was obtained by I .<><><> limes of dilution of
the stock solution into algal culture medium. The 72-hour \OI-C (Ixised 011 growth) was 40
mg/L and was acceptable for qualitative use because of the short exposure duration (less than 96
hours) and the missing exposure details.

G.4.5 Duckweed. Lemna gibba

Boudreau et al. (2002) performed a 7-day static acute algal growth inhibition test on

PFOS (potassium salt, 95% purity) with duckweed. / emna gibba. The study was part of a

Master's thesis at the I ni\ ei sity of (iuelph, Ontario, Canada and subsequently published in the

open literature as Boudreau et al. (2003a) The test followed protocols found in ASTM E1415-

91 (ASTM liw|). (iieenbeiu et al (ll^2) and Marwood et al. (2001). All treatment

concentrations were based 011 the PI-OS anion (without K) and solutions were prepared in

laboratory-grade distilled water. Duckweed was obtained from laboratory culture maintained

according to Marwood et al (2001), and originally acquired from University of Waterloo.

Toxicity testing consisted of five test treatments plus a negative control (0, 10, 20, 40, 80 and

160 mg/L) in 10 mL of Hunter's growing media in 60 x 15 mm polyethylene disposable petri

dishes. There were four replicates per treatment, but the number of plants and fronds per plant

were not reported. Tests were continuously illuminated with cool-white fluorescent light between

5,800 and 6,200 lux and incubated at 25ฑ1ฐC. Endpoints used to determine inhibition of growth

G-112


-------
were mean frond number and biomass, measured as wet weight. The most sensitive endpoint,
wet weight, had a reported NOEC of 6.6 mg/L and an ICso=31.1 mg/L. The plant values were
acceptable for quantitative use.

G-113


-------
Appendix H Other Estuarine/Marine PFOS Toxicity Studies

H.l Summary Table of Acceptable Qualitative Estuarine/Marine PFOS Toxicity Studies

Species (

Method'

Test
Diimlion

( hcmic;il /
Pu ii( \

pll

Temp.

(ฐC)

Siiliiiit^

r.iTcci

Chronic
Limits
(NOI'.C-

i.or.ci

Reported
I'-ITecl
(one.

Deficiencies

Reference

Bacterium,

Vibrio fischeri

S, M

15 min

PFOS-K

98%



18



1X50

(bioluminescence
inhibition)

-

500

Duration loo sliorl lor a
planl icsi. missing some
exposure details, non-
apical endpoint

knsal el al

(2010)



Cyanobacterium,

Anabaena sp.

S, M

24 hr

PFOS-K

98%

-

28

-

i:('50

(bioluniincsccnce
inhibition)

-

143.27

Duration too short for a
plant test, missing some
exposure details, non-
apical endpoint

Rosal et al.
(2010)



Dinoflagellate,

Pyrocystis lunula

S, M

24 hr

PFOS-K

98%

-

l<>

-

l ( 5d

(bioluniincsccnce
inhibition)

-

4.9

Duration too short for a
plant test

Hayman et
al. (2021)



Golden brown alga,
Isochrysis galbana

S,U

72 hr

PI ()S

-

20

-

IIC50

(growth inhibition)

-

37.5

Duration too short for a
plant test

Mhadhbi et
al. (2012)



Alga,

Ceratoneis closterium

S,U

72 hr

PF( )S-K
Unkiiou ii

-

-

33

NOEC

(growth)

4.16-
>4.16

4.16

Sediment and other PFAS
present in exposure

Simpson et
al. (2021)



Diatom,

Skeletonema costatum

S, M

% lir

HI ()S-K
8(> 3.20

Only one exposure
concentration

Desjardins
et al.
(2001)



Sandworm (adult),

Perinereis wilsoni

R, M

"d

PFOS-K
Unrcporied

8.1

17.1

36

NOEC

(survival)

0.000028-
>0.000028

0.000028

Only one exposure
concentration

Sakurai et
al. (2017)



Sea urchin (adult),

Glyptocidaris
crenularis

R, U

21 d + 7d
observation

I'lOS-K

8.1

13

30

NOEC

(mortality)

1.0->1.0

1.0

Not a true ELS test
(started with adults);
missing exposure details

Ding et al.
(2015)

Sea urchin (adult),

Glyptocidaris
crenularis

R, U

21 d + 7d
observation

PFOS-K

98%

8.1

13

30

LOEC

(SOD activity)

<0.01-
0.01

0.01

Not a true ELS test
(started with adults);
missing exposure details;
atypical endpoint

Ding et al.
(2015)

H-l


-------
Species (lilVst;iiio>

Method'

Test
Diimlion

( hcmic;il /
Pu ii( \

nil

Temp.
<ฐC)

S:dinil>
(pptl

r.iTcci

( limine
Limits
(NOIX -

i.or.ci

(mป/l.)

Reported
r.lTeel
(one.
(inii/1.)

Deficiencies

Reference



Purple sea urchin
(fertilized eggs),
Paracentrotus lividus

S,U

48 hr

PFOS
98%

-

20

-

EC 5ii

(growth inhibition)

-

20

Duration too short for an
acute test

Mhadhbi et
al. (2012)



Sea urchin (embryo),

Psammechinus
miliaris

R,Ub
(tissue)

72 hr

PFOS-K

>98%

8

19

'1

EC50

(morphological
abnormality)

-

(i 3999ฐ

Interpolated endpoint;
missing some exposure
details

Anselmo et
al. (2011)

Sea urchin (embryo),

Psammechinus
miliaris

R,Ub
(tissue)

16 d

PFOS-K

>98%

8

19

31

NOEC

(morphological
abnormalities, hatch
success,
development)

0.3999-
>0.3999

0.3999ฐ

Duration too short for
chronic test and too long
for acute test

Anselmo et
al. (2011)

Sea urchin (larva),

Psammechinus
miliaris

S,U

85 min.

PFOS-K

>98%

-

l<>

-

[C50

(cellular elllux
pump inhibition)

-

1.399ฐ

Duration too short for a
chronic test and too long
for an acute test, atypical
endpoint

Anselmo et
al. (2012)



Eastern oyster
(33.8mm),

Crassostrea virginica

S, M

96 hr

PF( )S-K
•Jii 4l>"„

~5-

S 1

::

20-21

EC.50

(shell deposition)

-

>3.0

Lack of replication;
atypical endpoint

Drottar and

Krueger

(2000c)

Eastern oyster
(adult, 70-100 mm),
Crassostrea virginica

S,U

48 hr

PI ()S
T",,

" 5

24

:u

LOEC

(cellular lysosomal
damage)

<3-3

3

Atypical endpoint

Aquilina-
Beck et al.
(2020)



Mediterranean mussel

(6.4 cm),

Mytilus

galloprovincialis

R, I

'lid

H< )S
Anal> deal
made

S 1

r.5

34.5

LOEC

(increase
micronuclei nuclear
aberrations in gills
cells)

<2-2

2

Atypical endpoint; missing
some exposure details

Nalbantlar
and Arslan
(2017)



Green mussel (adult),

Perna viridis

R, M

7 d - "d
observation

H< )S-K

-

25

30

EC50

(integrative
genotoxicity)

0.00095-
0.0097

0.033

Duration too short for a
chronic test and too long
for an acute test, atypical
endpoint

Liu et al.
(2014a)

Green mussel (adult),

Perna viridis

R, M

7 d

PFOS-K

98%

-

25

25

MATC

(CAT and SOD
activity)

0.106-
0.968

0.3203

Duration too short for a
chronic test and too long
for an acute test, atypical
endpoint

Liu et al.
(2014b)

H-2


-------
Speeies (lil'eMiiue)

Method-1

Test
l)ii r;il ion

( Ik-iii ic;il /
Pu lit \

pll

Temp.

<ฐC)

Siilinilt
ippl)

r.iTcci

('limine
Limits
(NOIX -

i.or.ci

(inii/l.)

Reported
l-'.ITeel
(one.

• mji/l-l

Delieieneies

Referenee

(iivcii mussel
mm;,

Perna viridis

k. M

7d

H< )S-K

98%



25

25

\l \T(

(rclalivc condition
factor)

(1(1096-
() 106

0.0319

Duration loo short lor a
chronic test and too long
for an acute test, atypical
endpoint

(Liu et al.
2014c)

Green mussel,

Perna viridis

R,M

7 d + 7 d
observation

PFOS-K

98%

8

25

30

MATC

(liemocyte cell
viability)

().()()')(ฆ-
II IIX.

0 0319

Duration too short for a
chronic test and too long
for an acute test, atypical
endpoint

Liu and
Gin (2018)



White sunset shell
(15.0-20.3 mm).

Soletellina alba

S,M

28 d

PFOS-K
Unknown

8

19

30

\oi:c

(survi val)

0.S5-
>0.85

0.85

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)



Bivalve
(8.1-18.9 mm),
Tellina deltoidalis

S,M

28 d

PFOS-K
Unknown

8

l<>

}<>

M \T(

(growth - weight)

0.22-0.28

0.2482

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)



Mysid (juvenile),

Americamysis bahia

S,M

96 hr

PFOS-K
90 4ซr„

S 1-

s:

2"! 5-
25 ^

2d

LC50

-

3.6

Percent recovery of test
substance is low

Drottar and

Krueger

(2000f)



Copepod (adult),

Nitocra spinipes

S,M

10 d

PF( )S-K
Unkimw ii

S 1

21

"!()

NOEC
(reproduction)

2.0->2.0

2.0

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)

Copepod (adult),

Nitocra spinipes

S,M

28 d

PF( )S-K

1 IlkllOWII

S 1

21

.0

NOEC
(survival)

0.48-
>0.48

0.48

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)



Copepod (adult,
female),

Tigriopus japonicus

R, I

lu d

PFOS
Unknot n

-

25

32

MATC

(reproduction)

0.1-0.25

0.1581

Difficult to determine test
methodology

Han et al.
(2015)



Amphipod (adult),

Melita plumulosa

S,M

10 d

H< )S-K
I nknown

-

21

30

EC10

(reproduction)

-

0.9

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)



Smooth sentinel crab
(6-15 mm carapace),

Macrophthalmus sp.

S,M

28 d

PFOS-K
Unknown

8

19

30

NOEC

(survival)

0.85-
>0.85

0.85

Other PFAS measured in
the sediment and water

Simpson et
al. (2021)



H-3


-------
Species (lil'eMiiue)

Method-1

Test
Diinition

( Ik-iii ic;il /
Purilv

pll

Temp.

<ฐC)

Siilinilt
(|)|)ll

r.iTcci

( limine
Limits
(NOIX -

i.or.ci

(inป/l.)

Reported

r.ricci

(one.

(mii/l.)

Delieieiieies

Rel'erenee

Chinese inilleu ercih
ill S'lgi,

Eriocheir sinensis

k, I

:i d

H< >S-k

>98%

8.1

is-::

(i ;

\i vrc

(lolal licmocyte

COUlll)

(i<>1-0.1

0.03162

Duration loo short lor a
chronic test and too long
for an acute test

/hang el al.
(2015)



Mud crab (3cm),
Macrophthalmus
japonicus

R, U

96 hr

PFOS
98%

-

20

30

LC50

-

0.03

Only three exposure
concentrations, atypical
source of organisms

Park et al.
(2015)

Mud crab (3cm),
Macrophthalmus
japonicus

R, U

7 d

PFOS
98%

-

20

30

i.oi:c

(mortality)

<0.(1(11-
0.OO1

0.001

Only three exposure
concentrations, atypical
source of organisms

Park et al.
(2015)



Marine medaka
(embryo, 2 dpf),

Oryzias melastigma

R, U

8 d

PFOS

98%

-

28

30

\\ \T(

(sinus venosus
bulbils arteriosus
distance)

4.0-16

8

Duration too short for a
chronic test and too long
for an acute test, only
three exposure
concentrations

Huang et
al. (2011)

Marine medaka
(embryo, 2 dpf),

Oryzias melastigma

R, U

8 d

PFOS

98%

-

28

30

1.()EC

(decrease heart rate)

<1-1

1

Duration too short for a
chronic test and too long
for an acute test, only
three exposure
concentrations

Huang et
al. (2011)

Marine medaka
(embryo),

Oryzias melastigma

R, M

8 d

PI OS

98%

-

28

30

NOEC

(embryo mortality)

16->16

16

Duration too short for a
chronic test and too long
for an acute test

Fang et al.
(2012)

Marine medaka
(embryo),

Oryzias melastigma

R, M

8 d

H '< )S

W„

-

28

30

LOEC

(malformation)

<1-1

1

Duration too short for a
chronic test and too long
for an acute test

Fang et al.
(2012)

Marine medaka
(embryo),

Oryzias melastigma

R, U

:i d

PFOS
98%

-

:x

30

MATC

(increase hatching
rate, decrease
hatching time)

1.0-4

2.00

Duration too short for a
chronic test, low control
hatch success, only three
exposure concentrations

Wu et al.
(2012)

Marine medaka
(embryo),

Oryzias melastigma

R, U

<21 d + 7d
observation

I'FOS

-

28

30

MATC

(larval survival)

1.0-4

2.00

Duration too short for a
chronic test, low control
hatch success, only three
exposure concentrations

Wu et al.
(2012)



Atlantic Cod
(juvenile),

Gadus morhua

F,Ub
(tissue)

5 d
(1 hr/day)

PFOS
Technical
grade

7.7

10

33.8

NOEC

(survival, growth)

0.2->0.2

0.20

Duration too short for a
chronic test and too long
for an acute test, only two
exposure concentrations

Preus-
Olsen et al.
(2014)



H-4


-------
Species (li Ic-static I

Method-1

Test
Duration

Chemical /
Pu lit \

l'ซซ

l oin p.

(ฐC)

Salinilt

1. ITcct

(limine
Limits
(NOIX -

i.or.ci

(inป/l.)

Ko|)(iitc'(l

i.nvct

(one.
(inป/l.)

Dolicioncios

Reference

l.lackl'Ock I'lsll
(5 mo. old;,

Sebastes schlegelli

R, L

od

H< )S

99%

S 11-

8.2

8.0-12

10

\oi:c

(survival, growth)

I->1

1

Duralion loo sliorl lor a
chronic test and too long
for an acute test, only two
exposure concentrations

Jeon et al.
(2010)

Blackrock fish
(5 mo. old),

Sebastes schlegelli

R, U

6 d

PFOS

99%

8.0-
8.2

8.0-12

17.5

NOEC

(survival, growth)

1- 1

1

Duration too short for a
chronic test and too long
for an acute test, only two
exposure concentrations

Jeon et al.
(2010)

Blackrock fish
(5 mo. old),

Sebastes schlegelli

R, U

6 d

PFOS

99%

8.0-
8.2

8.0-12

25

NOFC

(survival, growth)

1- 1

1

Duration too short for a
chronic test and too long
for an acute test, only two
exposure concentrations

Jeon et al.
(2010)

Blackrock fish
(5 mo. old),

Sebastes schlegelli

R, U

6 d

PFOS

99%

8.0-
8.2

S 11-1 2

U

\<>i:c

(survival, growth)

1->1

1

Duration too short for a
chronic test and too long
for an acute test, only two
exposure concentrations

Jeon et al.
(2010)



Turbot (embryo),

Scophthalmus
maximus
(formerly, Psetta
maxima)

R, U

6 d

PFOS
9X"„

-

18

-

I.C50

-

0.11

Duration too short for a
chronic test and too long
for an acute test

Mhadhbi et
al. (2012)

a S=static, R=renewal, F=flow-through, U=unmeasured. \1 measured. I total. I) dissolved. Diet=dietary, MT=maternaltransfer

b Study did not measure water concentrations, but there are measured concentrations from analysis of the tissue of organisms.

0 Reported in moles converted to gram based on a molecular ueiulil of 5(>u H u mol i l*l-'( )Si. 538.22 g/mol (PFOS-K); 629.4 g/mol (PFOS-TEA).

H-5


-------
H.2 Summary of Acute PFOS Toxicity Studies Used Qualitatively

H.2.1 Saltwater Invertebrates
H.2.1.1 Echinoderms

Mhadhbi et al. (2012) conducted a 48-hour static, unmeasured acute test with PFOS

(98% purity) on the sea urchin, Paracentrotus lividus (a non-North American species). A stock
solution of PFOS was made either with filtered sea water from the Ria of Vigo (Iberian
Peninsula) for low exposure concentrations, or with DMSO for high PI-'OS concentrations (at a
final maximum DMSO concentration of 0.01% (v/v) in the test medium). I lo\\e\ or, they did not
indicate what was considered a high test concentration. If a DMSO was used a sol\ cut control
was also included with the test. Sea urchins embryos were exposed to one of six nominal PFOS
treatments (0.5, 1, 2, 5, 10 and 20 mg/L) Four hundred fertilized eggs (within 30 minutes of
fertilization) were transferred to glass vials containing I mL of test solutions with four
replicates per PFOS treatment and five replicates per control Vials were incubated at 20ฐC in the
dark for 48 hours. At test termination samples were fixed in formalin and 35 larvae per vial was
measured for growth (length) The 4S-hour FCso (growth inhibition) was 20.0 mg/L and was
acceptable for tjnalitali\ e use due to the atypical acute endpoint and short test duration.

Ansel mo et al. (2011) conducted a 10-day renewal toxicity test on the potassium salt of
heptadecalluorooctane sulfonic acid (PFOS, CAS #: 2795-39-3, >98% purity) with the sea
urchin, Psammcchmns miliaris (a non-North American species). Acute endpoints were
extrapolated from this longer test at 72 hpf, after feeding had been initiated. Sea urchins were
collected from the Eastern Scheldt (The Netherlands) and maintained for at least two months
before use. The eggs and sperm used in the study were collected from the freshly dissected
gonads of a single pair of adult individuals. Fertilized eggs were randomly divided into glass
beakers containing 500 mL of fresh seawater or aged artificial seawater (this specific water used

H-6


-------
was not defined), and spiked with the appropriate test concentrations or a solvent control DMSO
at 0.1% v/v. The number of organisms per beaker was defined as ฑ 0.5 larvae per mL, with two
samples of 20 larvae each at each test measurement (16 and 72 hpf). Beakers were held at
19ฑ1ฐC with a photoperiod of 16-hr: 8-hr (light:dark); no other water quality parameters were
reported. A primary stock solution was prepared with DMSO, and nominal test concentrations
were 93, 186, 372 and 743 nMPFOS (or 0.0501, 0.1001, 0.2002. n 3^9 mg/L PFOS, when
converted based on a molecular weight of 538.22 g/mol). While test concentrations were not
measured in the solutions, measured larval tissue concentrations demonstrated a concentration-
response relationship. At 72 hpf, the highest test concentration <">.3999 mg/L PI 'OS-K had no
effect on morphological abnormalities. The 3-day NOEC of 0.3999 mg/L PFOS was acceptable
for qualitative use because of the short test duration and atypical end point.

Anselmo et al. (2012) also exposed I'sammcchiniis miharis to PFOS for 85 minutes
under static, unmeasured conditions Again, sea urchins were collected from the Eastern Scheldt
(The Netherlands) and maintained for at least two months before use. The eggs and sperm used
were collected from the freshly dissected gonads of a single pair of adult individuals. Tests were
initiated with lar\ ae I K-2<) hours post-fertilization (hpf) in the gastrula stage. PFOS stock
solutions were made with DMSO (0.5% v/v) and diluted 20 times in artificial sea water (Instant
Ocean) and 25 |.il of each dilution was added to 225 |il of ASW present in the respective well.
All nominal test concentrations (DMSO control, 0.2, 2, 20, 80, 160 and 320 |iM PFOS, or
0.1076, 1.076, 10.76, 43.06, 86.12, and 172.2 mg/L PFOS when converted based on a molecular
weight of 538.22 g/mol) were tested in triplicate and each test was replicated twice. The only
water quality parameter reported was temperature, 19 ฑ 1ฐC. A total of 12-15 larvae at the
gastrula stage were placed in each well of a 24 well plate. The 24 well plate was then placed on a

H-7


-------
rocking shaker (30 rpm) for 45 minutes exposure, then calcein-AM was added to obtain a final
concentration of 2.5 |iM covered with foil and incubated for an additional 40 minutes. The
accumulation of the cellular fluorescent calcein was a measure of cellular efflux pump inhibition
and calcein accumulation increased in a concentration-dependent manner. The 172.2 mg/L PFOS
concentration was noted as being acutely toxic to the sea urchin and not used in the curve fitting.
No additional details were provided about the toxic response. The S5-minute ICso based on
cellular efflux pump inhibition was 1.399 mg/L PFOS-K. The value was acceptable for
qualitative use only, due to the atypical endpoint and exposure duration.

H.2.1.2 Mollusks

Drottar and Krueger (2000c) reported the results of a 96-hour static, measured test on
the effects of PFOS (potassium salt, 90.4l)"(> purity) on the eastern oyster, Crassostrea virginica.
The GLP test was conducted at Wildlife International, Ltd. in Easton, MD in October, 1998,
using a protocol based on procedures outlined in I S EPA, OPPTS Number 850.1025; and
ASTM E729-88a I 11>XK) ()\ slers (27 X-41.5 mm) were obtained from P. Cummins Oyster
Company, Baltimore. Ml) and held lor 12 days in the same water used for testing before
exposure. The unfiltered natural seawater used for testing was collected at Indian River Inlet,
Delaware and diluted to a salinity of approximately 20 ppt with well water. This water was
supplemented with an algal suspension continuously during holding and testing to supplement
the oyster diet and enhance their condition and growth. Test chambers were 52 L polyethylene
tanks containing 40 I. of consistently aerated test solutions. A primary stock solution was
prepared in dilution water at 9.1 mg/L. It was mixed for -24 hours prior to use. After mixing, the
primary stock was proportionally diluted with dilution water to prepare the four additional test
concentrations. The test employed one replicate of 20 oysters each in five measured test
concentrations plus a negative control. Nominal concentrations were 0 (negative control), 1.2,

H-8


-------
2.0, 3.3, 5.5 and 9.1 mg/L. Mean measured concentrations were less <0.115 mg/L, 0.36, 0.40,
1.3, 1.9, and 3.0 mg/L, respectively. There was poor percent recovery across treatments (only 29-
39% of nominal). Analyses of test solutions were performed at Wildlife International, Ltd. using
high performance liquid chromatography with mass spectrometric detection (HPLC/MS).
Dissolved oxygen ranged from 6.1-7.7 mg/L; pH ranged from 7.5-8.1; test temperature ranged
from 22.0-22.7ฐC across all treatments. There was no mortality in I he oysters in the negative
control and across all treatments and all appeared healthy and normal. The locus of the acute
exposure was shell growth. There was significant inhibition of shell growth in the highest test
concentration, but that concentration failed to reduce at least a 50% inhibition in growth. Thus,
the 96-hour ECso was >3.0 mg/L. This acute value was acceptable for qualitative use because of
the lack of replication, atypical endpoint lor an acute exposure and (he poor percent recovery of
the test material in the test exposures.

Aquilinsi-liork ol ;il. (2020) conducted slum-term, sublethal exposures of two and seven
days to examine the effects of a technical mixture of PFOS (linear and branched isomers) in
adult Eastern oysters. ('/
-------
PFOS (Santa Cruz Biotechnology, CAS # 1763-23-1, purity >97%) was prepared in deionized
water at 10,000 mg/L. Individual oysters were exposed to 1 L of PFOS treatment in seawater at 0
(control), 3,30, and 300 mg/L in a glass beaker. After 48 hours, oysters were shucked and whole
tissue wet weight (ww) was recorded. The hepatopancreas (HP) was dissected, weighed, and
divided into three sections: two of which were frozen in liquid nitrogen for downstream
biomarker analysis, and the third section was processed immediately lor lysosomal
destabilization. PFOS 48-hour exposure experiments and biomarker analyses were repeated three
times with four or six replicates per treatment group, twice in the fall and once in the winter. A
combined total of 16 oysters per treatment group were tested Water quality para meters were
recorded at the beginning and end of the treatment and were within acceptable test conditions
(mean values ฑ standard deviation: Temp 24 ^ i)4ฐC:D.O 4 3 1 2 2 mg/L; Salinity: 20 ฑ 0.2
ppt; and pH: 7.5 ฑ 0.3). Biomarker analysis (lysosomal destabilization, lipid peroxidation, and
glutathione assays) in oyster tissue ie\ ealed no significant damage to lipid membranes or the
glutathione phase II enzyme system up to 3<)i) mg/L PFOS; however, significant cellular
lysosomal damage was oltsei\ed at 3 mu I. (I.OF.C; lowest treatment level). Apical
measurements (sui \ i\ al. growth) were not reported or measured during either test. Thus, the 48-
hour I .()!ฆ(' for cellular lysosomal damage was 3 mg/L. The value is considered qualitatively
based on atypical test endpoints.

H.2.1.3 Crustaceans

Drottar and Krueger (2000f) conducted a static, measured 96-hour acute toxicity test

with PFOS (potassium salt, 90.49% purity) on the mysid, Americamysis bahia (formerly known

as Mysidopsis bahia). The GLP test was conducted at Wildlife International, Ltd. in Easton, MD

in March 1999, using a protocol based on procedures outlined in U.S. EPA, OPPTS Number

850.1035; and ASTM E729-88a (1988). Mysids were obtained as juveniles from in-house

H-10


-------
cultures and were fed brine shrimp daily to prevent cannibalism. The unfiltered natural seawater
used for testing was collected at Indian River Inlet, Delaware and diluted to a salinity of
approximately 20 ppt with well water. Test chambers were 2 L polyethylene buckets containing
approximately 1 L test solutions. A primary stock solution was prepared in dilution water at 8.2
mg/L and was mixed for -22 hours prior to use. After mixing, the primary stock was
proportionally diluted with dilution water to prepare the four additional test concentrations. The
test employed two replicates of 10 mysids each in five measured test concern rations plus a
negative control. Nominal concentrations were 0 (negative control), 1.1. I Si. 3 n. 4.9 and 8.2
mg/L. Mean measured concentrations were less <0.1 15 mg/I.. 0.57, 1.1, 1.9, 3.0, and 5.4 mg/L,
respectively. There was poor test material percent recovery across treatments (at 96 hours only
36-71% of nominal). Analyses of test solutions were performed at Wildlife International, Ltd.
using high performance liquid chromatography with mass specliometric detection (HPLC/MS).
Dissolved oxygen ranged from (•> 0-8.8 mg/L; pH ranged from 8.1-8.2; test temperature ranged
from 23.5-25.4cC across all treatments There was no mortality in the negative control and the
two lowest test concentrations and mysids appeared healthy and normal. The 96-hour LCso was
3.6 mg'I. PFOS and was acceptable for qualitative use

H.2.2 Stillwater Fish
H. 2.2.1 (iac /lis morhua

Preus-Olscn ol ;il. (2014) conducted a flow-through pulsed exposure with PFOS

(technical grade) on the Atlantic cod (Gadus morhua). Juvenile fish were exposed to one of three

nominal PFOS treatments (0, 0.1 and 0.2 mg/L PFOS) for one hour per day for five consecutive

days and then further challenged to three different CO2 levels [normocapnia, moderate (0.3%)

and high (0.9%)]. The focus of this assessment was on the normal CO2 levels, normocapnia.

Juvenile fish were purchased from Atlantic Cod Juveniles (Risa, Norway) and acclimated to

H-ll


-------
laboratory conditions (10ฐC, 12-hr: 12-hr light:dark photoperiod) in flow-through tanks with
circulating seawater for 14 days. After acclimation, 120 fish were transferred to each tank and
exposed to one of three concentrations of PFOS, afterwards groups of 40 fish per treatment were
exposed to one of the three CO2 conditions. At 3, 6 and 9 days after CO2 exposure, five fish were
sampled for PFOS burden, steroid hormone and gene expression. Throughout the exposure pH
was maintained at 7.7, and salinity was 33.8 ppt in the normocapnia condition. While water
PFOS concentrations were not measured, tissue concentration in the sampled lish increased with
increasing PFOS concentrations. Mortality and growth data were not pro\ ided by the study
authors, but they noted that no significant differences in survival and growth maintenance
(length, weight, condition factor) between exposure groups and sampling days were observed.
The 5-day NOEC, 0.2 mg/L, based on sui \ i\ al and growth, was acceptable for qualitative use
only, due to the pulsed exposure regime.

H. 2.2.2 Psetta maxima

Mhadhbi et al. (2012) conducted a b-day renewal, unmeasured acute test with PFOS

(98% purity) on the turboi. Scophilialmiis maximns, (formerly, Psetta maxima; a non-North

American species) A slock solution of PFOS was made either with filtered sea water from the

Ria of Yiuo (Iberian Peninsula) for low exposure concentrations, or with DMSO for high PFOS

concentrations (a final maximum DMSO concentration of 0.01% (v/v) in the test medium).

However, they did not indicate what was considered a high test concentration. If a DMSO was

used a solvent control was also included. Fish were exposed to one of ten nominal PFOS

treatments (0.015, 0.03, 0.075, 0.15, 0.3, 0.325, 0.6, 1.2, 2.5 and 5 mg/L). Turbot eyed eggs from

a single stock of adults were supplied by a nearby fish hatchery (PESCANOVA Insuina) and

acclimated to laboratory conditions before use. At 72 hours post-fertilization (hpf), the floating

fertilized eggs were collected and the non-fertilized eggs at the bottom discarded. Embryos that

H-12


-------
had reached the blastula stage were used for testing. Fifty normal embryos were added to glass
beakers containing 500 mL of test solution. Each treatment had four replicates and were
incubated in the dark for six days at 18ฐC with no food or aeration provided. Dead embryos and
larvae were removed daily. Endpoints included dead embryos, malformation, hatch success at 48
hours and larvae survival (missing heartbeat and a non-detached tail) at six days. The 6-day LCso
of 0.11 mg/L PFOS was acceptable for qualitative use because of the atypical acute test duration.

H.3 Summary of Chronic PFOS Toxicity Studies Usetl Qualitatively

H.3.1 Saltwater Invertebrates
H. 3.1.1 Worms

Sakurai et al. (2017) report the results of a 7-day renew al, measured test of PFOS (purity
not provided) with the marine sandworm. Pcrincreis wilsoni. The focus on the study was on the
uptake and depuration kinetics of PFOS and was not a toxicity test. Sandworms were obtained
from as aquaculture farm and acclimated to laboratory conditions for four days without feeding.
About eight sandworms were held in polypropylene containers with holes in bottom and packed
with gravel [from hollom to top with approximately 25 mm high layer of gravel (6-8 mm), 0.5
mm mesh polyethylene netting, and approximately 70 mm high layer of gravel, (3-6 mm)].
Dilution water (filtered natural seawaler) llowed freely through the gravel and the gravel was too
large to he ingested There were 26 replicates for the exposure treatment and eight replicates for
the control. A methanol stock solution of PFOS added to seawater and diluted to achieve 32 ng/L
(or 0.000032 mg/L) PI'OS for the exposure treatment. The control received a methanol-spiked
seawater solution (0.35 mg/L), to mimic the same methanol concentration as the PFOS
treatment. Organisms were not fed during the exposure period and the lighting cycle was
approximately 10 hours light and 14 hours dark. Solutions were renewed daily with water levels
in the tanks being varied to mimic tidal action. Measured PFOS concentrations (0.000028 mg/L)

H-13


-------
were similar to nominal concentrations during the exposure period. Water quality conditions

were monitored daily: 91% D.O. saturation, 8.1 pH, 17.1ฐC water temperature, and 36%o

salinity. Mortality was 2.2% and 6.8% for the exposure and control treatment respectively, and

growth did not differ between treatments. As this study was focused on uptake and depuration,

the toxicity endpoints were not statistically analyzed in the paper. Since there was only one

exposure treatment and an atypical exposure period, the NOEC ('< > ooooZS nig/L PFOS) based on

survival was acceptable for qualitative use only.

H.3.1.2 Echinoderms

Ding et al. (2015) conducted aPFOS-K (lW(. puril\) 21-day renewal, unmeasured

toxicity test with the sea urchin, Glyptocidaris crcnnlaris (a non-\orth American species).

Adults, about 20 months old, were purchased from Dalian I lailiao fishery Company and held in

a recirculating aquaculture system for two months before use in testing. Healthy individuals, 5

cm in diameter and SO 3 u wet weight. were exposed to one olTour nominal PFOS

concentrations (< >. < > < > I. < > I and I mu I.) The number of sea urchins per treatment was not

identified, nor any details about treatment replication. Sea urchins were held in 5 L plastic

barrels with .1 I. of test solutions lor 21 days at a temperature of 13 ฑ 1ฐC, salinity of 30 ฑ 2 psu,

pH about S I and a pholoperiod of 12 Test solutions were 50% renewed every two days and sea

urchins were fed / ammaria /a/>onica every other day. After the 21-day exposure period test

organisms were then held for another seven days under the same test conditions. Mortality and

sublethal toxicity effects were observed daily, enzymatic activity was measured on days 1, 3, 5,

7, 14, 21, and 28 and DNA methylation was measured on days 7, 13,21, and 28. No mortality

was observed throughout the exposure and observation period in any treatment. However, the

SOD activity in the coelomic fluid decreased significantly in the lowest test treatment (0.01

mg/L PFOS) when compared to the control at 21 days. SOD activities increased in the

H-14


-------
depuration period in all PFOS treatments, but were still significantly lower than the controls. The
21-day + 7-day observation LOEC based on SOD activity of 0.01 mg/L and the NOEC based on
morality of 1 mg/L were both acceptable for qualitative use. This study was qualitative because
of the lack of details about the number of organisms, the short test duration, and starting age of
organisms.

Anselmo et al. (2011) performed a 16-day renewal, unmeasured PFOS-K toxicity test
(>98% purity) with embryos of the sea urchin, Psammechinus miliaris (a n on-North American
species). Sea urchins were collected from the Eastern Scheldt (The Netherlands) and maintained
for at least two months before use. The eggs and sperm used were collected from llie freshly
dissected gonads of a single pair of adult individuals. Fertilized eggs were randomly divided into
glass beakers containing 500 mL of fresh seawater or aged artificial seawater (this specific water
used was not defined), and spiked with the appropriate test concentrations or a solvent control
(DMSO at 0,1ฐ/.. \ \ ) The number of organisms per beaker was defined as ฑ 0.5 larvae per mL,
with two samples of 2<) lar\ ae each at each test measurement (72 hpf and 16 dpf). Mollusks

Nalhanllar and Arslan (2U17) conducted a 30-day renewal, unmeasured toxicity test on
PFOS (analytical grade) with the Mediterranean mussel, Mytilusgalloprovincialis. Mussels were
collected from an area free of wastewater inputs off the Turkish Aegean coast. They were
acclimated to laboratory conditions over six days in artificial seawater (5.1 ฑ0.1 mg/L dissolved
oxygen, 8.1 ฑ 0.1 pi I and 34 5 ฑ 0.2 psu), which was continuously aerated at a temperature of
17.5 ฑ 1ฐC. A total of 10 mussels per replicate, with a mean body length of 6.4 cm, were added
to glass aquaria with four replicates per treatment. PFOS was dissolved in DMSO and five test
concentrations were selected based on 1/10 and <1/10 the 96-hour LCso (2, 3, 4, 5 and 6 mg/L
PFOS). The amount of DMSO used was not provided, but experiments included both a dilution

H-15


-------
water control and solvent control. Mussels were exposed to PFOS for 30 days, with feeding
(Chlorella sp.), and water renewals every two days. At the end of the exposure period gill and
hemolymph cell samples were taken and examined for nuclear aberrations (frequency of
micronuclei and binuclei). Survival of mussels were not reported for the controls or any
treatment. The most sensitive endpoint was an increase in the frequency of micronuclei
aberrations in the gill cells with all PFOS treatments containing significantly more aberrations
than the control and solvent control. The 30-day LOEC (based on micronuclei nuclear
aberrations in gills cells) was 2 mg/L PFOS and was acceptable for qualitaii\ e use, due to the
atypical exposure endpoint.

Liu et al. (2014a, b, c) and Liu and Gin (2018) conducted a series of 7-day renewal,
measured experiments with perfluorooctanesullbnate potassium sail (PFOS-K, 98% purity) on
the green mussel, Perna viridis (a non-North American species). All of these studies employed a
similar test design. In.il with each publication providing different level of test details. In Liu et al.
(2014a), green mussels were obtained from a local fish farm and acclimated to laboratory
conditions prior to PI OS exposure Adult organisms were exposed in 70 L polypropylene tanks
in artificial seawater at a temperature of 25ฐC and at salinity of 30 ppt. Mussels were exposed to
one of fi\e nominal PFOS concentrations (0.0001, 0.001, 0.010, 0.1 and 1.0 mg/L) or control.
Each tank contained (->0-65 mussels, with two tanks per exposure concentration or control.

During exposures mussels were fed with microalgae and each tank was cleaned and refilled
every two days. After seven days of exposure and seven days of depuration, various biomarkers
were measured. The ECso (integrative genotoxicity) was reported as 0.033 mg/L PFOS and was
based on three genotoxic endpoints (DNA fragmentation and single strand breaks (comet assay),

H-16


-------
chromosomal breaks (micronucleus test) and apoptosis (DNA diffusion assay). The atypical
duration and endpoint resulted in the value to be considered acceptable for qualitative use only.

In Liu et al. (2014b), the oxidative damage of PFOS-K (98% purity) to green mussels
was assessed after seven days under similar conditions as Liu et al. (2104a). Green mussels were
obtained from a local fish farm in Singapore and acclimated to laboratory conditions for one
week prior to exposures. Organisms (60-65 mm) were exposed in polypropylene tanks
containing artificial seawater at a temperature of 25ฐC and at salinity of 25 ppl Mussels were
exposed to one of six nominal PFOS concentrations (0.0001, 0.001, 0.010, 0.1. I n and 10.0
mg/L) or control. Nominal concentrations were similar to measured concentration (t) 00012,
0.0011, 0.0096, 0.106, 0.968 and 10.156 mg/L, respectively) and no PFOS was detected in the
controls. Each treatment was replicated with an unknown number of mussels per replicate.

Again, during the exposure mussels were fed with micioaluae and each tank was cleaned and
refilled every two days The most sensitive parameters to PI OS were activation of antioxidant
enzymes (catalase |( \T| and superoxide dismutase [SOD]), which is an adaptive response to the
excessive reacti\ e oxygen species Significant effects were observed at 0.968 mg/L PFOS, but
notal ii I no mg I. The 7-day M.VI'C (CAT and SOD activity) was 0.3203 mg/L and was
acceptable lor qualitative use based on the atypical endpoint and duration.

Liu ol al. (2014c) utilized a similar test design and the same nominal PFOS
concentrations as I .in et al (2014a). However, in this study the test tanks were reported as 50 L
in size and containing 40 mussels per each exposure tank, and the 7-day exposure was not
followed by a 7-day depuration phase (similar to Liu et al. 2104b). The 7-day NOEC and LOEC
based on the relative condition factor (relationship between weight and length) was 0.0096 and

H-17


-------
0.106 mg/L, respectively. The 7-day MATC of 0.0319 mg/L was acceptable for qualitative use
because of the atypical test duration.

Liu and Gin (2018) employed the same test design and measured PFOS concentrations
as Liu et al. (2014a). The most sensitive biomarker endpoint reported was hemocyte cell viability
with a reported NOEC and LOEC of 0.0096 and 0.106 mg/L, respectively. Again, the MATC of
0.0319 mg/L PFOS was acceptable for qualitative use because of the atypical test duration.

Simpson et al. (2021) evaluated the chronic effects of perlluorooctane sulfonic acid
potassium salt (PFOS-K, purchased from Sigma Aldrich) on the deposit-feeding hi valve, Tellina
deltoidalis, (8.1-18.9 mm shell length) using 28-day growth and survival tests. The bivalves were
collected from the estuarine mud flats of the Lane Cove River, adjacent to Boronia Park, Hunters
Hill, New South Wales (NSW). Clean seawater (salinity 33 2 ppi) for culturing was sourced
from the southeast coast of New South Wales (\S\\ ). Australia, and clean sediments were from
Bonnet Bay, NSW Organisms were acclimated l<> days before starting experiments. Bivalves
were randomly assigned to a specific replicate of each treatment, labelled, weighed (wet mass)
and sized (dimensions In scanning as below ) and then distributed to the corresponding test
vessel The test was ami prised of three treatments plus a solvent control. The tests were
undertaken with Iaboratory-controlled conditions (dissolved oxygen >85% saturation, pH
8.0ฑ0.1, salinity 3d 0.5 ppt, temperature 19ฑ1ฐC and ammonia <10 mg NH3-N/L) over 28 days.
Sediments (PFOS-spiked and control) were re-homogenized immediately prior to being added to
test vials (80 g whole sediment per 250 mL polypropylene vial) and then equilibrated for 48 h
before tests were started. Bivalves were exposed in triplicate and fed once a week during the
exposure. At the end of the exposure the sediments were gently sieved to isolate bivalves. The
differences in wet mass and size (length and surface area) of bivalves measured at start and

H-18


-------
completion of tests (for surviving organisms) were used to assess growth. Bivalve growth was
expressed as average percent change in wet mass, shell length and surface area. There was no
effect on survival of the bivalve in the treatment with the highest concentration of PFOS in the
overlying water of 0.28 mg/L PFOS (NOEC). The author-reported 28-d growth (wet mass; most
sensitive endpoint) NOEC and LOEC were 0.22 and 0.28 mg/L PFOS, resulting in a 28-d growth
MATC of 0.2482 mg/L. Given the presence of other PFAS measured in the sediments and
overlying water at very low concentrations, this study was considered a mixture. The other PFAS
made up a much smaller percentage of the total PFAS and were generally in low concentration,
and thus, the result from the study is considered qualitative use only.

Simpson et al. (2021) evaluated the chronic effects of perfluorooctane sulfonic acid
potassium salt (PFOS-K, purchased from Sigma Aldrich) on the deposit-feeding bivalve,
Soletellina alba (15.0-20.3 mm shell length) using a 2S-day sur\ ival test. The bivalves were
collected from the estuarine mud llats of the Lane Cove Ri\ ei\ adjacent to Boronia Park, Hunters
Hill, New South Wales (NSW ) Clean seawater (salinity 33 ฑ 2 ppt) for culturing was sourced
from the southeast coast of New South Wales (NSW), Australia, and clean sediments were from
Bonnet Bay. NSW Organisms were acclimated 10 days before starting experiments. Bivalves
were randomly assigned to a specific replicate of each treatment, labelled, and then distributed to
the corresponding test vessel The test was comprised of a single treatment plus a solvent control
under laboratory-controlled conditions (dissolved oxygen >85% saturation, pH 8.0 ฑ 0.1, salinity
30 ฑ 0.5 ppt, temperature 19 ฑ 1ฐC and ammonia <10 mg NH3-N/L) over 28 days. Sediments
(PFOS-spiked and control) were re-homogenized immediately prior to use. The exposure
occurred in 2 L HDPE wide-mouth bottles (Nalgene) each containing approximately 500 g of
sediment and 1 L of overlying seawater. Bivalves were exposed in triplicate and fed once a week

H-19


-------
during the exposure. At the end of the exposure bivalves were isolated gently from the
sediments. Survival of bivalves in each treatment were calculated as the mean mortality of each
replicate while dead organisms were determined as no movement in 1-2 min by gentle touch.
There was no effect on survival of the bivalve in the treatment with a measured overlying water
concentration of 0.85 mg/L PFOS (NOEC). Given the presence of other PFAS measured in the
sediments and overlying water at very low concentrations, this study was considered a mixture.
The other PFAS made up a much smaller percentage of the total lJl.\S and were generally in low
concentration, and thus, the result from the study is considered qualitative use only.

Aquilina-Beck et al. (2020) conducted short-term, sublethal exposures of two and seven
days to examine the effects of a technical mixture of PFOS (linear and branched isomers) in
adult eastern oysters, Crassostrea virgin ica Adult oysters (7o-1 no nun in length) were collected
in the fall and winter seasons from a reference site commonly used as a control site at the mouth
of Leadenwah Creek at North l-disto River on Wadmalaw Island, SC. Seawater for testing was
collected from Charleston I larhor estuary, filtered (5 (jxn), UV-sterilized, activated carbon
filtered (5 [im). and diluted with deionixed water to 20 ppt salinity. Cleaned oysters were placed
in a controlled laboratory aquatic recirculating system to acclimate for 14 days at 25ฐC, 20 ppt
salinity, and 10-hour light: 8-hour dark cycle. Oysters were fed 10 mL of commercial shellfish
diet daily until the day of exposure. Oyster length and width were measured and recorded before
exposures. For the 7-day hiouptake and depuration exposure, 24 oysters collected in the fall were
kept in individual glass beakers containing 1-L volume of PFOS (Santa Cruz Biotechnology,
CAS 1763-23-1, purity >97%) concentrations at 0 (control), 0.3, and 3 mg/L. Eight replicates
were included for each treatment and control group. Each exposure concentration and control
seawater were replaced daily for seven days. Mortality was assessed daily, and water-quality

H-20


-------
measurements were recorded. Water-quality parameters for all treatments during exposures were
within acceptable test conditions (mean values ฑ standard deviation: Temp.: 25.3 ฑ 0.5ฐC; D.O.:
5.1 ฑ 1.4 mg/L; Salinity: 20.0 ฑ 0.4 ppt; and pH: 7.7 ฑ 0.3). After seven days of exposure, half of
the oysters from each treatment group were shucked and whole tissue was flash frozen in liquid
nitrogen. The remaining four oysters were placed into clean 20 ppt seawater and allowed to
depurate for 48 hours. At the end of 48-hour depuration, oysters u ere shucked, flash frozen in
liquid nitrogen, and stored at -80ฐC. All 24 oysters were processed lor chemical analysis.
Additionally, 1 mL water samples were taken for PFOS analysis from each treatment group on
day two of the seven day exposure and 24 hours later (day three) to represent a I -day dose. The
BAF calculated on day seven of the exposure revealed thai oysters had incorporated 50 times and
116 times the level of PFOS in the 0.3 and 3 mu I. PFOS treatments, respectively. Since the
study is less than 28 days and the authors did not demonstrate that steady-state was achieved, the
BAFs are considered qualitati\ e only

H.3.1.3 Crustaceans

A 21-day renewal, unmeasured PI 'OS-K (>98% purity) toxicity test with the Chinese

mitten crah. I.nochcir sinensis (a non-North American species) was evaluated by Zhang et al.

(2015) I leal thy crabs (1 l.SlJ 1.2 g) were purchased from a commercial crab hatchery on

Chongminu Island (Shanghai. China) and acclimated to laboratory conditions for two weeks in

120 L plastic tanks I )nri nu acclimation and testing pre-aerated municipal tap water was kept at

18-22ฐC, 8.3-8.6 pH, 0.3%o salinity, and >6.5 mg/L dissolved oxygen. After acclimation, 20

crabs per replicate were transferred to tanks and exposed to five nominal concentrations (0, 0.01,

0.1, 1.0 and 10 mg/L PFOS-K). There were three replicates for each treatment with a total of 60

crabs per treatment. Three crabs were randomly sampled from each replicate at day 0, 1, 4, 7, 14

and 21 and total hemocyte count, lysozyme activity levels, phenol oxidase activity levels, acid

H-21


-------
phosphatase activity levels, alkaline phosphatase activity levels, lysozyme gene mRNA
expression levels, hepatopancreas-specific C-type lectin gene expression levels and
prophenoloxidase activating factor in the hepatopancreas mRNA expression levels were
measured in the hemolymph. Survival on control and treatments was not provided. The total
hemocyte count was the most sensitive endpoint with a reported NOEC and LOEC of 0.01 and
0.1 mg/L, respectively. The 21-day MATC, based on total hemocyte count, of 0.03162 mg/L
PFOS was acceptable for qualitative use because of the short test duration and atypical endpoint.

Park et al. (2015) conducted a 7-day renewal, unmeasured toxicity test on PFOS (98%
purity) with the mud crab, Macrophthalmus japonicus (a non-North American species). The test
organisms (average length of 3 cm) were purchased from the Yeosu Aquatic Products Market
(Jeonnam, Korea) and acclimated for at least se\ en days to laboratory conditions. Crabs were
acclimated to a temperature of 20 ฑ 1ฐC, salinity of3<)"(1(). and an alternating 12-hrs: 12-hrs
light:dark schedule with daily water changes and constant aeration. Adult, non-damaged
specimens were then used lor testing in glass containers and natural seawater (source of dilution
water not provided) A primary stock solution was prepared with analytical grade acetone. In
acetone controls. sol\ ent was added at an actual test concentration of <0.1% acetone. Fifteen
crabs per replicate were exposed to one of three PFOS concentrations (0.001, 0.010 or 0.03
mg/L) or a sol\ ent control luich treatment was replicated three times. Exposures lasted seven
days, but survival was also recorded at 96 hours. The author did not calculate an LCso, but at 96
hours there was 64% survival in the highest test concentration. Crabs in the solvent control had
96% survival over the entire test duration. The 96-hour LCso was >0.03 mg/L and was acceptable
for qualitative use since no point estimate was calculated and only three exposures were used.

H-22


-------
The main focus of the study was to determine the effects of PFOS on the molecular transcription
of antioxidant and detoxification signaling.

Simpson et al. (2021) evaluated the chronic effects of perfluorooctane sulfonic acid
potassium salt (PFOS-K, purchased from Sigma Aldrich) on the smooth sentinel crab,
Macrophthalmus sp., (carapace width of 6-15 mm) using a 28-day survival test. The crabs were
collected from the estuarine mud flats of the Lane Cove River, adjacent to Boronia Park, Hunters
Hill, New South Wales (NSW). Clean seawater (salinity 33ฑ2 ppt) lor ailiuiing was sourced
from the southeast coast of New South Wales (NSW), Australia, and clean sediments were from
Bonnet Bay, NSW. Organisms were acclimated 10 days before starting experiments and were
held in trays (16 x 12 cm) with approximately 3 cm depth of sediment and 4 cm of seawater,
under laboratory-controlled conditions. Crabs were randomly assigned to a specific replicate of
each treatment, labelled, and then distributed to the corresponding test vessel. The test was
comprised of a single treatment plus a solvent control under laboratory-controlled conditions
(dissolved oxygen 85" <> saturation, pi I Si) i) I. salinity 30 ฑ 0.5 ppt, temperature 19 ฑ 1ฐC and
ammonia <10 mg \ I h-\ I.) o\ er 28 days Sedi ments (PFOS-spiked and control) were re-
homogenized immediately prior to use The exposure occurred in 2-L HDPE wide-mouth bottles
(Nalgene) each containing approximately 500 g of sediment and 1 L of overlying seawater.

Crabs were exposed in triplicate and fed once a week during the exposure. Survival of crabs in
each treatment were calculated as the mean mortality of each replicate while dead organisms
were determined as no movement in 1-2 min by gentle touch. There was no effect on survival of
the bivalve in the treatment with a measured overlying water concentration of 0.85 mg/L PFOS
(NOEC). Given the presence of other PFAS measured in the sediments and overlying water at
very low concentrations, this study was considered a mixture. The other PFAS made up a much

H-23


-------
smaller percentage of the total PFAS and were generally in low concentration, and thus, the
result from the study is considered qualitative use only.

Simpson et al. (2021) evaluated the chronic effects of perfluorooctane sulfonic acid
potassium salt (PFOS-K, purchased from Sigma Aldrich) on the epibenthic amphipod Melila
plumulosa from previously-established cultures. Clean seawater (salinity 33 ฑ 2 ppt) for
culturing was sourced from the southeast coast of New South Wales (NSW), Australia, and clean
sediments were from Bonnet Bay, NSW. A total of sixteen different PI 'OS-spiked sediments
were prepared, each having varying PFOS concentrations and sediment properties, by adding a
mass of PFOS-K into a 250 mL HDPE container with 5-l<~> ml. of ethanol added to dissolve it.
Filtered NSW seawater was added (10 mL/mL methanol) and PFOS-K methanol/seawater
suspension immediately poured onto the lop of llie wet sediment and homogenized. The PFOS
concentrations in all sediments and porewalers of selected sediments were determined prior to
toxicity testing. The non-spiked controls were prepared in the same manner. PFOS treatments
used for toxicity tests were prepared from dilutions of the PFOS-spiked sediments with clean
sediment(95% (->3 iim) and sand (<>"<> 03 inn) that were then equilibrated for a further 1-2
weeks before testing A I D-day amphipod survival-reproduction test was used to measure adult
(male and female) survival and the number of embryos and <1 day old juveniles in the second
brood. Sediments were re-homogenized immediately prior to being added to test vials (40 g
whole sediment and 2<><> ml. seawater per 250 mL vial) and then equilibrated for 48 hours before
tests were started. Each sediment tested had four replicates. Filtered seawater (<0.45 |iin, 30%o)
was added, and each beaker was incubated at 21ฐC with aeration for 1-week prior to
commencing the tests. The overlying water was not changed prior to tests commencing and was
not renewed during the tests. Six gravid females (gravid for <24 hour) and six males were

H-24


-------
randomly assigned to each beaker. Treatments were fed at a rate of 0.5 mg Sera Micron fish food
per amphipod twice a week. The sediments were renewed after five days. On day 10, the adults
were removed and counted (survival), and the number of embryos per female was counted by
microscopy. The total number of embryos and 
-------
of New South Wales (NSW), Australia, and clean sediments were from Bonnet Bay, NSW. A
total of six different PFOS-spiked sediments were prepared, each having varying PFOS
concentrations and sediment properties, by adding a mass of PFOS-K into a 250 mL HDPE
container with 5-10 mL of ethanol added to dissolve it. Filtered NSW seawater was added (10
mL/mL methanol) and PFOS-K methanol/seawater suspension immediately poured onto the top
of the wet sediment and homogenized. The PFOS concentrations in all sediments and porewaters
of selected sediments were determined prior to toxicity testing. The non-spiked controls were
prepared in the same manner. PFOS treatments used for toxicity tests were prepared from
dilutions of the PFOS-spiked sediments with clean sediment (95% <63 (j,m) and sand (0% <63
(j,m) that were then equilibrated for a further 1-2 weeks before testing. A 10-day copepod
survival-reproduction test was used to measure adult survival and reproduction and survival of
offspring. The females of N. spinipes are ileroparous. producing several broods after only one
mating encounter, and tests were initiated with gra\ id females collected directly from the
cultures. Sediments were homogenized, but not sieved, immediately prior to being added to test
vials (0.5 g sediment per I cm diameter l<> ml. \ ial, with five to six replicates per sediment).
Autocla\ ed filtered seawater (pi I S I. salinity 30%o) was added, and each vial was incubated at
21ฐC o\ ernight to allow sediments to settle. The following day, five gravid females (three to five
weeks old) were randomly assigned to each vial. Treatments were fed twice a week. The tests
were static. After in days, the total offspring (total nauplii, first juvenile life-stage, and
copepodite, second life-stage) in each vial was recorded by microscopy. Subsamples of overlying
seawater were taken on day one, and at three or four other times during the tests were combined
to analyze the time-weighted average concentration of PFOS for two of the five spiked-sediment
tests. A 28-day copepod (survival only) test was also run using the same conditions as the

H-26


-------
standardized 10-day copepod test but was conducted in 50 mL centrifuge vials containing 2 g of
sediment and 48 mL of autoclaved filtered seawater and only one nominal test concentration, 50
mg/kg PFOS spiked sediment. On day 28, the sediment was gently sieved (44 (j,m) and the
surviving adults were counted. The author-reported 10-d reproduction (progeny count) NOEC
was 2.0 mg/L, and the 28-d survival NOEC was 0.48 mg/L. Given the presence of other PFAS
measured in the sediments and overlying water at very low concentrations, this study was
considered a mixture. The other PFAS made up a much smaller percentage of the total PFAS and
were generally in low concentration, and thus, the result from the study is considered qualitative
use only.

H.3.2 Saltwater Fish

H. 3.2.1 Oryzias melastigma

A series of PFOS experiments with murine medaka embryos were conducted by Huang et

al. (2011), Fangetal. (2012) andWuetal. (2012). In all of these experiments medaka were

exposed to the same nominal test concentrations and under similar water quality parameters.

Huang et al. (2011) conducted an S-day renewal, unmeasured PFOS (98% purity) toxicity test

with Orynus mc/asiiifina embryos I 'ei tili/ed eggs were acclimated in artificial seawater for two

days to laboratory conditions (salinity ppt, 28 ฑ 1ฐC, and 14-hr:10-hr light:dark

photoperiod) PI OS was dissolved in DMSO and the control and each exposure group contained

0.1% DMSO. l-mlnyos (2 dpi") were exposed to four nominal concentrations (solvent control, 1,

4 and 16 mg/L PFOS). Three replicates of 70 embryos each were held in 90 mm petri dishes

containing 20 mL of solutions for eight days with solutions renewed daily. Twelve embryos from

each replicate were collected on day 2, 4, 6 and 8 and the cardiac morphology and heart rates

were recorded. The distance between the sinus venosus (SV) and bulbus arteriosus (BA) region

of the heart was also measured. The heart rate significantly decreased in all PFOS treatments

H-27


-------
when compared to the solvent control at 10 days post fertilization (exposure day eight).
Additionally, the SV-BA distance was increased at 10 dpf in the highest test concentration (16
mg/L PFOS) compared to other treatments. The 8-day LOEC (based on heart rate) of 1 mg/L and
the 8-day MATC (based on SV-BA distance) of 8.0 mg/L were both acceptable for qualitative
use (due to the short test duration).

Fang et al. (2012) also conducted an 8-day renewal, measured toxicity test on PFOS
(98% purity) with embryos of marine medaka, Oryzias me/asfigma. Fertilized eggs were cultured
in artificial seawater for two days (2 dpf) and then exposed to one of four nominal treatments
(0.1% DMSO control or 1, 4, and 16 mg/L PFOS). Three replicates of 70 embryos each were
held in 90 mm petri dishes containing 20 mL of solutions lor eight days (10 dpf). Limited water
quality conditions were reported by the authors (salinity of ppt. 28 I 'C, and 14-hr:10-hr
light:dark photoperiod). Test treatments ^ere renewed daily and PFOS concentrations were
measured at 4 dpf and l<> dpi" by I.C/MS. Measured concentrations were only reported
graphically, but from \ isual inspection measured concentrations were close to nominal
concentrations Similarly, tissue concentrations at 4 dpf reflected a concentration response
manner, with increasing tissue concentrations at higher test treatments. At 10 dpf larva mortality
was recorded and the percentage of malformations observed. There was no significant effect on
mortality in any concentration despite increasing embryo mortality from 2.0% in the DMSO
control to 14.6% in the highest test treatment (16 mg/L PFOS). However, a significant increase
in larval malformations (bent spine and edema) was observed in all PFOS treatments. Bent spine
malformations were observed in 29.0% of the larva in the DMSO control, but increased to 68.2%
at 1 mg/L PFOS. The atypical duration (8 days) deemed the LOEC based on malformations of 1
mg/L PFOS as acceptable for qualitative use only.

H-28


-------
Wu et al. (2012) conducted another renewal, unmeasured toxicity test with PFOS (98%
purity) on marine medaka embryos. Exposure lasted up to 21 days (time it took embryos to hatch
in the DMSO controls) plus a seven day observation period in clean water with hatched larva.
For each treatment (three replicates), 50-70 embryos (2 dpf) were collected from in-house
cultures and distributed into 90 mm petri dishes containing 20 mL artificial seawater and at a
salinity of 30%o, 28 ฑ 1ฐC under a 14-hr: 10-hr (light:dark) photo period In all nominal treatments
(solvent control, 1, 4 and 16 mg/L PFOS), the concentration of DMSO was n I" o. Embryos were
observed and the media was renewed daily. After embryos hatched (varied across treatments),
larvae were held for an additional seven days in clean 3<~>%n artificial seawater without food.
PFOS had a significant effect on increasing the hatching rale, decreasing the hatching time and
decreasing the larval survival in the obsei \ ation period at 4 mu I. PFOS. The ~21-day MATC
(based on larval survival) of 2 mg/L PFOS (ueomean of I and 4 mg/L) was acceptable for
qualitative use only due to the short exposure duration for an early life stage fish test.

H. 3.2.2 Sclnisics scltlc^cli

Jeon rl al. (201 Or) performed a o-day renewal, unmeasured toxicity test on

perfluorooclane sulfonic acid potassium salt (PFOS-K, 99% purity) with 5-month old blackrock

fish, Scbcisics schlegeli (a non-North American species). Fish were obtained from the National

Fishery Science Institute (1 Justin, Korea) and acclimatized to four different salinities (10, 17.5,

25, and 34 ppt) for two weeks before exposure. The salinity of test solutions was adjusted by

mixing filtered natural seawater and groundwater. A 3.0 mg/L PFOS stock was made with HPLC

grade methanol and diluted to two PFOS concentrations (0.1 and 1 mg/L). Twenty fish were

added to each tank with 180 L test solutions renewed on day two and four. Treatments and

control were not replicated and fish were not fed during the exposure. Across all exposures, pH

ranged from 8.0-8.2 and temperature from 8-12ฐC; no other water quality parameters were

H-29


-------
reported. There were no significant differences in total length, weight and survival (no mortality
observed in any of the exposures) over the 6-day exposure. The NOEC (survival and growth)
was 1 mg/L at each test salinity (10, 17.5, 25 and 34 ppt) and was acceptable for qualitative use
based on the atypical test duration.

H.4 Summary of Plant PFOS Toxicity Studies Used Qualitatively

H.4.1 Bacterium. Vibrio fischeri

Rosal et al. (2010) conducted a 15-minute static, measured bioluminescence inhibition

test with perfluorooctane sulfonate potassium salt (PFOS-K, 98% purity) on the bacterium,

Vibrio fischeri following ISO 11348-3 standard protocol. While limited details are provided

about the exposure, the authors stated they were following the standard protocol. The experiment

used a commercially available Biofix Lunii lesl (Macherey-Nauel. Germany), where the

bacterium is supplied freeze-dried. It was reconstituted and incubated at 3ฐC for five minutes

before use. The experiments employed a 0 34 M \a('l (2ฐ.. u \ ) test medium and conducted at

18ฐC. No solvents were used to make test solutions. Specific test concentrations were not

provided, but the authors noted that nominal and measured concentrations did not have

significant de\ iations The I 5-minute I X';.. based on bioluminescence inhibition was >500 mg/L.

At this concentration there was 12".. luminescence. The test was acceptable for qualitative use

only because ol" the short test duration and lack of exposure details.

H.4.2 Cvanobacteria. Anabaena sp.

Rosal et al. (2010) also conducted a 24-hour static, measured bioluminescence inhibition

test with PFOS-K (98% purity) on the cyanobacteria, Anabaena sp. Limited details are provided

about the exposure, but the authors stated they were following the test design in Rodea-

Palomares et al. (2009). The cyanobacterium Anabaena, CPB4337 strain, was grown at 28ฐC on

a rotary shaker in 50 mL AA/8 media supplemented with nitrate (5 mM) in 125 ml Erlenmeyer

H-30


-------
flasks and 10 mg/mL of neomycin sulphate. No solvents were used to make test solutions.
Specific test concentrations were not provided, but the authors noted that nominal and measured
concentrations did not have significant deviations. The 24-hour bioluminescence inhibition EC so
was 143.27 mg/L, and was acceptable for qualitative use only because of the short test duration
and lack of exposure details.

H.4.3 Pinoflapellate. Pyrocvstis lunula

Hayman et al. (2021) conducted a short-term, sublethal 24 hour exposure to examine the

effects of PFOS on the bioluminescent dinoflagellate (Pyrocystis lunula) following ASTM

E1924-97 (ASTM 2004). Dilution water was 0.45 |uin liltered seawater collected from North San

Diego Bay, CA and spiked PFOS. Spiking consisted of I lie addition of stock solutions of

perfluorooctanesulfonic acid potassium sail (PI'OS-K, 98% puriiy. CAS # 2795-39-3) dissolved

in methanol; highest methanol concentration of <) S"() (\ \ ) Concentrations of PFOS for the

toxicity tests were determined from a range finding study Measured concentrations for PFOS

were 0 (control and sol\ent control). i) SS. I I, 1.0,2.0,2.5, 10, 34, and 120 mg/L.

Approximately 3,Odd cells of/', lunula were added to 2.5 mL of test solution in acrylic test

cuvettes, with six replicates per treatment concentration. P. lunula were exposed for 24 hours in

a 19CC incubator with a re\ersed (e.g., dark during the typical "day" period) 12-hr: 12-hr

light:dark cycle Test cuvettes were removed from the incubator after 24 hours and after being in

the dark period for approximately three hours, inserted and analyzed in a specialized

spectrometer (QwikLite 200 Biosensor System, Assure Controls, Carlsbad, CA) and the light

output was recorded. Less light output relative to concurrently evaluated controls is indicative of

an adverse effect. The 24-hr bioluminescence EC50 fori5, lunula was determined to be 4.9 mg/L

PFOS. The value is considered qualitatively only because of the short exposure period (less than

96 hours).

H-31


-------
H.4.4 Golden brown alga. Isochrysis galbana

A 72-hour static, unmeasured algal growth inhibition test on PFOS (98% purity) with

Isochrysis galbana was performed by Mhadhbi et al. (2012) following OECD (2006) test

methodology. Golden brown alga were provided by Estacion de Ciencias Marinas de Toralla

(ECIMAT). The cultures were maintained in 250 mL glass Erlenmeyer flasks with autoclaved

filtered sea water and EDTA-free f/2 culture medium. PFOS stock solutions were prepared in

DMSO and added to the dilution water with a maximum DTVISO concentration of 0.01% (v/v).

Nominal test concentrations were solvent control, 3.75, 7.5, 15, 30 and 60 mu I. PTOS. Each

flask was inoculated at a density of 10,000 cells/nil., with the algae in the exponential growth

phase. Each treatment was replicated three times. Flasks u civ kept at 20ฐC with a 24 hour light

period, and manually shaken daily. Cell counts were carried out e\ cry 24 hours, with a reported

72-hour ECso based on growth inhibition of 37 5 mu I. PI-OS The short test duration (<96 hours)

made the effect concentration acceptable for qualitative use only

H.4.5 Alga. Ceratoncis ('lostcrium

Simpson et al. (2021) c\ alualcd the toxic effects of perfluorooctane sulfonic acid

potassium salt (PI 'OS-K. purchased IV0111 Sigma Aldrich) on the temperate marine benthic

microalga. ( \-ratoneis closicrmm (pre\ iously named Nitzschia closterium) from previously-

established cultures using a 72-hr growth test Clean seawater (salinity 33ฑ2 ppt) for culturing

was sourced from the southeast coast of New South Wales (NSW), Australia, and clean

sediments were from I Jon net Bay, NSW. Toxicity was assessed by measuring the decrease in

rate of growth (growth rate inhibition) of the microalga attached to the surface of a PFOS-spiked

sediments in three tests, each with different sediment PFOS concentrations and sediment

properties. The benthic microalgae were added to the surface of a sediment subsample and total

chlorophyll extracted as a surrogate for algal cell density. To each 50 mL triplicate centrifuge

H-32


-------
tube, 2 g of sediment was added along with overlying water consisting of 20 mL filtered
seawater and nutrients (0.2 mL of 0.26 mM NaNCte and 0.2 mL of 1.3 mM KH2PO4). After a day
overlying water was exchanged and 2-6 x 104 cells/mL of algae was added to initiate the test.
Seawater only and sediment without algae were also tested as controls. PFOS concentrations in
the overlying water were assumed based on the PFOS measured in the overlying water from
other tests in the study. There was no effect on growth of the microaluae in the treatment with
the highest estimated overlying water concentration of 4 I (ฆ> mu I. PFOS The author-reported 72-
hour survival NOEC was 4.16 mg/L. Given the presence of other PFAS measured in the
sediments and overlying water at very low concentrations, this study was considered a mixture.
The other PFAS made up a much smaller percentage of the total PFAS and were generally in low
concentration, and thus, the result from the study is considered t|LiaIilative use only.

H.4.6 Diatom. Skeletonema costatum

Desjardins ct nl. (2001) performed a sialic, measured 96-hour growth inhibition study on

the marine diatom Skclcioncma cosiamm Protocol from U.S. Environmental Protection Agency,

OPPTS Number 85<) 54<)i) was loll owed Aluae was cultured at Wildlife International Ltd. in

saltwater alual medium Diatoms were exposed to one control and one test concentration (3.46

mg/L. PI-OS nominal, 3 2d mu I. PI7OS measured). Diatoms were added to sterile 250 mL

Erlenmeyer ll asks. Ill led with 100 mL of either test or control medium with a salinity of 30%o,

and plugged with loam sloppersl. Test chambers were kept at 20 ฑ 2 ฐC on a mechanical shaker

table set to 100 rpm, with a 14-hr: 10-hr light:dark cycle. Single cell counts were taken every 24

hours throughout the duration of the test, and group cell counts were taken at 72 hours and 96

hours. While there was some minor growth inhibition in the treatment group, it was not enough

to show a statistical difference when compared to the control group.

H-33


-------
Appendix I Acute to Chronic Ratios

1.1 Acute to Chronic Ratios from Quantitatively Acceptable Toxicity Tests.

Spocics

( hcmiciil /
Piiriu

Acme
Method-'

Chronic
Method-'

Acule
Test
Dui'iilion

Chronic

lesl
Dui'iilion

( hronic I'.ITecl

Acule
I'.ITecl
(one.
(niii/l.)

Chronic
I'.ITecl
(one.

(mป/l.)

A( R*

SMACK1'

Reference

Fatmucket,

Lampsilis siliquoidea

PFOS

>98%

S, M

S, M

24 hr

36 d

MATC

(metamorphosis

success)

I<> 5

uun.x

933.3

933.3

Hazelton et al.

(2012),	Hazelton

(2013)



Snail,

Physella heterostropha
pomilia

(formerly, Physa
pomilia)

PFOS-K

>98%

S, M

R, M

96 hr

44 d

IX lu

(clulch size)

161.8

8.831

18.32

18.32

Funkhouser (2014)



Rotifer,

Brachionus calyciflorus

PFOS

>98%

S,Ub

R,Ub

24 hr

I p In
I5X III'

i.oir

(reduced nel
reproductive rale)

(.1.8

0.25

247.2

>247.2

Zhang et al. (2013)



Cladoceran,

Daphnia carinata

PFOS-K

>98%

S,U

R. T

4S hr

:i d

\1ATC

(days to first brood)

11.56

0.003162

3,656

3,656b

Logeshwaran et al.
(2021)



Cladoceran,

Daphnia magna

PFOS-K
90.49%

S, M

R, M

4S hr

:i d

EC10

(survival)

58.51

11.19

5.229

-

Drottar and
Krueger (2000a)

Cladoceran,

Daphnia magna

PFOS-K

95%

s. u

k. 1J

4S hr

:i d

EC10

(survival)

67.2

16.35

4.110

-

Boudreau et al.
(2003a)

Cladoceran,

Daphnia magna

PFOS
Unreported

S. 1

R, U

4S hr

:i d

EC10

(# of young/adult)

35.46

0.7885

44.97

-

Ji et al. (2008)

Cladoceran,

Daphnia magna

PFOS-K

>98%

S. 1

k, L

4S hr

:i d

EC10

(total neonates/female)

63.84d

2.919

21.87

-

Li (2009, 2010)

Cladoceran,

Daphnia magna

PFOS-K

99%

S, M

k.M

4S hr

21 d

EC10

(reproduction)

78.09

2.26

34.55

-

Yang et al. (2014)

Cladoceran,

Daphnia magna

PFOS
98%

s,u

k. 1

4S hr

21 d

EC10

(number of
offspring/brood/female)

23.41

0.001712

13,674b

-

Lu et al. (2015)

Cladoceran,

Daphnia magna

PFOS-K

>98%

s,u

R, U

48 hr

21 d

EC10

(survival)

94.58

3.596

26.30

-

Liang et al. (2017)

Cladoceran,

Daphnia magna

PFOS-K

98%

s,u

R, U

48 hr

21 d

EC10

(growth-length)

22.43

0.9093

24.67

17.35

Yang et al. (2019)

1-1


-------
Spocics

( hcmic;il /
Piiriu

Acme
Method-'

Chronic
Method'

Acule
Test
Dui'iilion

Chronic

Tesl
Diii'iilion

Chronic I'.ITecl

Acnle
I'.ITecl
(one.

(iii^/l.)

Chi'onic
I'.ITecl
(one.
(mii/l.)

A( K*

SMACK1'

Reference



Cladoceran,

Moina macrocopa

PFOS
Unreported

S,U

R, U

48 hr

7 d

EC10

(# of young/starting
adult)

r :u

0.1789

96.14

96.14

Ji et al. (2008)



Crayfish,

Procambarus fallax f.
virginalis

PFOS-K

>98%

S, M

R, M

96 hr

28 d

T.c:n

S"

0 !(ฆ"

358.5

358.5

Funkhouser (2014)



Zebrafish,

Danio rerio

PFOS-K
unknown/
PFOS 96%

R, U

R, U

96 hr

LC

IX lu

(l 'l oil spring: " o
sur\i\al)

17

O.O1(j50

1,030

1,030

Wang et al. (2011),
(Wang et al.

2013b)



Fathead minnow,

Pimephales promelas

PFOS-K
90.49%

S, M

F, M

96 hr

4" d

rem

(survival)

•J.020

0.4732

19.06

19.06

Drottar and
Krueger (2000d)

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, I) dissnl\al. I)iel dicl;n \. M l mala rial transfer
b Values appears to be an outlier and is not used SMACR calculation.

0 Values in bold are used in the SMACR and FACR calculations
d Geometric mean of three LC50s.

1-2


-------
1.2 Acute to Chronic Ratios from Qualitatively Acceptable Toxicity Tests.

Species

Anile / (lironie
( hemie;il iiml
Puriu

Aeule
Method'1

(lironie
Method"

Aeule
lesl
l)iir;ili(in

( lironie

lesl
Diinilion

Aeule
I'.ITeel

( lironie HITeel

Aeule
I'.ITeel
(one.

(mป/l.)

(lironie
I'.ITeel
Cone.
(mป/l.)

ACR

Rel'erenees



Planaria,

Dugesia japonica

PFOS-K

>99%

R,U

R, U

96 hr

10 d

l.( 50

I.OLC

(regeneration:
decreased appearance
of auricles)

29.46

0.5

58.92

Yuanetal. (2014)



Snail,

Lymnaea stagnalis

PFOS
Unreported

S, M

R,M

96 lu-

:i d

T.C50

MATC

(survival)

171.5

4.243

40.41

Olson (2017)



Midge,

Chironomus sp.

PFOS-K (99%) /
PFOS Unreported

S, M

S,M

ge hr

-36 d
ristof 10

generations)

L( 5(1

LOEC

(F1 developmental
time, adult weight,
exuvia length)

182.12

0.004

45,530

Yang et al. (2014);
Marziali et al.
(2019)

Midge,

Chironomus sp.

PFOS-K (99%) /
PFOS-K (95%)

S, M

S,M

96 In-

Life
( 5i) di

l.( 5(1

EC10

(total emergence)

182.12

0.089

2,039

Yang et al. (2014);
MacDonald et al.
(2004)



Yellow fever mosquito,

Aedes aegypti

PFOS
Unreported

S. 1

k.l

4S hr

4: d

LC50

MATC

(average time to
emergence)

1.18

0.079

14.94

Olson (2017)



Rainbow trout,

Oncorhynchus mykiss

PFOS-K (9X"„i
PFOS (89".,)

k. \1

S.l

<>(. hr

14 d

LC50

LOEC

(decrease LSI)

2.5

1.0

2.500

Sharpe et al.
(2010); Oakes et
al. 2005



Zebrafish,

Danio rerio

PFOS

>97%

S. V

k. 1

<>(. hr

6 d

LC50

EC50

(uninflated swim
bladder)

58.47

2.29

25.53

(Hagenaars et al.
2011),Hagenaars
et al. (2014)

Zebrafish,

Danio rerio

PFOS-K

98%

k. \1

k. \1

96 hr

21 d

LC50

LOEC

(reduced fecundity)

22.2

0.5

44.40

Sharpe et al.
(2010)

Zebrafish,

Danio rerio

PFOS
98%

s, I

k,U

96 hr

70 d

LC50

MATC

(increased
malformation &
decreased survival of
F1 fish)

3.502

0.02236

156.6

Du et al. (2016);
Du et al. (2009)



1-3


-------
Species

Anile / (lironie
( hemieiil iind
Puriu

Aeule
Method'1

(lironie
Melliod-'

Aeule
lesl
Diii'iilion

(lironie

lesl
Diinilion

Aeule
I'.ITeel

( lironie HITeel

Aeule
I'.ITeel
(one.

(111^/1.)

( lironie
I'.ITeel
(one.
(inป/l.)

ACR

Uel'erenees

African clawed frog,
Xenopus laevis

PFOS-K (86.9%) /
PFOS-K (86.9%)

R, M

R,M

96 hr

96 hr

LC50

LOEC

(growth)

15.49

8.26

1.875

Palmer and
Krueger (2001)

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D=dissolved, Diet=dietar>. M l maternal transfer

1-4


-------
Appendix J Unused PFOS Toxicity Studies

J.l Summary Table of Unused PFOS Toxicity Studies

Author

( iliilion

Kciison I misod

Arukwe, A. and A.S. Mortensen

2011. Lipid peroxidation and oxidative stress responses of salmon fed a diet
containing perfluorooctane sulfonic- or pcrfluorooclaiic carhowlic acids.
Comp. Biochem. Physiol. Part C 154: 288-295.

Force-fed (oral gavage); only one exposure
concentration

Arukwe, A., M.V. Cangialosi, R.J. Letcher, E.
Rocha, A.S. Mortensen

2013. Changes in morphometry and association between wholc-hody fatty
acids and steroid hormone profiles in relation to bioacciinuilaiiou patterns in
salmon larvae exposed to perfluorooclanc sulfonic or perfluorooctane
carboxylic acids. Aquat. Toxicol. 130-131: 219-230.

Only one exposure concentration

Balbi, T., C. Ciacci, E. Grasselli, A. Smerilli, A.
Voci, and L. Canesi

2017. Utilization of Mytilus digestive gland cells for the in vitro screening of
potential metabolic disruptors in aquatic m\ ertebrates. Comp. Biochem.
Physiol. PartC. 191:26-35.

In vitro (excised cells)

Bilbao, E., D. Raingeard, 0. Diaz de Cerio, M.
Ortiz-Zarragoitia, P. Ruiz, U. Izagirre, A. Orbea, I.
Marigomez, M.P. Cajaraville and I. Cancio

2010. Effects of exposure to Prestige-like hea\ > fuel oil and to
perfluorooctane sulfonate on com eutional bioniarkeis and target gene
transcription in the thicklip me> mullet Chclon lahiosns. \quat. Toxicol. 98:
282-296.

Only one exposure concentration; the
number of fish was not reported

Blanc, M., A. Karrman, P. Kukucka, N. Scherbak
and S. Keiter

2017. Mixture-specific gene expression mi zehralish (Danio rerio) embryos
exposed to perfluorooctane sulfonic acid (Pl '( )Si. perfluorohexanoic acid
(PFHxA) and 3.3'.4.4'.5-pcntachlorobiphcn\ 1 (PCB126). Sci. Total Environ.
590: 249-257.

Mixture (PFOS, PFHxA and PCB126)

Blanc, M., J. Ruegg, N. Scherbak, and S.H. Keiter

2(i I<>. Environmental chemicals differentially affect epigenetic-related
mechanisms in the zebrafish liver (zf-1) cell line and in zebrafish embryos.
Aquat. Toxicol.215:105272-9999.

Control absent from test

Chen, J., L. Zheng, L. Tian, N. Wang, L. Lei, Y.
Wang, Q. Dong, C. Huang, and D. Yang

2018. Chronic PFOS exposure disrupts thyroid structure and function in
zebrafish. Bull. Environ. Contain. Toxicol. 101: 75-79.

Only one treatment concentration; severe
lack of procedural details

Cheng, J., S. Lv, S. Nie, J. Liu, S. Tong. \ Kang. Y.
Xiao, 0- Dong, C. Huang and D. Yang

2016. Chronic perfluorooctane sulfonate (PFOS) exposure induces hepatic
steatosis in zebrafish. Aquat. Toxicol. 176: 45-52.

Only one exposure concentration;
unmeasured chronic exposure

Consoer, D.M.

2017 \ mechanistic investigation of perfluoroalkyl acid kinetics in rainbow
trout i( )iicorhvnchus mvkiss). A dissertation submitted to the faculty of the
Univeisiis ofMinnesota.

Injected toxicant; only one exposure
concentration

Cui, Y., W. Liu, W. Xie, W. Yu, C. Wang and 11
Chen

2015 1 ii\ estigation of the effects of perfluorooctanoic acid (PFOA) and
pcrfliioi'ooclane sulfonate (PFOS) on apoptosis and cell cycle in a zebrafish
11 )anio rerio) liver cell line. Int. J. Environ. Res. Public Health 12(12): 15673-
15682.

Excised cells (liver cell line)

Diaz de Cerio, 0., E. Bilbao, M.P. Cajaraville and I.
Cancio

2012. Regulation of xenobiotic transporter genes in liver and brain of juvenile
tliicklip grey mullets (Chelon labrosus) after exposure to Prestige-like fuel oil
and to perfluorooctane sulfonate. Gene 498: 50-58.

Only one exposure concentration

J-l


-------
Author

( iliilion

Kciison I misod

Dorts, J., P. Kestemont, P.A. Marchand, W.
D'Hollander, M.L. Thezenas, M. Raes and F.
Silvestre

2011. Ecotoxicoproteomics in gills of the sentinel fish species, Cottus gobio,
exposed to perfluorooctane sulfonate (PFOS). Aquat. Toxicol. 103: 1-8.

Only two exposure concentrations, not
North American species

Du, J., S. Wang, H. You and Z. Liu

2016b. Effects of ZnO nanoparticles on perfluorooctane sulfonate induced
thyroid-disrupting on zebrafish larvae. J. Environ. Sci. 47: 153-164.

Only 72-75% control survival in 14-day test

Du, J., J. Tang, S. Xu, J. Ge, Y. Dong, H. Li, and M.
Jin

2018. Parental transfer of perfluorooctane sulfonate and ZnO nanoparticles
chronic co-exposure and inhibition of growth in F1 offspring Rcgul. Toxicol.
Pharmacol. 98: 41-49.

Excessive control mortality in the F0
generation

Fang, C., Q. Huang, T. Ye, Y. Chen, L. Liu, M.
Kang, Y. Lin, H. Shen, and S. Dong

2013. Embryonic exposure to PFOS induces immunosuppression in (lie fish
larvae of marine medaka. Ecotox. Environ. Safclv 92: 104-111.

Excessive control mortality (-60% control
survival)

Fernandez-Sanjuan, M., M. Faria, S. Lacorte and C.
Barata

2013. Bioaccumulation and effects of pcrfluorinated compounds < Pl ( m hi
zebra mussels (Dreissena poly mo rplia) Fnviron Sci. Pollut. Res. 20:2<><> 1
2669.

Mixture

Gorrochategui, E., S. Lacorte, R. Tucker and F.L.
Martin

2016. Perfluoroalkylated substance effects hi Xenopus laevis A6 kidney
epithelial cells determined by ATR-FTTR sped roscopv and chemometric
analysis. Chem. Res. Toxicol. 29: 924-932.

The tests were performed on cell cultures
obtained from an outside source. Whole
organisms were not investigated.

Hagenaars A., I.J. Meyer, D. Herzke, B.G. Pardo, P.
Martinez, M. Pabon, W. De Coen, and D. Knapen

2011. The search for alternative aqueous film forming foams (AFFF) with a
low environmental impact: Physiological and Iranscriplomic effects of two
Forafacฎ fluorosurfaclanls in lurboi. \quat Toxicol. 104: 168-176.

Only one exposure concetration; missing
detail (focus is on other chemicals)

Hoff, P.T., W. VanDongen, E.L. Esmans, R. Blust,
W.M. De Coen

2003. Evaluation of the loxicological effects of perfluorooctane sulfonic acid
in the common carp (Cvprinus carpio). Aqual Toxicol. 62 (4): 349-359.

Exposure was from a single intra-peritoneal
injection

Hoff, P.T., K. Van Campenhout, K. Van de Vijver,
A. Covaci, L. Bervoets, L. Moens, G. Huyskens, G.
Goemans, C. Belpaire, R. Blust and W. De Coen

2005. Perfluorooclanc sulfonic acid and organohalogcn pollutants in liver of
three freshwater fish species in Flanders (Belgium): relationships with
biochemical and organismal effects. Environ. Pollut. 137: 324-333.

Field exposure, but concentrations were not
measured so no B AFs could be calculated

Honda, M., A. Muta, T. Akasaka, Y. Inoue, Y.
Shimasaki, K. Kanna, N. Okino, and Y. Oshima

2014. Identification of perfluorooctane sulfonate binding protein in the plasma
of tiger puffcrfish Takiliimi i iihripes. Ecotox. Environ. Safety. 104: 409-413.

Injected toxicant; only one exposure
concentration

Honda, M., A. Muta, A. Shimazaki, T. Akasaka. M.
Yoshikuni, Y. Shimasaki, and Y. Oshima

2018. H igh conccntrat ions of perfluorooctane sulfonate in mucus of tiger
puffer fish Takifugu rubripes: a laboratory exposure study. Environ. Sci.
Pollut. Res 25: 1551-1558.

Injected toxicant

Huang, T.S., P.A. Olsvik, A. Krovcl. H S Tung and
B.E. Torstensen

2009 Stress-induced expression of protein disulfide isomerase associated 3
(PDI \') mi Atlantic salmon (Salmo salarL.). Comp. Biochem. Physiol. PartB
Biocliem Mol. Biol. 154(4): 435-442.

In vitro (cultured hepatocytes)

Huang, Q., S. Dong, C. Fang, X. Wu, 1 Ye and Y
Lin

2012 1 )eep sequencing-based transcriptome profiling analysis of Oryzias
melasimma exposed to PFOS. Aquat. Toxicol. 120-12: 54-58.

Only one or two exposure concentrations

Huang, Q., Y. Chen, Y. Chi, Y. Lin, H. Zhang, (
Fang and S. Dong

2(il5. Inununotoxic effects of perfluorooctane sulfonate and di(2-ethylhexyl)
plilhalatc on the marine fish Oryzias melastigma. Fish Shell. Immunol. 44:
302-306.

Only two exposure concentrations

Jacobson, T., K. Holmstrom, G. Yang, A.T. Ford, U.
BergerandB. Sundelin

2010. Perfluorooctane sulfonate accumulation and parasite infestation in a
field population of the amphipod Monoporeia affinis after microcosm
exposure. Aquat. Toxicol. 98(1): 99-106.

Dilution water not characterized, mixture

J-2


-------
Author

( iliilion

Kciison I mi sod

Jantzen, C.E., K.M. Annunziato and K.R. Cooper

2016. Behavioral, morphometric, and gene expression effects in adult
zebrafish (Danio rerio) embryonically exposed to PFOA, PFOS, and PFNA.
Aquatic Toxicology. 180:123-130.

Single concentration test where exposure to
PFOS was of an acute (117 hours) duration
but endpoints were measured at 6 months of
age.

Keiter S., K. Burkhardt-Medicke, P. Wellner, B.
Kais, H. Farber, D. Skutlarek, M. Engwall, T.
Braunbeck, S.H. Keiter, T. Luckenbach

2016. Does perfluorooctane sulfonate (PFOS) act as ehemoscnsili/cr in
zebrafish embryos? Science of the Total Enviroi imciil 548-549:317-324.

Mixture

Khan, E.A., X. Zhang, E.M. Hanna, F. Yadetie, I.
Jonassen, A. Goksoyr, and A. Arukwe

2021. application of quantitative transcriplomics in cvaliialiim I lie ox vivo
effects of per- and polyfluoroalkyl substances on Atlantic cod <( nidus morhua)
ovarian physiology. Sci. Total Environ.755(l): 11 p.

In-vitro study

Kim, S., K. Ji, S. Lee, J. Lee, J. Kim, S. Kim, Y. Kho
and K. Choi

2011. Perfluorooctane sulfonic acid exposure increases cadmium in\icil> mi
early life stage of zebrafish, Danio rerio Environ Tuvicol. Chem. 30(4). X~0-
877.

Only one exposure concentration; atypical
duration (7 days)

Kovacevic, V., A.J. Simpson, and M.J. Simpson

2019. The concentration of dissolved oruamc mallei' impacts the metabolic
response inDaphnia niauua exposed to 17a-olli> u\ leslradiol and
perfluorooctane sulfonate 1 !a>ln\ia>l. Environ. Sal' l~<> 468-478.

Only one treatment concentration
(examined across a gradient of dissolved
organic matter concentrations); endpoints
measured were a suite of metabolic
changes; atypical design for this test
organism

Krovel, A.V., L. Softeland, B. Torstensen, and P.A.
Olsvik

2008. Transcriptional effects ol Pl '( )S m isolated hepatocs tes from Atlantic
salmon Salmo salarL. Conipai,ili\ e l.kiclieiiiisii^ and Physiology, Part C.
148: 14-22

In vitro

Lee, W. and Y. Kagami

2010. Effects nf peifluorooctanoic acid and perfluorooctane sulfonate on gene
expression profiles in medaka (Oryxias latipes). Abstracts. Toxicol. Letters
196S: S3~-S,ol.

Abstract only, cannot judge against data
quality objectives

Li, M.H.

2011. Cliauues of eholmcslcnisc and carboxylesterase activities in male
uuppics. Poeeilu reticulata. after exposure to ammonium perfluorooctanoate,
hul not to pei'1'likn'noclanc sulfonate. Fresenius Environ. Bull. 20(8a): 2065-
2
-------
Author

( iliilion

Kciison I misod

\1ariin. J.W., S.A. Mabury, K.R. Solomon and
D.C.G. Muir

2003a. Bioconcentration and tissue distribution of perfluorinated acids in
rainbow trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 22: 196-204.

Bioaccumulation (steady state no
documented); only 12 days

Martin, J.W., S.A. Mabury, K.R. Solomon and
D.C.G. Muir

2003b. Dietary accumulation of perfluorinated acids in juvenile rainbow trout
(Oncorhynchus mykiss). Environ. Toxicol. Chem. 22(1): 189-195.

Mixture

Martin, J.W., S.A. Mabury, K.R. Solomon and
D.C.G. Muir

2013. Progress toward understanding the bioaccumulalion of perfluorinated
alkyl acids. Environ. Toxicol. Chem. 32( 11): 2421-2423

Review paper

Mortensen, A.S., R.J. Letcher, M.V. Cangialosi, S.
Chu, and A. Arukwe

2011. Tissue bioaccumulation patterns, xenobiol ^biotransformation and
steroid hormone levels in Atlantic salmon (Salmo salar) led a diel containing
perfluoroactane sulfonic orperfluorooclanc carboxvlie acids, ( homosphere
83: 1035-1044.

One dietary dosage level provided over a 6-
day period; not intended as a toxicity test

Mylroie, J.E., M.S. Wilbanks, A.N. Kimble, K.T. To,
C.S. Cox, S.J. Mcleod, K.A. Gust, D.W. Moore, E.J.
Perkins, and N. Garcia-Reyero

2021. Perfluorooctanesulfonic acid n id need toxicity onzebrafish emhr\os in
the presence or absence of the chorion Lnviron Toxicol. Chem. 40(31 ~8u-
791.

Use of dilution medium (estradiol media) to
prepare stock solutions inconsistent with
EPA test guidelines

Oh, J.H., H.B. Moon and E.S. Choe

2013. Alterations in differentially expressed uenes after repeated exposure to
perfluorooctanoate andperfluorooctanesulfonale in liver of Oryzias latipes.
Arch. Environ. Contain. Toxicol. 64(3): 475-48 ^

Only one exposure concentration, no
concentration-response observed, not North
American species

Pablos, M.V., P. Garcia-Hortiguela and C. Fernandez

2015. Acute and chronic loxicity of emerging aniianiiiiaiiis, alone or in
combination, in Chlorella \ uluaris and Daphnia niauna 1 !nviron. Sci. Pollut.
Res. 22: 5417-5424.

Mixture

Popovic, M, R. Zaja, K. Fent and T. Smital

2014. Interaction of em iroiinienial amianiiiiaiiis with /ebrafish organic anion
transporting polypeptide. ();iip 1 d 1 (Sleo Id h Toxicol. Appl. Pharmacol.
280(1): 149-158.

Excised cells

Prosser, R.S., K. Mahon, P.K. Sibley, D. Poirier and
T. Watson-Leung

2016. Bioaccumulalion of pcrfluorinaled carboxylates and sulfonates and
poly chlorinated biphcnvls in laboratory-cultured Hexagenia spp., Lumbriculus
variegatus and Pimephales promclas fro in field-collected sediments. Sci. Total
Environ. 543: 715-726.

Mixture (filed collected sediment, contained
PFAS mixtures and PCBs)

Roland, K., P. Kestemont, L. Henuset, M. A.
Pierrard, M. Raes, M. Dieu and F. Silveslre

2013. Protcomic responses of peripheral blood mononuclear cells in the
European eel (Anguilla anguilla) lifter perfluorooctane sulfonate exposure.
Aquat. Toxicol. 128/129: 43-52.

In vitro (excised cells)

Shi, X.,Y. Du, P.K.S. Lam, R.S.S. Wuand li.Zhou

2008 Developmental toxicity and alteration of gene expression in zebrafish
embrwis exposed to PFOS. Toxicol. Appl. Pharmacol. 230(1): 23-32.

Excessive control mortality

Shi, X., L.W.Y. Yeung, P.K.S. Lam. R S S Wii and
B. Zhou

2009h Protein profiles in /.ebrafish (Danio rerio) embryos exposed to
perfhiorooclane sulfonate. Toxicol. Sci. 110(2): 334-340.

Only one exposure concentration; atypical
duration (8 days)

Stevenson, C.N., L.A. MacManus-Spencer, T
Luckenbach, R.G. Luthy and D. Epel

2006. New perspectives on pefluorochemical ecotoxicology: inhibition and
induction of an efflux transporter in marine mussel, Mytilus californianus.
1 Jiviron. Sci. Technol. 40: 5580-5585.

Excised cells (mussel gill tissue)

Thienpont, B., A. Tingaud-Sequeira, E. Prats, C.
Barata, P.J. Babin and D. Raldua

2011. Zebrafish eleutheroembryos provide a suitable vertebrate model for
screening chemicals that impair thyroid hormone synthesis. Environ. Sci.
Technol. 45(17): 7525-7532.

Only one exposure concentration; atypical
duration (3 days)

J-4


-------
Author

( ilalion

Reason I nusod

Qiu, X., N. Iwasaki, K. Chen, Y. Shimasaki and Y.
Oshima

2019. Tributyltin and perfluorooctane sulfonate play a synergistic role in
promoting excess fat accumulation in Japanese medaka (Oryzias latipes) via in
ovo exposure. Chemosphere. 220: 687-695.

Injectedt toxicant into eggs, not North
American species

Wagner, N.D., A.J. Simpson and M.J. Simpson

2016. Metabolomic responses to sublethal contaminant exposure in neonate
and adult Daphnia magna. Environ. Toxicol. Chcm. 36(4): 938-946.

Only one exposure concentration

Wagner, N.D., A.J. Simpson and M.J. Simpson

2018. Sublethal metabolic responses to contaminant mixture toxicity in
Daphnia magna. Environ. Toxicol. Chem. 37(9): 2448-245".

Only one exposure concentration

Wang, S., C. Zhuang, J. Du, C. Wu, and H. You

2017. The presence of MWCNTs reduces developmental toxicils of PFOS in
early life stage of zebrafish. Environ Pollul. 222: 201-209.

The 96 hour LC50 reported in the
publication is the same as the value in Du et
al. 2016 (no details provided about this test)

Xia, X., X. Chen, X. Zhao, H. Chen and M. Shen

2012. Effects of carbon nanotubes. chars, and ash onbioaccumulaliou of
perfluorochemicals by Chironomus pluniosus larvae in sediment. Env iron Sci.
Technol. 46: 12467-12475.

Mixture (PFCs mixed in sediment)

Xia, X., A.H. Rabearisoa, X. Jiang and Z. Dai

2013. Bioaccumulation of perfluoroalks 1 substances by Daphnia magna in
water with different types and concentral inns of protein. Environ. Sci.
Technol. 47: 10955-10963.

Bioaccumulation (steady state not
documented); only 3 days; test was
unmeasured

Xia, X., Z. Dai, A.H. Rabearisoa, P. Zhao and X.
Jiang

2015a. Comparing humic substance and protein compound effects on the
bioaccumulation of perfluoroalky 1 substances b\ 1 )aphuia magna in water.
Chemosphere 119: 978-986.

Bioaccumulation (steady state not
documented); only 3 days; test was
unmeasured

Xia, X., A.H. Rabaerisoa, Z. Dai, X. Jiang, P. Zhao
and H. Wang

2015b. Inhibition effect of Na+ and Ca2 on the bioaccumulation of
perfluoroalkyl substances bv Daphnia magna in the presence of protein.
Environ. Toxicol. Chem. 34(2): 429-436.

Bioaccumulation (steady state not
documented); only 3 days; test was
unmeasured

Zhang, L., Y.Y. Li, T. Chen, W. Xia, Y. Zhou, Y.J.
Wan, Z.Q. Lv, G.Q. Li and S.Q. Xu

201 la. Abnormal dc\ elopmcnt of motor neurons in perfluorooctane
sulphonaic c\posal /chrafish embryos. Ecotoxicol. 20: 643-652.

Static, unmeasured exposure to single-
concentration (1 mg/L) from 6 hours post-
fertilization to 120 days post-fertilization

Zhang, L., Y.Y. Li, H.C. Zeng, J. Wei, Y.J. Wan. .1
Chen, S.Q. Xu

2u I | h MicrokW expression changes during zebrafish development induced
h> pciTliioiooclauc sulfonate .1 \ppl. Toxicol. 31: 210-222.

Poor control survival (>80% at 24 hour and
increasing)

J-5


-------
Appendix K EPA Methodology for Fitting Concentration-Response Data
and Calculating Effect Concentrations

Toxicity values, including LCso and ECio values, were independently-calculated from the
data presented in the toxicity studies meeting the inclusion criteria described above (see Section
2.10) and when adequate concentrations-response data were published in the study or could be
obtained from authors. When concentration-response data were not presented in toxicity studies,
concentration-response data were requested from study authors lo independently calculate
toxicity values. In cases where study authors did not respond to EPA's request lor data or were
unable to locate concentration-response data, the toxicity values were not independently-
calculated by EPA, and the reported toxicity values were retained for criteria deviation. EPA also
retained author-reported effect concentrations when data availability did not support effect
concentration calculation by EPA. This retention was done to be consistent with use of author-
reported toxicity values in pre\ ions criteria documents and retain informative toxicity values
(that would have otherw ise not been used only 011 the basis of lacking the underlying C-R data).
Where concentration-response data were a\ ailable, they were analyzed using the statistical
software program R (\ ersion .1 (•> 2) and the associated dose-response curve (drc) package.

I11 some cases, the author reported toxicity values were different than the corresponding
effect concentrations calculated by EPA. Overall, the magnitude of such discrepancies were
limited and largely occurred for several potential reasons such as: (1) instances where authors
were presumed to calculate effect concentrations using replicate level data but EPA only had
access to treatment mean data; (2) the model selected to fit a particular set of C-R data, and; (3)
the software used to fit a model to C-R data and calculate an effect concentration.

K-l


-------
K.1 Fitting Concentration Response Data in R

Concentration-response data were obtained from quantitatively-acceptable toxicity studies

when reported data were available. In many scenarios, toxicity studies report treatment-level
mean concentrations and mean organismal responses; however, individual-replicate data may
also be reported. When fitting C-R curves, replicate-level data was preferred over treatment-level
data, if both types of data were available. Within R, the drc package can fit a variety of
mathematical models to each set of C-R data.

K.l.l Fitting Acute Mortality Data
K. 1.1.1 Dichotomous Data

Dichotomous data are binary in nature (eg. Ii\ e dead or <> I) and are typical of survival

experiments. They are usually represented as a proportion snr\ i\ ed

K.1.2 Fitting Chronic Growth. Reproduction, and Slii\ i\ al Data
K. 1.2.1 Continuous Data

Continuous data take on any value along the real number line (e.g., biomass).

K. 1.2.2 Count / kna

Count data lake on only integer \ allies (e g . number of eggs hatched).

K. 1.2.3 Dichotomous / kua

Dichotomous data are hi nary in nature (e.g., live/dead or 0/1) and are typical of survival

experiments They are usually represented as a proportion survived.

K.2 Determining Most Robust Model Fit for Each C-R curve

The R drc package was used to fit a variety of models to each individual C-R dataset. A

single model was then selected from these candidate models to serve as the representative C-R
model. The selected model represented the most statistically-robust model available. To
determine the most-statistically-robust model for a C-R dataset, all individual model fits were
assessed on a suite of statistical metrics.

K-2


-------
K.2.1 Selecting Candidate Models

Initially, models were ranked according to the Akaike information criteria (AIC). The AIC

provides a measure of the amount of information lost for a given model by balancing goodness

of fit with model parsimony. The models with the lowest AIC, relative to other models based on

the same data, tend to be optimal. In some instances, however, the model with the lowest AIC

possessed a questionable characteristic that suggested said model was not the most appropriate.

Rather than selecting a model based solely on the lowest ATC. the initial ranking step was only

used to identify a subset of candidate models that were more closely examined before selecting a

model fit for each C-R dataset.

K.2.2 Assessment of Candidate Models to Determine the Most Appropriate Model

Candidate models (i.e., models with low AIC scores relative to other models produced for

a particular C-R dataset) were further evaluated based on additional statistical metrics to

determine a single, statistically robust curve for each quantitatively-acceptable toxicity test.

These additional statistical metrics were evaluated relative to the other candidate curve fits

produced for each C-R dataset Of these statistical metrics, residual standard errors, confidence

intervals relati\ e to effects concentration estimates, and confidence bands carried the most

weight in determining the most appropriate model to be representative of an individual C-R

dataset These additional statistical metrics included:

K.2.2.1 Comihitv.sy>ii of i\-siiiual standard errors

As with AIC. smaller values were desirable. Residual standard errors were judged

relative to other models.

K. 2.2.2 Width of confidence intervals for EC estimates

Confidence intervals were assessed on standard error relative to estimate and confirming

that the intervals were non-negative. Judged in absolute and relative to other models.

K-3


-------
K. 2.2.3 Width of confidence bands around the fitted model

A general visual inspection of the confidence bands for the fitted model. Wide bands in

the area of interest were undesirable. Judged in absolute and relative to other models.

K. 2.2.4 P-values of parameters estimates and goodness of fit tests

Hypothesis tests of parameter values to determine whether an estimate is significantly

different from zero. Goodness of fit tests were used to judge the o\ era 11 performance of the

model fit. Typically, the level of significance was set at 0.05. There may have been occasional

instances where the 0.05 criterion may not be met, but there was little recourse lor choosing

another model. Judged in absolute terms.

K.2.2.5 Residual plots

Residuals were examined for homoscedasticity and biasedness. Judged in absolute and

relative to other models.

K. 2.2.6 Overly influential observations

Observations were judged hased on Cook's distance and leverage. When an observation

was deemed overly inlluenlial. it was not reasonable to refit the model and exclude any overly

influential obser\ ations ui\ en the limited data a\ ailable with typical C-R curves. Judged in

absolute terms

K.3 Determining Curve Acceptability for use in Criteria Derivation

The final cui\ e fits selected for each of the quantitatively-acceptable toxicity tests were

further evaluated and classified to determine whether the curves were: 1) quantitatively-
acceptable for use, 2) qualitatively acceptable for use, or 3) unacceptable. To determine curve
acceptability for use in deriving an effect concentration, each individual curve was considered
based on the statistical metrics described above and assessed visually to compare how the
calculated effect concentration aligned with the underlying raw C-R data. Instead of evaluating

K-4


-------
curves fits relative to other curve fits for the same data (as was previously described to select the
most-robust curve for each test), curve fit metrics were used to assign each curve a score:

•	Quantitatively Acceptable Model. Model performed well on most/all statistical
metrics and resultant effect concentrations were typically used in a quantitative
manner.

•	Qualitatively Acceptable Model. Model generally performed well on statistical
metrics; however, the model presented some characteristic* s) that called estimates
into question. Such models were considered with caution. These problems may have
consisted of any number of issues such as a parameter with a high p-\ al lie, poor
goodness of fit p-value, wide confidence bands for fit or estimate interval, or
residuals that indicate model assumptions are not met Broadly, effect concentrations
from models that were deemed qualilali\ ely acceptable were not used numerically in
criteria deri\ alion if quantitatively acceptable models for different endpoints or tests
from the same publication were a\ ailable If quantitatively acceptable models for
different endpoints or tests from the same publication were not available, effect
concentrations from the qualitati\ ely acceptable model were used numerically in
criteria derivation 011 a case-by-case basis.

•	Unacceptable Model. Model poorly fit the data. These models were not used for
criteria deri\alion.

No single statistical metric can determine a given model's validity or appropriateness. Metrics
should be considered as a whole. As such, there is a slightly subjective component to these
evaluations. That said, this assessment scheme was developed to aid in evaluating models as to
their quantitative or qualitative attributes in a transparent and relatively repeatable manner.

K-5


-------
Appendix L Derivation of Acute Protective PFOS Benchmarks for

Estuarine/Marine Waters through a New Approach Method
(NAM): WeblCE

The 1985 Guidelines for Deriving Numerical National Water Quality Criteria for the
Protection of Aquatic Organisms and Their Uses (1985 Guidelines; Stephen et al. 1985)
recommend that data for a minimum of eight families be available lo fulfill taxonomic minimum
data requirements (MDRs) to calculate criteria values, including to calculate csluarine/marine
aquatic life criteria. Acute estuarine/marine test data are currently available lor only five of the
eight family MDRs (the dataset was missing another family in the Phylum Chordata. a family in
a phylum other than Chordata, and any other family); thus, l-l\\ was not able to derive an acute
estuarine/marine criterion element for PFOS Imscd on the WX5 Cinidelines MDR specifications
(Section 3.2.1.2). However, EPA was able to dc\ clop a draft acute Pl'OS protective benchmark
for aquatic life using a New Approach Methods (\.\MS) process, via the application of
Interspecies Correlation Fstimation (ICE) models (Raimondo et al. 2010). Although not a
criterion based on ] lM5 Guidelines MDR specifications, because of gaps in available data for
several of the taxonomic MDRs listed in the 1985 Guidelines for the derivation of aquatic life
criteria, this benchmark represents an aquatic life value derived to be protective of aquatic
communities The ICI v model predictions supplement the available test dataset to fulfill the
missing MDRs and allow the derivation of acute estuarine/marine benchmark recommendations
for aquatic life using procedures consistent with those in the 1985 Guidelines. This is important
as it provides an approach by which values that are protective of aquatic life communities can be
developed, even when MDRs are not fulfilled by PFOS test data. This approach is consistent
with both the 1985 Guidelines "good science" clause, EPA's interest in providing useful
information to states and tribes regarding protective values for aquatic life, and EPA's intention

L-l


-------
to reduce the use of animal testing via application of NAM S (https ://www. epa. gov/chemical-
research/epa-new-approach-methods-work-plan-reducing-use-animals-chemical-testing).

L.l Introduction to Web-ICE

ICE models, developed by EPA's Office of Research and Development, are log-linear
regressions of the acute toxicity (EC50/LC50) of two species across a range of chemicals, thus
representing the relationship of inherent sensitivity between those species (Raimondo et al.
2010). Each model is derived from an extensive, standardized database of acute toxicity values
by pairing each species with every other species for which acceptable toxicity data are available.
Once developed, ICE models can be used predict the sensiii\ iiy of an untested taxon (predicted
taxa are represented by the y-axis) from the known, measured sensitivity of a surrogate species
(represented by the x-axis) (Figure L-l).

ICE models have been developed lor a broad range of different of chemicals (e.g., metals
and other inorganics, pesticides. sol\ cuts, and reactive chemicals) and across a wide range of
toxicity values. There are approximately 3.4<)i) significant ICE models for aquatic animal and
plant species in the most recent \ersion of wcb-ICI- (v3.3, www3.epa.gov/webice, last updated
June 2') I (Raimondo et al 2<) I 5)

Models were validated using leave-one-out cross validation, which formed the basis for
the analyses of uncertainly and prediction robustness. For this process, each datapoint within the
model (representing the relative sensitivity of two species for a particular chemical) is
systematically removed, one at a time. The model is then redeveloped with the remaining data
(following each removal) and the removed value of the surrogate species is entered into the
model. The estimated value for the predicted species is then compared to the measured value for
that species (Raimondo et al 2010; Willming et al. 2016).

L-2


-------
ICE models have high prediction accuracy when values are derived from models with
robust parameters (e.g., mean square error, R2), that fall within a defined range of acceptability,
and with close prediction confidence intervals that facilitate evaluating the fit of the underlying
data (Brill et al. 2016; Raimondo et al. 2010; Willming et al. 2016). Results of these analyses
provides the basis of the user guidance for selecting ICE predicted toxicity with high confidence
(Box 1).

ICE models have undergone extensive peer review and their use has been supported for
multiple applications, including direct toxicity estimation for endangered species (NRC 2013)
Willming et al. 2016) and development of Species Sensitivity Distributions (SSI)s) (Awkerman
et al. 2014; Bejarano et al. 2017; Dyer et al. 2006; Dyer el al 2<>08; Raimondo et al. 2010. The
application of ICE-predicted values to de\ clop protective aquatic life values by multiple
independent, international groups confirms thai \ allies de\ eloped from ICE-generated SSDs
provides a level of protection thai is consistent with using measured laboratory data (Dyer et al.
2008; (Feng et al. 2<)|3. l-'ojui et al 2" 12a; Fojut et al. 2012b; Palumbo et al. 2012; Wang et al.
2020; Wu et al 2<~>1 5. Wu et al 2d I (•>. /hang et al. 2017). A recent external review of ICE
models additionally supports their use in regulatory applications based on the reliability of
underlying data, model trail spa rency. statistical robustness, predictive reliability, proof of
principle, applicability to probabilistic approaches, and reproducibility of model accuracy by
numerous independent research teams (Bejarano and Wheeler 2020).

L-3


-------
5

c
_g

4-

2

-1 -

0	2	4

6

ฃ

Rainbow trout (log LC50)

Figure L-l. Example ICE Model for Rainbow Trout (surrogate) and Atlantic Salmon
(predicted).

Each model datapoint is a common chemical that was tested in both species to develop a log-linear regression.

Box 1. ICE Model User Guidance Recommendedfor
Listed Species (Willming et al 2016):

Close taxonomic distance (within class)

•	Low MSE(<~ 0.95)

•	High R2 (>~ 0.6)

•	High slope (>~ 0.6)

•	Prediction confidence intervals should be used to
evaluate the prediction using professional
judgement for the application (Raimondo et al. in
prep).

•	For models between vertebrates and invertebrates,
using those with lower MSE or MOA-speciftc
models (not available for PFAS) has been
recommended for listed species predictions
(Willming et al. 2016).

L-4


-------
L.2 Application of Web-ICE with PFOS

ICE models are developed using a diversity of compounds (e.g., metals and other
inorganics, pesticides, solvents, and reactive chemicals) across a wide range of toxicity values;
however, PFAS are not included in web-ICE v3.3 due to the lack of available PFAS toxicity data
at the web-ICE v3.3 was created. PFAS acute values (typically reported as mg/L) can be greater
than those used to develop an ICE model (ICE database toxicity range IE"4 lo IE8 (J,g/L) such
that the input PFAS value of the surrogate would be outside the model domain In these cases, a
user can either enter the value as [j,g/L and allow the model to extrapolate beyond its range or
enter the toxicity as a "scaled" value (i.e., enter and estimate the value as mg/L) The principal
assumptions of ICE models are: 1) they represent the relationship of inherent sensitivity between
two species, which is conserved across chemicals, mechanisms of action, and ranges of toxicity;
and 2) the nature of a contaminant that was tested on the surrogate reflects the nature of the
contaminant in the predicted species (e.g., effect concentration (ECso) or lethal concentration
(LCso), percentage of acti\ e ingredient, technical grade; Raimondo et al. 2010). While neither of
these assumptions are \ iolated l\\ either extrapolating beyond the range of the model or using
scaled toxicity data, the uncertainty of using ICI- models in either manner had not been
thoroughly e\ aluated. Additionally, since PFAS were not included in the database used to
develop web-ICI- \ 3 3, the \ alidation of ICE models to accurately and specifically predict to
these compounds has not been previously explored. We address both these topics in the sections
below.

L.2.1 Prediction Accuracy of Web-ICE for Scaled Toxicity and Values Beyond the Model

Domain

The accuracy of using scaled toxicity data as input into ICE models was evaluated using
an analysis with the existing ICE models (v3.3) and is described in detail in Raimondo et al. (in

L-5


-------
prep). Briefly, ICE models containing a minimum of 10 datapoints and spanning at least five
orders of magnitude were separated into two subsets: 1) a lower subset that contained all paired
chemical data corresponding to values below the 75th percentile of surrogate species values; and
2) an upper subset containing paired chemical data above the 75th percentile of surrogate values.
The lower subset was used to develop "truncated" ICE models. The surrogate values in the upper
subset were converted to mg/L and entered into the truncated TCI- model. The predicted mg/L
value was compared to the respective value of the measured predicted species. Prediction
accuracy was determined as the fold difference (maximum of the predicted/measured and
measured/predicted) between the predicted and 1 Ik- measured value, consistent with previously
published evaluations of ICE models (Raimondo et al. 2<) I <). Willming et al. 2016). Accuracy of
using scaled toxicity as input into ICE models was compared to o\ crall ICE prediction accuracy
as previously reported and prediction accuracy of the rcspecti\ e upper subset data points that
were entered into the models as [iu I. (i.e . \nines heyoiul the model domain). Atotal of 3104
datapoints from 398 models were c\ aluated A match-paired comparison showed that the
average fold differences of toxicity \ allies predicted using scaled toxicity was not significantly
different than the respect i\ c a\ crauc fold differences of all cross-validated data points reported in
Willminu et al (2016) (Wilcoxon paired rank sum test, V = 42741, p-value 0.11). Additionally,
Raimondo et al (2d 10) and Wi liming et al (2016) showed a consistent and reproducible
relationship between the taxononiic distance of the predicted and surrogate species, which was
also reproduced using scaled values; the percentage of datapoints predicted using scaled toxicity
was within 5-fold of the measured value for over 94% of all validated datapoints for species pairs
within the same order, with a reduction in accuracy coinciding with decreasing taxonomic
relatedness (Raimondo et al. in prep). Comparison of scaled values with those predicted from

L-6


-------
[j,g/L values beyond the model domain showed that predicted values varied by a factor of 10 for
models with slopes ranging from 0.66 - 1.33. Toxicity values predicted from models with slopes
within this range had a median fold difference of 2.4 using mg/L values and 2.8 using [j,g/L
values (Wilcoxon paired rank sum test, V = 1334749, p-value 0.77). These results and a detailed
review of ICE model assumptions are provided in (Raimondo In prep).

L.2.2 Direct Comparison of Web-ICE and Measured Toxicity Values

Since limited PFOS toxicity test data are available for estuarine murine species, the
ability of ICE models to predict PFOS toxicity was evaluated using direct com pari sons of
freshwater species sensitivity as reported in the draft criteria document and predicted by web-
ICE. In this comparison, the measured species mean acute values (SMAVs) for PFOS reported in
Appendix A.l and Appendix B.lwere used as \ allies for surrogate species to predict to all
possible species that also had a measured PI-OS SM.W reported The available SMAVs for
PFOS that could be used as ICI- surrogate values along with the number of ICE models (i.e.,
potential predicted species) corresponding to each surrogate are shown in Table L-l.

Table L-l. Surrogate Species Measured Values for PFOS and Corresponding Number of
TCT. Models for Kacli Surrogate.

I\ปr cviniplc. I here ;ne 5 ' speeies for u hieli /' '.i/'lmi.i ni.i^u.i enn predict to\icit\

Broad

Species

pros

SM.W
(mg/1.)

Number of l(T.

Taxon

(0111111011 Name

Scientific

Models

Amphibian

Bullfrog

Lithobates catesbeiana a

133.3

9

Amphibian

African clawed frog

Xenopus laevis

15.99

2

Crustacean

Mysid

Americamysis bahia

4.914

28

Crustacean

Cladoceran

Daphnia magna

48.87

53

Fish

Zebrafish

Danio rerio

24.44

2 (juvenile models)
6 (embryo models)

Fish

Rainbow trout

Oncorhynchus mykiss

7.515

77

L-7


-------
lirosul
Tsixon

*1

(0111111011 N;ime

H'CicS
Scicnl il'ic

pros

SMAV
(mg/l.)

N il in her of l( T.
Models

Fish

Fathead minnow

Pimephales promelas

6.95

74

Mollusc

Fatmucket

Lampsilis siliquoidea

16.5

29

Mollusc

Black sandshell

Ligumia recta

13.5

1

a Lithobates catesbeianus was used in web-ICE

Table L-2 shows direct comparisons for PFOS measured and ICE-predicted values. The
regressions for these comparisons are provided in the Appendix L.2.6 Comparisons are limited
by the number of measured toxicity values and models available. To be included in this
comparison, a measured value was needed for both species in an ICE model pair I-'or direct
comparison of predicted and measured PFOS values, the measured SMAV of the surrogate
species is entered into a model for which the measured SMAV lor ihe intended predicted species
is also known. The PFOS toxicity predicted l\v this model is then compared to the measured
SMAV for the predicted species as listed in Appendix A I. Appendix B.l and Table L-l. This
allows both species of an ICI- model to ser\ c as either the predicted or surrogate species. The
exception to this was in cases in\ ol\ inu zebrafish embryos, as web-ICE v3.3 only included
models for which zebra fish embryos were used as surrogates. Accuracy of ICE predictions are
presented as the "fold-difference" between the measured and the predicted species, such that fold
difference is the maximum of the ratio of the predicted LCso/measured LCso or measured
LCso/predicted LO- Analyses of ICE prediction accuracy have shown that ICE models over-
and under-estimate toxicity values randomly, i.e., there is no systematic bias associated with the
models (Table Z.2., Raimondo et al. 2010, Raimondo et al. in prep). For accuracy assessments,
the fold difference provides a simplified metric to easily see how close predictions are to
measured values at a glance. A 5-fold difference has been demonstrated to be the average

L-8


-------
interlaboratory variability of acute aquatic toxicity tests and represents a conservative amount of
variance under standardized test conditions for a given life stage (Fairbrother 2008) Raimondo et
al. 2010). This inter-test variation can increase significantly where experimental variables differ
between tests; however, all ICE models are based on standardized life stages to minimize
extraneous variability (Raimondo et al. 2010).

These comparisons are consistent with web-ICE user guidance (Raimondo et al. 2015),
previously published reports on ICE model accuracy (Raimondo et al 2<> I <). Willming et al.
2016), and the above presented uncertainty analysis of using scaled toxicity as model input. ICE
models predict with acceptable accuracy for PFOS u lien invertebrates were used to predict to
invertebrate species and vertebrates were used to predict to \ cilebrate species in these
comparisons. Models validated across a wide range of species, chemicals, and toxicity values
show an acceptable level of prediction accuracy (>9')",, \ allies predicted within 5-fold of
measured value) when adhering lo the model guidance listed in Box 1 (Raimondo et al. 2010;
Willming et al. 2<) I (ฆ>)

The results summarized in Sections 1.21 and L.2.2 and described more thoroughly in
Raimondo et al. (in prep) demonstrate that the relationship of inherent sensitivity represented by
ICE models is preserved across ta\a. chemicals, and range of toxicity values when using robust
ICE models While the current analysis uses freshwater species to predict to estuarine/marine
species, previous model \ alidation and uncertainty analyses did not indicate the habitat of the
species to be an influential source of ICE model uncertainty (Raimondo et al. 2010; Willming et
al. 2016).

L-9


-------
Table L-2. Comparison of ICE-predicted and measured values of PFOS for species using both scaled values (entered as mg/L) and values
potentially beyond the model domain (entered as jig/L) (Raimondo et al. in prep).

Measured SMAVs are for the predicted species as listed in Appendix A. 1, Appendix B. 1 and Table L-l. Footnotes indicate where predictions or models do not meet one or more





To\ici(> \'sillies Polcnlisill.t lic.tond Model Domsiin

Scsilcd To\ici(\ \ sillies

Prcdiclcd Species

SiiitoปsiIc Species

Mcsisurcd
SMAY

-l( 1.
Predicted

Confidence
lnler\sil

(inii/l.)

l-old
Difference

Mull frog
(Lithobates
catesbeianus)

Daphnid

(.Daphnia magna)

133,300

56338.67

11646.07 -
272542.13

2.4

1 ^ ^ ^

59.73

6.99-510.48

2.2a



Fathead minnow
(Pimephales promelas)



8356.68

3748.61 - I8(.:>J 28

16.0



13.26

4.13-42.57

10.1



Rainbow trout
(Oncorhynchus mykiss)



15140.53

8139.36 - 281<> ^ 81

8.8



33.9

13.63 -84.26

3.9

African clawed frog
(Xenopus laevis)

Fathead minnow
(Pimephales promelas)

15,990

7034.49

800 65 - 61804. '5



15.99

18.93

0.306- 1170.65

1.2ab

Mysid

(Americamysis bahia)

Daphnid

(Daphnia magna)

4,914

8774.84

4^88 ()() - I 54 (4

1 8

4.914

27.16

18.60 -39.67

5.5



Fathead minnow
(Pimephales promelas)



359.91

1 35.34 - 957.15

13. T



0.481

0.104-2.21

10.2ฐ



Rainbow trout
(Oncorhynchus mykiss)



i r:

702.88 - 1955.47

4.2ฐ



2.01

1.08-3.75

2.4ฐ

Daphnid

(Daphnia magna)

Bullfrog

(Lithobates catesbeianus)

4X.X~<)

XI'Ui. ()4

17394.84 -
386042.67

1.7

48.87

199.47

32.95 - 1207.24

4.1



Fathead minnow
(Pimephales promelas)



li.T 85

1 14^.29 - 2508.22

28.8ฐ



3.29

1.36-7.96

14.9ฐ



Fatmucket

(Lampsilis silic/iioidea)



:^i:: 84

"<.U 81 -70030.01

2.1



7.73

1.46-40.85

6.3b



Mysid

(.Americani vsis bahia)



<,o<>(, "5

3829.31 - 9706.79

8.0



21.29

13.73 - 33.02

2.3



Rainbow trout
(Oncorhynchus mvkiss)



:""5 45

2007.74 - 3836.72

17.6ฐ



8.83

5.26 - 14.80

5.5ฐ



Zebrafish embryo
(Danio rerio - embrvo)



^>26.87

910.19- 16941.77

12.4ฐ



2.48

0.143 -42.99

-yabc

L-10


-------




Toxicity Y;ducs Polcnlhdl.t lic.tond Model Dninnin

Sc;ded l"ซi\icil> Y;dues

Prcdiclcd Species

Siiitoป;i1c Species

Measured
SMAY

(.11 Si/I.)

\\cl>-l< 1.
Prcdiclcd

(.llii/l.)

')5Vii Confidence
1 iilei*\ids Uiii/I.)

l-old
DilTcrence

Measured
SMAY
(ni!ป/l.)

\\el>-l( 1.
Predicled
(lllii/l.)

Confidence
ln(cr\;il

(niii/l.)

l-old
Difference

Rainbow Ironl

(Oncorhynchus mykiss)

Bullfrog

(Lithobates catesbeianus)

7,515

823^5.25

;s:4" 4S -

177501.32

1 1 o

7.515

39.73

16.29 -96.91

5.3



Daphnid

(Daphnia magna)



21354.25

14550.83 - 31338.69

2.8C



236.65

174.72 - 320.52

31.5ฐ



Fathead minnow
(Pimephales promelas)



2771.13

2136.90 - 3593.60

2 "



3.43

2.01 -5.85

2.2



Fatmucket

(Lampsilis siliquoidea)



48028.61

3264.96 - 706515.68

6.4ac



13.14

1.03 - 167.64

^ 'yabc



Mysid

(Americamysis bahia)



6169.68

3855.11) - wn K) 1

1.2



68.63

45.85 - 102.73

9.1



Zebrafish embryo
(Danio rerio - embryo)



10047.33

3682.52 - 2~412 l>5

1.3



2.97

0.513 - 17.19

2.5b

Fathead minnow
(Pimephales promelas)

African clawed frog
(Xenopus laevis)

6,950

16080.14

1020 G1 - 253332

9 oa

6.95

7.89

0.071 -868.40

l.lab



Bullfrog

{Lithobates catesbeianus)



121541.51

44-4 20-
v,,,:o4 os

r.5



91.08

33.84 -245.08

13.1



Daphnid

(.Daphnia magna)



45004.86

2^()^5 "2 - (>Kh< 12 SS

(> 5"



687.68

456.02 -
1037.02

98.9ฐ



Fatmucket

(Lampsilis siliquoidea)



1 l(.(.(.ซJ,9

I'M" 15 -

698863.06

lo.8ฐ



595.88

48.52 -7317.09

85.7abc



Mysid

(.Americamysis bahia)



1

5348.62 - 34952.76

2.0ฐ



254.88

118.25 -549.38

36.7ฐ



Rainbow trout
(Oncorhynchus mvkiss)



I4424.T

1 K>2S .()- 18867.80

2.1



36.58

23.89 -56.02

5.3



Zebrafish embryo
(Danio rerio - embryo)



2_x



22"'ซi 4

8979.24 - 57844.79

1.4



132.33

45.71 -383.08

8.0



Fathead minnow
(Pimephales promelas)



~ 17.52

149.82 -3436.35

23.0ฐ



3.21

0.065 - 158.39

5.1abc



Rainbow trout
(Oncorhynchus mykiss)



1585.37

485.38 -5178.24

10.4ฐ



44.11

9.18-211.95

2.7ฐ

Black sandshell
(Ligumia recta)

Fatmucket

(Lampsilis siliquoidea)

13,500

19191.22

4438.79 -82973.68

1.4

13.5

26.59

1.49-472.22

2.0ab

L-ll


-------
a Confidence interval >1.5 order magnitude
b Input data outside model range

0 Guidance for model mean square error, R2, and/or slope not met.

L-12


-------
L.2.3 Prediction of Estuarine/Marine Species Sensitivity to PFOS

A value of PFOS sensitivity was predicted with web-ICE v3.3 for all possible species
using all available surrogate species (Table L-l). Predicted values were obtained by entering all
available surrogate species into the web-ICE SSD generator, which predicts to all possible
species from all available surrogates simultaneously and exports results into an excel
spreadsheet. Web-ICE results were generated using both mg/I. and iiu L values to evaluate the
full set of possible predictions using both units of measure against the model domain, confidence
intervals, and model parameters. First, all available models were evaluated based on the
parameter (MSE, R2, slope) guidance in Box 1, which are the same for an ICE species pair
regardless of input value (Table L-3). Models that did not meet the parameter criteria in Box 1
were rejected in this first pass. In the next slep. \ allies that ^ere predicted using [j,g/L were
evaluated against the model domain and selected lor the next tier of evaluation when the
surrogate value was within the range of data used to de-\ clop the model. If the surrogate value
reported as (a,g/I. was beyond the model domain, the mg/L value was evaluated if it was within
the model domain and if the model slope was between 0.66-1.33 (Raimondo (Inet al. in prep).
Cases in which both units were outside the model domain were not included quantitatively, but
the value with the narrowest confidence intervals was included for qualitative considerations.
Values (using either iiu I. or mg/L input value) were excluded quantitatively from the SMAVs
but retained for qualilali\ e consideration if an evaluation of confidence intervals, model
parameters, and the model domain indicated the relationship between surrogate and predicted
species was not informed by robust underlying data. At this stage, specific predictions should be
based on holistic evaluation of all available information provided by the model, confidence
interval, and data used to develop the model. Decisions to exclude a prediction from the SMAV

L-13


-------
are clarified in footnotes. Because the sensitivity of a single species can be predicted by multiple
surrogates, we calculated the SMAV where multiple robust models were available for a predicted
species. Each predicted species was then assigned to the appropriate saltwater MDRs as defined
in the 1985 Guidelines.

Saltwater MDRs:

a.	Family in the phylum Chordata

b.	Family in the phylum Chordata

c.	Either the Mysidae or Penaeidae family

d.	Family in a phylum other than Arthropoda or Chordala

e.	Family in a phylum other than Chordata

f.	Family in a phylum other than Chordata

g.	Family in a phylum other than Chordata

h.	Any other family

The acute sensitivity of estuarine marine species lo PTOS is presented in Table L-4. A
total of 36 models representing 19 estuarine murine species were a\ ailable in web-ICE to predict
the toxicity of PFOS to saltwater species (Table L-3). Of these. 12 models were initially rejected
based on model parameters not meeting the guidance in Box 1, reducing the number of predicted
species to 17 represented by 24 models I'lirther evaluation of ICE predictions resulted in 12
SMAYs The range of sensiti\ ily lor the predicted taxa is consistent with the range of sensitivity
of freshwater species for this compound

L-14


-------
Table L-3. All ICE Models Available in web-ICE v3.3 for Saltwater Predicted Species Based on Surrogates with Measured PFOS.

Model parameters are used to evaluate prediction robustness. Cross-validation success is the percentage of all model data that were predicted within 5-fold of the measured value
through leave-one-out cross-validation (Willming et al. 2016). Taxonomic distance describes the relationship between surrogate and predicted species (e.g., 1 = shared genus, 2 =



































Dimw





Menu

Model

*ซim riiซiili-

( I'ovs













ฆ .I





*>(|ii:irc

Minimum

Model

\ iiliilsiiioii













1' ivrllnlll





Iitmi

\ :iluc

M;i\illllllll



1 :i\Miiiiiiiic



1*1vilified N|irdrป

*>IIITii!!iilr



Inlrri r|>l

i\ 21

K;

|> \illllr

i mm-: i

ill!! 1 i

\ illllr (ill! 1 i

l""l

l)i-l;illii-

1 -i- ill ( rilcrhi

Acartia tonsa

Daphnia magna

0.59

1.31

2

0.91

0.0443

0.17

2.24

38514.70

50

5

Rejected

Allorchestes compressa

Daphnia magna

0.83

1.59

3

0.8

0.039

0.12

5.00

184.54

100

5

Accepted

Allorchestes compressa

Pimephales promelas

0.84

0.15

3

0.96

0.0028

0.02

163.05

26895.72

100

6

Accepted

Americamysis bahia

Daphnia magna

0.83

0.02

160

0.68

<0.001

0.93

0.07

840000.00

64

5

Accepted

Americamysis bahia

Oncorhynchus mykiss

0.92

-0.5

150

0.6

<0.001

1.08

0.06

1100000.00

57

6

Rejected

Americamysis bahia

Pimephales promelas

0.95

-1.12

46

0.55

<0.001

1.75

2.27

70200000.00

35

6

Rejected

Chelon labrosus

Lampsilis siliquoidea

1.27

1.5

1

0.99

0.0403

0

19.01

281.00

na

6

Accepted qualitatively

Chelon macrolepis

Pimephales promelas

1.51

-1.04

2

0.97

0.0114

0.05

26.00

2533.38

100

4

Accepted qualitatively

Crassostrea virginica

Americamysis bahia

0.44

1.76

114

0.34

<0.001

0.88

0.003

117648.20

55

6

Rejected

Crassostrea virginica

Daphnia magna

0.44

1.54

116

0.28

<0.001

1.08

0.08

137171.43

58

6

Rejected

Crassostrea virginica

Lampsilis siliquoidea

0.82

-0.28

3

0.95

0.0041

0.06

30.00

22000.00

100

4

Accepted

Crassostrea virginica

Oncorhynchus mykiss

0.59

0.97

120

0.5

<0.001

0.68

0.02

570000.00

68

6

Rejected

Crassostrea virginica

Pimephales promelas

0.75

0.44

24

0.61

<0.001

0.68

1.24

206300.75

69

6

Accepted

Cyprinodon bovinus

Oncorhynchus mykiss

0.72

0.8

2

0.91

0.0427

0.08

4.93

1637.92

100

4

Accepted qualitatively

Cyprinodon bovinus

Pimephales promelas

0.67

0.65

2

0.99

0.0043

0

10.49

7847.42

100

4

Accepted

Cyprinodon variegatus

Americamysis bahia

0.57

1.88

88

0.56

<0.001

0.67

0.003

182000.00

64

6

Rejected

Cyprinodon variegatus

Daphnia magna

0.53

1.79

84

0.49

<0.001

0.72

0.08

304000.00

64

6

Rejected

Cyprinodon variegatus

Lampsilis siliquoidea

0.72

0.76

1

0.99

0.0392

0

30.00

22000.00

na

6

Accepted qualitatively

Cyprinodon variegatus

Oncorhynchus mykiss

0.75

0.9

87

0.65

<0.001

0.56

0.82

12700000.00

75

4

Accepted

Cyprinodon variegatus

Pimephales promelas

0.69

0.98

24

0.74

<0.001

0.43

2.27

16500000.00

77

4

Accepted

Farfantepenaeus duorarum

Americamysis bahia

1.03

0.06

6

0.81

0.0022

0.55

0.01

720.00

50

4

Accepted

Farfantepenaeus duorarum

Daphnia magna

1.08

0.14

16

0.76

<0.001

1.32

0.04

65686.02

44

5

Rejected

Farfantepenaeus duorarum

Oncorhynchus mykiss

1.2

-1.36

15

0.72

<0.001

1.54

0.57

221000.00

47

6

Rejected

Fenneropenaeus merguiensis

Daphnia magna

0.82

1.43

4

0.66

0.0473

0.4

5.00

1251.41

67

5

Accepted

Gasterosteus aculeatus

Oncorhynchus mykiss

1.05

0.29

4

0.9

0.0038

0.18

0.61

890.00

83

4

Accepted

Hydroides elegans

Daphnia magna

0.49

1.59

2

0.96

0.0182

0.01

5.00

1251.41

100

6

Rejected

Hydroides elegans

Oncorhynchus mykiss

0.2

2.3

1

0.99

0.0179

0

1.84

13390.93

na

6

Rejected

Litopenaeus stylirostris

Americamysis bahia

1.04

0.01

5

0.6

0.0401

0.29

0.58

24.09

57

4

Accepted

Menidia menidia

Oncorhynchus mykiss

1.28

-1.4

3

0.94

0.005

0.23

11.24

91000.00

60

4

Accepted qualitatively

Menidia peninsulae

Americamysis bahia

0.63

0.91

3

0.88

0.0162

0.32

0.01

1160.00

80

6

Accepted qualitatively

L-15


-------
I'lVlliclrll N|irdrป

*>IIITii!!iilr

*ปlii|U-

Inlrri r|>l

Dimw
ฆ .I

1-1vcilnm

|N 2 i

K;

|> Mllllr

Menu

*>(|ii:irc
Iitmi

iMM.i

Model
Minimum
\ nliir
(ll!i 1 1

MimI.I
M;i\illllllll
\ nliir i n;! 1 i

( I'ovs
\ iiliilsiiioii
simtss

1 ilMIIIMIIlic

Di'-liimr

1 -i- ill ( rilcrhi

\ lenidia prninsnlai'

Oncorhynchus mykiss

1.01

-0.36

2

0.91

0.0421

0.35

0.82

1 ()( )( ) 00

50

4

Accepted qualitatively

Metamysidopsis insularis

Daphnia magna

0.86

0.93

3

0.94

0.0057

0.18

6.97

317472.74

80

5

Accepted

Metamysidopsis insularis

Lampsilis siliquoidea

1.03

0.62

2

0.99

0.0027

0.02

19.01

87705.88

75

6

Accepted

Mugil cephalus

Oncorhynchus mykiss

1.44

-0.37

3

0.89

0.0144

0.12

0.82

29.18

100

4

Accepted qualitatively

Tigriopus japonicus

Pimephales promelas

0.81

1.12

5

0.76

0.0103

0.11

195.14

27000.00

86

6

Accepted

Tisbe battagliai

Daphnia magna

0.86

1.25

2

0.94

0.0289

0.08

0.61

184.54

100

5

Accepted

L-16


-------
Table L-4. ICE-Estimated Species Sensitivity to PFOS.

Values in bold and underlined are used for SMAV.









Ksti milled











1 ii put

Toxicity

<)5% Confidence



Com 111011 Nn 1110

Scientific

SlllTOปiltO

I nil

(mป/L)

lntor\ ills (mป/l.)

SMAV

Calanoid copepod

Acartia tonsa

Daphnia magna

LI, 1.

i: mi

0.66-244.8

NA

Amphipod

Allorchestes compressa

Daphnia magna

m, 1.

1020.12

310.42 - 3352.37

49.69





Pimephales promelas

LI, 1.

2.42

1.29-4.54



Mysid

Americamysis bahia

Daphnia magna

LI, 1.

8.77

4.99- 15.44

8.77





Oncorhynchus mykiss

(ig/L

1.17

0.70 - 1.96







Pimephales promelas

Liir/I,

0.36c

0.14-0.96



Thicklip mullet

Chelon labrosus

Lampsilis siliquoidea

in, 1.

1 I44.93ab

126.12 - 10393.70

NA

Bigscale mullet

Chelon macrolepis

Pimephales promelas

LI, L

61.79ab

4.94-772.16

NA

Eastern oyster

Crassostrea virginica

Americamysis bahia

(ig/L

:.52c

1.45-4.37

1.886





Daphnia magna

Mg/L

4 iw

1.97-8.91







Lampsilis siliquoidea

LI, 1.

1.56

0.44-5.55







Oncorhynchus mykiss

LI, 1.

2.or

1.30-3.10







Pimephales promelas

f-ig/L

2.28

0.78-6.67



Leon springs pupfish

Cyprinodon bovinus

Oncorhynchus mykiss

mg/L

21.5T

3.20 -236.94

1.82





Pimephales promelas

l-ig/L

1.82

0.78-4.24



Sheepshead minnow

Cyprinodon varietal us

Americamysis bahia

(ig/L

9.87c

5.58 - 17.46

5.769





Daphnia magna

(ig/L

19.69c

9.49-40.84







Lampsilis siliquoidea

(ig/L

6.76a

0.56-81.92







Oncorhynchus mykiss

(ig/L

7.08

4.53 - 11.06







Pimephales promelas

(ig/L

4.70

2.32-9.52



Pink shrimp

Farfantepenaeus duorarum

.1 mericamysis bahia

mg/L

6.02

1.34-26.97

6.02





Daphnia magna

(ig/L

162.46c

14.13 - 1868.00







Oncorhynchus mykiss

(ig/L

2.12c

0.38 - 11.71



Banana prawn

Fenneropenaeus merguiensis

Daphnia magna

mg/L

688.07

124.18 - 3812.56

688.07

Threespine stickleback

Gasterosteus aculeatus

Oncorhynchus mykiss

mg/L

16.46

5.22-51.84

16.46

Polychaete

Flydroides elegans

Daphnia magna

(ig/L

8.20bc

1.29-52.12

NA





Oncorhynchus mykiss

(ig/L

1.28c

0.89 - 1.83



Blue shrimp

Litopenaeus stylirostris

Americamysis bahia

mg/L

5.41

1.59- 18.41

5.41

Atlantic silverside

Menidia menidia

Oncorhynchus mykiss

(ig/L

3.97a

0.52 - 30.32

NA

L-17


-------








Ksti milled











1 ii put

Toxicity

95% Confidence



Com moil Nn 1110

Scientific

SlllTOปJllO

I nil

(niซ/l.)

Inlonills (mป/l.)

S\1 AY

Tulcw alor si 1 \ cisidc

Menidia penin.su/ae

Americamvsis hahia

mg 1.

:: fo

3 47 -147 11

\ A





Oncorhynchus mykiss

mg/ L

3.35

U.U95 - 11S.0U



Mysid

Metamysidopsis insularis

Daphnia magna

mg/L

245.18

45.26 - 1328.00

152.2





Lampsilis siliquoidea

l-ig/L

94.52

27.87-320.53



Striped mullet

Mugil cephalus

Oncorhynchus mykiss

mg/L

7.66d

2.17-27.01

NA

Harpacticoid copepod

Tigriopus japonicus

Pimephales promelas

(.ig/L

18.04

7.20-45.24

18.04

Harpacticoid copepod

Tisbe battagliai

Daphnia magna

mg/L

522.86

11)3 X4- 2632.73

522.86

Calanoid copepod

Acartia tonsa

Daphnia magna

Llg 1,

12.66abc

ii hh . 244.8

NA

NA = Not Available

a Both confidence intervals >1.5 order magnitude
b Input data outside model range

0 Guidance for model mean square error, R2, and/or slope not met
d Does not meet slope criteria for using scaled toxicity (0.66-1.33)

L-18


-------
L.2.4 Derivation of Acute Water Quality Benchmark for Estuarine/Marine Water

The web-ICE predicted acute dataset for PFOS contains 15 genera, representing the eight
MDR groups that would be necessary for developing an estuarine/marine criterion. EPA fulfilled
these eight MDRs by integrating the acceptable quantitative study data (discussed in Section
3.1.1.2) with data derived using web-ICE to support calculating a protective benchmark. In
scenarios where both empirical LCso values and estimated LC;.. \ allies were available for the
same species, only the empirical data were used to derive the species mean acute value. The
ranked GMAVs for these combined data along with the MDR met by each G\l.\V is
summarized in Table L-5. From this dataset, an acute benchmark was calculated using
procedures consistent with the 1985 Guidelines and with those used for the derivation of
freshwater criteria values for PFOS. GMAVs lor the lour most sensitive genera were within a
factor of 1.7 of each other (Table L-6). The cstuarine murine FAY (the 5th percentile of the genus
sensitivity distribution) lor PTOS is I 096 mg/l. (Table I .-(•>) The FAV is lower than all of the
GMAVs for both the tested species and for \ allies derived using web-ICE. The FAV was then
divided by two to obtain a concentration yielding a minimal effects acute benchmark. The
FAV'2. which is the estuarine marine acute water column benchmark magnitude, is 0.55 mg/L
PFOS (rounded to two signillcant figures) and is expected to be protective of 95% of
estuarine/marine genera potentially exposed to PFOS under short-term conditions of one-hour of
duration, if the one-hour a\ crime magnitude is not exceeded more than once in three years
(Figure L-2). This draft acute benchmark for estuarine/marine aquatic life is lower than the
recommended acute freshwater criterion (3.0 mg/L), suggesting that estuarine/marine species
may be more acutely sensitive to PFOS and emphasizing the importance of having a separate
benchmark value for the protection of estuarine/marine aquatic life.

L-19


-------
Table L-5. Ranked Estuarine/Marine Genus Mean Acute Values.

Values in bold were derived from empirical toxicity tests with the species

MDR
(.roup

N;i mi'

Species Milesla^e)

SM.W

CiM.W

Rank

Pcrccnlile

D

Mediterranean mussel

Mytilus galloprovincialis

1.1

1.1

1

0.06

F

Purple sea urchin

Strongylocentrotus purpuratus

1.7

1.7

2

0.13

E

Sea urchin

Paracentrotus lividus

1.795

1.795

3

0.19

D

Eastern oyster

Crassostrea virginica

1.886

1.886

4

0.25

C

Mysid

Americamysis bahia

4.914

4.914

5

0.31

A

Leon springs pupfish

Cyprinodon bovinus

i s:

5.225

6

0.38

Sheepshead minnow

Cyprinodon variegatus

> 15

F

Blue shrimp

Litopenaeus stylirostris

5.41

5 41

7

0.44

F

Pink shrimp

Farfantepenaeus duorarum

6.02

<> u:

8

0.50

C

Mysid

Siriella armata

6.9

6.9

9

0.56

B

Threespine stickleback

Gasterosteus aculeatus

16.46

lo.4<>

10

0.63

G

Harpacticoid copepod

Tigriopus japonicus

18.04

18.04

11

0.69

E

Amphipod

Allorchestes compressa

49.69

49.69

12

0.75

C

Mysid

Metamysidopsis insularis

152.2

152.2

13

0.81

H

Harpacticoid copepod

Tisbe battagliai

522.9

522.9

14

0.88

F

Banana prawn

Fenneropenaeus nierguiensis

(.XX 1

688.1

15

0.94

MDR Groups

a.	Family in the phylum Choixlata

b.	Family in the phylum Choixlata

c.	Either the Mysidae or Penaeidae family

d.	Family in a phylum other than Arthropoda or Chordata

e.	Family in a phylum other than Choixlata

f.	Family in a phylum other (hail Choixlata
g Family in a phylum other than Chordata
h Any oilier family

L-20


-------
Table L-6. Estuarine/Marine Final Acute Value and Protective Aquatic Acute Benchmark.

Bold values represent genera for which empirical toxicity data were available.	

Calculated Estuarine/Marine FAV based on 4 lowest values; n=15

Rank

Genus

GMAV
(mg/L)

ln(GMAV)

ln(GMAV)2

P=R/(N+1)

sqrt(P)

1

Mytilus

1.1

0.10

0.01

0.063

0.250

2

Strongylocentrotus

1.7

0.53

0.28

0.125

0.354

3

Paracentrotus

1.795

0.59

0.34

0.188

0.433

4

Crassostrea

1.886

0.63

0.40

0.250

0.500



E (Sum):

1.85

1.04

0.63

1.54

S2 =

L =

A =
FAV =
PVAL=

5.30 S = slope
-0.423 L = X-axis intercept
0.092 A = InFAV
1.096 P = cumulative probability
0.55 mg/L PFOS (rounded to two significant figures)





1.00
0.90
0.80



ซ 0.70

0.60 --

ฆa 0.50
s

0i

s

eu
u

s

0.40
0.30
0.20
0.10
0.00

"



Fenneropenaeus ~

"



Tisbe battagliai ~

--



~ Metamysidopsis

ฆ



~ Allorchestes



~ Tigriopus



O Gasterosteus

ฆ

ฆ Siriella





~ Farfantepenaens

ฆ

~ Litopenaeus

• Fish (Empirical and WeblCE)

-

# Cyprinodon

O Fish (WeblCE)



ฆ Americamysis

ฆ Invertebrate (Empirical)

ฆ

A Crassostrea

~ Invertebrate (WeblCE)



ฆ Paracentrotus

A Mollusk (Empirical)

:

ฆ Strongylocentrotus

A Mollusk (WeblCE)

-

~ Mytilus

	Acute Benchmark







0.1

1	10	100

Genus Mean Acute Value (mg/L PFOS)

1000

Figure L-2. Ranked Estuarine/Marine Acute PFOS GMAVs used for the Aquatic Life
Acute Benchmark Calculation.

L-21


-------
L.2.5 Estuarine Marine/Benchmark Uncertainty

Epistemic uncertainty of individual ICE estimates used for SMAV calculation was

quantified through the calculation of corresponding 95% confidence intervals for each ICE

estimate. Of the individual models and resultant ICE-estimated LCso values estimates from the

available and quantitatively acceptable models (see bolded and underlined values in Table L-4; n

=16), the range of individual 95% CIs (i.e., 95% CI range = upper ^5" o CI - lower 95% CI) as a

percent of the corresponding LCso estimate (i.e., = [95% CT range l.('5<) estimate]* 100) ranged

from 92.23% to 536.05%. The ICE model with the lowest 95% CI range relali\ c to the LCso

estimate (i.e., 92.23%) employed Oncorhynchus myluss as the predictor species and ('yprinodon

variegatus as the predicted species. The ICE model ^ilh the largest 95% CI range relative to the

LCso estimate (i.e., 536.04%) employed / hi/>lniia magna as the predictor species and

Fenneropenaeus merguiensis as the predicted species I i fleen of the I (•> ICE-predicted values in

Table L-4 that ^ere used for S\f.\Y calculation had 95".. CI ranges that were greater than the

corresponding I.estimate (i e. ^5"..CI range was > 100% of the LC50 estimate). The

relatively wide ranging ^5".. CIs demonstrate the underlying uncertainty in the PFOS

estuarine marine benchmark

Six of the 15 GMAVs used to deri\e the acute PFOS estuarine/marine benchmark were

based on empirical toxicity tests. Interestingly, the six GMAVs based on empirical data were not

evenly distributed across the GSD, with all empirical data falling below the 60th percentile of

sensitivity (Table L-2) Also, three of the four most sensitive GMAVs in the GSD (Figure L-2)

were based on empirical data and five of the six most sensitive GMAVs were based empirical

acute values, meaning final estuarine/benchmark magnitude was primarily based on relatively

certain empirical toxicity tests and the inherent uncertainty in the ICE models had little influence

on the final acute estuarine/marine benchmark magnitude.

L-22


-------
It is unclear if ICE-estimated data were typically greater than empirical data because of a
simple coincidence or a systematic mechanistic reason. A systematic mechanistic reason why
ICE-estimated acute values were greater than empirical acute values could be attributed to the
use of freshwater species to predict to estuarine/marine species in the ICE regressions. For
example, estuarine/marine LCso values from quantitatively acceptable studies (Appendix B.l)
were typically lower than acute LCso values for freshwater species ( Appendix A.l). The apparent
increase in PFOS toxicity in estuarine/marine environments relative lo fresh waters may represent
a unique toxicological consideration of PFOS (and possibly other PFAS) lha l was not a
toxicological attribute of the other chemicals used to build the supporting ICE models, which
would result in artificially high PFOS LCso estimates for estuarine/marine species.

The estuarine/marine benchmark slill appeal s adequately protective based on the
available high quality empirical data (Appendix 1} I) The acule PFOS estuarine/marine
benchmark (i.e.. <> 55 mu I.) is two times lower than the lowest GMAV (i.e., 1.1 mg/L), which
was based on empirical data lor lyiilns. EPA further evaluated the appropriateness of the
estuarine/marine benchmark In comparing it to empirical, but qualitatively acceptable, data for
estuarine marine species l-l\\ specifically focused on qualitatively-acceptable estuarine/marine
tests reported in Table H. 1 that (1) tested an animal species; (2) exposed test organisms to a
PFOS for a continuous exposure duration that was reasonably similar to standard acute
exposures (e.g., 4K hours to seven days); (3) reported acute apical effects; and (4) reported effect
concentrations that were lower than the acute estuarine/marine benchmark final acute value (i.e.,
1.096 mg/L). EPA identified three individual tests in Table H.l as meeting the previous criteria:
1. Park et al. (2015) conducted a seven day test with the mud crab, Macrophthalmus
japonicus. Exposures lasted seven days, but survival was also recorded at 96 hours. The

L-23


-------
author did not calculate an LCso, but at 96 hours there was 36% mortality in the highest
test concentration (i.e., 0.03 mg/L). Therefore, the 96-hour LCso was >0.03 mg/L. The
test was not used quantitatively because an LCso could not be calculated based on the
three exposure concentrations used. Overall, 36% mortality after 96 hours in the 0.03
mg/L treatment suggests this species may be sensitive to acute PFOS exposures relative
to the acute estuarine/marine benchmark. However, the source of the organisms (fish
market) could be problematic as there is no mention of potential pre\ ions exposure or
measures of PFOS in test organisms at any point during the experiment

2.	Mhadhbi et al. (2012) conducted a 6-day test with the turbot, Scophthalmns maximus.
Endpoints included dead embryos, malformation, hatch success at 48 hours and larvae
survival (missing heartbeat and a non-detached tail) al six days The 6-day LCso of 0.11
mg/L PFOS was not acceptable for acute hench mark deri\ ation because of the relatively
long exposure duration Nevertheless, the 0-day LC;.. is nearly an order of magnitude
lower than the acute estuai ine marine benchmark final acute value (i.e., 1.096 mg/L) and
five times lower than the acute estuai ine marine benchmark, suggesting S. maximus is
sensiti\ e to acute PI OA exposures at concentrations below the acute estuarine/marine
benchmark

3.	Jeon et al (2010c) performed a 6-day test on blackrock fish, Sebastes schlegeli. There
were no significant differences in total length, weight and survival (no mortality observed
in any of the exposures) over the 6-day exposure. The NOEC (survival and growth) was 1
mg/L at each test salinity (10, 17.5, 25 and 34 ppt), which is less than the acute
estuarine/marine benchmark final acute value (i.e., 1.096 mg/L). The lack of effects

L-24


-------
observed at 1.0 mg/L preclude this test from providing meaningful information about the
protectiveness of the acute estuarine/marine benchmark.

Results from Mhadhbi et al. (2012), which was determined to only be acceptable for
qualitative use, suggests S. mctximus is sensitive to acute PFOA exposures at concentrations
below the acute estuarine/marine benchmark but at an exposure duration that was 50% longer
than the standard 96-hour exposure duration from quantitatively acceptable tests. Additionally,
results of quantitatively acceptable empirical toxicity studies with estuarine/marine organisms do
not provide any evidence that the aquatic estuarine/marine community will experience
unacceptable acute effects at the acute estuarine/marine PFOA benchmark.

L.2.6 ICE Regressions Supporting the Acute Estuarine/Marine Benchmark

-2	0	2	4	6

Americamysis bahia
(Log LC50)

Figure L-3. Americamysis bahia (X-axis) and Daphnia magna (Y-axis) regression model
used for ICE predicted values.

L-25


-------
-2	0	2	4

Americamysis bahia
(Log LC50)

Figure L-4. Americamysis bahia (X-axis) and Oncorhynchus mykiss (Y-axis) regression
model used for ICE predicted values.

-2	0	2	4	6

Americamysis bahia
(Log LC50)

Figure L-5. Americamysis bahia (X-axis) and Pimephales promelas (Y-axis) regression
model used for ICE predicted values.

L-26


-------
2	3	4	5	6	7

Danio rerio- embryo
[Log LC50)

Figure L-6. Danio rerio -embryo (X-axis) and Daphnia magna (Y-axis) regression model
used for ICE predicted values.

— O
CO Li"

? q

;=

O

Danio rerio- embryo
(Log LC50)

Figure L-7. Danio rerio - embryo (X-axis) and Oncorhynchus mykiss (Y-axis) regression
model used for ICE predicted values.

L-27


-------
Danio rerio- embryo
[Log LC50)

Figure L-8. Danio rerio - embryo (X-axis) and Pimephales promelas (Y-axis) regression
model used for ICE predicted values.

0	2	4	6

Daphnia magna
(Log LC50)

Figure L-9. Daphnia magna (X-axis) and Americamysis bahia (Y-axis) regression model
used for ICE predicted values.

L-28


-------
0

2

4

6

Daphnia magna
(Log LC50)

Figure L-10. Daphnia magna (X-axis) and Lampsilis siliquoidea (Y-axis) regression model
used for ICE predicted values.

0	2	4	6

Daphnia magna
(Log LC50)

Figure L-ll. Daphnia magna (X-axis) and Lithohates catesbeianus (Y-axis) regression
model used for ICE predicted values.

L-29


-------
-2	0	2	4	6

Daphnia magna
(Log LC50)

Figure L-12. Daphnia magna (X-axis) and Oncorhynchus mykiss (Y-axis) regression model
used for ICE predicted values.

ra

CD

E

O o

-s

ill

Daphnia magna
(Log LC50)

Figure L-13. Daphnia magna (X-axis) and Pimephales promelas (Y-axis) regression model
used for ICE predicted values.

L-30


-------
Lampsilis siliquoidea
(Log LC50)

Figure L-14. Lampsilis siliquoidea (X-axis) and Daphnia magna (Y-axis) regression model
used for ICE predicted values.

CD

"o O
p m

i"

b o

Lampsilis siliquoidea
(Log LC50)

Figure L-15. Lampsilis siliquoidea (X-axis) and Ligumia recta (Y-axis) regression model
used for ICE predicted values.

L-31


-------
Lampsilis siliquoidea
(Log LC50)

Figure L-16. Lampsilis siliquoidea (X-axis) and Oncorhynchus mykiss (Y-axis) regression
model used for ICE predicted values.

i--

CO



u=<

CD





0.1





E





o

o



Cl_



kO

CO

o

1



GD

	1



03

CD
O



Cl

_l

ฆ"=t

E





CL





cn -

2	3	4	5	6

Lampsilis siliquoidea
(Log LC50)

Figure L-17. Lampsilis siliquoidea (X-axis) and Pimephalespromelas (Y-axis) regression
model used for ICE predicted values.

L-32


-------
Ligumia recta
(Log LC50)

Figure L-18. Ligumia recta (X-axis) and Lampsilis siliquoidea (Y-axis) regression model
used for ICE predicted values.

2	4	6

Lithobates catesbeianus
(Log LC50)

Figure L-19. Lithobates catesbeianus (X-axis) and Daphnia magna (Y-axis) regression
model used for ICE predicted values.

L-33


-------
2

4

6

Lithobates catesbeianus
(Log LC50)

Figure L-20. Lithobates catesbeianus (X-axis) and Oncorhynchus mykiss (Y-axis) regression
model used for ICE predicted values.

2	4	6

Lithobates catesbeianus
(Log LC50)

Figure L-21. Lithobates catesbeianus (X-axis) and Pimephales promelas (Y-axis) regression
model used for ICE predicted values.

L-34


-------
0	2	4	6

Oncorhynchus mykiss
(Log LC50)

Figure L-22. Oncorhynchus mykiss (X-axis) and A mericamysis bahia (Y-axis) regression
model used for ICE predicted values.

Oncorhynchus mykiss
(Log LC50)

Figure L-23. Oncorhynchus mykiss (X-axis) and Daphn'ui magna (Y-axis) regression model
used for ICE predicted values.

L-35


-------
CD
Q)

"O

O	_

=>	o

cr	Li"

=	O ,

co 	|

.52	o

—	o

ra
_l

CM -

1	2	3	4	5

Oncorhynchus mykiss
(Log LC50)

Figure L-24. Oncorhynchus mykiss (X-axis) and Lampsilis siliquoidea (Y-axis) regression
model used for ICE predicted values.

C
CD

-------
0

2

4

6

Oncorhynchus mykiss
(Log LC50)

Figure L-26. Oncorhynchus mykiss (X-axis) and Pimephales promelas (Y-axis) regression
model used for ICE predicted values.

2	4	6	8

Pimephales promelas
(Log LC50)

Figure L-27. Pimephales promelas (X-axis) and Americamysis bahia (Y-axis) regression
model used for ICE predicted values.

L-37


-------
0

2

4

6

Pimephales prornelas
(Log LC50)

Figure L-28. Pimephales prornelas (X-axis) and Daphnia magna (Y-axis) regression model
used for ICE predicted values.

Pimephales prornelas
(Log LC50)

Figure L-29. Pimephales prornelas (X-axis) and Lampsilis siliquoidea (Y-axis) regression
model used for ICE predicted values.

L-38


-------
0	2	4	6

Pimephales prornelas
(Log LC50)

Figure L-30. Pimephales prornelas (X-axis) and Lithobates catesbeianus (Y-axis) regression
model used for ICE predicted values.

0	2	4	6

Pimephales prornelas
(Log LC50)

Figure L-31. Pimephales prornelas (X-axis) and Oncorhynchus mykiss (Y-axis) regression
model used for ICE predicted values.

L-39


-------
2

3

4

5

Pimephales prornelas
(Log LC50)

Figure L-32. Pimephales prornelas (X-axis) and Xenopus laevis (Y-axis) regression model
used for ICE predicted values.

Xenopus laevis
(Log LC50)

Figure L-33. Xenopus laevis (X-axis) and Pimephales prornelas (Y-axis) regression model
used for ICE predicted values.

L-40


-------
Appendix M Environmental Fate of PFOS in the Aquatic Environment

Natural degradation of PFOS has not been observed. As described above in Section 2.2
above, under environmental conditions, PFOS does not photolyze, hydrolyze, or biodegrade and
is thermally stable. For these reasons, PFOS is considered to be highly persistent in the
environment (Beach et al. 2006; OECD 2002).

M.l Photolysis

PFOS does not appear to photolyze (OECD 2002). No experimental e\ ideuce of direct
and indirect photolysis was available (Hatfield 2o<) I) The indirect photolylie luill-life of PFOS
using an iron oxide photo-initiator matrix model was estimated to be > 3.7 years al 25ฐC. This
half-life was based on the analytical melhod of detection (Giesv et al. 2010).

M.2 Hydrolysis

No hydrolytic loss of PFOS was obsei\ ed in a -N day study under experimental

conditions of 50 (' and pi I conditions of 1.5. 5, 7, 9, or 11 (Hatfield 2001b). Instead, the half-life
of PFOS was estimated to lx- 41 years at 25 'C. However, this estimate was influenced by the
analytical limit ol"(.|iiantillcation and that no loss of PFOS was actually detected (Giesy et al.
2010)

M.3 Biodegradation

Several studies have demonstrated that PFOS does not biodegrade under aerobic or

anaerobic conditions ((iledhill and Markley 2000 a-c; Key et al. 1998; Kurume Laboratory 2002;
Lange 2001; Remde and Debus 1996; Saez et al. 2008). Results from a study conducted by
Kurume Laboratory in 2002 showed no biodegradation of PFOS after 28 days as measured by
net oxygen demand, loss of total organic carbon, and loss of parent material. Key et al. (1998)
demonstrated that even under the sulfur-limiting conditions, PFOS did not degrade. Similarly,

M-l


-------
Saez et al. (2008) observed no PFOS degradation under aerobic or anaerobic conditions in
municipal sewage sludge. In contrast, Schroder (2003) reported that PFOS was anaerobically
degraded; however, the reported results are uncertain as the results could likely be attributed to
sorption and there was a lack of increased fluoride concentrations reported (Fromel and Knepper
2010).

The persistence of PFOS has been attributed to the strong C-l hond. Additionally, there
have been limited indications that naturally occurring defluorinating enzymes exist that can
break a C-F bond, which is likely due to the rarity of fluorinated molecules in nature (Fromel and
Knepper 2010). To date, no laboratory data exist that demonstrates the PFOS undergoes
significant biodegradation in environmental conditions (Beach et al. 2006; Giesy et al. 2010;
OECD 2002).

M.4 Thermal Stability

Based on carhon-siilfur (C-S) bond energy, which is weaker than the carbon-carbon (C-
C) or the C-F bond energies. PI-OS is considered lo have relatively low thermal stability. Thus,
PFOS would more easily lueakdow n under incineration conditions and would be nearly
completely destroyed when incinerated (Bench et al 2006; Giesy et al. 2010).

M.5 Ailsorption/Desorption

In general. PFOS may adsorb to sediments (with a Kd greater than 1 mL/g; Giesy et al.

2010). However, this sorption to sediment is limited and PFOS has a Koc of 2.57 indicated that
PFOS is relatively mobile in water and the physicochemical characteristics of the sediment
ultimately influence the sorption of PFOS (Ahrens et al. 2011; Beach et al. 2006; Giesy et al.
2011; Higgins and Luthy 2006). Sediment characteristics have a strong influence on the
partitioning of PFOS (You et al. 2010). Specifically, organic content was found to have a

M-2


-------
significant influence on the partitioning of PFOS. Density of the sediment was also found to be
an important factor influencing partitioning (Ahrens et al. 2011). PFOS has a high affinity to
bind to organic carbon with log Koc values ranging between 2.57 and 3.8 cm3/g (Higgins and
Luthy 2006 and (Ahrens et al. 2010);Ahrens et al. 2010; respectively). A sorption mechanism
could be a salting-out and calcium-bridging effect, as PFOS sorption to sediment increased with
increased salinity, pH, and calcium (You et al. 2010). Thus, the sorption of PFOS is a
complicated process that is partially dependent on other factors such as metal anion
concentrations, pH, temperature, and salinity; however, the strong relationship between PFOS
concentrations and organic carbon in soil, sediment, and sludge indicates that these other factors
have a minor influence on PFOS sorption (Ahrens et al. 2011, Chen et al. 2012; Higgins and
Luthy 2006; You et al. 2010).

M-3


-------
Appendix N Occurrence of PFOS in Abiotic Media

N.l Summary of Measured Perfluorooctane Sulfonate Concentrations in
Surface Waters Across the United States.

Modified from: Jarvis et al. (2021).				





Arilhinelic



K;nii>e of







Mciiii PI OS

Medinn PI OS

PI OS







(oiuvnlnilion

(oiuvnlnilion

('oiHvntnKion



Sliiu*

\\ silorhodv1

(iii>/I.):

(nii/l.):

(nii/l.)

UcfemuT





3.77

3

2 S - 5.5

Sinclair et al.
(2006)



Lake Erie

31.3

32.5

21 5 38 5

Boul anger et
al. (2004)



2.84

2.63

2.49- 3 41

De Silva et
al. (2011)





4.5

42

4.0 - 5 3

I 'urdui et al.
(2008)



Lake Huron

2.25

1.96

0.239 - 5.46

De Silva et
al. (2011)



1

1 5

1 2-2.7

Furdui et al.





1 . / J)

(2007)











Simcik and





2.03

2 i >3

0.93-3.13

Dorweiler



1 .tike Michiuan







(2005)





: do

1.96

1.73-2.36

De Silva et
al. (2011)





not |ii'o\ ided

4.9

2.9-30

(Sinclair et
al. 2006)





55.4

59.8

16.5-85.5

Boul anger et
al. (2004)



Lake Ontario

5 lHi

5.63

2.60-9.48

De Silva et
al. (2011)





8.69

6.6

3.6-37.6

Furdui et al.
(2008)





2.20

not provided

not provided

Houde et al.
(2008)





0.255

0.236

0.095-0.395

De Silva et
al. (2011)



Lake Superior

0.233

0.3

0.1-0.3

Furdui et al.
(2008)





0.246

0.124

0.074-0.996

Scott et al.
(2010)

Alabama

Waterbody
near Decatur

58,016

41,027

9- 150,000

OECD
(2002)

N-l


-------




Arilhinelic



K;nii>e of







Mcsiii PI OS

Medinn PI OS

PI OS







C oiuvnlnilion

(oiuvnlnilion

C oiuvnlnilion



Sliiu*

\\ silorhodv1



(nii/l.):

(nii/l.)

UcfemuT



Waterbody in
Decatur

2.5 
-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnKion
(nซ/l.)2

Medinn PI OS
C oiuvnlnilion

(nii/l.):

R;nii>e of
PIOS
( on con l r:iliซปii

(nii/l.)

UcfemuT



Eagle River

0.68

0.68

0.68



East Plum
Creek

<0.43

<0.43

<0.43

Erie Lake

3.70

3.70

3.70

Fairmount
Reservoir

<2.50

<2.50

<2.50

Fountain
Creek

16.9

20.0

3 50 -24.0

Fraser River

1.00

1.00

| do

Gore Creek

0.98

0.98

0 WS

Gunnison
River

0.71

0.71

0.71

Horsetooth
Reservoir

0.51

i)5l

0.51

Jackson Creek

<0.44

<0.44

<0.44

Jerry Creek

< 0 4S5

^ n 4S5

< 0.48 -
<0.49

Kannah Creek
Flowline


-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnilion

(nii/l.):

Medinn PI OS
(oiuvnlnilion

(nii/l.):

K;nii>e of
PI OS
C oiuvnlnilion

(nii/l)

UcfemuT



Purgatoire
River

0.47

0.47

0.47



Ralston
Reservoir

<0.46

<0.46

<0.46

Rio Grande

<0.47

<0.47

<0.47

Roaring Fork
River

<0.50

<0.50

0.50

San Juan River

<0.44

<0.44

<0.44

Sand Creek

30.3

30.3

(-> 5<) - 54.0

S every Creek

<0.47

<0.47

<0.47

Somerville
Flowline

<0.48

<0.48

<0.48

South Boulder
Creek

0.50

0.50

0.50

South Platte
River

10.5

11.5

3 80 - 16.0

St. Vrain River

3 W

3.90

3.90

Strontia
Springs

< 0.5 1

<0.51

0.51

Taylor River

<0.45

0 45

0.45

Uncompahgrc
River (delta)

i) 54

0.54

0.54

Wei ton
Reservoir

2 60

2.60

2.60

While Ri\er

() 4o

<0.46

<0.46

Yampa River

0 47

<0.47

<0.47

Del aw arc.
New Jersey.
PennsyK iiniii

Del aw arc
River

3.98

3.5

0.97 - 6.92

Pan et al.
(2018)

Florida

Waterbody in
Pcnsacola

16.29

2.5 
-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnilion

Medinn PI OS
(oiuvnlnilion

(nii/l.):

K;nii>e of
PI OS
C oiuvnlnilion

(nii/l.)

UcfemuT



Pond in
Columbus

<2.5

<2.5

<2.5



Conasauga
River

162.1

192

< 1.5-321

Konwick et
al. (2008)

Altamaha
River

2.63

2.6

2.6-2.7

Streams and
ponds in
Dalton

70.36

70.73

in 5-1 19.5

Oostanaula
River

150.3

151

148 - 152

Lasier et al.
(2011)

Louisiana

Waterbodies
(locations of
concern) near
Barksdale
A.F.B.

776.7

195 o

10-7,07d

(Cochran
2015); Lanza
et al. (2017)

Reference
waterbodies
near Barksdale
A.F.B.

< ID

in

10

Michigan

Raisin River

3.5

3 5

3.5

Kannan et al.
(2005)

St Clair River

2.6

2

1.9-3.9

SiskiniI Luke

i) 283

0.283

0.277-0.289

Scott et al.
(2010)

Minnesota

I'PIW
Mississippi
Ri\ er

528.9

<2

<2-18,200

Newsted et
al. (2017)

Lake of the

Isles

2.47

2.47

2.47

Simcik and

Dorweiler

(2005)

Lake Calhoun

50.4

50.4

50.4

Lake Harriet

22.1

22.1

22.1

Minnesota
Ri\ er

9.21

9.21

9.21

Lake

Tettegouche

0.23

0.23

0.23

Lake

Nipisiquit

<0.27

<0.27

<0.27

Lake Loiten

<0.27

<0.27

<0.27

Little Trout
Lake

1.2

1.2

1.2

N-5


-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnilion

Medinn PI OS
(oiuvnlnilion

(nii/l.):

K;nii>e of
PI OS
C oiuvnlnilion

(nii/l.)

UcfemuT

New Jersey

Echo Lake
Reservoir

<2

<2

<2

NJDEP
(2019)

Passaic River

13.1

13.1

13.0-13.2

Raritan River

6.9

6.9

6.9

Metedeconk
River

1.65

1.65

<2-2.8

Pine Lake

102

102

102

Horicon Lake

10

10

10

Little Pine
Lake

100

100

| do

Mirror Lake

72.9

72.9

72.9

Woodbury
Creek

6.4

6.4

6.4

Fen wick Creek

3.1

3.1

3.1

Cohansey
River

<2

< 2

<2

Harbortown
Road

1

1 93

1.93

Zhang et al.
(2016)

Passaic River

4

4 i)7

u.244 - 9.99

New Mexico

Alamogordo
Domestic
Water Sys.

< 1

^ 1

< 1

New Mexico
Environment
Department
(2021)

Animas River

o 799

0.625

<0.89- 1.5

Canadian
Ri \ er

o S4S

0.9

<0.89- 1.2

Cloud Country
Estates \\ I A

0.93

<0.93

<0.93

Gila River

<0.93

<0.93

<0.93

Holloman
A1 B Golf
Course Pond 1

1,220

1,220

1,220

1 lolloman
ALB Golf
Course Pond 2

878

878

878

Holloman
AFB Lagoon
G

310

310

310

Holloman
AFB Outfall

951

951

951

N-6


-------
Sliiu*

\\ silorhodv1

Arilhmclic
Mcsiii PI OS
C oiuvnlnilion
(nซ/l.)2

Medinn PI OS
(oiuvnlnilion

(nii/l.):

K;nii>e of
PI OS
C oiuvnlnilion

(nii/l.)

UcfemuT



Holloman
AFB Sewage
Lagoon

2,200

2,200

2,200



Karr Canyon
Estates

<0.93

<0.93

<0.93

La Luz
MDWCA

< 1.3

< 1.3

1.3

Lake
Holloman

4,033

4,50i)

1.700- 5,900

Mountain

Orchard

MDWCA

<0.93

<0.93

i) W3

Pecos River

1.223

1.50

<0.94 - 1.70

Rio Chama

<0.98

ฆ l)l)S

< 0.96 - < 1

Rio Grande

1.052

0.474

< 0.465 -
2.90

Rio Puerco

4 35

4 35

3 10-5.60

San Juan River

<115

1 15

: 1.06-
< 1.24

Tularosa
Water System

0.723

0.723

<0.89- 1.0

New York

Washington
Park Lake

1 67

1.77

<0.25-2.88

Kim and

Kannan

(2007)

Rensselaer
I.akc

7 1 1

6.58

5.85-9.3

Iroquois Lake

not provided

not provided

not provided

Unnamed lake
1 outside
Albany, ]\Y

not provided

not provided

not provided

1	n named lake

2	outside
Allninv, NY

not provided

not provided

not provided

Niagara River

5.17

5.5

3.3-6.7

Sinclair et al.
(2006)

Finger Lakes

not provided

1.6

1.3-2.6

Lake

Onondaga

681

756

198- 1,090

Lake Oneida

3.5

3.5

3.5

Erie Canal

8.37

6.4

5.7 - 13

Hudson River

not provided

1.7

1.5-3.4

Lake

Champlain

not provided

2.7

0.8-7.7

N-7


-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnilion

(nii/l.):

Medinn PI OS
C oiuvnlnilion

(nii/l.):

K;nii>e of
PI OS
C oiuvnlnilion

(nii/l.)

UcfemuT



Lower NY
Harbor

0.755

0.755

0.755

Zhang et al.
(2016)



Staten Island

1.66

1.66

1.66



Hudson River

1.81

1.81

0.79-2.84



North
Carolina

Cape Fear
River

31.2

28.9

1-132

Nakayama et
al. (2007)



Narragansett
Bay

2.2

2.2



Benskin et al.
(2012)



Allen Cove
Inflow

1.20

1.20

1 2d





Bristol Harbor

0.508

0.46

0.437 0.020





Brook at Mill
Cove

9.80

9.80

9.80





Buckeye
Brook

4.13

4 13

4.13





Chickasheen
Brook


-------
Sliiu*

\\ silorhodv1

Arilhinelic
Mcsiii PI OS
C oiuvnlnilion

Mcdiiin PI OS
(oiuvnlnilion

(nii/l.):

Kiiniic ol'
PI OS
C oiuvnlnilion

(nii/l.)

UcfemuT



Woonasquatucket
River

14.6

14.6

5.87-23.2



South
Carolina

Charleston
Harbor

12.0

not provided

not provided

Houde et al.
(2006)

Tennessee

Waterbody
near Cleveland

2.5 
-------
northern Great Lakes. These northern Great Lakes (i.e., Lakes Huron, Michigan, and Superior)
have a maximum observed concentration of 5.46 ng/L, which was observed in Lake Huron
(Remucal 2019). However, current measured PFOS concentrations were not from sampling sites
around urbanized areas (such as Chicago and Detroit) and may not be representative of the
potential sources of PFOS related to these areas. The measured concentrations of PFOS in the
surface waters of Lakes Huron and Michigan range between 0 24 and 5.46 ng/L (De Silva et al.
2011; Furdui et al. 2008) and 0.93 and 3.13 ng/L (De Silva et al. 201 I. Simcik and Dorweiler
2005), respectively. In contrast, measured PFOS concentrations observed in I .tike Superior were
considerably lower and range between 0.074 and <> ng/I. (De Silva et al. 201 I. I'urdui et al.
2008; Scott et al. 2010). The higher PFOS concentrations in Lakes Erie and Ontario are likely
due to higher levels of industrial activities and urbanization around these lakes (Boulanger et al.
2004; Remucal 2019) and could also be associated with the sampling locations. A mass balance
constructed for Lake Ontario In lioulanger et al. (2004) indicated wastewater effluent was the
major source of PI OS to the lake In contrast, inputs from Canadian tributaries and atmospheric
deposition of PI-OS. and other PI AS that may be transformed into PFOS, were the major
contributing sources of PI OS to l.ake Superior. Inputs from Canadian tributaries and
atmospheric deposition were estimated to contribute 57 and 32% of PFOS inputs into Lake
Superior, respecti\ely (Scott et al. 2010).

N.3 PFOS occurrence and concentrations in the southeastern U.S.

Measured PFOS concentrations in southeastern U.S. surface waters were similar to those

measured in Lakes Erie and Ontario, with some of the highest observed concentrations occurring
in waterbodies near areas with PFOS manufacturing. In 2001, the 3M Company conducted a
multi-city study measuring PFOS concentrations across waterbodies with known manufacturing

N-10


-------
and/or industrial uses of PFOS (3MCompany 2001). In the 3M Company's 2001 report, PFOS
concentrations from sites with known PFOS discharges were compared to PFOS concentrations
measured in waterbodies with no known sources of any PFAS (3MCompany 2001). In this
comparison study, cities with known PFOS exposure were Mobile and Decatur, Alabama,
Columbus, Georgia, and Pensacola, Florida. Measured PFOS concentrations ranged from not
detected (reported detection limit of 2.5 ng/L; 3MCompany 2001) io SO ng/L in the cities with
known PFOS discharges. These PFOS concentrations were compared lo those measured in
control cities. These control cities were Cleveland, Tennessee and Port St. Lucie. Mori da and
PFOS concentrations ranged from not detected to 137.5 ng/I. (3MCompany 2001) The PFOS
concentrations measured in Cleveland, Tennessee were below the limit of quantification (25
ng/L) and were lower than the PFOS concentrations observed in the cities with known PFOS
exposure, as was expected in the report for the control cities However, PFOS concentrations
around Port St. Lucie. Morida. the other control city, were unexplainably similar to, and at times
higher than, the waterbodies with known PFOS discharges. The sources of PFOS near Port St.
Lucie, Florida remain unknown. howe\er. observed PFOS concentration suggest the presence of
a potential nianulactui inu industrial source or the use of AFFF in this area (3MCompany 2001).

Water samples were collected from ponds near all of the sampling sites except those in
Cleveland. Tennessee. PFOS concentrations in these additional pond sites were similar to those
measured in Mobile. Alabama (ranging between 32 and 33 ng/L), lower than those observed in
Columbus, Georgia (as PFOS was not detected with a detection limit of 2.5 ng/L), and higher
than those measured in Decatur, Alabama (ranging between 108 and 111 ng/L) and in Port St.
Lucie, Florida (ranging between 1,830 and 48,200 ng/L). Samples collected from the pond site
near Port St. Lucie, Florida had some of the highest measured PFOS concentrations in publicly

N-ll


-------
available literature with the maximum concentration of 48,200 ng/L. In the report, the 3M
Company conducted additional sampling at the pond site in Port St. Lucie, Florida and
determined that the measured PFOS concentrations at this site were more variable than the initial
measurements indicated and were lower than the previous measurements, ranging between below
detection (i.e., < 2.5 ng/L) and 2,340 ng/L. Aside from the samples collected in Port St. Lucie,
Florida, this report demonstrated that measured PFOS concenlraiions in surface waters tend to be
higher in areas with PFOS manufacturing and/or industrial use (3MConi|Xiny 2001).

In separate studies, PFOS and PFOA concentrations were measured in surface waters by
Hansen et al. (2002) near Decatur, Alabama, and k on wick et al (2008) in Georgia I lansen et al.
(2002) studied a stretch of the Tennessee River near Decatur, Alabama, and Konwick et al.
(2008) focused on the Conasauga River in Georgia. both areas with known PFOS discharge and
use. In Hansen et al. (2002), discharge from a lluorocliemical manufacturing facility entered the
Tennessee River towards llie middle of the study area. In contrast, Konwick et al. (2008)
compared the PI OS concenlraiions measured in the Conasauga River with those from sites with
no known exposure along the Allamaha Ri\ er In both studies, mean PFOS concentrations were
higher in the study areas with PI'OS sources. Specifically, Hansen et al. (2002) observed mean
PFOS concenlraiions upstream of the fluorochemical manufacturing facility were 30.85 ng/L
(ranging between I(ฆ>  ng/L) and were 103.9 ng/L (ranging between 30.3 and 144 ng/L)
downstream of the lluorochemical manufacturing facility. Similarly, Konwick et al. (2008)
observed higher measured PFOS concentrations in the Conasauga River, which ranged from
below the limit of detection (i.e., 1.5 ng/L) to 321 ng/L, compared to those in the Altamaha
River, ranging between 2.6 and 2.7 ng/L. Consistent with the report from the 3M Company
summarized above, effluents from manufacturing facilities, WWTP, and carpet mill effluents

N-12


-------
were determined to be the source of increased PFOS concentrations in both the Tennessee and
Conasauga Rivers (Hansen et al. 2002; Konwick et al. 2008; respectively). These PFOS
concentrations are relatively consistent with those measured in Alabama and Georgia as reported
by the 3M Company (3MCompany 2001).

Nakayama et al. (2007) and Cochran (2015) measured PFAS, including PFOS, in the
Cape Fear Drainage Basin in North Carolina and waterbodies on Barksdale Air Force Base in
Bossier City, Louisiana, respectively. PFOA and PFOS were found to be the dominant PFAS
detected in both studies. Nakayama et al. (2007) detected PFOS in 97.5% of all samples above
the limit of quantification of 1 ng/L. PFOS concentrations in the Cape Fear Drainage Basin
ranged between < 1 (the lower limit of quantification) and 132 ng/L with a mean concentration
of 31.2 ng/L. As in other studies summarized abo\ e. lower H AS concentrations, including
PFOS, were found in the upland tributaries and concentrations were highest in the middle
reaches of the Cape I'ear Drainage Basin, nearer expected sources. Wastewater treatment plant
effluents were idenlilied as the source of PFAS to the study area. AFFF usage at the Department
of Defense base in l'ayelle\ ille. North Carolina and the land application of contaminated
biosolids likely contributed as well (Nakayama et al. 2007). Cochran (2015) detected PFOS in
79% of ulI water samples collected and concentrations ranged between below the limit of
quantification (i e.. 10 ng/L) and 7,070 ng/L, with an average concentration of 776.7 ng/L. PFOS
concentrations collected in Barksdale Air Force Base varied based on proximity to fire training
areas. Cochran (2015) attributed the evaluated PFOS concentrations to runoff and ground
infiltration of AFFF formerly used on the base during firefighting and/or training.

N-13


-------
N.4 PFOS occurrence and concentrations in the midwestern U.S.

Similar PFOS concentrations were reported in the publicly available literature for

waterbodies in urban areas across the midwestern U.S., with lower PFOS concentrations reported
in remote areas in the same states (Newsted et al. 2017; Simcik and Dorweiler 2005). In
Minnesota, Simcik and Dorweiler (2005) observed PFOS concentrations ranged between 2.4 and
50.4 ng/L in urban areas near Minneapolis and between less than the limit of quantification (i.e.,
0.27 ng/L) and 1.2 ng/L in remote areas in northern Minnesota. Additionall\. Newsted et al.
(2017) reported an average PFOS concentration of 528.9 ng/L (ranging between Mow limit of
quantification and 18,200 ng/L; limit of quantification not provided) in surface waters collected
from the Upper Mississippi River near the Minneapolis St Paul, Minnesota metropolitan area.
The source of PFOS at these urban sites was ami huted to manufacturing (3M plant), runoff, and
wastewater discharge (Newsted et al. 2017, Simcik and Dorweiler 2005).

N.5 PFOS occurrence and concentrations in the northeastern U.S.

Several studies measured PI OS concentrations in surface waters in the northeastern U.S.

that are com para hie to those reported in Minnesota (NJDEP 2019; Sinclair et al. 2006). Sinclair

et al. (2<)<)6) measured PI OS in \ arious waterbodies across New York state and observed a

median concentration of 75o ng/L in surface waters collected from the Superfund site at Lake

Onondaga ( ranging between N8 and 1,090 ng/L; Table 1) and attributed these elevated

concentrations to sc\ eral industries located along Lake Onondaga. All other observed

concentrations of PFOS in New York, including sites along the Niagara River, the Finger Lakes,

Lakes Oneida and Champlain, the Erie Canal, and the Hudson River, had lower median PFOS

concentrations ranging between 0.8 and 13 (Sinclair et al. 2006).

The New Jersey Department of Environmental Protection (NJ DEP) measured PFOS in

surface water samples collected from 14 different sites across New Jersey. PFOS concentrations

N-14


-------
ranged from below the detection limit of 2.0 ng/L and 102 ng/L (NJDEP 2019). Individual
samples collected along Pine, Little Pine, and Mirror Lakes had measured PFOS concentrations
of 102, 100, and 72.9 ng/L, respectively. All other observed concentrations of PFOS in New
Jersey freshwaters were below 15 ng/L. NJDEP attributed the elevated concentrations of PFOS
observed at Pine, Little Pine, and Mirror Lakes to the use of AFFF in training and/or fire-fighting
on the Department of Defense (DoD) Joint Base McGuire-Lix-T.akeluirst (NJDEP 2019).

N.6 PFOS occurrence and concentrations in the western I'.S.

PFOS concentrations in surface waters of western U.S. states are consistent with the

lower-end concentrations (less than 100 ng/L) measured in eastern states; howe\ er. the
monitoring data for PFOS was limited in the western U.S. I-Mumlee et al. (2008) measured PFOS
concentrations in Coyote Creek and a tributary of I "pper Sih er Creek in San Jose, California and
found concentrations to be similar to those measured in eastern states. Concentrations of PFOS
in Coyote Creek ranged from 4 S to 25 ng/L and concentrations in Upper Silver Creek ranged
from 27 to 56 nu I. The source of PI OS to these aquatic systems was unknown, however,
Plumlee et al (2<)0S) stated that a combination of atmospheric deposition of volatile precursors
and surface runoff were likely sources of I'FOS to both Coyote and Upper Silver Creeks.

I .astly. Dinglasan-Panlilio et al. (2014) measured PFOS concentrations in surface waters
along the Puuel Sound in Washington, as well as Clayoquot and Barkley Sounds in British
Columbia, Canada PI-OS concentrations measured by Dinglasan-Panlilio et al. (2014) were
lower than those observed from sites in eastern states (such as those summarized above for
Alabama, Florida, and North Carolina with known manufacturing and/or industrial use of PFOS.
Concentrations ranged from 0.2 to 5.9 ng/L in Puget Sound and 0.25 to 0.7 ng/L in Clayoquot
and Barkley Sound, British Columbia. These concentrations are consistent with those reported in

N-15


-------
the publicly available literature for remote areas, such as in Minnesota (Simcik and Dorweiler
2005) and in New York (Sinclair et al. 2006), as summarized above. The study authors indicated
specific regional sources and atmospheric deposition were likely PFOS sources to these remote
areas (Dinglasan-Panlilio et al. 2014).

N.7 Comparison of PFOS occurrence in the U.S. to global surface waters

Similar to surface waters in the U.S., generally PFOS and PI OA were the most

commonly detected PFAS in surface waters around the world (Ahrens 2<) I I) Ona global scale,
PFOS concentrations in surface waters generally range between picogram liter and
nanogram/liter with some concentrations in the milligram liter range. However. PI-OS
occurrence data were limited for surface waters in Africa and South America. Based on the
currently available data, PFOS concentrations in the U.S. were iclali\el\ similar to those
reported in studies with sampling sites in other countries (ilobal surface water PFOS
concentrations reported in the public literature ranged between not detected and 2,100,000 ng/L
(Jarvis et al. 2021) These global surface water concentrations are summarized in Jarvis et al.
2021 to provide a comparison with those obser\ ed in the U.S.

()\erall. the currently a\ ailahlc data on PI OS occurrence in ambient surface waters show
the widespread distribution and variability of PFOS concentrations in surface waters around the
world and that surrounding land use has a large influence on PFOS concentrations in surface
waters. In general, urbanized areas with high population densities tended to have elevated PFOS
concentrations in surface waters (Jarvis et al. 2021). Like in the U.S., PFOS concentrations in
surface waters around the world vary widely and current information on the environmental
distribution of PFOS in surface waters around the world is relatively limited.

N-16


-------
N.8 PFOS Occurrence and Detection in Aquatic Sediments

PFOS has been detected in sediments of aquatic environments across various countries

(Lau et al. 2007). Typically, in the U.S., soil and sediment measurements of PFOS occur in the
|ig/kg dry weight (dw) range with measured concentrations in the public literature ranging from
not detected (with a detection limit of 0.08 |ig/kg dw) to 31.38 |ig/kg dw (3MCompany 2001;
Cochran 2015). Anderson et al. (2016), measured concentrations of PI-AS in sediment across ten
U.S. Air Force bases where there is a known history of use of Al IT. and found that PFOS
concentrations were detected in 94% of samples. The median concentration of PI-OS across all
sample sites was 31.0 |ig/kg, with a maximum concentration of 190,000 |ig/kg (Anderson et al.
2016). Arias et al. (2015) measured PFOS in sediment from an evaporation pond used to collect
the wastewater arising from fire-fighting exercises at an Australian military air base. Despite the
discontinued use of PFOS/PFOA-based foams six years earlier, the PFOS sediment
concentration was 38,000,000 ug^g, a million limes higher than the average global values for
sediments (0.28 - 3.8 jug/kg PFOS) reported by the authors (Arias et al. 2015).

These obsei \ ed concentrations were similar to other sediment concentrations in areas
with known periluoiinaled chemical discharges and manufacturing. (Lasier et al. 2011) measured
PFOS in sediment from the Coosa Ri\ er, Georgia watershed, upstream, and downstream of a
land-application site of municipal/industrial wastewater with sediment concentrations ranging
from less than the method detection limit (MDL) to 1.73 |ig/kg dw upstream of the land-
application and 1.66 - 2d 18 |ig/kg dw PFOS downstream. Giesy and Newsted (2001), as
presented in OECD (2002), measured PFOS in sediments collected from locations upstream and
downstream of the 3M facility in Decatur, Alabama. The two closest sites downstream of the 3M
facility had significantly greater concentrations (1,299 and 5,930 |ig/kg ww) than the two
upstream sites (-0.18 and 0.98 |ig/kg ww; OECD 2002).

N-17


-------
Other sediment concentrations across the U.S. were much lower: < 4 |ig/kg across sites in
Puget Sound, Washington, San Francisco and Monterey Bay California, the Niagara River in
New York, and Lake Michigan. These concentrations appeared to be similar to other sediment
concentrations across the globe (Table N-l).

Table N-l. Global Sediment Concentration of PFOS.

Location

PI-OS concentration

Reference

Tokyo Bay, Japan

0.29-0.36 |ig/kg dw

(Ahrens et al. 2010)

Ariake Sea, Japan

0.11 |ig/kg ww

Nakata et al. (2006)

Toronto, Canada

<0.1-2.2 |ig/kg ww

Vedagiri ct al. (2018)

Lake Ontario, Canada

10 |ig/kg dw

(ECCC 2d IS)

Lake Ontario, Niagara Basin

27-47 iig/kg

Meyers et al. (2012)

Lake Ontario, Mississauga Basin

4.4-19 |ig/kg

Meyers etal. (2<) 12)

Lake Ontario, Rochester Basin

8.1-49 |ig/kg

Meyers et al. (2012)

Resolute Lake, Canada

24-85 |ig/kg ww

Butt et al. (2010)

Gufunes Bay, Iceland

5<) iig/kg ww

Butt et al. (2010); Kallenborn et al.
2<) 10: Butt et al. (2010)

Faroe Islands

5ii-i) 11 |ig/kg

Unit et al. 2010; Kallenborn et al.
2010,(Butt et al. 2010)

Urban reservoir, Singapore

2.8-3 (•> nu kg dw

Nguyen et al. (2016)

N.9 PFOS Occurrence and Detection in Air and Rain

Air concentrations of PI OS in the atmosphere varied widely across the globe. In an urban

area in Allniny. \ Y. pei'lluorinuled acids were measured in air samples in both the gas and
particulate phase in May and July 2006 (Kim and Kannan 2007). PFOS in the gas phase had a
mean concentration of 1.70 pu m3 (range: 0.94-3.0) and in the particulate phase had a mean
concentration of 0 M pu in' (range: 0.35-1.16) (Kim and Kannan 2007). Kim and Kanaan (2007)
also reported mean PFOS concentrations of 0.36 ng/L and 0.62 ng/L in rain and snow,
respectively.

Above Lake Ontario, concentrations of PFOS in the particulate phase measured in air

samples over the lake were higher than those observed by Kim and Kannan (2007) near Albany,

NY. The mean concentration of PFOS at Lake Ontario was 6.4 ฑ3.3 pg/m3 (Boulanger et al.

N-l 8


-------
2005); with a range of concentrations from detected to 8.1 pg/m3 (Martins et al. 2010). In an
urban area in Minneapolis, Minnesota, PFOS was measured in both the particulate and gas
phase. PFOS in the particulate phase ranged from 2.1 - 7.9 pg/m3 and the gas phase ranged from
1.8 - 5.0 pg/m3 in across the five samples (MPCA 20072008).

In Canada, PFOS air concentrations measured in 2009 showed widespread distribution
with remote sites having similar concentrations as urban sites (T.CCC 2018). Using passive
samplers PFOS concentrations were detected in Toronto, Ontario (8 pg m'). an agricultural site
in Saskatchewan (5 pg/m3), Whistler, British Columbia (4 pg/m3), and Alert. \ Niinavut (2
pg/m3) (EC 2013).

Other reported concentrations of PFOS in air samples included Sydney, Florida (3.4
pg/m3), Tudor Hill, Bermuda(6.1 pg/m3). \lalin I lead. Ireland (3 3 pg'm3), andHilo, Hawaii
(6.6 pg/m3) are similar to the concentrations reported in Canada (l-CCC 2018) and Japan (Sasaki
et al. 2003). The annual geometric mean concentration of PFOS in air samples collected monthly
from 2001-2002 in the tow n of 0\ amazaki and Fukuchiyama City were 5.3 and 0.6 pg/m3,
respectively (Sasaki et a I 2< n >3)

Across Europe. PI-OS air concentrations were reported to be variable. In the particulate
phase PI OS concentrations ranged from <1.8-46 pg/m3 (Martin et al. 2010). Most locations
had low ( I -2 pu m') to less than the reported Minimum Detection Limit (MDL) and included
Hazelrigg, United kingdom. Kjeller Norway, and Mace Head, Ireland (Barber et al. 2007). The
highest concentrations were reported in Manchester, United Kingdom. Similarly, high
concentrations were reported for another urban area, 150 pg/m3 for Paris, France (ECCC 2018).

Even in the Arctic, PFOS, its precursors, and degradation products, have been detected in
air samples in Resolute Bay, Nunavut, Canada, during the summer of 2004 (Stock et al. 2007).

N-19


-------
PFOS in the filter samples were 1-2 orders of magnitude greater than other compounds, with a
mean concentration of 5.9 pg/m3 (Butt et al. 2010). These concentrations are greater than PFOS
concentrations measured in the particle phase of air samples measured in Zeppelinstasjonen,
Svalbard, Norway (Butt et al. 2010). PFOS was measured in September and December, 2006 and
August and December, 2007, with mean concentrations of 0.11 pg/m3 (range: 0.03 - 0.50 pg/m3)
and 0.18 pg/m3 (range: 0.02 - 0.97 pg/m3), respectively (Research 2<)o7a: Research 2007b).

N.10 PFOS Occurrence and Detection in Groundwater

Similar to surface water, PFOS and PFOA are the dominant PFAS detected in

groundwater. Generally, PFOS concentrations tend to occur in the ng/L range, with some
elevated detections in the |ig/L range (Ahrens 2011; Xiao 2017). Concentrations of PFOS were
detected in groundwater samples across Minnesota in 2<~)06 and 2<)()7. approximately five years
after the 3M Corporation phased out PFOS production in Minnesota in 2002 (MPCA 2007). Data
collected from shallow aquifers across Minnesota in both inhaii and agricultural areas were
likely affected by a \ arielv of different contamination sources (i.e., industrial and municipal
stormwater. pesticides, land application of contaminated biosolids and atmospheric deposition)
and indicated that pei lluorinated chemicals are present in areas beyond the disposal sites and
aquifers associated with these disposal sites (MPCA 2007). Groundwater samples of PFOS
ranged from <>.(ป( >222 - 
-------
2011; Moody and Field 2000). The use of AFFF in particular has been identified as an important
source of groundwater contamination with PFAS (Moody and Field 2000). This contamination is
often persistent, lasting for many years after the release (Xiao 2017). The transformation of
PFOS precursor compounds (see Section 2.3) by soil micro-organisms may be a contributing
source of PFOS in groundwater (Xiao 2017).

Groundwater samples from wells in the area of a know n 111111110 were measured in 1998
and 1999. Samples were taken at the Wurtsmith Air Force Base in northeastern Michigan, a base
where fire-training exercises were conducted from the 1950's until the base was decommissioned
in 1993. PFOS concentrations ranged from 4.0 to 1 10 |ig/I. depending on the proximity to the
training pad, demonstrating that PFOS is still present in measurable quantities for at least five or
more years after the use of AFFF (Moody el al 2<)i)3) These \ allies are consistent with ten other
U.S. Air Force bases where there is a know 11 historic use of Al l I to extinguish hydrocarbon-
based fires but were 1101 acti\e lire-training areas. Anderson el al. (2016) measured groundwater
samples between March and September 2014 at the ten locations with PFOS concentrations
detected in 96ฐ „ of samples The median groundwater concentration of PFOS across all sites was
2.17 iig I., with a maximum concentration of 8,970 |ig/L (Anderson etal. 2016). Other reported
groundwater concentrations at other U.S. military installations summarized by (Cousins et al.
2016) include Tvndall Air l-'orce Base (147 - 2,300 |ig/L; Schultz et al. 2004), Fallon Naval Air
station (< LOD - 3S<) nu I.. Schultz et al. 2004) and Ellsworth Air Base (5 - 75 |ig/L; McGuire et
al. 2014). Similar concentrations are reported at other airports and bases globally, including at a
fire training area in Cologne, Germany (0.02 - 8.35 |ig/L; WeiB et al. 2012); air force base F18 in
Sweden (< 0.001 - 42.2 |ig/L; Filipovic et al. 2015) and the Jersey airport in the United Kingdom
(10 - 98 |ig/L; Rumsby et al. 2009).

N-21


-------
N.ll PFOS Occurrence and Detection in Ice

Very little information was provided about PFOS concentrations in ice. Saez et al. (2008)

found PFOS in a Russian Artie ice core sampled in 2007. The PFOS concentration reported was
0.0053 ng/L (Martin et al. 2010).

N-22


-------
Appendix O Meta-Analysis of Nominal Test Concentrations Compared to
Corresponding Measured Test Concentrations

A substantial number of PFOS toxicity tests reported only nominal, or unmeasured,

PFOS concentrations (roughly 71% of the acute and 50% of the chronic freshwater quantitative
aquatic life toxicity tests, and 43% of the acute and 33% of the chronic saltwater quantitative
aquatic life toxicity tests). Therefore, EPA examined whether nominal and measured
concentrations were typically in close agreement with each oilier among the current toxicity
literature for aquatic life. The studies used in the analysis were the aquatic lilc toxicity tests with
measured PFOS concentrations that were considered quantitatively (Appendix A through
Appendix D) and qualitatively acceptable (Appendix G through Appendix H) for both freshwater
and saltwater. Approximately 36% of the I So of the freshwater toxicity tests and 49% of the 41
saltwater toxicity tests measured PFOS concentrations

O.l Summary of Available Freshwater and Saltwater Data for PFOS

Among the PI-OS freshwater toxicity tests used quantitatively and qualitatively in the

criteria derh at ion. (o had measured concentrations. However, several tests only reported
measured concentrations graphical l\. reported summary statistics of the measured concentrations
(e.g., measured concentrations averaged 88% ฑ 7% of their nominal concentrations), reported the
same measured concentrations across several tests (i.e., acute test by Hazelton et al. 2012;
Hazelton, 2013), or lacked the detail to be part of this analysis. Therefore, 49 unique toxicity
tests were used, yielding 477 pairs of measured and nominal concentrations. These pairs
excluded controls since PFOS was rarely detected in controls. Specifically, there were only four
instances where PFOS was detected in controls. However, the measured concentrations were low
at 0.00008 mg/L (Keiter et al. 2012), 0.00001 mg/L (Roland et al. 2014), 0.000006-0.00006
mg/L (Foguth et al. 2020) and 0.002 mg/L (Hazelton et al. 2012). Based on current occurrence

O-l


-------
data, PFOS concentrations in the first two studies are consistent with ambient surface water
concentrations that are considered to be reference sites, which range between below detection
and 0.000138 mg/L (see Section 2.4.1).

Similarly, among the PFOS saltwater toxicity tests considered for quantitatively and
qualitatively use, 20 had measured concentrations. However, for similar reasons as the
freshwater toxicity tests summarized above, 11 of the tests could not he part of this analysis.
Therefore, nine toxicity tests were used, yielding 171 pairs of measured and nominal
concentrations. Of these, 58 pairs were from two acute tests, conducted by the same
investigators. Lastly, the saltwater pairs excluded controls since PI-OS was not detected or not
reported in controls from the saltwater studies.

0.2 Methods of Meta-Analysis to C o m pa re of \ o m i n a I and Measured PFOS
Concentrations

EPA grouped the data by classifications of water type (salt or freshwater) and
experimental conditions The experimental conditions included: (1) acute and chronic test
duration; (2) whether test organisms were fed or unfed; (3) test vessel material (glass or plastic);
(4) use of sol\ ent or no sol\ eni, and (5) the presence of a substrate. These data classifications
were used to observe if differences between nominal and measured concentrations of PFOS
could be linked to these ex peri mental conditions, as has been indicated in some toxicity literature
for PFAS (Boudreau et al 2<><>3a and b; Hansen et al. 2001; Martin et al. 2004).

Once grouped by the classifications, paired nominal and measured concentrations were
compared by water type and across the experimental conditions mentioned above through linear
correlation. The linear correlation evaluations were followed by comparisons of measured
concentrations as a percent of nominal in relation to a threshold of greater than ฑ 20% or 30% to
better understand the magnitude and trend of any discrepancies identified in the linear

0-2


-------
correlations. Lastly, changes in PFOS concentration in a test solution over time were evaluated
for the studies that measured PFOS concentrations in a test solution at the time it was introduced
(new solution) and at a later time (old solution) in the exposure duration. All methods pertaining
to these analyses are included below.

0.2.1 Linear Correlation Analysis

The linear correlation analysis plotted nominal concentrations on the X-axis and

corresponding measured concentrations on the Y-axis and assessed correlation between the

paired concentrations across the classifications of water type (freshwater or saltwater) and

experimental conditions. Additionally, the geometric means of the ratios between measured and

nominal concentrations and the median percent differences were calculated across the

classifications of water type and experimental conditions The geometric means of the ratios

between measured and nominal concentrations were calculated In di\ iding the measured

concentration by the nominal. The median percent differences were calculated as the absolute

value of the difference between the nominal and measured concentration divided by the nominal

concentration multi pi ied In I'm

0.2.2 Assessment of Measured Concentrations as a Percent of Nominal

The assessment of measured concentrations as a percent (or relative ratio) of nominal was

used to identify the proportion of paired nominal and measured PFOS concentrations that were

outside a threshold of greater than either ฑ 20% or 30%. Measured concentrations within 20% of

the corresponding nominal concentrations were considered in close agreement within one

another based on EPA's Office of Chemical Safety and Pollution Prevention (OCSPP)'s

Ecological Effects Test Guidelines. For example, U.S. EPA (2016c) states, "measured

concentration of test substance at each treatment level remains within plus or minus (ฑ) 20% of

the time-weighted average concentration for the duration of the test." Similarly, U.S. EPA

0-3


-------
(1996) states, "In any case there must be evidence that test concentrations remained at least 80
percent of the nominal concentrations throughout the test or that mean measured concentrations
are an accurate representation of exposure levels maintained throughout the test period
Finally, the Organization for Economic Cooperation and Development (OECD 2019) defines a
stable exposure concentration as, "A condition in which the exposure concentration remains
within 80-120% of nominal or mean measured values over the enure exposure period
Recently, in a study of key considerations for accurate exposures in ecotoxicological assessments
of perfluorinated carboxylates and sulfonates, Rewerts et al. (2021) used a threshold of ฑ 30% to
agree with nominal concentration for both stock and exposure solutions, as specified by the
guidelines in the consolidated Quality Systems Manual for Environmental Laboratories set by
the U.S. Department of Defense and the I S Department of I -neiuy (C oats et al. 2017). The
assessment of measured concentrations as a percent of nominal was used to understand the
magnitude and trend of PI-OS concentrations from the toxicity literature that were outside either
one of these two thresholds Similar to the linear correlations above, the assessment of measured
concentrations as a percent of nominal was conducted across the classifications of water type
(salt or freshwater) and experimental conditions

0.2.3 Assessment of Measured PFOS Concentrations in Test Solution Over Time

The assessment of measured PFOS concentrations in a test solution over time compared

concentrations from studies that measured PFOS in a solution at the time it was introduced (new

solution) and at a later time (> 2 days) in the test (old solution). Thus, this comparison was

limited to measurements made on the same solution at different times and excludes

measurements on the same treatment but not the same solution. There were five freshwater

studies (Drottar and Krueger 2000e; Drottar and Krueger 2000i; Sanderson et al. 2002) that

allowed a direct comparison of the change in PFOS concentration in a test solution over time

0-4


-------
(see greater detail in Appendices A through G). This assessment excludes Yang et al. (2014),
which measured PFOS concentrations on the same treatment, but not the same solution and
Palmer and Krueger (2001), which measured PFOS from different renewal solutions of the same
treatment. Unlike the two previous analyses, this assessment was not conducted across the
classifications of water type (salt or freshwater) and experimental conditi ons since the available
dataset was relatively limited.

0.3 Results of Meta-Analysis to Compare of Nominal and Measured PFOS
Concentrations

0.3.1 Examination of Freshwater and Saltwater Data with Discrepancies between Nominal and
Measured Concentrations

Of the 477 freshwater pairs of measured and nominal concentrations evaluated in this
meta-analysis, 85 were greater than the 2' >".. threshold and 57 w civ greater that the 30%
threshold described by Rewerts et al. (2021) Of these (->3 pairs are from several tests and
discussed in more detail in TaMe ()-5 below Any study not summarized in Table 0-5 had
nominal and measured concentrations that were within the 20% exceedance threshold.

Evaluation of saltwater data was consistent with the approach described for the
freshwater data Discussion of similar studies with large discrepancies are described in Table 0-6
(also see greater detail in Appendices A through G). Of the 171 pairs of measured and nominal
concentrations e\ aluated in this meta-analysis, 112 were greater than the 20% threshold and 77
were greater than the 3< >" <ป threshold used by Rewerts et al. (2021). In this examination of the
saltwater data, it was apparent that none of the experimental factors could explain the differences
between nominal and measured concentrations. Instead, apparent systematic dosing issues were
indicated as the cause for the observed differences. Similar to the freshwater dataset, any
saltwater study not summarized in Table 0-6 had nominal and measured concentrations that
were within the 20% exceedance threshold.

0-5


-------
0.3.2 Comparison of Paired Nominal and Measured PFOS Concentrations through Linear
Correlation

0.3.2.1 Linear Correlation of Nominal and Measured PFOS Concentrations by Water Type
Overall, the comparison of nominal and measured concentrations from the freshwater

PFOS toxicity literature indicated that measured and nominal concentrations were in close

agreement. Specifically, the ratio of measured to nominal concentrations from the freshwater

dataset showed little bias with a geometric mean value of 0.9676.

Figure 0-1A also displays the strong correlation (0.9770) of the 477 pairs of nominal and

measured concentrations, with the pairs mostly falling in a tight range (close lo I <>) The median

percent difference between pairs was 6.923%.

In contrast, the comparison of measured to nominal concentrations in saltwater showed

greater differences, with most measured concentrations being lower than the nominal

concentrations indicated. Specifically, while I'iuiiie ()-115 displays the strong correlation

(0.9735) of the 171 pairs of nominal and measured concentrations, the ratio of measured to

nominal concentrations from the saltwater dataset showed bias with a geometric mean value of

0.6468. Additionally, the median percent difference between measured and nominal

concentration was 2X These results indicate that measured and nominal concentrations

from saltwater tests were not in close agreement, with most measured concentrations being lower

than nominal These disparities were examined further in the subsequent analysis below to better

understand the magnitude of these apparent differences in relation to the 20% threshold (see

Section 0.3.3 below)

0-6


-------
Nominal Concentration (mg/L)

Nominal Concentration (mg/L)

Figure O-l. Comparison of PFOS measured and nominal concentrations for freshwater (A)
and saltwater (B) data.

Many points overlie each other, such that the dark areas represent multiple values, over-plotted.

0.3.2.2 Linear Correlation of PFOS Nominal and Measured Concentrations by Experimental
Condition

The nominal and measured concentration pairs were further compared in terms of
experimental conditions, keeping the data separated by water type. In general, strong correlations
were observed across all experimental conditions in freshwater (correlations > 0.95) with only

0-7


-------
slightly weaker correlation in saltwater tests (correlations > 0.84). Additionally, the freshwater
nominal and measured concentrations were in close agreement across experimental conditions.
Ratios between paired concentrations were generally in a tight range (close to 1.0) with a
geometric mean ratio of > 0.90, which indicated experimental conditions play little role in the
observed differences between nominal and measured PFOS concentrations in freshwater. The
exception was the inclusion of substrate in freshwater tests where llie geometric mean value of
the ratios between measured and corresponding nominal concentrations was only 0.7455.

In contrast to freshwater, the saltwater nominal and measured concentrations were found
to be in less agreement across experimental conditions compared to the freshwater dataset, with
the pairs falling outside a tight range and geometric mean \ allies of the ratios between measured
and nominal concentrations ranging between n 5<)5(-> and <> NI I 5 (TaMe ()-l).

0-8


-------
Table O-l. Comparison of Pairs Nominal and Measured PFOS Concentrations across
Experimental Conditions.		



Freshwater

Saltwater







Geometric







Geometric









Mean of

Median





Mean of

Median



# Paired



Measured/

Percent

# Paired



Measured/

Percent

1 CiiiHliiiiui

Obs.

Correlation

Nominal

Difference

Obs.

Correlation

Nominal

Difference

\ciilc

211

0.9978

0.9608

5.544%

82

0.9706

0.5056

39.52%

( limine

266

0.9570

0.9731

8.035%

89

0.9911

0.8115

19.30%

I ii led

213

0.9978

0.9663

5.556%

25

0.8430

0.7701

21.50%

led

264

0.9569

0.9687

8.000%

146

0.9002

0.6277

29.52%

Sol\ cm

78

0.9981

0.9035

8.147%

31

0.9455

0.7824

20.81%

\ii sii|\ cut

399

0.9645

0.9807

6.552%

140

0.8(01

0.6201

31.85%

Suhsirale

49

0.9951

0.7455

48.40%

1

a

\iป substrate

428

0.9826

0.9970

5.795%

170

0.9547

0.6456

28.13%

(ilass

155

0.9993

1.031

6.957%

72

0.9959

0.7707

22.20%

Masiic

245

0.9567

0.9405

5.932%

99

0.9705

0 5(ii)4

36.34%

I nspeeil'ied
material

77

0.9980

0.9317

9.403%

0

a

aNot evaluated due to a lack of data for these experiment condition categories.

0.3.2.3 Comparisons by Test Duratioi i and (h ;<;< // nsinal 1 ฆ ccc In /( V >i iditions

First, freshwater and saltwater nominal and measured concentrations were compared by

exposure duration (i e . acute or chronic) and oruanismal feeding. In the freshwater dataset, the

acute pairs were the same as the unfed pairs and the chronic pairs are the same as the fed pairs,

with the exception of two data points from the unfed chronic mussel test of Hazelton et al.

(2012) Coin ersely. while pairs for saltwater acute and chronic tests were more evenly split 82

vs. 8l) pairs, respectively, most pairs were from fed tests, despite being acute tests.

The comparison of nominal and measured concentrations in freshwater indicated that

measured and nominal concentrations under acute, chronic, unfed, and fed conditions displayed

strong correlations (> 0.95; Table O-l). Further, the freshwater nominal and measured

concentrations were found to be in close agreement, with the pairs mostly falling in a tight range

(close to 1.0) with geometric mean values of the ratios between measured and nominal

concentrations being > 0.96, indicating that test duration and organismal feeding play little role

in the observed differences between nominal and measured concentrations of PFOS in

0-9


-------
freshwater. Additionally, the median percent differences for the pairs across these experimental
conditions were equal to or < 8% (Table 0-1).

In contrast, the saltwater nominal and measured concentrations were in less agreement,
with test pairs under acute, chronic, unfed, and fed conditions falling in a highly variable range
with geometric mean values of the ratios between measured and nominal concentrations being
anywhere from 0.5056 to 0.8115. Additionally, the median percent differences for the pairs
across these experimental conditions were relatively high (> 19%; Table ()-1) The saltwater
comparison of nominal and measured concentrations indicates that these experimental conditions
(acute/chronic and unfed/fed) may influence the observed differences between measured and
nominal concentrations.

O-IO


-------
1000
100
10
1

0.1

0.0001

0.00001

A: Freshwater Acute

0.00001 0.0001

0.001 0.01 0.1	1

Nominal Concentration (nig/L)

1000
100
10
1

0.1
0.01
0.001

0.00001

B: Freshwater Chronic



O

oy
o a

0.00001 0.0001

0.001 0.01 0.1	1

Nominal Concentration (mg/L)

10

100

1000

100

0.1

0.01

0.0001

0.00001

C: Saltwater Acute

0.1

0.01

0.0001

0.00001

I): Saltwater Chronic

O0

8

0.001 0.01	0.1

Nominal Concentration (mg/L)

0.00001 0.0001

0.001 0.01	0.1

Nominal Concentration (mg/L)

10

Figure 0-2. Comparison of PFOS measured and nominal concentrations in freshwater (top) and saltwater (bottom) tests with
acute (A and C, respectively) and chronic durations (B and D, respectively).

In the freshwater datasets acute is the same as unfed, and chronic the same as fed. with the exception of two data points. Many points overlie each other, such that the dark areas
represent multiple values, over-plotted.

O-ll


-------
0.3.2.4 Comparisons by Test Vessel Material

The potential influence of additional experimental conditions related to test vessel

material (e.g., plastic, glass, or unspecified test vessels) were also considered. This comparison
considers the influence of the material from which the exposure chambers were constructed for
freshwater and saltwater tests. The "unspecified" designation were those tests that did not specify
the material used to construct the exposure chambers and appears to be similar to the other two
test vessels types (Table 0-1); thus, this experimental condition was not included in Figure 0-3
below.

The comparison of nominal and measured PI-OS concentrations in freshwater tests
indicated that measured and nominal concentrations lor tests using various test vessel types
displayed strong correlations (> 0.95; Table O-1) Additionally, the freshwater nominal and
measured concentrations were found to be in close agreement, with pairs mostly falling in a tight
range (close to 1.0) with geometric mean of the ratios between measured and nominal
concentrations bei ng <) The median percent differences for the pairs across these
experimental conditions were l<>"<> (Table O-l) These results indicate that test vessel material
played little role in the obser\ ed differences between nominal and measured concentrations of
PFOS in lYeshwaler tests.

In contrast, the saltwater comparison of nominal and measured concentrations indicated
that tests using various test vessel types of plastic and glass were in less agreement and nominal
to measured ratios displayed higher variability than in corresponding test vessel materials from
freshwater tests. The geometric mean ratio of measured and nominal PFOS concentrations from
saltwater tests using glass was 0.7707. Pairs from saltwater tests conducted in plastic test vessels
showed the highest degree of variability, with a geometric mean of 0.5694. Additionally, the

0-12


-------
median percent differences for the pairs across test vessel materials were relatively high (> 22%;
Table 0-1).

0-13


-------
100 -

0.001

0.0001 -I

0.00001 0.0001

0.001 0.01 0.1	1

Nominal Concentration (mg/L)

0.00001 0.0001

0.001 0.01 0.1	1

Nominal Concentration (mg/L)

0.001

0.0001

0.01

0.001

0.0001

0.001 0.01	0.1

Nominal Concentration (mg/L)

1	10

0.00001 0.0001

0.001 0.01 0.1
Nominal Concentration (mg/L)

Figure 0-3. Comparison of PFOS measured and nominal concentrations in freshwater (top) and saltwater (bottom) tests
conducted in glass (A and C, respectively) and plastic test vessels (B and D, respectively).

Many points overlie each other, such that the dark areas represent multiple values, over-plotted.

0-14


-------
0.3.2.5 Comparisons by Presence of Solvent

The potential influence of the use of a solvent was also considered as part of the

experimental conditions that might impact potential differences between nominal and measured

concentrations in the PFOS toxicity literature. Measured and nominal concentrations for

freshwater tests using solvent and not using solvent both displayed strong correlations of >0.99

and > 0.96, respectively (Table 0-1). The nominal and measured concentrations were in

agreement in tests with solvent (geometric mean of nominal to measured ratios was 0.9035), and

in even closer agreement in tests without solvent (geometric mean of nominal to measured ratios

was 0.9807).

The pairs from saltwater tests with or without sol\ cut were in less agreement than
freshwater pairs, with a geometric mean of the ratios between measured and nominal
concentrations being 0.7824 with solvent and 0.0201 without solvent. Further, the median
percent difference for the pairs with solvent was 2.1% while the median percent difference for
saltwater tests without a soKent was 32%.

0.3.2.6 Comparisons by /'reseiice of Substrate

I .astly, the potential inlluence of the presence of a substrate was considered as part of the

experimental conditions that might impact potential differences between nominal and measured

concentrations in the PFOS toxicity literature. This experimental condition was considered to be

potentially important since the presence of substrate could possibly increase the likelihood of

disparities between measured and nominal given the potential sorption of PFOS under certain

environmental conditions (see Section 2.2), but there are too few data to be conclusive.

Generally, strong correlations (> 0.98; Table O-l) were observed with freshwater pairs from test

with and without substrate. However, the geometric mean of the ratios between measured and

nominal concentrations were 0.9970 (no substrate) or 0.7455 (substrate) indicating that the

0-15


-------
presence of substrate may be removing the PFOS from test solutions. Similarly, the median
percent differences of 5.795% for tests without substrate and 48.40% with substrate indicate
disparities between measured and nominal concentrations that may be associated with the
presence of substrate.

There were insufficient data for the same comparison in saltwater because only one
observation of paired and nominal concentrations was available lor saltwater tests with substrate.
Pairs from saltwater tests without substrate were disparate, with a geometric mean of the ratios
between measured and nominal concentrations being 0.6456 and a median percent difference of
28.13%.Drottar and Krueger (2000e); Drottar and Krueger (2<~><~>0f)

0.3.3 Assessment of Measured Concentrations as a Percent of Nominal in Relation to the 20%

Threshold

0.3.3.1 Assessment of Measured Concentrations as a Percent <>J Sominal based on Water
Type

In general, the freshwater nominal and measured concentrations were found to be in close
agreement, with limited instances (only about 18%) of measured concentrations differing from
paired nominal concentrations In more than 2<)"() It should be noted that the majority of these
are from studies detailed in TaMe ()-5 Mow A smaller portion, approximately 12%, of
measured concentrations differed from paired nominal concentrations by more than 30%. (Figure
0-4). The magnitude of measured concentrations as a percent (presented as relative ratio in
Figure 0-4 below ) of nominal varied by a minimum of 6.8% to a maximum of 452% of nominal
(Table 0-2). In contrast, while the saltwater nominal and measured concentrations were found to
be in relatively close agreement in the linear correlation, a much higher proportion (65.5%) of
nominal and measured concentration pairs were outside the 20% threshold and 45% outside the
30%) threshold (Figure 0-4). The magnitude of measured concentrations as a percent of nominal
varied by a minimum of 12.45%) to a maximum of 120% of nominal (Table 0-2). However, as

0-16


-------
noted previously, the saltwater dataset was much smaller and the systematic discrepancies
between measured and nominal concentrations are harder to decipher.

Table 0-2. Proportion of Paired Nominal and Measured PFOS Concentrations Outside the
20% Threshold.



l-'lVsllHilkT

Saltwater







Min. of

Max. of





Min. of









Measured

Measured





Measured

Max. of





Proportio

Concentration

Concentration



Proportion

Concentration

Measured



#

n Outside

as a Percent

as a Percent

#

Outside +/-

as a Percent

Concentration

l'!\perimenlal

Paired

+/- 20%

of Nominal

of Nominal

Paired

20%

of Nominal

as a Percent of

Condition

Obs.

Threshold

(%)

(%)

Obs.

Threshold

(%)

Nominal (%)

All

477

0.1782

6.8

452

171

0.6550

12.45

120

Acute

211

0.0806

54.7

142

82

0.8659

12.45

93.00

Chronic

266

0.2556

6.8

452

89

0.4607

45.76

120

Unfed

213

0.0892

54.7

452

25

0.5200

55.50

93.00

l-'ed

264

0.2500

6.8

288

146

0.6781

12.45

120

Solvent

78

0.1282

45

124

31

0.5161

55.50

93.00

No Solvent

399

0.1880

6.8

452

140

0.6857

12.45

120

Substrate

49

0.7347

6.8

288

1

_i

No Substrate

428

0.1145

45

452

170

0.6588

12.45

120

Glass

155

0.1290

54.7

452

72

0.5694

45.76

103.3

Plastic

245

0.2204

6.8

288

99

0.7172

12.45

120

Unspecified

77

0.1429

45

136

0

_i

1	Not evaluated due to a lack of data for these experiment amdilimi calcunries.

2	Bolded values represenl lesi amdilimis with a high proportion of measured concentrations that were not within
20% threshold.

0-17


-------
A: Freshwater

O

120
100

•o	o—

ฐ-V. ฐs	^"_.1g^sVs

% #

o-

0.001

B: Saltwater

o.ioo

Nominal Concentration (mgL)

10.000

1,000.000

130
120

100

> 70

8 0

0.001

0.100

Nominal Concentration (mg'L)

10.000

Figure 0-4. Assessment of measured concentrations as a percent of nominal in relation to a
20% and 30% threshold for freshwater (A) and saltwater (B) data.

The horizontal line coding: 100% (dash line); +/- 20% (solid lines); +/- 30% (dash/dot lines). Data within these
thresholds were considered to result in close agreement of measured and nominal concentrations. Many points
overlie each other, such that the dark areas represent multiple values, over-plotted.

0.3.3.2 Assessment ofMeasured Concentrations as a Percent of Nominal based on
Experimental Conditions

In freshwater across all test conditions, 18% paired nominal and measured concentrations
were outside the 20% threshold. In the various subsets of conditions, the proportion of these
values outside this threshold ranged from 8 to 73% (excluding test with substrate, the maximum
percent of values outside the threshold was < 26%). Conversely, all saltwater pairs were

0-18


-------
observed to fall outside the 20% threshold across the experimental conditions, with all
experimental conditions having > 46% of pairs outside the 20% threshold.

To examine these observed differences further, key experimental conditions with either
noted differences documented in the linear correlation analysis above or previously documented
potential influence from the toxicity literature (Boudreau et al. 2003a and b; Hansen et al. 2001;
Martin et al. 2004) are presented below in greater detail. These experimental conditions include
(1) test duration in saltwater tests; (2) test vessel material in both freshwater and saltwater tests;
and (3) the presence of substrate in freshwater tests. The nominal and measured concentrations
for all other experimental conditions were found to be in close agreement (Table ()-2) or the
observed differences were considered to be influenced by individual studies, not the individual
experimental condition. Thus, these additional experimental conditions were not considered
further.

0.3.3.3 Assessment based on I est Duration in Saltwater lests

Similar to the linear correlation analysis above, the nominal and measured concentrations

from acute saltwater tests were disparate, with the majority (> 86%) of measured concentrations
as a percent of nominal (presented as relative ratio in Figure 0-5 below) falling outside the 20%
threshold and 73% outside the 3U% threshold. Additionally, the magnitude of the discrepancies
outside the 2<)% threshold ranged between 12.45 and 93.00% (Table 0-2). In contrast, while the
nominal and measured concentrations from chronic saltwater tests were found to be in closer
agreement in the linear correlation analysis, a relatively high proportion (46.07%) of measured
concentrations as a percent of nominal were outside the 20% thresholds and 19.10% outside the
30%) threshold (Figure 0-5B). Further, the magnitude of exceedances outside the 20% threshold
ranged between 45.76 and 120% (Table 0-2). Indicating that measured concentrations were

0-19


-------
sometimes much lower than corresponding nominal concentrations in both acute and chronic
saltwater tests.

A: Saltwater Acute

130

120

100"

so

u

> 70

1.0

B: Saltwater Chronic

3.0

10.0

Nominal Concentration (mg'L)

130
120

o 100--

>

"5 SO1

70

0.001

0.100

Nominal Concentration (mg'L)

10.000

Figure 0-5. Assessment of measured concentrations as a percent of nominal in relation to a
20% and 30% threshold for saltwater acute (A) and chronic (B) tests.

The horizontal line coding: 100% (dash line); +/-20% (solid lines); +/-30% (dash/dot lines). Data within these
thresholds were considered to result in close agreement of measured and nominal concentrations. Many points
overlie each other, such that the dark areas represent multiple values, over-plotted.

O-20


-------
0.3.3.4 Assessment based on Test Vessel Material in Freshwater and Saltwater Tests

Similar to the linear correlation analysis above, the freshwater nominal and measured

concentrations from tests using various test vessel types were found to be in close agreement,
with < 22% outside the 20% threshold and < 17% outside the 30% threshold (Figure 0-6). Tests
conducted in glass test vessels had the widest magnitude of measured concentrations varied with
respect to nominal by a minimum of 54.7% to a maximum of 452" n of nominal (Table 0-2).
These results further indicate that test vessel material played link- role in the observed
differences between nominal and measured concentrations of PFOS in freshwater tests.

As for the saltwater nominal and measured concentrations from tests using \ arious test
vessel types, relatively high proportions (56.94 and 71,72"..) of measured concentrations as a
percent of nominal were outside the 20% threshold for glass and plastic test vessels, respectively
(Figure 0-6B). Further, tests conducted in plastic test \ essels had the widest magnitude of
exceedances outside the 20% threshold, which ranged between 12.45 and 120% (Table 0-2).
Indicating that measured concentrations were typically lower than corresponding nominal
concentrations in saltwater tests However, as previously mentioned, the saltwater data were
relati\ ely limited, and it was dill! cult to discern if these observed differences were the result of
the test \ essel material, water type, or were di rectly related to the systematic discrepancies in
individual studies

0-21


-------
A: Freshwater Glass

B: Freshwater Various Plastic

*120-
100-

- tea..

120-ฆ
100-'
80-ฆ



-t"8S	T-

1.00

Nominal Concentration (mg'L)

1.00

Nominal Concentration (mg L)

C: Saltwater Glass

1.00

Nominal Concentration (rag/L)

D: Saltwater Various Plastic

2 so
.ฃ70

0.001	0.100

Nominal Concentration (mg'L)

10.000

Figure 0-6. Assessment of measured concentrations as a percent of nominal in relation to a 20% and 30% threshold for
freshwater (top) and saltwater (bottom) tests conducted in glass (A and C; respectively) and various plastic (B and D;
respectively) test vessels.

The horizontal line coding: 100% (dash line); +/-20% (solid lines); +/-30 i.e. (dash/dot lines). Data within these thresholds were considered to result in close
agreement of measured and nominal concentrations. Many points overlie each other, such that the dark areas represent multiple values, over-plotted.

0-22


-------
0.3.3.5 Assessment based on Presence of Substrate in Freshwater Tests

The nominal and measured concentrations from freshwater tests with the presence of a

substrate indicate a discrepancy, with 73.47% of measured concentrations as a percent of

nominal falling outside the 20% threshold and 63.27%) falling outside the 30% threshold.

Additionally, the magnitude of the exceedances outside the 20% threshold ranged between 6.8

and 288%) (Table 0-2). In contrast, the nominal and measured concentrations from freshwater

tests without the presence of a substrate were found to be in much close agreement, with few

(11.45%)) of the measured concentrations as a percent of nominal outside the 2<ปฐ <ป threshold and

6.07%) outside the 30% threshold (Figure 0-7B) I low ever, the magnitude of exceedances

outside the 20% threshold ranged between 45 and 452ฐ/.. in freshwater tests without the presence

of substrate (Table 0-2). It should be noted that measured and nominal concentration pairs for

studies with a substrate present were limited, especially lor saltwater tests in which there was

only a single pair. It is because of this data limitation that a similar assessment was not

conducted for saltvs ater tests

0-23


-------
A: Freshwater Substrate

o

o

o

...8	

o

o











o















w





o o

ฎ 0



o O



o o

ooo

o
8

8

o

1

o
o

o o

GO

O

0.01	1.00	100.00

Nominal Concentration (mg'L)

B: Freshwater No Substrate



o















o

ฐ o ฐ

o ฐ 0 0ฐ





O o



41^8 {#$- - |g - t-

o

	o	



	

ฐ on O ฃ % 0ฎ

o
o

o



o

o	r.	_.o	^	

ฎ o O ฐ

0.001	0.100	10.000	1,000.000

Nominal Concentration (mg L)

Figure 0-7. Assessment of measured concentrations as a percent of nominal in relation to a
20% and 30% threshold for freshwater tests with substrate (A) and without substrate (B).

The horizontal line coding: 100% (dash line); +/-20% (solid lines); +/-30% (dash/dot lines). Data within these
thresholds were considered to result in close agreement of measured and nominal concentrations. Many points
overlie each other, such that the dark areas represent multiple values, over-plotted.

0.3.4 Assessment of Measured PFOS Concentrations in Test Solution Over Time

Figure 0-8 compares the measured PFOS concentrations from old test solution versus the

new test solution and a strong correlation (0.9938) was observed. The measured PFOS

concentrations in old test solutions appear to correspond well to those in the new test solutions

0-24


-------
with the median ratio between new and old solutions of 1% (that is, the old solution 1% lower
than the new) and the geometric mean of the ratios between measured and nominal
concentrations of 0.9517. These results are limited considering the level of PFOS data that is
currently available. However, these results indicate that concentrations of PFOS in water were
stable over time.

Figure 0-8. PFOS exposure concentrations measured at the end of the renewal period or
static test (old solutions) vs. concentrations measured in the same solution immediately
after it was introduced into the exposure chamber (new solutions).

0.4 Conclusions of Measured Meta-Analysis to Compare of Nominal and
Measured PFOS Concentrations

The comparison of nominal and measured concentrations in the current PFOS toxicity

literature generally displayed a high degree of linear correlation and close agreement based on

geometric means of the ratios and the median percent differences between measured and nominal

concentrations. Assessment of measured concentrations as a percent (or relative ratio) of nominal

0-25


-------
also indicated that PFOS concentrations were generally in close agreement, with greater than
82% of the freshwater data within 20% threshold that is consistent with the test acceptability
threshold identified by EPA's OCSPP's Ecological Effects Test Guidelines. Recent PFAS
literature has indicated standard variability between nominal and corresponding measured
concentrations may even be as high as 30% and only 12% of paired values were outside this
threshold. For example, Rewerts et al. (2021) indicated that, nominal and measured
concentrations for both stock and exposure solutions should to fall within the margin of
100 ฑ 30%), as specified by the guidelines in the consolidated Quality Systems Manual for
Environmental Laboratories set by the U.S. Department of Defense and the U.S. Department of
Energy (Coats et al. 2017). Further, Rewerts et al. (2021) concluded the variability between
measured and nominal concentrations may he influenced by solution homogenization and
subsampling procedures, noting storage container type may influence agreement between
measured and nominal PTOS concentrations based on the concerns stated in previous literature.
However, it should he noted that container type (as glass or plastic) did not influence the
observed differences between measured and nominal PFOS concentrations in the analysis
presented here

Specifically, these analyses indicated that nominal and measured concentrations of PFOS
were in close agreement across freshwater data when grouped by water type and the following
experimental conditions (I) acute and chronic test duration; (2) whether test organisms were fed
or unfed; (3) test vessel material (glass or plastic); and (4) use of solvent or no solvent. Thus, the
results of this meta-analysis indicate that these experimental conditions had little influence on
any observed discrepancies between nominal and measured concentrations of PFOS. Instances
where measured concentrations were not found to be in close agreement with nominal

0-26


-------
concentrations (either in the linear correlation analysis or the assessment of measured
concentrations as a percent of nominal related to a 20% exceedance threshold) were isolated to a
few studies. In these cases, suspected dosing errors, unexplained phenomena, and/or presence of
substrate may have contributed to observed differences. Therefore, dosing errors and differences
in experimental design were not considered to be systemic issues across PFOS toxicity tests
since discrepancies were only observed in a small subset of the obser\ ed pairs of measured and
nominal concentrations. The presence of substrate caused disparities between measured and
nominal concentrations. However, in these tests measured concentrations were not systematically
less than or greater than nominal concentrations Therefore. PFOS could bind to substrate,
effectively removing it from the water column, or added substrate may even be acting as a source
of PFOS in certain instances. However, expected discrepancies between nominal and measured
concentrations in freshwater tests with substrate are of minimal concern to the final acute and
chronic freshwater PI-OS criteria magnitudes. No unmeasured tests with substrate were used to
quantitatively derive the PI OS acute criterion magnitude. Only one unmeasured test with
substrate (Spachmo and Arukwe 2<>l 2) was used to derive the chronic water column-based
criterion maunitude In this test. Spachmo and Arukwe (2012) used tank bed gravel to simulate a
riverbed en\ ironment for hatching Atlantic salmon. Spachmo and Arukwe (2012) observed no
effects at the single treatment concentration (i.e., 0.10 mg/L PFOS). Because no other chronic
toxicity data were a\ ailable for members of the genus Salmo, the 49-day growth-based NOEC of
0.10 mg/L served directly as the Salmo salar SMCV and the Salmo GMCV. The Salmo GMCV
was the sixth most sensitive GMCV, which had a minimal impact on the chronic water-column
criterion magnitude since it was not among the four most sensitive genera (see section C.2.6).

0-27


-------
Compared to freshwater tests there was much less agreement between nominal and
measured PFOS concentrations in saltwater tests bot as a whole and across experimental
conditions. Overall, results of this meta-analysis indicated that measured concentrations from
saltwater tests were systematically lower than nominal concentrations. Therefore, it could be
hypothesized (but not confirmed) that the PFOS saltwater benchmark is underproductive because
two unmeasured tests were included in the derivation process, including the acute 72-hr. acute
test on Paracentrotus lividus (Gunduz et al. 2013), which was the basis of ilie third most
sensitive estuarine/marine GMAV (see Appendix I.)

Based on the results of this meta-analysis and the general close agreement between
measured and nominal concentrations of PFOS, both measured and unmeasured PFOS toxicity
tests were used to derive the aquatic life criteria Additionally, use of both measured and
unmeasured PFOS toxicity tests is further supported In the high stability of PFOS indicated in
this meta-analysis (Section ().> 4) and current literature (see Section 1.2.1).

0-28


-------
Table 0-3. Freshwater Nominal and Measured Concentrations for PFOS.













Nomiiiid

Moiisuivd



( lioni ic;il



IVsl Vessel







(one.

(one.



/ Pu ri( \

Test Diimlion

Miilcriiil

Sซil\out I sod

\le;isiireinen( Mel hod

Time Measured

(mji/l.)



Reference











0 hours

5.7

5.47













4X hours

5.7

5.18













')(> hours

5.7

5.24













ii hours

5.7

4.93













48 hours

5.7

5.70













96 hours

5.7

5.26













0 hours

11

11.4













48 hours

11

11.2













96 hours

11

10.9













0 hours

11

10.1













48 hours

11

10.5













96 hours

11

15.4













0 hours

23

19.0



PFOS-K
90.49%







1 huh perlormaiiee liquid
chronialogiaph} willi
mass spectrometric
dclcclion

48 hours

23

18.7

Drottar and

96 hours

Polyethylene

None

96 hours

23

22.9

Krueger







0 hours

23

16.8

(2000d)









48 hours

23

18.7













96 hours

23

22.4













0 hours

46

37.2













48 hours

46

37.1













96 hours

46

48.2













0 hours

46

40.6













48 hours

46

39.5













96 hours

46

40.5













0 hours

91

69.0













48 hours

91

81.3













96 hours

91

88.2













0 hours

91

74.7













48 hours

91

77.6



0-29


-------












Nomiiiiil

Moiisuivd



( homiciil



Test Vessel







(tine.

(one.



/ Piiriu

Tcsl Dunilion

Miiloriiil

Snl\en( I sed

Measurement Method

1 inu' Moiisuml

(inii/l.)

(inii/l.i

UcTciyik'o











'H< hours

91

85.7















1) IkiIM's

0.005

0.0054





24 hours





HighPerformanee I.K|ind
Chromaloerapliv / Mas>s>
Spcciminelrv

II IkiIM's

0.05

0.0514

Hazelton et

PFOS

(glochidia) and

Glass

None

II IkiIM's

0.5

0.456

al. (2012),

>98%

96 hours

(assumed)

U liours

5

4.68

Hazelton



(juvenile)





0 hours

50

47.2

(2013)











0 hours

500

490















0 hours

12

10.5













24 hours

12

11.5













48 hours

12

10.9













0 hours

12

10.6













24 hours

12

12.5













48 hours

12

12.0













0 hours

20

17.2













24 hours

20

22.8













48 hours

20

21.4











1 hull IVi'luiiiiaiice Liquid
( lii'niiialnmapliy / Mass
S|xvli'niiictry

0 hours

20

18.1

Drottar and

Krueger

(2000a)

PFOS-K

48 hours

I'lasiic

None

24 hours

20

21.6

90.49%

48 hours

20

18.8









0 hours

33

30.2











24 hours

33

34.0













48 hours

33

31.3













0 hours

33

34.1













24 hours

33

36.1













48 hours

33

34.0













0 hours

55

50.5













24 hours

55

57.0













48 hours

55

56.8













0 hours

55

49.9



0-3 0


-------
( hcmic;il
/ Piiriu

Tesl Dumlion

Tesl Vessel
Miiloriiil

Sซil\onl I sed

Mc;iMiremenl Method

Tiiiu' Measured

Nomiiiiil
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











24 hours

55

63.0



4X hours

55

56.4

i) hours

91

87.6

24 hours

91

90.1

4X hours

91

88.7

U ho ui'!>

91

102

24 hours

91

84.4

48 hours

91

92.4



PFOS-K

99%

48 hours

Glass
(assumed)

DMSO

Midi Performance 1 .k|iikI
( hromatography/Mass
S|xvli'ninclrv

0 hours

20

19.8

Yang et al.
(2014)

48 hours

20

18.43

0 hours

377.91

372.35

48 hours

377.91

341.74



PFOS-K

86.9%

96 hours

(ilass

None

Liquid Chromatography /
Tandem Mass
Spectrometry

0 hours

1.82

2.58

Palmer and

Krueger

(2001)

96 hours

1.82

1.42

0 hours

3.07

3.94

96 hours

3.07

1.72

0 hours

5.19

6.62

96 hours

5.19

2.84

0 hours

8.64

10.7

96 hours

8.64

5.09

0 hours

14.4

18.5

96 hours

14.4

10.8

0 hours

24

26.9

96 hours

24

22.3



PFOS-K

86.9%

96 hours

Glass

None

Liquid Chromatography /
Tandem Mass
Spectrometry

0 hours

1.82

1.77

Palmer and

Krueger

(2001)

96 hours

1.82

2.04

0 hours

3.07

3.59

96 hours

3.07

2.49

0-31


-------
( homiciil
/ Piiriu

Tesl Dumlion

Tesl Vessel
Miiloriiil

Sol\enl I sed

Mc;iMiremenl Method

Tiiiu' Moiisuml

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











ii hours

5.19

5.45



')(> hours

5.19

4.18

i) hours

8.64

8.43

')(> hours

8.64

7.51

ii Ikmiis

14.4

14.5

9o houi b

14.4

12.1

0 hours

24

23.0

96 hours

24

23.1



PFOS-K

86.9%

96 hours

Glass



1 .iqiiid ( hionialouiaphv /
1 aiidcm Mass
S|Kvliniaetry

0 hours

1.82

1.77

Palmer and

Krueger

(2001)

96 hours

1.82

2.08

0 hours

3.07

3.59

96 hours

3.07

2.94

0 hours

5.19

5.45

96 hours

5.19

5.05

0 hours

8.64

8.43

96 hours

8.64

8.09

0 hours

14.4

14.5

96 hours

14.4

13.5

0 hours

24

23.0

96 hours

24

24.7



PFOS

>98%

36 days

(ilass

\iปne

High Performance Liquid
Chromatography / Mass
Spectrometry

10, 11 days

0.001

0.00452

Hazelton et
al. (2012)

10, 11 days

0.1

0.0695



PFOS-K
90.49%

21 days

Plastic

None

Reverse Phase High
Performance Liquid
Chromatography

0 days

1.4

1.78

Drottar and

Krueger

(2000b)

2 days

1.4

1.58

11 days

1.4

1.38

14 days

1.4

1.36

18 days

1.4

1.38

0-32


-------
( hoiniciil
/ Piiriu

Tesl l)ii r;iliซui

Tesl Vessel
Miiloi'iiil

Snl\en( I sed

Mc;iMiremenl Mel hori

Time Measured

Nomiiiiil
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











2 1 days

1.4

1.50













ii das s

1.4

1.72













: d;i\s

1.4

1.56













1 1 d;i\ s

1.4

1.47













I4d;i\s

1.4

1.32













18 da\ s

1.4

1.43













21 davs

1.4

1.45













0 days

2.9

3.20













2 days

2.9

3.01













11 days

2.9

2.75













14 days

2.9

2.85













18 days

2.9

2.79













21 days

2.9

2.81













0 days

2.9

3.05













2 days

2.9

3.07













11 days

2.9

2.77













14 days

2.9

2.71













18 days

2.9

2.81













21 days

2.9

2.82













0 days

5.7

5.97













2 days

5.7

5.65













11 days

5.7

5.63













14 days

5.7

5.36













18 days

5.7

5.58













21 days

5.7

5.24













0 days

5.7

5.87













2 days

5.7

5.72













11 days

5.7

5.59













14 days

5.7

5.39













18 days

5.7

5.75













21 days

5.7

5.37



0-33


-------












Nomiiiiil

Moiisuivd



( homiciil



Test Vessel







(ttnc.

(one.



/ Piiriu

Tcsl l)ii r;iliซui

Miiloriiil

Snl\en( I sed

Measurement Method

Tiiiu' Measured

(inii/l.)

(inii/l.)

UcTciyik'o











11 days

11

11.5













: da\s

11

11.6













1 1 dass

11

11.3













14 das s

11

11.2













IS da\s

11

11.8













21 da\ s

11

11.5













0 davs

11

11.5













2 days

11

11.8













11 days

11

11.3













14 days

11

11.6













18 days

11

11.6













21 days

11

11.3













0 days

23

24.2













2 days

23

24.0













11 days

23

22.8













14 days

23

23.6













18 days

23

24.8













0 days

23

23.1













2 days

23

24.6













11 days

23

22.5













14 days

23

23.1













18 days

23

25.0













0 days

46

47.3













2 days

46

49.1













0 days

46

48.0













2 days

46

49.4













High Performance Liquid
Chromatography / Mass
Spectrometry

After renewal

2

1.98



PFOS-K

21 days

Glass

DMSO

Before renewal

2

1.74

Yang et al.

99%

(assumed)

After renewal

7.43

7.54

(2014)









Before renewal

7.43

6.78



0-34


-------
( homiciil
/ Piiriu

Tesl Dumlion

Tesl Vessel
Miiloriiil

Sol\enl I sed

Mc;iMiremenl Method

Tiiiu' Measured

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference



PFOS-K

95%

20 days

Polypropylene

None

Liquid Chromalographv /
Tandem Mass
Speclromelry

20 dav

0.001

0.0023

MacDonald
et al. (2004)

20 dav

0.005

0.0144

20 dav

0.01

0.0217

2() da>

0.05

0.0949

20 da>

0.1

0.149



PFOS-K

95%

10 days

Polypropylene

None

Liquid Chromalnmaphs
Tandem Mass
Speclrometiy

10 day

0.001

0.0008

MacDonald
et al. (2004)

10 day

0.005

0.0046

10 day

0.01

0.0115

10 day

0.02

0.0241

10 day

0.04

0.0491

10 day

0.08

0.0962

10 day

0.15

0.1501



PFOS-K
90.49%

47 days

(ilass

Nunc

111*1.( MS Verification

0 days

0.14

0.147

Drottar and

Krueger

(2000c)

4 days

0.14

0.141

7 days

0.14

0.144

14 days

0.14

0.134

21 days

0.14

0.153

28 days

0.14

0.160

35 days

0.14

0.179

42 days

0.14

0.157

47 days

0.14

0.147

0 days

0.14

0.160

4 days

0.14

0.140

7days

0.14

0.148

14 days

0.14

0.135

21 days

0.14

0.143

28 days

0.14

0.158

35 days

0.14

0.173

0-35


-------
( hoiniciil
/ Piiriu

Tesl l)ii r;iliซui

Tesl Vessel
Miiloi'iiil

Snl\en( I sed

Mc;iMiremenl Mel hori

Time Measured

Nomiiiiil
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











42 days

0.14

0.160













4" das s

0.14

0.155













ii d;i\ s

0.29

0.287













4 d;i\ s

0.29

0.270













~ d;i\ s

0.29

0.292













14 da\ s

0.29

0.269













21 davs

0.29

0.307













28 days

0.29

0.343













35 days

0.29

0.311













42 days

0.29

0.319













47 days

0.29

0.296













0 days

0.29

0.277













4 days

0.29

0.289













7days

0.29

0.296













14 days

0.29

0.266













21 days

0.29

0.315













28 days

0.29

0.341













35 days

0.29

0.325













42 days

0.29

0.313













47 days

0.29

0.276













0 days

0.57

0.571













4 days

0.57

0.619













7 days

0.57

0.597













14 days

0.57

0.539













21 days

0.57

0.608













28 days

0.57

0.639













35 days

0.57

0.646













42 days

0.57

0.575













47 days

0.57

0.545













0 days

0.57

0.576













4 days

0.57

0.659



0-36


-------
( hoiniciil
/ Piiriu

Tesl l)ii r;iliซui

Tesl Vessel
Miiloi'iiil

Snl\en( I sed

Mc;iMiremenl Mel hori

Time Measured

Nomiiiiil
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











"days

0.57

0.642













14 davs

0.57

0.535













2 1 davs

0.57

0.580













2X davs

0.57

0.617













.'5 das s

0.57

0.644













42 da\ s

0.57

0.576













47 davs

0.57

0.543













0 days



1.14













4 days



1.21













7 days



1.13













14 days



1.03













21 days



1.19













28 days



1.3













35 days



1.3













42 days



1.14













47 days



1.13













0 days



1.13













4 days



1.25













7days



1.23













14 days



1.1













21 days



1.24













28 days



1.31













35 days



1.31













42 days



1.19













47 days



1.09













0 days

2.3

2.21













4 days

2.3

2.52













7 days

2.3

2.43













0 days

2.3

2.27













4 days

2.3

2.46













7days

2.3

2.38



0-37


-------
( homiciil
/ Piiriu

Tesl Dumlion

Tesl Vessel
Miiloriiil

Sol\enl I sed

Mc;iMiremenl Method

Tiiiu' Moiisuml

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











11 days

4.6

4.56



4 da\ s

4.6

4.79

~ da\ s

4.6

4.46

ii da\ s

4.6

4.4

4 das s

4.6

4.79

7da>s>

4.6

4.76



PFOS

>98%

21 days (FO) +
24 days (Fl)

Glass

None

Liquid Chronialnmaphv /
Electrospray lniii/alinn
Mass Specli'ninch's

Not reported

0.03

0.0276

Ankley et al
(2005)

Not reported

0.1

0.101

Not reported.

0.3

0.281

Not reported.

1

0.818



PFOS
98%

110 days

(ilass

\iปne

Liquid Chromatography /
Eleclrospra> Ionization/
Mass Spectrometry

Vveraged over
110 days

0.03

0.04

Ankley et al.
(2004)

Averaged over
110 days

0.1

0.13

Averaged over
110 days

0.3

0.36

Averaged over
110 days

1

0.97

Averaged over
110 days

3

3.55

Averaged over
110 days

10

12.5



PFOS
98%

4 months

\nl rcpuricd

DMSO

Liquid chromatography
coupled to tandem mass

spectrometry with
electrospray ionization

0 hours after water
change

0.0001

0.00009

Lou et al.
(2013)

24 hours after
water change

0.0001

0.00006

48 hours after
water change

0.0001

0.000045

0 hours after water
change

0.001

0.001

0-38


-------
( homiciil
/ Piiriu

Tesl Dumlion

Tesl Vessel
Miiloriiil

Sol\enl I sed

Mc;iMiremenl Method

Tiiiu' Measured

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











24 hours after
water change

0.001

0.00045



4N hours after
w ale r change

0.001

0.00045

(i hours after water
change

0.1

0.1117

24 lioui'b alter
u ater change

0.1

0.0898

4X hours after
w ater change

0.1

0.0891

0 hours after water
change

1

0.716

24 hours after
water change

1

0.661

48 hours after
water change

1

0.632



PFOS-K
Unknown

35 days

Polyvinyl
chloride

Nolle

Hiah-pcrformance liquid
chromatography ion
chromatography

1 day

1

1.33

Sanderson et
al. (2002)

8 days

1

1.15

35 days

1

1.08

1 day

10

12.3

35 days

10

11.7

1 day

30

33.9

35 days

30

29.8



PFOS-K

99%

96 hours

\oi reported.

I)\1S0

High Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

60

58.94

Yang et al.
(2014)

Before Renewal

60

51.42

After Renewal

149.3

150.84

Before Renewal

149.3

127.73



PFOS-K

99%

96 hours

Not repoileil

DMSO

High Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

100

99.85

Yang et al.
(2014)

Before Renewal

100

86.5

After Renewal

371.29

368.24

0-39


-------
( homiciil
/ Piiriu

Tesl l)u ration

Tesl Vessel
Material

Sol\enl I sed

Measurement Method

Tiiiu' Measured

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(nili/l.)

Reference











1 ieloi'c Renewal

371.29

328.84





PFOS-K

99%

96 hours

Not reported.

DMSO

High Performance Liquid
Chromatography / Mass
Spec tro met rv

\lier Renewal

6

5.95

Yang et al.
(2014)

1 ieloi'c Renewal

6

4.88

After Renewal

113.37

109.22

Before Renewal

113.37

97.85



PFOS-K

99%

96 hours

Not reported.

DMSO

HighPerlorn lance Liquid
Chromatography Mass
Spectromet i-\

After Renewal

100

99.46

Yang et al.
(2014)

Before Renewal

100

92.24

After Renewal

371.29

369.29

1 '.efore Renewal

371.29

340.45





28 days

Not reported.

None (assumed)

Re\eised Phase Liquid
Chronialomaphs w ilh
Elecl rospray ionizalion
Mass Spectrometry

Average (See
Note)

0.001

0.00081

Roland et al.
(2014)

Average (See
Note)

0.01

0.011



PFOS-K

99%

96 hours

Not reported.

DMSO

High Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

20

19.2

Yang et al.
(2014)

Before Renewal

20

17.5

After Renewal

209.72

203.43

Before Renewal

209.72

183.36



PFOS

>98%

96 hours

(ilass

DMSO

High Performance Liquid
Chromatography / Mass
Spectrometry

See Note

0.500

0.5201352

Feng et al.
(2015)

See Note

5.001

5.0913234



PFOS
100.3%

96 hours

Glass u n h
Teflon bloel
components

N,N-
di met hv lformamide

Liquid Chromatography /
Mass Spectrometry

See Note

0.05

0.045

Kim et al.
(2010)

See Note

0.5

0.62

See Note

5

5.395

See Note

50

48.242

0-40


-------
( hcmic;il
/ Piiriu

Tesl l)u ration

Tesl Vessel
Material

Sol\ent I sซl

Measurement Method

Tiiiu' Moiisuml

Nomiiiiil
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference

PFOS

>96%

114 hours

Not provided.

DMSO

Liquid Chromatography /
Mass Specl ro met ry

11 days

0.5001

0.5051

Huang et al.
(2010)

5 davs

0.5001

0.5064

ii davs

2.0005

2.0403

5 davs

2.0005

2.1603

0 das s

4.001

4.0123

5 da\ s

4.001

4.2228



PFOS

>98%

24 hours

Plexiglass

ETOH

Liquid Chromatography /
Mass Specl i\> meli}

Not reported.

0.0950247

0.0550143

Huang et al.
(2016)

Not reported.

0.950247

0.600156

Not reported.

9.50247

9.00234



PFOS-K

99%

96 hours

Not reported.

DMSO

1 hull IVrl on nance Liquid
( hioiiiatomaphs Mass
Specliomcli'N

\flcr Renewal

30

29.1

Yang et al.
(2014)

1 '.cfore Renewal

30

26.12

After Renewal

227.81

220.98

Before Renewal

227.81

195.2



PFOS-K

99%

30 days

Not reported.

DMSO

Higli Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

1.5

1.48

Yang et al.
(2014)

Before Renewal

1.5

1.34

After Renewal

5.57

5.73

Before Renewal

5.57

4.98



PFOS-K

99%

96 hours

Not reported

DMSO

High Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

20

21.16

Yang et al.
(2014)

Before Renewal

20

17.92

After Renewal

151.88

150.03

Before Renewal

151.88

137.93



PFOS-K

99%

30 days

Not reported

DMSO

High Performance Liquid
Chromatography / Mass
Spectrometry

After Renewal

1.5

1.47

Yang et al.
(2014)

Before Renewal

1.5

1.31

After Renewal

5.57

5.46

Before Renewal

5.57

4.85



0-41


-------
( homiciil
/ Piiriu

Tcsl l)ii r;iliซui

Test Vessel
Miiloriiil

Snl\en( I sed

Measurement Method

1 inu' Moiisuml

Nomiiiid
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

UcTciyik'o

PFOS-K

>98%

14 days

Polypropylene

None

l.(-MS/MS

ii hours

50

64.14

Funkhouser
(2014)

')(> hours

50

75.61

i) hours

75

110.2

')(> hours

75

135.4

ii Ikmiis

100

131.8

9o houi b

100

174.2

0 hours

125

172.4

96 hours

125

223.9

0 hours

150

236.3

96 hours

150

313.3

0 hours

175

170.8

96 hours

175

231.3



PFOS

>98%

40 days

Plastic

Wine

l.( -MS MS

40 days

10

7.662

Hoover et al.
(2017)

40 days

100

76.34

40 days

1000

877.6



PFOS-K

86%

285 days

Plastic

None

Ion cinematography

6-33 days

3

2.681294118

Boudreau et
al. (2003b)

6-33 days

10

9.998625

6-33 days

30

29.85411765



PFOS-K

>98%

330 days

(ilass

Nunc

l.(-MS/MS

30-60 days

0.0006

0.0007568

Keiter et al.
(2012)

30-60 days

0.1

0.12907375

30-40 days

0.3

0.259083333



PFOS-K
90.49%

96 hr

l\>l\elh\ lone

None

HPLC/MS

0 hr

3.6

3.16

Drottar and

Krueger

(2000)

48 hr

3.6

3.08

96 hr

3.6

3.46

0 hr

3.6

3.53

48 hr

3.6

3.22

96 hr

3.6

3.13

0-42


-------
( hoiniciil
/ Piiriu

Tesl l)ii r;iliซui

Tesl Vessel
Miiloi'iiil

Snl\en( I sed

Mc;iMiremenl Mel hod

Time Measured

Nomiiiid
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











0 hr

5.9

6.05













4S hr

5.9

5.48













l><> hr

5.9

5.7













u hr

5.9

5.07













4S hr

5.9

5.89













96 lir

5.9

5.55













0 hr

9.9

8.99













48 lir

9.9

9.88













96 hr

9.9

9.7













0 hr

9.9

9.47













48 hr

9.9

9.33













96 hr

9.9

9.52













0 hr

16

18.2













48 hr

16

15













96 hr

16

14.8













0 hr

16

19.3













48 hr

16

15.6













96 hr

16

16.2













0 hr

27

28.5













48 hr

27

27













96 hr

27

26.8













0 hr

27

28.5













48 hr

27

27.8













96 hr

27

26.6















0 hr

3.1

3.15













48 hr

3.1

2.9



PFOS-K

96 hr

Polyclli\ lone

None

HPLC

96 hr

3.1

2.83

Palmer et al.

86.9%

0 hr

3.1

3.02

(2002)











48 hr

3.1

3.01













96 hr

3.1

2.97



0-43


-------
( hoiniciil
/ Piiriu

"I'c-sl l)ii r;iliซui

Test Vessel
Miiloi'iiil

Snl\en( I sed

MeiisuremeiH Mel hod

Time Measured

Nomiiiid
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference











0 hr

6.3

6.22













4X hr

6.3

6.16













hr

6.3

6.15













u hr

6.3

6.21













4S hr

6.3

6.43













96 lir

6.3

6.6













0 hr

13

13.2













48 lir

13

12.7













96 hr

13

13.1













0 hr

13

12.1













48 hr

13

12.3













96 hr

13

12.6













0 hr

25

25













48 hr

25

24.3













96 hr

25

25.7













0 hr

25

25.7













48 hr

25

25.7













96 hr

25

26.2













0 hr

50

49.7













48 hr

50

51.1













96 hr

50

49.6













0 hr

50

49.8













48 hr

50

51.5













96 hr

50

50.8





PFOS-K

96 hr

Not pan ided

l)\IS()

l.( -MS/MS

0 hr

20

21.8806

Huang et al.

>98%

24 hr

20

22.3311

(2021)

0-44


-------
( homiciil
/ Piiriu

"I'c-sl l)ii r;iliซui

Test Vessel
Miiloriiil

Sol\enl I sed

\le;isu lenient Method

Tiiiu' Measured

Nomiiiid
Cone,
(inii/l.)

Moiisuivd
(one.
(inii/l.)

Reference

PFOS-K

Not
reported

33 d

Not provided

Acetone

Radiometric analysis,

iml reported

0.13

0.12

3M Co.
(2000)

iml reported

0.25

0.28

iml reported

0.5

0.45

iml reported

1

1

not reporied

2

1.9



PFOS

>98%

150 dpost
metamorphosis

Not provided

None

I.C-MS

time weighted
a\ crage (initiation
;ind end)

0.06

0.05

Fort et al.
(2019)

lime weighted
a\ crage (initiation
and end)

0.13

0.13

lime weighted
a\ crage (initiation
and end)

0.25

0.31

time weighted
average (initiation
and end)

0.5

0.59

time weighted
average (initiation
and end)

1

1.05



PFOS

>98%

150 dposl
metamorphosis

Not pro\ idcd

None

I.C-MS

time weighted
average (initiation
and 150 day post
metamorphosis)

0.05

0.05

Fort et al.
(2019)

time weighted
average (initiation
and 150 day post
metamorphosis)

0.12

0.13

time weighted
average (initiation
and 150 day post
metamorphosis)

0.29

0.31

0-45


-------
( homiciil
/ Piiriu

Tcsl Duration

Test Vessel
Material

Sซil\onl I swl

Measurement Method

Tiiiu' Moiisuml

Nomiiiiil
( ttllC.
(iiiu/l.)

Moiisuivd
(one.
(inii/l.)

UcTciyik'o











I ime weighted
a\ crage (initiation
and 150 day post
metamorphosis)

0.62

0.59



lime wcidilcd
average < iniiiaiion
and 150 dav post
metamorphosis)

1.1

1.05



PFOS-K

86.9%

96 hours

Glass

None

Liquid

( hi'onialouraphs landeni
Mass S|xvli'nmcli'\

Ohr

1.82

2.58

Palmer and

Krueger

(2001)

96 hr

1.82

1.42

Ohr

3.07

3.94

96 hr

3.07

1.72

Ohr

5.19

6.62

96 hr

5.19

2.84

Ohr

8.64

10.7

96 hr

8.64

5.09

Ohr

14.4

18.5

96 hr

14.4

10.8

Ohr

24

26.9

96 hr

24

22.3



PFOS-K

95%

21 days

l\i|> pi\>p\ lone

\mie

Liquid
( hiomatography/Tandem
Mass Spectrometry

20 day

0.001

0.0023

MacDonald
et al. (2004)

0.005

0.0144

0.01

0.0217

0.05

0.0949

0.1

0.149

0-46


-------












Nomiiiid

Moiisuivd



( heniieal



Test Vessel







Cone.

(one.



/ Piiriu

"I'c-sl l)ii r;iliซui

Material

Sol\enl I sed

Measurement Method

Time Measured

(inii/l.)

(inii/l.)

Reference











u hours after water
change

0.0001

0.00009













24 hours after
v\ ater change

0.0001

0.00006













48 hours after
water change

0.0001

0.000045













0 hours after water

0.001

0.001













change











Liquid chromatography
coupled to tandem mass

spectrometry wiili
elect rospray ionization
(IIH.C LSI MS MS;

24 hours after
water change

0.001

0.00045



PFOS

4 months

Not reported.

DMSO

48 hours after
water change

0.001

0.00045

Lou et al.

98%

ii hours after water
change

0.1

0.1117

(2013)









24 hours after

0.1

0.0898













water change













48 hours after

0.1

0.0891













water change













u hours after water

1

0.716













change













24 hours after

1

0.661













water change













48 hours after

1

0.632













water change





PFOS

>98%







Liquid

Not reported.

0.0950247

0.0550143

Huang et al.
(2016)

24 hours

I'leMulass

ETOH

Chromatography/Mass

Not reported.

0.950247

0.600156







Spectrometry

Not reported.

9.50247

9.00234













Ohr

50

64.14



PFOS-K

>98%









96 hr

50

75.61

Funkhouser
(2014)

14 days

Polypropv lone

None

LC-MS/MS

Ohr

75

110.2









96 hr

75

135.4











Ohr

100

131.8



0-47


-------
( homiciil
/ Piiriu

Tcsl Dunilion

Test Vessel
Miiloriiil

Snl\en( I sed

Measurement Method

Time Measured

Nomiiiid
(ttnc.

(inii/l.)

Moiisuivd
(one.
(inii/l.)

UcTciyik'o











% hr

100

174.2



u hr

125

172.4

hr

125

223.9

u hr

150

236.3

hr

150

313.3

0 111'

175

170.8

96 hr

175

231.3



PFOS

>98%

40 days

Plastic

None

LC-MS/MS

40 days

10

7.662

Hoover et al.
(2017)

40 days

100

76.34

40 days

1000

877.6



PFOS
Not
reported

116 d

Not provided

None (assumed)

l.( \1S MS

0 hr

0.01

0.0127

Foguth et al.
(2020)

0 hr

0.01

0.0118

0 hr

0.01

0.0136

0 hr

0.01

0.0133



PFOS
98%

10 day

iidpi:

\mie

l.( MS

1 day

0.33

0.0759

McCarthy et
al. (2021)

10 day

0.33

0.0225

1 day

33

17.7

10 day

33

13.5

1 day

100

51.6

10 day

100

47.6

1 day

350

217

10 day

350

216



PFOS
98%

16 day

hdpi:

None

LC/MS

10 day

0.001

0.000376

McCarthy et
al. (2021)

15 day

0.001

0.000477

20 day

0.001

0.000489

10 day

0.005

0.00182

15 day

0.005

0.00208

0-48


-------
( hoiniciil
/ Piiriu

Tesl l)ii r;iliซui

Tesl Vessel
Material

Snl\en( I sed

Measurement Mclhori

Time Measured

Nomiiiiil
(ttnc.

(inii/l.)

Measured
(one.
(inii/l.)

Reference











20 day

0.005

0.00238













In das

0.01

0.00369













15 das

0.01

0.00406













2(i das

0.01

0.00485













In das

0.05

0.0199













15 dav

0.05

0.0241













20 dav

0.05

0.0253













10 day

0.1

0.0443













15 day

0.1

0.0459













20 day

0.1

0.0486



0-49


-------
Table 0-4. Saltwater Nominal and Measured Concentrations for PFOS.













Nomin;il

Measured



( hemie;il

Tesl

Tesl Vessel

Sol\enl





Cone.

Cone.



/ Pin-ill

Dnnilion

Msilerisil

I sod

Measurement Method

Time Measured

(inii/l.)

(inii/l.l

Referenee











II hours

1.1

U.575













48 hours

1.1

0.605













96 hours

1.1

0.391













0 hours

1 1

0.622













48 hours

I.I

0.64













96 hours

1 1

0.58













i i hours

1.8

1.12













48 hours

1.8

1.1













')(> hours

1.8

1.04













(i hours

1.8

1.19













48 hours

1.8

1.09













96 hours

1.8

1.13











High-Perlo n i u) nee
Liquid Chromalographs
with Mass
S|xviromelric
Dclcclion

ii hours

3

1.92











48 hours

3

1.92



PFOS-K

96 hours

Polyethylene

None

')(> hours

3

1.79

Drottar and

90.49%

(i hours

3

1.99

Krueger (2000f)









48 hours

3

1.91











96 hours

3

1.91













0 hours

4.9

3.05













48 hours

4.9

2.96













96 hours

4.9

3.11













0 hours

4.9

2.66













48 hours

4.9

3.35













96 hours

4.9

3.11













0 hours

8.2

5.82













48 hours

8.2

3.58













96 hours

8.2

5.22













0 hours

8.2

5.78













48 hours

8.2

5.85













96 hours

8.2

5.86



O-50


-------












Nomiiiid

Mo.isiiml



( homiciil

losl

1 osl Vessel

Sol\enl





(one.

(one.



/ I'm l it \

Duration

Makrial

I sod

Measurement Method

Time Measured

(inii/l.)

(inii/l.)

UiTcmuv











ii das s

I) 0X<>

I) (Ki'U













14 davs

0.086

0.0606













28 davs

0.086

0.0515













7 davs

0.086

0.0478













21 davs

0.086

0.0554













35 davs

0.086

0.058













0 davs

0.086

0.0578













14 davs

0.086

0.0614













2X davs

0.086

0.0569













" days

0.086

0.0619













2 1 davs

0.086

0.0509













"5 davs

0.086

0.0514













U das s

0.17

0.125













14 das s

0.17

0.124













2X days

0.17

0.122











High-Pcrformancc
I.k|iikI ( hinmalographv
w ilh Mass
S|xvli'nmclric
Detection

" davs

0.17

0.0778











2 1 davs

0.17

0.097



PFOS-K

35 days

Glass

Nunc

"o days

0.17

0.124

Drottar and

90.49%

U days

0.17

0.114

Krueger (2000g)









14 days

0.17

0.127











28 days

0.17

0.128













7 days

0.17

0.125













21 days

0.17

0.112













35 days

0.17

0.119













0 days

0.34

0.289













14 days

0.34

0.276













28 days

0.34

0.262













7 days

0.34

0.231













21 days

0.34

0.227













35 days

0.34

0.278













0 days

0.34

0.286













14 days

0.34

0.253













28 days

0.34

0.271













7 days

0.34

0.197













21 days

0.34

0.212













35 days

0.34

0.251



0-51


-------












Nominal

Me.isured



( hemie;il

Tesl

1 es( Vessel

Sol\enl





(one.

(one.



/ Piiriu

l)ii r;it ion

Miilei'iiil

I sod

Meiisui'enieiil Melhori

l ime Measured

(inii/l.)

(inii/l.)

Ueferenee











ii das s

ii (i'j

(i 5(>2













14 davs

0.69

0.543













28 davs

0.69

0.529













7 davs

0.69

0.581













21 davs

0 69

0.516













35 davs

()(,<>

0.556













0 davs

().<.<>

0.659













14 davs

0.09

0.542













2X davs

0.69

0.544













" days

0.69

0.45













2 1 davs

0.69

0.528













"o davs

0.69

0.583













ii das s

1.4

1.23













14 da\s

1.4

1.35













2X davs

1.4

1.39













" davs

1.4

1.13













2 1 davs

1.4

1.23













'5 davs

1.4

1.26













ii days

1.4

1.32













14 days

1.4

1.27













2S days

1.4

1.39













" days

1.4

1.2













2 1 days

1.4

1.15













35 days

1.4

1.2













11 days

2.7

2.56













14 days

2.7

2.54













7 days

2.7

2.58













0 days

2.7

2.79













14 days

2.7

2.69













7 days

2.7

2.3





PFOS-K
Unknown

7 days

Polypropylene

Mclhanol

1 .iquid chromatograph
connected to a triple-
quadrapole type tandem
mass spectrometer

Average of day 1, 3,5, 7

0.000032

0.000028

Sakurai et al.
(2017)



0-52


-------












Nomiiiid

Me.isnred



( homiciil

losl

1 osl Vessel

Sol\enl





(one.

(one.



/ I'm l it \

l)ii r;it ion

Miiloriiil

I sod

Measurement Method

Time Measured

(inii/l.)

(inii/l.)

UiTcmuv











U hours

1.2

U.331













0 hours

1.2

0.353













48 hours

1.2

0.341













48 hours

1.2

0.429













0 hours

20

0.622













0 hours

: u

0.633













48 hours

: u

0.299













48 hours

2.0

0.313













96 hours

2.0

0.249













')<< hours

2.0

0.257













0 hours

3.3

1.36











High-Performancc
Liquid Chromalomaphs
with Mas>!>
Speclromelnc
Deleclion

ii hours

3.3

1.15











48 hours

3.3

0.924



PFOS-K

96 hr

Polyethylene

None

48 hours

3.3

0.878

Drottar and

90.49%

')(> hours

3.3

1.58

Krueger (2000h)









')(> hours

3.3

1.72











(i hours

5.5

2.42













(i hours

5.5

2.53













48 hours

5.5

2.02













48 hours

5.5

2.24













96 hours

5.5

1.45













96 hours

5.5

0.970













0 hours

9.1

3.44













0 hours

9.1

3.01













48 hours

9.1

3.74













48 hours

9.1

3.57













96 hours

9.1

1.99













96 hours

9.1

2.19















Every 24 hours

0.0001

0.00009



PFOS-K

98%







Liquid Chromatography

Every 24 hours

0.001

0.00095



7 days

Polypropylene

None

- tandem mass

Every 24 hours

0.01

0.0097

Liu et al. (2014a)







spectrometry

Every 24 hours

0.1

0.096













Every 24 hours

1

0.989





0-53


-------












Nomiiiid

Mo.isiiml



( homiciil

losl

Tesl Vessel

Sol\enl





(one.

(one.



/ I'm l it \

l)ii r;it ion

Miiloriiil

I sod

Measurement Method

Time Measured

(inii/l.)

(inii/l.)

UiTcmuv











Samples used lor water
concenlralion dala were
taken every 2 da> s.
Twelve samples per
concentration were taken

(i ( ii ii 11

(i (Kin 12











Liquid Chromatography
- tandem mass
spectrometry

0.001

0.0011

Liu et al.
(2014b,c); Lui
and Gin (2018)

PFOS-K

7 days

Polypropylene

None

0.01

0.0096

98%

0.1

0.106









1

0.968











III

10.156















2 davs

1

0.916











Liquid Chromatography
- tandem mass
spectrometry

2 days

4

3.053



PFOS

8 days

Glass

DMSO

2 days

16

11.76

Fang et al. (2012)

98%

S davs

1

0.916









S das s

4

3.664













8 das s

16

12.67















ii day

20

16.4













24 hr (old)

20

15.2













24 hr inew)

20

16.2













48 h (old)

20

13.4













48 hr (new)

20

17.4













72 hr (old)

20

15.2













72 hr (new)

20

17













96 hr

20

13.7



PFOS-K

86.9%









0 day

20

15.7

Palmer et al.
(2002)

96 hr

l\>l\eth\ lone

Methanol

HPI.C'MS

24 hr (old)

20

15.3









24 hr (new)

20

15.7











48 h (old)

20

11.3













48 hr (new)

20

18.2













72 hr (old)

20

18













72 hr (new)

20

16.6













96 hr

20

11.1













0 day

20

16













24 hr (old)

20

15.2













24 hr (new)

20

15



0-54


-------
( hcmic;il
/ Piiriu

Tesl
l)ii r;it ion

1 cs( Vessel
Miilei'iiil

Sol\cnl
I sod

Measurement Method

l ime Measured

Nomiiiiil
(one.
(inii/l.)

Measured
(one.
(inii/l.)

Reference











4S h (old)

"Ml

12."



48 hr(new )

20

18.6

72 hr (old)

20

16.2

72 hr(new )

20

17.4

96 hr

2d

13.4



PFOS-K

98%

7 d

Polypropylene

None

LC-MS

0 dav

0.0001

0.00012

Liuetal. (2013)

ii day

0.001

0.0011

ii dav

0.01

0.0096

0 dav

0.1

0.106

ii day

1

0.968

0-55


-------
Table 0-5. Freshwater PFOS Toxicity Studies with Systematic Discrepancies between Nominal and Measured Concentrations

that were > 2C

1%.

Sliid\

losl
Diinilioii

lost Method

S;ini|)k- Collection
l"lV(|IK'IIO

PI-OS
An;il> tic;il
Mi'liiori1

Notes on Discivpiincics

Foguth et al.
(2020)

Chronic
(116-d)

Static

Not reported.
Assumed at the
beginning of the
mesocosm test after
spiking with PFOS

LC-MS/MS

All measured test concentrations were higher than the nominal concentrations,
by a factor of approximately 1.3, with three of the measured values in the four
replicate mesocosms dosed with PFOS falling outside of the 20% exceedance
threshold. The systematic discrepancies in the mesocosm test indicate an
apparent dosing or perhaps test design issue related to use of the leaf litter in
the mesocosms.

Funkhouser
(2014)

Chronic
(14-d)

Renewal
(every 3-4 d)

Before and after each
renewal

LC-MS/MS

All measured concentrations were higher than the nominal concentrations with
all treatments falling outside of the 20% exceedance threshold. These
systematic discrepancies indicate an apparent dosing issue. Particularly since
these systematic discrepancies were also consistent with other tests reported in
the thesis (that were not included in the meta-analysis), all of which were
higher than the nominal concentrations and outside the 20% exceedance
threshold.

Hoover et al.
(2017)

Chronic
(40-d)

Renewal
(every 4 d)

End of test duration
(at 40 days)

I.C-MS/MS

All measured test concentrations were lower than the nominal concentrations
with two of the three treatments falling outside of the 20% exceedance
threshold. These systematic discrepancies indicate an apparent dosing issue.

Huang et al.
(2016)

Acute
(24-hr)

Static

Not reported.

I.C-MS

All of the measured PFOS concentrations were lower than the nominal
concentrations with two of the three measured concentrations falling outside of
the 20%) exceedance threshold. These systematic discrepancies indicate an
apparent dosing issue

Lou et al.
(2013)

Chronic
(4-month)

Renewal
(every 48 hr)

Bel ore and after each
renew al

HPLC-ESI-
MS/MS

Nine of the twelve measured test concentrations were lower than the nominal
concentrations with seven of the twelve measured concentrations falling
outside of the 20% exceedance threshold. These systematic discrepancies
indicate an apparent dosing issue.

MacDonald et
al. (2004)

Chronic
(20-d)

Renewal
(every 48 hr)

Lnd of lest duration
(20 days)

LC-MS/MS

All measured test concentrations were higher than the nominal concentrations,
by a factor of 2, with all of the five treatments falling outside of the 20%
exceedance threshold. However, similar systematic discrepancies were not
observed in the 10-day test that was also reported in this paper. In the 10-day
test the measured concentrations were not systematically higher or lower than
the nominal concentrations and only two of the five treatments were slightly
outside the 20% exceedance threshold (at 120.5 and 122.15%). The systematic
discrepancies in the 20-day test indicate an apparent dosing issue.

McCarthy et al.
(2021)

Chronic
(10-d)

Renewal
(every 48-72 hr)

Al days 1 and 10

LC-MS

All measured test concentrations were lower than the nominal concentrations,
by over a factor of 2, with all of the four treatments falling outside of the 20%
exceedance threshold. The systematic discrepancies in the 10-day test indicate
an apparent dosing or perhaps test design issue related to use of natural field-
collected sediment in exposure chambers.

0-56


-------
Siiul\

lis!
l)ur;ilion

lost Method

S;ini|)k- ( olk-clion

l'lV(|IK'IIC\

PI-OS
An;il> tic;il
Mi'liiod1

Noll's on l)isi'i'i'|)iiiii'ii's

McCarthy el al.
(2021)

Chronic
(16-d)

Renew al

(every 24 hr)

At days 10, 15 and 20

LC-MS

All measured test concentrations were lower than the nominal concentrations,
by over a factor of 2, with all of the five treatments falling outside of the 20%
exceedance threshold. Similar systematic discrepancies were observed in the
10-day test that was also reported in this paper. The systematic discrepancies in
the 16-day test indicate an apparent dosing or perhaps test design issue related
to use of natural field-collected sediment in exposure chambers.

Palmer and
Krueger (2001)

Acute
(96-hr)

Renewal
(every 24 hr)

Before and after each
renewal

LC-MS/MS

Three, independent assays of this test were conducted instead of simultaneous
replicates. All measured test concentrations from the first assay were
systematically higher than the nominal concentrations after renewal and
systematically lower before the next renewal and all were outside the 20%
exceedance threshold. These same systematic differences were not observed in
the other two assays and all treatments in second and third assays were within
the 20% exceedance threshold. The systematic discrepancies in the first assay
indicate an apparent dosing issue.

Ankley et al.
(2004)

Chronic
(110-d)

Flow-through

Water samples were
collected and
averaged over the
exposure period

HPLC-ESI-
MS/MS

Three of the six treatment groups had measured concentrations higher than
nominal and were outside the 20% exceedance threshold. Two of the three
remaining treatment groups were also higher than the nominal concentrations
and were not outside the 20% exceedance threshold. One of the treatment
groups was lower than the nominal concentration and was also not outside the
20% exceedance threshold. Therefore, the observed exceedances were not
considered to be systematic across the treatment groups and the three treatment
groups outside the 20%) exceedance threshold were fairly close to the threshold
(at 133, 130, and 125%).

Drottar and

Krueger

(2000b)

Chronic
(21-d)

Renewal
(every 3 days
per week)

On days 0,2,11,14,
18, and 21

RP-HPLC

Two of the sixty-three measured test concentrations were outside the 20%)
exceedance threshold. However, these differences were not considered to be
systematic across the treatment groups and the two differences were fairly close
to the 20%o exceedance threshold (at 127 and 123%).

Drottar and
Krueger (2000c)

Chronic (47-

d)

Flow-through

Water samples were
collected on days 0.
4,7.14.21,28,35.
42.and 47

HPLC/MS

Two of the eighty-three measured test concentrations were outside the 20%
exceedance threshold. However, these differences were not considered to be
systematic across the treatment groups and the two differences were fairly close
to the 20%o exceedance threshold (at 128 and 124%).

Drottar and

Krueger

(2000d)

Acute
(96-hr)

Renewal
(even 48 hr)

Before and after each
renew al

HPLC/MS

Three of the thirty measured test concentrations were outside the 20%
exceedance threshold. However, these differences were not considered to be
systematic across the treatment groups and the three exceedances were fairly
close to the 20%o exceedance threshold (at 73, 75, and 140%).

0-57


-------
Siuch

Tesl
Diinilion

lost Method

S;ini|)k- ( ollccl ion

l'lV(|IK'IIC\

PI-OS
An;il> tic;il
Mc-UkkI1

Soles on Discivpiincios

iiazellon et al.
(2012)

Chronic
(36-d)

Renew al

(every 24 hr)

on days 10 and 11

LC-MS

All measured test concentrations were outside the 20% exceedance threshold.
However, the measured concentrations were not systematically higher or lower
than the nominal concentrations between the two treatment groups.

Specifically, the measured concentrations from the low treatment group was
higher (452% higher) than the nominal concentration and the high treatment
was lower (69.5%) than the nominal concentration. While there appears to be a
dosing issue for the chronic test in Hazelton et al. 2012, the discrepancies are
not systematic (both lower or higher than the nominal concentrations) and there
is no apparent justification for the observed discrepancies.

Keiter et al.
(2016)

Chronic
(330-d)

Flow-through

On days 30 and 60

LC-MS/MS

Two of the three treatment groups had measured concentrations higher than
nominal and was outside the 20% exceedance threshold. The remaining
treatment group was lower than the nominal concentrations and was not outside
the 20%) exceedance threshold. While there appears to be a dosing issue in
Keiter et al. 2016. the discrepancies are not systematic (all lower or higher than
the nominal concentrations) and there was no apparent justification for the
observed discrepancies.

Sanderson et al.
(2002)

Chronic
(35-d)

Static

On days 1, 8, and ^

1IPI.C-1C

Six out of the seven treatment groups had measured concentrations higher than
nominal and two of the seven were outside the 20% exceedance threshold.
Therefore, while the observed exceedances could be considered to be
systematic across the treatment groups, since only two treatment groups were
outside the 20%o exceedance threshold (and were fairly close to the threshold at
123 and 133%).

1PFOS Analytic Methods: Liquid Chromatography/Tandem Mass Spectrometry (LC-MS/MS), Liquid Chromatography / Mass Spectrometry (LC-MS), Liquid Chromatography
Coupled to Tandem Mass Spectrometry with Electro spray Ionization (I IPLC-ES1-MS/MS). Reverse Phase High Performance Liquid Chromatography (RP-HPLC), and High-
Performance Liquid Chromatography with Mass Spectrometry Detection (I IPI.C/MS)

0-58


-------
Table 0-6. Saltwater PFOS Toxicity Studies with Systematic Discrepancies between Nominal and Measured Concentrations
that were > 20%.

Studies are listed alphabetically.				

S(iul\

losl
Diinilion

Tcsl Mel hod

Siimplo ( olk-clion

llV(|IK'IK'\

PI-OS
Mel hod1

\oies on Diserepiineies

Drollar and

Krueger

(2000f)

Acute

(96-hr)

Sialic

Al 0, 48, and 96

HPLC/MS

All measured lesl concentrations were onlside ihe 20% exceedance
threshold and were syslematic across the treatment groups.

Drottar and

Krueger

(2000g)

Chronic
(35 days)

Flow-through

AtO, 7,14, 21, 28 and 35 days

HPLC/MS

The majority of measured lest concentrations at low concentration (i.e., <1
mg/L) were slightly outside the 20% exceedance threshold although the
level of quantitation was adequate for the range of concentrations tested.
Possible analytic quantitation issue.

Drottar and

Krueger

(2000h)

Acute
(96-hr)

Static

At 0, 48, and 96 hours

HPLC/MS

All measured test concentrations were outside the 20% exceedance
threshold and were systematic across the treatment groups.

Palmer et al.
(2002)

Acute
(96-hr)

Renewal
(every 24 hr)

At 0, 24, 48, 72, and 96 hours

HPLC/MS

I Ialf of the measured test concentrations were outside the 20% exceedance
threshold across the treatment groups.

1 PFOS Analytic Methods: Liquid Chromatography/Tandem Mass Spectrometry (LC-MS/MS), Liquid Chromatography / Mass Spectrometry (LC-MS), Liquid Chromatography
Coupled to Tandem Mass Spectrometry with Electrospray Ionization (HPLC-ESI-MS/MS). Reverse Phase High Performance Liquid Chromatography (RP-HPLC), and High-
Performance Liquid Chromatography with Mass Spectrometric Detection (HPLC/MS

0-59


-------
Appendix P Bioaccumulation Factors (BAFs) Used to Calculate PFOS Tissue Values

P.l Summary Table of PFOS BAFs used to calculate tissue criteria and supplemental fish tissue values

Com 111011 N:i 1110

Scientific \:imc

Tissue

Log
IJAI

lood

4()4S

1 1 l(.~

liiuli

1 ~ Sues mi si\ m;i|oi' ii\ ei's. Korea

l.am el al (2o|4i

mandarin

Sinipcrca scherzeri

Blood

4.So.

"3oi:

high

1 ~ Siles in s>i\ major n\ en>. Korea

Lam el al. (2014)

common carp

Cyprinus carpio

Blood

3.860

7244

high

Xiaoqing River. China

Pan et al. (2017)

Crucian carp

Carassius carassius

Blood

3.984

9638

hiuli

Tangxum Lake. China

Shi et al. (2015)

Crucian carp

Carassius carassius

Blood

4.301

19999

limli

Xiaoqing River, China

Shi et al. (2015)

Crucian carp

Carassius carassius

Blood

4.352

22484

hiuli

Gaobeidian Lake, China

Shi et al. (2020)

Crucian carp

Carassius carassius

Blood

4.328

21275

h.gh

Yubci River, China

Shi et al. (2020)

Crucian carp

Carassius carassius

Blood

4.904

80168

high

Beijing Airport, China

Wanget al. (2016)

European perch

Perca fluviatilis

Blood

4.763

58	

medium

Lake Halmsjon, near Stockholm,
Sweden

Wang et al. (2016)

conger eel

Conger myriaster

Blood

^ 544

3500

medium

Tokyo Bay

Taniyasu et al.
(2003)

Japanese stingfish

Sebastiscus marmoratus

Blood

^ "i:

5154

medium

Tokyo Bay

Taniyasu et al.
(2003)

rockfish

Sebastes inermis

1 Jlood

^ 'J "4

942"

medium

Tokyo Bay

Taniyasu et al.
(2003)

black seabream

Acanthopagrus schlege/i

1>lood

4 150

14 nx

medium

Osaka Bay

Taniyasu et al.
(2003)

Japanese scad

Trachurus japonicus

1 Jlood

4 I5()

14138

medium

Osaka Bay

Taniyasu et al.
(2003)

White croaker

Argyrosomus argentatus

1 >lood

4.2'JI

19540

medium

Osaka Bay

Taniyasu et al.
(2003)

bluegill

Lepomis macrochirus

Blood

4.043

11053

medium

Lake Biwa

Taniyasu et al.
(2003)

largemouth bass

Micropterus salmoides

Blood

5.230

169737

medium

Lake Biwa

Taniyasu et al.
(2003)

carp

Cyprinus carpio

Blood

4.925

84211

medium

Lake Biwa

Taniyasu et al.
(2003)

Lefteye flounder

Paralichthys olivaceus

Blood

3.750

5625

medium

Ariake Bay

Taniyasu et al.
(2003)

P-l


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference



European chub

Leuciscus cephalus

Gonad

4.000

10000

high

()iuc Kin ci\ near Paris, France

Labadie and
Chevreuil (2011)

Crucian carp

Carassius carassius

Gonad

3.770

5888

high

l auuMiui Lake. China

Shi etal. (2015)

Crucian carp

Carassius carassius

Gonad

4.060

11482

high

Xiaoqing River. China

Shi etal. (2015)

Crucian carp

Carassius carassius

Gonad

3.904

8012

high

Gaobcidian Lake. China

Shi et al. (2020)

Crucian carp

Carassius carassius

Gonad

3.903

7990

high

Yubci River. China

Shi et al. (2020)

Crucian carp

Carassius carassius

Gonad

4.409

25645

hiph

Beijing Airport. China

Wanget al. (2016)

European perch

Perca fluviatilis

Gonad

4.204

16000

medium

Lake Halmsjon. near Stockholm,
Sweden

Ahrens et al. (2015)

chub

Leuciscus cephalus

Gonad

3.347

::::

medium

Rotcr Main, Upper Franconia,
(icrmanv

Becker et al. (2010)



common shiner

Notropis cornutus

Liver

4.100

125X<>

liiuli

Spium Elobicokc Creek, Toronto,
Canada

Awad et al. (2011)

European chub

Leuciscus cephalus

Liver

4 300

19953

liidi

Orge River, near Paris, France

Labadie and
Chevreuil (2011)

goldfish

Carassius auratus

Liver

^ (ฆ(ฆ()

4572

high

17 Sites in six major rivers, Korea

Lam et al. (2014)

mandarin

Siniperca scherzeri

Liver

4 W

24718

high

17 Sites in six major rivers, Korea

Lam et al. (2014)

common carp

Cyprinus carpio

Liver

i (Oil

4467

high

Xiaoqing River, China

Pan etal. (2017)

Bream

Parabramis pekinensis

Li\ or

i 5oo

3162

high

Pearl River Delta, China

Pan etal. (2014)

goldfish

Carassius auratus

Linci"

4 300

19953

high

Pearl River Delta, China

Pan etal. (2014)

Common carp

Cyprinus carpio

I.INCI"

4 400

25119

high

Pearl River Delta, China

Pan etal. (2014)

Chub

Hypophthalmichthys
molitrix

1 ,i\ cr

V'joo

7943

high

Pearl River Delta, China

Pan etal. (2014)

Tilapia

Tilapia aurea

Li\ ci"

3.500

3162

high

Pearl River Delta, China

Pan etal. (2014)

Snakehead

Ophicepha/us argus

Lin ci"

4.200

15849

high

Pearl River Delta, China

Pan etal. (2014)

Leather catfish

Clarias fuscus

Lin ci"

3.700

5012

high

Pearl River Delta, China

Pan etal. (2014)

grass carp

Ctenopharyngodon
idellus

Lin ci"

4.600

39811

high

Pearl River Delta, China

Pan etal. (2014)

Crucian carp

Carassius carassius

Liver

3.646

4426

high

Tangxum Lake, China

Shi etal. (2015)

Crucian carp

Carassius carassius

Liver

3.965

9226

high

Xiaoqing River, China

Shi etal. (2015)

Crucian carp

Carassius carassius

Liver

4.048

11180

high

Gaobeidian Lake, China

Shi et al. (2020)

Crucian carp

Carassius carassius

Liver

4.031

10735

high

Yubei River, China

Shi et al. (2020)

P-2


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

Loป
IJAI

(l./kป-
\\\\)

liAl

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

Sil\ or porch

Bidyanus bidyanus

l.i\er

4 415

		

hi—h

ShoalhaNen reuinu. \usiralia

1 creclkn s el al
(:u|<>)

Crucian carp

Carassius carassius

Liver

4.923

83753

high

Bcijiim \irpnri, China

Wanget al. (2016)

European perch

Perca fluviatilis

Liver

4.591

39000

medium

Lake 1 laluis|nu. near Stockholm,
Sweden

Ahrens et al. (2015)

chub

Leuciscus cephalus

Liver

3.659

4556

medium

Rolcr Main. Upper Franconia,
Germany

Becker et al. (2010)

Mozambique
tilapia

Oreochromis
mossambicus

Liver

2.640

436.8

medium

aMalikulu, N2 Bridge

Fauconier et al.
(2020)

Cape stumpnose

Rhabdosargus holubi

Liver

2.046

111.2

medium

a\ [atikulu, N2 Bridge

Fauconier et al.
(2020)

tilapia

tilapia

Liver

3.708

5 1 (18

medium

kev River, Taiwan

Lin et al. (2014)

tilapia

tilapia

Liver

3.621

4ISI

medium

kev River, Taiwan

Lin et al. (2014)

tilapia

tilapia

Liver

3.533

3409

medium

Kcv River, Taiwan

Lin et al. (2014)

Mud carp

Cirrhinus molitorella

Liver

4.400

25119

medium

Pearl River Delta, China

Panetal. (2014)

common seabass

Lateolabrax japonicus

Liver

^ 514

3269

medium

Tokyo Bay

Taniyasu et al.
(2003)

flatfish

Pleuronectidae

l.i\er

; S i5

6846

medium

Tokyo Bay

Taniyasu et al.
(2003)

Japanese stingfish

Sebastiscus marmoratus

l.i\er

V<4<>

44^

medium

Tokyo Bay

Taniyasu et al.
(2003)

rockfish

Sebastes inermis

l.i\er

^ V) |

:4(.:

medium

Tokyo Bay

Taniyasu et al.
(2003)

common seabass

Lateolabrax japonicus

l.i\er

: (.(.ฆ?

459.8

medium

Osaka Bay

Taniyasu et al.
(2003)

Japanese scad

Trachums japonicus

l.i\er

3.015

1034

medium

Osaka Bay

Taniyasu et al.
(2003)

White croaker

Argyrosomus argcnlalus

Lin or

3.207

1609

medium

Osaka Bay

Taniyasu et al.
(2003)

bluegill

Lepomis macrochirus

1 .i\ or

4.870

74211

medium

Lake Biwa

Taniyasu et al.
(2003)

largemouth bass

Micropterus salmoides

l.i\er

4.789

61579

medium

Lake Biwa

Taniyasu et al.
(2003)

carp

Cyprinus carpio

Liver

3.022

1053

medium

Lake Biwa

Taniyasu et al.
(2003)

P-3


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

l.cl'lc>e llnuudci'

I'araliclilhys olivaceus

I.INCI'

4

:^5x

medium

\riake l!a>

T;iiii\;isii cl al

sea mullet

Mugil cephalus

Liver

3.699

5000

medium

Sydney Harbour. Australia

Thompson et al.
(2011)



European perch

Perca fluviatilis

Muscle

3.531

3400

high

Lake Halmsjon. near Stockholm,
Sweden

Ahrens et al. (2015)

minnow

Hemiculter leucisculus

Muscle

3.785

6092

hi sih

Tailui Lake. China

Fang et al. (2014)

silver carp

Hypophthalmichthys
molitrix

Muscle

3.246

1761

hiuli

Taihu Lake, China

Fang et al. (2014)

white bait

Reganisalanx
brachyrostralis

Muscle

3.452

2835

high

1 ;nhu Lake, China

Fang et al. (2014)

Japanese crucian
carp

Carassius cuvieri

Muscle

4.193

155W

high

1 nihil Lake, China

Fang et al. (2014)

Lake Saury

Coilia mystus

Muscle

3.963

9190

lnuli

Taihu Lake, China

Fang et al. (2014)

common carp

Cyprinus carpio

Muscle

3.882

7623

hiuli

1 aihu Lake, China

Fang et al. (2014)

Mongolian culter

Culter mongolicus

Muscle

4 P9

15088

hiuli

Taihu Lake, China

Fang et al. (2014)

mudfish

Misgurnus
anguillicaudatus

Muscle

4.HU

10810

high

Taihu Lake, China

Fang et al. (2014)

Chinese bitterling

Rhodeus sinensis
Gunther

Muscle

3.809

6444

high

Taihu Lake, China

Fang et al. (2014)

Goby

Ctenogobius giurinus

Muscle

3.788

6144

high

Taihu Lake, China

Fang et al. (2014)

eel

Anguilla anguilla

Muscle

3.510

3236

high

Netherlands

Kwadijk et al.
(2010)

European chub

Leuciscus cephalus

Muscle

3.400

2512

high

Orge River, near Paris, France

Labadie and
Chevreuil (2011)

Juvenile char

Salvelinus alpinus

Muscle

3.274

1878

high

Meretta Lake, Canadian High Arctic

Lescord et al.
(2015)

Juvenile char

Salvelinus alpinus

Muscle

3.016

1038

high

Resolute Lake, Canadian High Arctic

Lescord et al.
(2015)

Juvenile char

Salvelinus alpinus

Muscle

4.033

10800

high

Char Lake, Canadian High Arctic

Lescord et al.
(2015)

Adult char

Salvelinus alpinus

Muscle

2.767

585.4

high

Meretta Lake, Canadian High Arctic

Lescord et al.
(2015)

P-4


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

\dull cli;ii

Sa/ve/inus a/pinus

Muscle

. (O'

4500

I"-'1

Resolute Like. ( ;iii;kIi;iii 1 IiuIi \rclic

l.escnid el ;il
(2o151

Adult char

Salvelinus alpinus

Muscle

4.602

40000

high

( 'liar Like. Canadian High Arctic

Lescord et al.
(2015)

Anchovy

Engraulis encrasicolus

Muscle

3.246

1761

high

Girondc csluarv. SW France

Munoz et al. (2017)

Mullet

Liza ramada

Muscle

3.089

1226

high

Girondc csluarv. SW France

Munoz et al. (2017)

Meagre

Argyrosomus regius

Muscle

3.397

2496

high

Girondc csluarv. SW France

Munoz et al. (2017)

Common seabass

Dicentrarchus labrax

Muscle

3.513

3257

hieh

Girondc csluarv. SW France

Munoz et al. (2017)

Spotted seabass

Dicentrarchus punctatus

Muscle

3.404

2535

limli

Girondc csluarv. SW France

Munoz et al. (2017)

Spotted seabass

Dicentrarchus punctatus

Muscle

3.766

5830

limli

Gironde estuary, SW France

Munoz et al. (2017)

common carp

Cyprinus carpio

Muscle

2.730

537.0

high

Xiaoqing River, China

Panetal. (2014)

Bream

Parabramis pekinensis

Muscle

2.600

398.1

high

I'earl RiverDelta, China

Panetal. (2014)

goldfish

Carassius auratus

Muscle

3.200

1585

high

Pearl RiverDelta, China

Panetal. (2014)

Common carp

Cyprinus carpio

Muscle

3.200

1585

liiuh

Pearl RiverDelta, China

Panetal. (2014)

Chub

Hypophthalmichthys
molitrix

Muscle

2.800

631.0

high

Pearl RiverDelta, China

Panetal. (2014)

Tilapia

Tilapia aurea

Muscle

2 400

251.2

liidi

Pearl River Delta, China

Panetal. (2014)

Snakehead

Ophicephalus argus

Muscle

2 (.no

398.1

high

Pearl River Delta, China

Panetal. (2014)

Leather catfish

Clarias fuscus

Muscle

: 4(10

251.2

high

Pearl River Delta, China

Panetal. (2014)

grass carp

Ctenopharyngodon
idellus

Muscle

^ 400

2512

high

Pearl River Delta, China

Panetal. (2014)

Crucian carp

Carassius carassius

Muscle

2.8 /0

741.3

high

Tangxum Lake, China

Shi et al. (2015)

Crucian carp

Carassius carassius

Muscle

3.195

1567

high

Xiaoqing River, China

Shi et al. (2015)

Crucian carp

Carassius carassius

Muscle

} 053

1130

high

Gaobeidian Lake, China

Shi et al. (2020)

Crucian carp

Carassius carassius

Muscle

3.0(.7

1167

high

Yubei River, China

Shi et al. (2020)

Silver perch

Bidyanus bidvanus

Muscle

3.778

6000

high

Shoalhaven region, Australia

Terechovs et al.
(2019)

Crucian carp

Carassius carassius

Muscle

4.701

50234

high

Beijing Airport, China

Wang et al. (2016)

chub

Leuciscus cephalus

Muscle

2.683

481.5

medium

Roter Main, Upper Franconia,
Germany

Becker et al. (2010)

goby

Gobio gobio

Muscle

3.472

2963

medium

Roter Main, Upper Franconia,
Germany

Becker et al. (2010)

Black Crappie

Pomoxis nigromaculatus

Muscle

3.200

1585

medium

Lake Niapenco, Ontario, Canada.

Bhavsar et al.
(2016)

P-5


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

lii'ou ii 1 >iiIIIic;k.I

. Inwiiirus nehubsus

Muscle

: <><><>

~'U ^

medium

Lake \iapeucn. Ouiai'in. Canada

l!lia\ sarel al
i:o|(.i

Channel Catfish

Ictalurus punctatus

Muscle

3.500

3162

medium

Lake Niapcnco. Ontario, Canada.

Bhavsar et al.
(2016)

Common Carp

Cyprinus carpio

Muscle

3.900

7943

medium

Lake Niapcnco. Ontario. Canada.

Bhavsar et al.
(2016)

Largemouth Bass

Micropterus salmoides

Muscle

3.700

5012

medium

Lake Niapcnco. Ontario. Canada.

Bhavsar et al.
(2016)

Northern Pike

Esox lucius

Muscle

3.000

1000

medium

Lake Niapenco, Ontario, Canada.

Bhavsar et al.
(2016)

Pumpkinseed

Lepomis gibbosus

Muscle

2.800

6"1 0

medium

1 .ake Niapenco, Ontario, Canada.

Bhavsar et al.
(2016)

Smallmouth Bass

Micropterus dolomieu

Muscle

3.800

^ 1 ()

medium

Lake Niapenco, Ontario, Canada.

Bhavsar et al.
(2016)

White Crappie

Pomoxis annularis

Muscle

3.000

Kiiiii

medium

Lake Viapenco, Ontario, Canada.

Bhavsar et al.
(2016)

Yellow Perch

Perca flavescens

Muscle

2 win

794.3

medium

Lake Niapenco, Ontario, Canada.

Bhavsar et al.
(2016)

Mozambique
tilapia

Oreochromis
mossambicus

Muscle

1 241

17.44

medium

aMatikulu, N2 Bridge

Fauconier et al.
(2020)

Cape stumpnose

Rhabdosargus holubi

Muscle

(i 'Muo

S "IS

medium

aMatikulu, N2 Bridge

Fauconier et al.
(2020)

Eel

Anguilla anguilla

Muscle

vIK.U

1 I4S

medium

Schiphol Amsterdam Airport

Kwadijk et al.
(2014)

Eel

Anguilla anguilla

Muscle

2.3 "u

234.4

medium

Schiphol Amsterdam Airport

Kwadijk et al.
(2014)

tilapia

tilapia

Muscle

2.389

245.0

medium

Key River, Taiwan

Lin et al. (2014)

tilapia

tilapia

Muscle

2.509

323.0

medium

Key River, Taiwan

Lin et al. (2014)

tilapia

tilapia

Muscle

2.328

213.0

medium

Key River, Taiwan

Lin et al. (2014)

Sprat

Sprattus sprattus

Muscle

2.908

808.7

medium

Gironde estuary, SW France

Munoz et al. (2017)

Mud carp

Cirrhinus molitorella

Muscle

3.400

2512

medium

Pearl River Delta, China

Panetal. (2014)

sea mullet

Mugil cephalus

Muscle

2.196

157.1

medium

Sydney Harbour, Australia

Thompson et al.
(2011)



P-6


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

common shiner

X<> tropis cor mi I us

WIS

^ M)

1W5

I"-'1

Spi iim l!ii
-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

Chameleon uoln

Tridenliger
Irigonocepha/us

WIS

. 'JOS

S()S(.

medium

(iuII I'ark - Xiamen Sea. China

Dai and /.lieim
< 2o | 'J i

Lake Trout

Salvelinus namaycush

WB

4.300

19953

medium

Lake Superior

Furdui et al. (2007)

Lake Trout

Salvelinus namaycush

WB

4.200

15849

medium

Lake Huron

Furdui et al. (2007)

Lake Trout

Salvelinus namaycush

WB

4.400

25119

medium

Lake Eric

Furdui et al. (2007)

Lake Trout

Salvelinus namaycush

WB

3.900

7943

medium

Lake Ontario

Furdui et al. (2007)

Lake Trout

Salvelinus namaycush

WB

3.800

6310

medium

Lake Michigan

Furdui et al. (2007)

herring

Clupea harengus
membras

WB

4.320

20893

medium

Baltic Sea

Gebbink et al.
(2016)

spat

Sprattus sprattus

WB

4.350

22387

medium

Baltic Sea

Gebbink et al.
(2016)

alewife

Alosa pseudoharengus

WB

4.380

:4(i(io

medium

1 .akc Ontario

Houde et al. (2008)

rainbow smelt

Osmerus mordax

WB

4.653

45(100

medium

1 .akc Ontario

Houde et al. (2008)

slimy sculpin

Cottus cognatus

WB

5.369

234000

medium

Lake Ontario

Houde et al. (2008)

lake trout

Salvelinus namaycush

WB

4.531

34000

medium

Lake Ontario

Houde et al. (2008)

Sea Bass

Lateolabrax

WB

2.585

384.6

medium

Oinuta River mouth and estuary,
Japan

Kobayashi et al
(2018)

Grey mullet

Mugil cephalus

WIS

3.016

1038

medium

Omuta River mouth and estuary,
Japan

Kobayashi et al
(2018)

Yellowfin goby

Acanthogobius
flavimanus

WIJ

: "(.i

5~<> '>

medium

Omuta River mouth and estuary,
Japan

Kobayashi et al
(2018)

Perch

Esox lucius

WIS

^

:u4

medium

Schiphol Amsterdam Airport

Kwadijk et al.
(2014)

Perch

Esox lucius

WIS

. XdO

b31u

medium

Schiphol Amsterdam Airport

Kwadijk et al.
(2014)

Small snakehead

Channa asiatica

WIS

3.lus

1283

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

Flag-tailed glass
perchlet

Ambassis miops

WB

2.889

774.6

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

goby

Pomatoschislus

wi:

3.380

2400

medium

Gironde estuary, SW France

Munoz et al. (2017)

Chinese icefish

Neosalanx tangkahkeii
taihuensis

WIS

3.355

2267

medium

Lake Chaohu, China

Panetal. (2019)

Common carp

Cyprinus carpio

WIS

3.000

1000

medium

Xerta, Ebro Delta, Spain

Pignotti et al. (2017)

Mullet

Liza

WB

0.680

4.786

medium

Xerta, Ebro Delta, Spain

Pignotti et al. (2017)

Roach

Rutilus rutilus

WB

2.300

199.5

medium

Xerta, Ebro Delta, Spain

Pignotti et al. (2017)

P-8


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

Loป
liAl

\\\\)

IJAI

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

kndd

Scardinius
ervlhrophlahmis

WIS

I <>oo

~K) 4'

mcd in in

\eila. 1 !hix> Delia. Spam

huikUli el al <2o|"i

European catfish

Silurus glanis

WB

2.000

100.0

medium

Xeria. Lbro Delia. Spain

Pignotti et al. (2017)

Ebro chub

Squalius laietanus

WB

2.000

100.0

medium

Xerla. Ebro Delia. Spain

Pignotti et al. (2017)

Bleak

Alburnus alburnus

WB

2.400

251.2

medium

Xcrla. Ebro Delta. Spain

Pignotti et al. (2017)

grass goby

Zosterisessor
ophiocephalus

WB

2.936

863.7

medium

AC Site. Orbclcll lagoon. Italy

Renzi et al. (2013)

grass goby

Zosterisessor
ophiocephalus

WB

2.752

565.0

medium

NC Site. Orbclcll lagoon. Italy

Renzi et al. (2013)

grass goby

Zosterisessor
ophiocephalus

WB

2.521

332.0

medium

1'(' Site, Orbetell lagoon, Italy

Renzi et al. (2013)



Manila clam

Ruditapes philippinarum

Invertb

3.601

39') 1

high

liaozhou Bay, China

Cui et al. (2019)

zooplankton

zooplankton

Invert

2.580

380 3

high

Tailui Lake, China

Fang et al. (2014)

zooplankton

zooplankton

Invert

2.813

650.0

high

Lake Ontario

Houde et al. (2008)

Microzooplankton

Microzooplankton

Invert

3.480

3017

liigh

17 Sites in six major rivers, Korea

Lam et al. (2014)

Mesozooplankton

Mesozooplankton

Invert

3.538

3450

high

17 Sites in six major rivers, Korea

Lam et al. (2014)

zooplankton

zooplankton

lii\ crl

^ 077

1195

high

Meretta Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankton

lii\ eil

1 1 1

: .OS

high

Resolute Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankton

lii\ eil

. .SO

24oo

high

Char Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankton

lii\ eil

. .SS

2444

high

Small Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankimi

ln\ eil

4.5o4

36667

high

North Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankton

Tn\c il

5.000

100000

high

9-Mile Lake, Canadian High Arctic

Lescord et al.
(2015)

zooplankton

zooplankton

ln\ eil

2.470

295.1

medium

Baltic Sea

Gebbink et al.
(2016)

zooplankton

zooplankton

ln\ crt

2.425

266.0

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

amphipod

Gammarus, Hyalella

1 nvert

3.779

6015

high

Welland River, Hamilton, Ontario,
Canada

De Solla et al.
(2012)

freshwater mussel

Unionidae

Invert

2.758

572.2

high

Taihu Lake, China

Fang et al. (2014)

P-9


-------
Com 111011 N:i 1110

Scientific \:imc

Tissue

1,0"
liAl
(l./kป-
\\\\)

liAl

(l./lvli-
\\\\)

k;inkin<>

Locution

Reference

pearl mussel

1 monidae

lii\ eil

i 005

10 | |

liiuli

Tailm Lake. China

laimelal (2o|4i

(.liiiuiiuiindi

Diplera

Invert

3.845

7000

h.gh

iVlerella Lake. Canadian High Arctic

l.escnid el al

(2015)

chironomids

Diptera

Invert

4.233

17115

high

Rcsolule Lake. Canadian High Arctic

Lescord et al.
(2015)

chironomids

Diptera

Invert

5.447

280000

high

Char Lake, Canadian High Arctic

Lescord et al.
(2015)

chironomids

Diptera

Invert

4.770

58889

hi till

Small Lake, Canadian High Arctic

Lescord et al.
(2015)

chironomids

Diptera

Invert

5.386

243333

liigh

\niih Lake, Canadian High Arctic

Lescord et al.
(2015)

chironomids

Diptera

Invert

5.740

55	

liigh

•J-Milc Lake, Canadian High Arctic

Lescord et al.
(2015)

worms

Capitellidae

Invert

2.961

913.93

hi'ill

Mai Po Marshes, Hong Kong

Loi et al. (2011)

Copepods

Copepoda

Invert

0.531

3.400

lnuli

Girondc Estuary, SW France

Munoz et al. (2019)

mysids

Mysidacea

Invert

0.591

3.900

hiuli

Gironde Estuary, SW France

Munoz et al. (2019)

white shrimp

Palaemon longirostris

ln\ eil

0.531

3.400

hiuli

Gironde Estuary, SW France

Munoz et al. (2019)

brown shrimp

Crangon crangon

Invert

o 5'J 1

3.900

high

Gironde Estuary, SW France

Munoz et al. (2019)

Oyster

Crassostrea gigas

ln\ crt

2 08(i

122.0

high

Gironde Estuary, SW France

Munoz et al. (2017)

snails

Bithynia tentaculata

lii\ eil

:.io<>

1284

high

Hogsmill River, Chertsey Bourne
River, Blackwater River

Wilkinson et al.
(2018)

amphipod

Gammarus pulex

lii\ eil

2 o"2

1 180

liigh

Hogsmill River, Chertsey Bourne
River, Blackwater River

Wilkinson et al.
(2018)

Pacific Oyster

Crassostrea gigas

lii\ eil

^ 808

6430

medium

Gulf Park - Xiamen Sea, China

Dai and Zheng
(2019)

Pacific Oyster

Crassostrea gigas

lii\ eil

3.621

4180

medium

Jimei Bridge - Xiamen Sea, China

Dai and Zheng
(2019)

Ghost crab

Ocypode stimpsoni

In\ eil

3.515

3270

medium

Fenglin - Xiamen Sea, China

Dai and Zheng
(2019)

Ghost crab

Ocypode stimpsoni

ln\ eil

3.627

4240

medium

Jimei Bridge - Xiamen Sea, China

Dai and Zheng
(2019)

Orange-striped
hermit crab

Clibanarius infraspinatus

ln\ ert

3.589

3879

medium

Jimei Bridge - Xiamen Sea, China

Dai and Zheng
(2019)

Snail

Gastropoda

Invert

1.183

15.26

medium

aMatikulu N2 Bridge

Fauconier et al.
(2020)

P-10


-------
Coin 111011 \:ime

Scientific \:ime

Tissue

1,0"
IJAI
(l./kป-
\\\\)

IJAI

(l./lvli-
\\\\)

Kiinkin^

Locution

Reference

mvsid

Mysis relicta

Invert

3.477

3000

medium

1 .ake ()ul;u'K<

1 loude el al t^uuxj

diporcin

Diporeia hoyi

Invert

4.505

32000

medium

1 .ake ()ni;ii'k>

Houde et al. (2008)

Snail

Cerithidea
rhizophorarum

Invert

0.430

2.692

medium

()imii;i ki\ or mouth and estuary,
Japan

Kobayashi et al
(2018)

waterlouse, water
boatmen,
amphipods,
roundworm

Isopoda, Hemiptera,
amphipoda, nematoda

Invert

2.974

942.0

medium

site A Stockholm \ilauda \irport

Koch et al. (2019)

Fresh water
amphipods

Amphipoda

Invert

2.957

905.0

medium

site R Ronneby Airport

Koch et al. (2019)

Mayflies,
Caddisflies,
Dragonflies,
Water boatmen,
Waterlouse, Fresh
water amphipods

Ephemeroptera,
Trichoptera, Odonata,
Hemiptera, Isopoda,
Amphipoda

Invert

2.728

5U<)

medium

site k the Kvarntorp area

Koch et al. (2019)

Gastropoda

Gastropoda

Invert

1.965

92.33

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

worms

Nereidae

Invert

1.893

78.25

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

worms

Sabellidae

Invert

2.562

364.6

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

Black tiger prawn

Penaeus monodon

Invert

2.344

220.7

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

Sand prawn

Metapenaeus ensis

Invert

2.457

286.4

medium

Mai Po Marshes, Hong Kong

Loi et al. (2011)

Copepods

Copepoda

Invert

3.140

1380

medium

Gironde Estuary, SW France

Munoz et al. (2017)

Mysids

Mysidacca

Invert

3.551

3560

medium

Gironde Estuary, SW France

Munoz et al. (2017)

Gammarids

Gammarus

Invert

3.377

2380

medium

Gironde Estuary, SW France

Munoz et al. (2017)

White shrimp

Palaemon longirostris

ln\ eil

3.44S

2803

medium

Gironde Estuary, SW France

Munoz et al. (2017)

Brown shrimp

Crangon crangon

Invert

3.N52

7110

medium

Gironde Estuary, SW France

Munoz et al. (2017)

bivalve

Mytilus galloprovincialis

Invert

3.701

5029

medium

AC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

bivalve

Mytilus galloprovincialis

Invert

3.436

2728

medium

NC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

bivalve

Mytilus galloprovincialis

Invert

3.056

1137

medium

FC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

crab

Carcinus aestuarii

Invert

3.210

1623

medium

AC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

crab

Carcinus aestuarii

ln\ ert

3.057

1140

medium

NC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

crab

Carcinus aestuarii

Iiin ert

2.742

551.5

medium

FC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

bivalve

Ruditapes decussatus

Invert

3.025

1059

medium

M Site, Orbetell lagoon, Italy

Renzi et al. (2013)

bivalve

Ruditapes decussatus

Invert

3.061

1150

medium

AC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

bivalve

Ruditapes decussatus

Invert

2.761

577.2

medium

NC Site, Orbetell lagoon, Italy

Renzi et al. (2013)

p-11


-------






1,0"
IJAI

IJAI













(l./kป-

(l./lvli-







Coin 111011 \:ime

Scientific \:ime

Tissue

\\\\)

\\\\)

Kiinkin^

Locution

Reference

bivalve

Ruditapes decussatus

Invert

2.593

392.2

medium

IT Sue. Orbclcll lagoon. Ilalv

Rcn/.i cl al. (2013)

prawn

Palaemon serratus

Invert

2.655

451.4

medium

AC Site. Orbclcll lagoon. Ilalv

Rcn/.i cl al. (2013)

prawn

Palaemon serratus

Invert

2.481

302.6

medium

NC Silc. Orbclcll lagoon, Italy

Renzi et al. (2013)

prawn

Palaemon serratus

Invert

2.273

187.5

medium

FC Silc. Orbclcll lagoon, Italy

Renzi et al. (2013)

rock oyster

Saccostrea commercialis

Invert

1.933

85.71

medium

Sydney Harbour, \uslralia

Thompson et al.
(2011)

a - Fish whole body
b - Invertebrate whole body

P-12


-------
P.2 Summary of PFOS BAFs used to calculate tissue criteria and
supplemental fish tissue values

Field measured BAFs used to calculate fish and invertebrate PFOS tissue criteria (fish
muscle, fish whole body, and invertebrate whole body) and supplemental fish tissue values
(blood, reproductive tissue, liver) are shown in Appendix P.l. Summary statistics for the BAFs
from this table used to derive tissue criteria and additional tissue \ allies (i e., lowest species-level
BAF from each site) are reported in Table 3-11 and Table Q-2, respecti\ civ Rankings for
individual BAFs were determined by Lawrence (2021), who devised a ranking system based on
five characteristics: 1) number of water samples. 2) number of tissue samples; 3) spatial
coordination of water and tissue samples; 4) temporal coordination of water and tissue samples;
and 5) general experimental design. For the first four characteristics, a score of one to three was
assigned, based on number of samples or how closely the water and tissue measurements were
paired. For the experimental design characteristic, a default \alue of zero was assigned; unless
the measured tissues were composites of mixed species, in which case it was assigned a three
(Lawrence 2<)21) These sub-scores were then summed and assigned a rank based on the final
score Studies with high quality rankings had scores of four or five, studies with medium quality
rankings had scores of five or six, and studies with low quality rankings had scores of seven or
higher (Lawrence 2< >21) Parameters for the scores assigned to the five characteristics are listed
in Table 2-2, and additional details can be found in Burkhard (2021). Only BAFs from studies
with high or medium quality rankings were included for the final BAF geometric mean
calculations used to derive tissue criteria (Table 3-12) and supplemental tissue values (Table
Q-3).

P-13


-------
P.3 PFOS BAFs References

Ahrens, L., K. Norstrom, T. Viktor, A.P. Cousins, S. Josefsson. 2015. Stockholm Arlanda
Airport as a source of per- and polyfluoroalkyl substances to water, sediment and fish.
Chemosphere 129: 33-38.

Awad, E., X. Zhang, S.P. Bhavsar, S. Petro, P.W. Crozier, E.J. Reiner, R. Fletcher, S.A.
Tittlemier, E. Braekevelt. 2011. Long-Term Environmental Fate of Perfluorinated Compounds
after Accidental Release at Toronto Airport. Environmental Science and Technology 45: 8081 -
8089.

Becker, A.M., S. Gerstmann, H. Frank. 2010. Perfluorooctanoic Acid and Perfluorooctane
Sulfonate in Two Fish Species Collected from the Roter Main River, Bavieiith, Germany.
Bulletin of Environmental Contamination and Toxicology 84: 132-135

Bhavsar, S.P., C. Fowler, S. Day, S. Petro, N. Gandhi, S.B. Gewurtz, C. Hao. \ /hao, K.G.
Drouillard, D. Morse. 2016. High levels, partitioning and fish consumption based water
guidelines of perfluoroalkyl acids downstream of a former firefighting training facility in
Canada. Environment international 94: 415-423.

Cui, W.J., J.X. Peng, Z.J. Tan, Y.X. Zhai. MV1. Guo, and H.J. Mini 2019. Pollution
characteristics of perfluorinated alkyl substances (PFASs) in seawater. sediments, and biological
samples from Jiaozhou Bay, China. Huanjing Kexue 4<)(lJ) 3990-3999.

Dai, Z. and F. Zheng. 2019. Distribution and bioaccumulation of perfluoroalkyl acids in Xiamen
coastal waters. Journal of Chemistry 36: 1-8.

De Silva, A. O., C. Spencer. li I ' Scott, S. Backus and D. C. Muir. 2011. Detection of a cyclic
perfluorinated acid, |->eilliioioelh\ lc\ clohexane sulfonate, in the Great Lakes of North America.
Environ Sci Teclinol 45( N) Si)(•<<)-X<)(•<(•<

DeSolla. S R., AO l)eSil\a. R.I I .etcher. 2012. Highly elevated levels of perfluorooctane
sulfonate and other pcriluoiinated acids found in biota and surface water downstream of an
international airport, Hamilton. Ontario, Canada. Environment International 39: 19-26.

Fang, S., X. Chen. S Zhao. Y Zhang, W. Jiang, L. Yang, L. Zhu. 2014. Trophic magnification
and isomer fractionation of perfluoroalkyl substances in the food web of Taihu Lake, China.
Environmental Science Technology 48: 2173-2182.

Fauconier, G., T. Groffen, V. Wepener, and L. Bervoets. 2020. Perfluorinated compounds in the
aquatic food chains of two subtropical estuaries. Sci. Total Environ. 719: 135047

Furdui, V.I., N.L. Stock, D A. Ellis, C.M. Butt, D M. Whittle, P.W. Crozier, E.J. Reiner, D.C.G.
Muir, S.A. Mabury. 2007. Spatial Distribution of Perfluoroalkyl Contaminants in Lake Trout
from the Great Lakes. Environ Sci Technol. 41(5): 1554-1559.

P-14


-------
Gebbink, W.A., A. Bignert, U. Berger. 2016. Perfluoroalkyl Acids (PFAAs) and Selected
Precursors in the Baltic Sea Environment: Do Precursors Play a Role in Food Web Accumulation
of PFAAs? Environ Sci Technol. 50(12): 6354-6362.

Houde M., G. Czub, J.M. Small, S. Backus, X. Wang, M. Alaee, D.C. Muir. 2008. Fractionation
and bioaccumulation of perfluorooctane sulfonate (PFOS) isomers in a Lake Ontario food web.
Environmental Science and Technology 42: 9397-9403.

Iwabuchi, K., N. Senzaki, S. Tsuda, H. Watanabe, I. Tamura, H. Takanobu, N. Tatarazako.
2015. Bioconcentration of perfluorinated compounds in wild medaka is related to octanol/water
partition coefficient. Fundam. Toxicol. Sci. 2(5): 201-208.

Kobayashi, J., Y. Maeda, Y. Imuta, F. Ishihara, N. Nakashima, T. Komorila. T Sakurai. 2018.
Bioaccumulation Patterns of Perfluoroalkyl Acids in an Estuary of the Ariake Sea. Japan.

Bulletin of Environmental Contamination and Toxicology 100: 536-540.

Koch, A., A. Karrman, L.W.Y. Yeung, M. Jonsson, L. Ahrens, and T. Wang. 2011->. Point source
characterization of per- and polyfluoroalkyl substances (PFASs) and extractable organofluorine
(EOF) in freshwater and aquatic invertebrates. Environmental Science Process and Impacts 21:
1887-1898.

Kwadijk, C., M.J.J. Kotterman, A. Koelmans 2<)|4 Partitioning of periluorooctanesulfonate and
perfluorohexanesulfonate in the aquatic environment a flu' ail accidental release of Aqueous Film
Forming Foam at Schiphol Amsterdam Airport I-n\ ironmental Toxicology and Chemistry 33:
1761-1765.

Kwadijk, C., P. Kor\ tar and A Koelmans 2010. Distribution of perfluorinated compounds in
aquatic systems in the Netherlands I-n\ iron mental science & technology. 44(10): 3746-3751.

Labadie, P. and M Che\ renil 2d I I Partitioning behaviour of perfluorinated alkyl contaminants
between water, sediment and fish in the Orge River (nearby Paris, France). Environmental
Pollution 159; 391-3l)7

Lam. \ -I I . (' -R Cho, J.-S l.ee, H.-Y. Soh, B.-C. Lee, J.-A. Lee, N. Tatarozako, K. Sasaki, N.
Saito, K. Iwabuchi. K. Kannan. H.-S. Cho. 2014. Perfluorinated alkyl substances in water,
sediment, plankton and fish from Korean rivers and lakes: A nationwide survey. Science of the
Total Environment 4l) I ~N2: 154-162.

Lescord, G. L., K. A. Kidd, A. O. De Silva, M. Williamson, C. Spencer, X. W. Wang and D. C.
G. Muir. 2015. Perfluorinated and polyfluorinated compounds in lake food webs from the
Canadian High Arctic. Environ. Sci. Technol. 49: 2694-2702.

Lin, A. Y.-C., S.C. Panchangam, Y.-T. Tsai, T.-H. Yu. 2014. Occurrence of perfluorinated
compounds in the aquatic environment as found in science park effluent, river water, rainwater,
sediments, and biotissues. Environmental monitoring and assessment 186: 3265-3275.

P-15


-------
Loi, E. I., L. W. Yeung, S. Taniyasu, P. K. Lam, K. Kannan andN. Yamashita. 2011. Trophic
Magnification of Poly- and Perfluorinated Compounds in a Subtropical Food Web. Environ. Sci.
Technol.(45): 5506-5513.

Munoz, G., H. Budzinski, M. Babut, H. Drouineau, M. Lauzent, K.L. Menach, J. Lobry, J.
Selleslagh, C. Simonnet-Laprade, P. Labadie. 2017. Evidence for the trophic transfer of
perfluoroalkylated substances in a temperate macrotidal estuary. Environmental Science &
Technology 51: 8450-8459.

Munoz, G., H. Budzinski, M. Babut, J. Lobry, J. Selleslagh, N. Tapic. P. Labadie. 2019.

Temporal variations of perfluoroalkyl substances partitioning between surface water, suspended
sediment, and biota in a macrotidal estuary. Chemosphere 233: 3 ll)-32(->

Pan, C.-G., J.-L. Zhao, Y.-S. Liu, Q.-Q. Zhang. 2014. Bioaccumulation and risk assessment of
per- and polyfluoroalkyl substances in wild freshwater fish from rivers in the Pearl River Delta
region, South China. Ecotoxicology and Environmental Safety 107: 192-199

Pan, Y., H. Zhang, Q. Cui, N. Sheng, L.W.Y. Yeung, Y. Guo, Y. Sun, J. Dai. 2d I 7 First Report
on the Occurrence and Bioaccumulation of Hexafluoropropylene Oxide Trimer Acid: An
Emerging Concern. Environmental Science and Technology 51: 9553-9560.

Pan, X., J. Ye, H. Zhang, J. Tang, and D. Pan. 2019. Occurrence, removal and bioaccumulation
of perfluoroalkyl substances in Lake Chaohu, China. Int. J. Environ. Res. Public Ftealth 16(10):
1692.

Pignotti, E., G. Casas, M. Llorca. A Tcllbuscher, D. Almeida, E. Dinello, M. Farre, D. Barcelo.
2017. Seasonal variations in the occurrence of perfluoroalkyl substances in water, sediment and
fish samples from Ebro Delia (Catalonia. Spain). Science of the Total Environment: 607-608:
933-943.

Renzi. M . (' Gucnanti. A Giowtni. G Perra, S.E. Focardi. 2013. Perfluorinated compounds:
Levels, trophic web enrichments and human dietary intakes in transitional water ecosystems.
Marine Pollution Bulletin 7(v 146-157.

Shi, Y., R Ycslcrurcn. Z. Zhou, X. Song, L. Xu, Y. Liang, Y. Cai. 2015. Tissue distribution and
whole body burden of the chlorinated polyfluoroalkyl ether sulfonic acid F-53B in crucian carp
(Carassius carassius) l-\ idence for a highly bioaccumulative contaminant of emerging concern.
Environmental Science and Technology 49:14156-14165.

Shi Y., R. Vestergren, T.H. Nost, Z. Zhou, Y. Cai. 2018. Probing the differential tissue
distribution and bioaccumulation behavior of per-and polyfluoroalkyl substances of varying
chain-lengths, isomeric structures and functional groups in crucian carp. Environmental Science
& Technology 52: 4592-4600.

Shi, Y., X. Song, Q. Ji, W. Li, S. He, Y. Cai. 2020. Tissue distribution and bioaccumulation of a
novel polyfluoroalkyl benzenesulfonate in crucian carp. Environment International 135: 105418.

P-16


-------
Taniyasu. S., K. Kannan, Y. Horii, N. Hanari, N. Yamashita. 2003. A survey of perfluorooctane
sulfonate and related perfluorinated organic compounds in water, fish, birds, and humans from
Japan. Environmental Science and Technology 37: 2634-2639.

Terechovs, A. K. E., A.J. Ansari, J.A. McDonald, S.J. Khan, F.I. Hai, N.A. Knott, J. Zhou, L.D.
Nghiem. 2019. Occurrence and bioconcentration of micropollutants in Silver Perch (Bidyanus
bidyanus) in a reclaimed water reservoir. Science of the Total Environment 650 (Part 1): 585-
593.

Thompson, J., A. Roach, G. Eaglesham, M.E. Bartkow, K. Edge. .1 I' Mueller. 2011.
Perfluorinated alkyl acids in water, sediment and wildlife from Sydney I larbour and
surroundings. Mar Pollut Bull. 62(12): 2869-2875.

Wang, Y., R. Vestergren, Y. Shi, D. Cao, L. Xu, X. Zhao, F. Wu. 2016. Identification, Tissue
Distribution, and Bioaccumulation Potential of Cyclic Perfluorinated Sulfonic Acids Isomers in
an Airport Impacted Ecosystem. Environmental Science and Technology 50: 10^23-10932.

Wilkinson, J.L., P.S. Hooda, J. Swinden, J. Barker, S. Barton. 2018. Spatial (bio) accumulation
of pharmaceuticals, illicit drugs, plasticisei s. peifluorinated compounds and metabolites in river
sediment, aquatic plants and benthic organisms linviron mental pollution 234: 864-875.

Zhou, Z., Y. Shi, L. Xu, Y. Cai. 2012. Perlluoiinalcd Compounds in Surface Water and
Organisms from Baivangdian T.ake in North China. Source Profiles, Bioaccumulation and
Potential Risk. Bulletin of lii\ iron mental Contamination and Toxicology 89: 519-524.

P-17


-------
Appendix Q Translation of Chronic Water Column Criterion into Other
Fish Tissue Types (liver, blood, reproductive tissues)

The PFOS aquatic life criteria (summarized in Section 3.3) includes chronic tissue
criteria for fish whole body, fish muscle, and invertebrate whole-body. Additional values for fish
liver, fish blood, and fish reproductive tissues were also calculated by transforming the chronic
water column criterion (i.e., 0.0084 mg/L) into representative tissue concentrations using tissue-
specific bioaccumulation factors (BAFs). Fish BAFs for liver. Mood, and reproductive tissues
were identified following the same approaches used to identify fish whole body, muscle, and
inverterbrate whole body BAFs, which are described in detail in Section 2.11.3 I Ikielly, BAFs
were determined from field measurements and calculated using the equation:

BAF = 4ฃi2ฃฃL	(Eq. Q-l)

(ฆwater

Where:

Cbiota = PFOS coiiccniralion in orgitiiisimtl lissucts)

Cwater = PF( )S LOIIlVIIIICIIIOII III WilkT

For further details on liAI's compilation and ranking, see Section 2.11.3.1 and Burkhard
(2021) liAl s based on reprodncti\ e tissues identified by Burkhard (2021) were further screened
to evaluate characteristics inlluence reproductive tissue BAFs. These characteristics included
timing of sample collection and organism sex, age, length and weight. However, since the data
were limited the inlluence of these characteristics could not be fully evaluated to determine their
potential influence on PFOS BAFs for reproductive tissues. Therefore, characteristics of timing
of sample collection and organism age, length or weight were currently not considered to be
influential given available data. Reproductive tissue BAFs were additionally screened to ensure
only BAFs based on adult females were considered, because female reproductive tissues are

Q-l


-------
most relevant to potential maternal transfer to offspring. This subset of reproductive-based BAFs
and corresponding species and sampling locations are described in Table Q-l.

Table Q-l. Characteristics of adult fish sampled for the calculation of PFOS reproductive
tissue BAFs.

All sampled fish were adults, and all reproductive tissues identified as gonad. Weights, lengths, and BAFs are

Author

Species

Collection
Dale

il

Sex

Age
<\r.)

Weight
(g-\v\v)

Length
(cm)

ISA 1'"

(1 -/Kg)

Mirciis el al

<-"15)

1 Mirupcan perch

(Perca (luviatiUs)

iu i: :ui:

3

l;

s.

\ k

\ k

		

Becker et al.
(2010)

European chub

(Leuciscus
cephalus)

8/28/2007

6

N.R.

4

rx 5

25.5

2,222

Labadie and

Chevreuil

(2011)

European chub

(Leuciscus
cephalus)

April 2010

5

3 M
2 F

N.R.

228.0 (M)
258.2 (F)

:x 5 (M)

:"siP)

10,000

Shi et al.
(2015, 2018)

Crucian carp

(Carassius
carassius)

July 2014'

30

24 F
6 M

N.R.

79.4	(F)

60.5	(M)

15(1 (F)
13.7 (M)

11,482

Shi et al.
(2015, 2018)

Crucian carp

(Carassius
carassius)

July 2014:

13

9 F
4 M

N.R.

352.3 (F)
0:11.7 (M)

24.6 (F)
24.8 (M)

5,888

Shi et al.
(2020)

Crucian carp

(Carassius
carassius)

N.R.

303

N.R

\ k

\ R.

N.R.

7,990

Shi et al.
(2020)

Crucian carp

(Carassius
carassius)

\ k

203

N.R.

N.R.

N.R.

N.R.

8,012

Wang et al.
(2016)

Crucian carp

(Carassius
carassius)

\pril :<)|4

S

N.R.

N.R.

(16.8-
65.1)5

(10.0-
14.7)5

25,645

'Xiaoqing River, China
2Tangxun Lake, China
3Yubei River, China
4Gaobeidian Lake, China5Range

The ilisirilmiions of fish liver, fish blood, and fish reproductive BAFs identified in the
literature used to calculate tissue-specific BAFs were determined in the same manner as
invertebrate, fish muscle, and fish whole body BAFs (Section 3.2.3.1). Briefly, distributions of
BAFs used to derive additional tissue values were based on the lowest species-level BAF
reported at a site. When more than one BAF was available for the same species at the same site,
the species-level BAF was calculated as the geometric mean of all BAFs for that species at that

Q-2


-------
site. Summary statistics for the PFOA BAFs used in the derivation of the additional tissue-based
values are presented below (Table Q-2) and individual BAFs are provided in Appendix P.

Table Q-2. Summary Statistics for PFOS BAFs in Additional Fish Tissues1.





(ieometrio

Median

20"'









Mean

HA I-"

('entile









ISA I'"

(l./kg-

liAl

Minimum

.Maximum





(l./kg-wet

wet

(1 ./kg-wet

(1 ,/kซ-wel

(1 ./kg-w e(

Category

11

weight)

weight)

weight)

weight)

weight)

Liver

19

5,708

4,572

2,462

Ill

83,753

Blood

11

14,355

11,167

6,273

3.5'111

80,168

Reproductive
Tissue

8

8,903

9,006

5,155

2,222

25,645

1- Based on the lowest species-level BAF measured at a site (i.e.. when two or more BAFs w civ a\ ailable for the
same species at the same site, the species-level geometric mean I! \l' was calculated, and the louesi species-level
BAF was used).

The chronic freshwater column criterion (see Section .1 2 I .1) was then translated into
tissue values using the 20th centile BAFs from the distributions of BAFs summarized in Table
Q-2 using the following equation

Tissue Value = Chronic Water Column Criterion x 20th Centile BAF (Eq. Q-2)

The resulting tissue \ allies that correspond to the 20th centile tissue-specific BAF used in the
equation Q-2 are reported in Table Q-3. The values reported in Table Q-3 represent tissue-based
concentrations that offer a le\ el of protection that is equal to the magnitude components of the
chronic water column criterion as well as the fish whole body, fish muscle, and invertebrate
whole-body tissue-based criteria; however, the tissue-based values reported in Table Q-3 are
only presented for comparative purposes and are not recommended criteria.

Q-3


-------
Table Q-3. PFOS Concentrations for Additional Fish Tissue.1,2

Category

PI-'OS Concent ration (nig/kg ww)

Liver

20.68

Blood

52.69

Reproductive Tissue

43.30

1	These PFOS concentrations are provided as supplemental information and are not intended to replace the PFOS fish tissue
criteria provided in Table .

2	Tissue criteria derived from the chronic water column concentration (CCC) with the use of bioaccumulation factors and are
expressed as wet weight (ww) concentrations.

Q-4


-------
Appendix R Example Data Evaluation Records (DERs)

The PFOS toxicity literature evaluated and used to derive the draft PFOS aquatic life
criteria was identified using the ECOTOXicology database (ECOTOX;
https://cfpub.epa.gov/ecotox/) as meeting data quality standards. ECOTOX is a source of high-
quality toxicity data for aquatic life, terrestrial plants, and wildlife. The database was created and
is maintained by the EPA, Office of Research and Development. Center for Computational
Toxicology and Exposure. The ECOTOX search generally begins with a comprehensive
chemical-specific literature search of the open literature conducted according to l-COTOX
Standard Operating Procedures (SOPs). The search terms are often comprised of chemical terms,
synonyms, degradates and verified Chemical Abstracts Service (CAS) numbers. After
developing the literature search strategy, l -COTOX curators conduct a series of searches, identify
potentially applicable studies based on title and abstract, acquire potentially applicable studies,
and then apply the applicability criteria for inclusion in ECOTOX. Applicability criteria for
inclusion into ECOTOX generally include:

I.	The toxic effects are related to single chemical exposure (unless the study is being
considered as part of a mixture effects assessment);

2 There is a biological effect 011 li\ 0. whole organisms or in vitro preparation including
gene chips or omics data on ad\ erse outcome pathways potentially of interest;

3.	Chemical test concentrations are reported;

4.	There is an explicit duration of exposure;

5.	Toxicology information that is relevant to OW is reported for the chemical of concern;

6.	The paper is published in the English language;

7.	The paper is a\ ailahle as a full article (not an abstract);

8.	The paper is publicly available;

9.	The paper is the primary source of the data;

10.	A calculated endpoint is reported or can be calculated using reported or available
information;

II.	Treatment(s) are compared to an acceptable control;

12.	The location of the study (e.g., laboratory vs. field) is reported; and

13.	The tested species is reported (with recognized nomenclature).

R-l


-------
Following inclusion in the ECOTOX database, toxicity studies are subsequently
evaluated by Office of Water. All studies were evaluated for data quality generally as described
by U.S.EPA (1985) in the 1985 Guidelines and in EPA's Office of Chemical Safety and
Pollution Prevention (OCSPP)'s Ecological Effects Test Guidelines (U.S.EPA 2016c), and EPA
OW's internal data quality SOP, which is consistent with OCSPP's data quality review approach
(U.S.EPA 2018). These toxicity data were further screened to ensure iliat the observed effects
could be primarily attributed to PFOS exposure. Office of Water completed a I )ER for each
species by chemical combination from the PFOS studies identified by ECOTOX I sample DERs
are presented here to convey the meticulous le\ el of e\ aluation. review, and documentation each
PFOS study identified by ECOTOX was subject to. Appendix 11.1 shows an example fish DER
and Appendix R.2 shows an example aquatic in\ ei lebrate Dl-R

R-2


-------
R.1 Example Fish DER

Part A: Overview
I. Test Information

Chemical name:

CAS name:

Purity:

Solubility in Water (units):

Controlled Experiment

(imanipulated)

CAS Number:
Storage conditions:

Field Study/Observation

(inot manipulated)

Date:

(I'hu e X by One)

Date:

IP A
IP A

Primary Reviewer: 	

Secondary Reviewer: 	

(At least one reviewer should be from EPA for sensitive taxa)

Citation: Indicate: author (s), year, study title, journal, volume, and pages.

(e.g., Slonim, A.R. 1973. Acute toxicity of beryllium sulfate to the common guppy. J. Wal. Pollul. Conlr. Fed. 45(10): 2110-2122)

Contractor (Place X by One)
Con tractor (Place X by One)

Companion Papers: Identify any companion papers associated i villi this paper using I he citation format above.

Were other DERs completed lor Companion I'sipcrs?

Yes

(Ifyes, list file names of
No DERs below)

Study Classification for Aquatic l.il'e Criteria Development: Place X by One Based on Highest Use

	 Acceptable for ()imniiuili\e I sc

	 Acceptable for Onalilali\c I sc

	 Not Acceptable lor I sc I nuscd

General Notes: Provide any necessary details regarding the study's use classification for all pertinent endpoints,
including non-apical endpoints within the study (e.g., note all study classifications for each endpoint if the use varies)

Major Deficiencies (note any slated exclusions): Check all that apply. Checking any of these items make the study "Not
Acceptable for Use"

Mixture (for controlled experiments only)

No Controls (for controlled experiments
only)

Excessive Control Mortality (> 10% for acute and > 20% for chronic)

. ,	. . .	.	Bioaccumulation: steady state not

Dilution water not adequately characterized	, ,

n J		reached

Dermal or Injection Exposure Pathway

Review paper or previously published without modification

R-3


-------
Other: (if any, list here)

POTENTIAL CHEMICAL MIXTURES : Describe any potential chemicals mixtures as characterized by study
authors (including any confirmation of chemical mixtures).

DESCRIPTION OF DILUTION WATER: Describe concerns with characterization of and/or major deficiencies
with dilution water.

General Notes:

Minor Deficiencies: List and describe any minor deficiencies or other concerns with test. These items may make the study
"Acceptablefor Qualitative Use" (exceptions may apply as noted)

For Field Studies/Observations: A field study/observation may be considered "Acceptable for Quantitative Use" if it

consisted of a range of exposure concentrations and the observed effects are justifiably contributed to a single chemical
exposure

	 Mixture (observed effects noljusliliabh con111 hutcd to single chemical exposure)

	 Uncharacterized Reference Sites/Condilions

POTENTIAL CHEMICAL MIXTl RES PRESENT A I Sill. I v.v ri be any potential chemicals mixtures present at
the site as characterized by study authors (including any confirmation of chemicals present at study site).

EXPOSURE VARIABILITY ACROSS STUDY S1TE(SV Describe any exposure variability across study site(s)
as characterized by study authors (i.e.. description of study design with reference and contaminated sites).

General Notes:

Reviewer's Comments: Provide additional comments that do not appear under other sections of the DER.

R-4


-------
ABSTRACT: Copy and paste abstract from publication.

SUMMARY: Fill out and modify as needed.
Acute:

Species (lilesliiiie)

Method'

Tesl
Diinilion

( hemiciil
/ Piiriu

pll

1 em p.

<ฐC)

Ihirdncss

(111li/l. ilS

CsiCO.0

or
S;ilini(\
(DDII

DOC
(lliu/l -)

r.ll'ecl

Reported
r.llecl
( oiicenlriilion
(mป/l.)

Verified
HITccl
( onccnlr;ilion
(niiป/l.)

Cliissificiilion























Quantitative /
Qualitative / Unused

a S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D=dissolved, Dict=diel;ii\. M l maternal transfer

Chronic:

Species < liTesliii^e)

Method-'

lesl
l)ui';ilioii

( hemiciil
/ Piiriu

pll

Temp.

<ฐC)

lliii'diiess

(111li/l. ilS

CaCO.o

or
S;ilini(\
(ppl)

DOC

(111^/1.)

Chronic
l.imils

Reported
C h ron ic
Value
(ni^/l. or

Verified
( lironic
Value
(illli/l. or

Chronic
Value
l.ndpoinl

( liissiliciilion

























Quantitative /
Qualitative /
Unused

a S=static, R=renewal, F=flow-throudi. 1 unmeasured. M measured. I luial. I) dissolved, Diet=dietary, MT=maternal transfer

R-5


-------
II. Results Provide results as reported in the publication (including supplemental materials). Include screen shots of tables and/or
figures reporting results from the article following tabulated data table in each associated results section for all studies. Complete
tabulated data tables for all studies for studies marked "Acceptable for Quantitative Use" and "Acceptable for Qualitative Use".

Water Quality Parameters: If only general summary data of water quality parameters is provided by study authors (i.e., no
specific details of water quality parameters on a treatment level is provided), summarize any information regarding water quality
parameters under General Notes below and indicate data not provided in Table A.II.l.

General Notes: For aquatic life criteria development, measured water quality parameters in the treatments nearest the toxicity
test endpoint(s), e.g., LC50, EC20, etc., are most relevant.

Table A.II.1. Measured Water Quality Parameters in Test Solutions.

Dissolved oxygen, temperature, pH and [other parameters (hardness, saliniis. !)()( )| 111 lesi solutions during the /Ay-day
exposure of [test organism] to [concentration of treatments)] of [test substance / under I static renewal (flow-through]
conditions.

Parameler

Treatment

Mean

Range

Dissolved
Oxygen

(% saturation
or mg/L)

[1]





[2]





j





j





Temperature
(O

[I]





[2]





j





j





pi"

HI





12/





J





J





Oilier (e.g..
hardness,
salinity, DOC)

11J





12]





j





J





R-6


-------
Chemical Concentrations: Summarize the concentration verification data from test solutions/media. Expand table to include
measured concentration data for each media type (i.e., water, diet, muscle, liver, blood, etc.).

General Notes: Provide any necessary detail regarding the measured concentrations, including any identified cause for
substantial differences between nominal and measured concentrations, if samples were collected on separate days (and if so provide
details), and any potential cross contamination.

Table A.II.2. Measured (and Nominal) Chemical Concentrations in Test Solutions/Media.

[Analytical Method] verification of test and control concentrations during an [X]-da> exposure of [test organism] to [test
snbsiaiice] Milder [sialic renewal 'flou-ihioimh] conditions





|Mi'iin|





Nil in her of

ISiandard





Nominal

Mc.isiiivd





Samples

l)c\ ialion or





(oiicciilralioii

( (IIHTIIII'illioil

Nil nihi l- olฐ

Noii-

Ik'low \
-------
Mortality: Briefly summarize mortality results (if any).

General Notes: Comment on concentrations response relationship and slope of response ifprovided. Compare mortality in
treatments with control group and/or the reference chemical.

Table A.II.3. Mean Percent [Mortality or Survival].

Mean percent mortality [or number of immobilized, survival] of [test organism] exposed to [test substance] for [test duration]
under [static/renewal/flow-through] conditions.



|\k';in "A,

|M;iihI;ii'(I l)o\ iiilion

1 iviilmonl

Mor(;ilil\|

or Sliiiuliii'd I'Iitoi'I

Control





[1]





[2]





[3]





[4]





[5]





[6]





[LCx]



NOEC



LOEC



a Use superscript to identify the values reported U> he simnlicaiilly different from control.

R-8


-------
Growth: Briefly summarize growth results (if any).

General Notes: Comment on concentrations response relationship and slope of response ifprovided. Compare growth endpoints
in treatments with control group and/or the reference chemical.

Table A.II.4. Mean [Growth].

Mean growth [length and/or weight] of [test organism] exposed to [test substance] for [test duration] under
[static/renewal/flow-through] conditions.



Menu (ii'owlh



Mosul IVrmil





| l.cMiiilh/W oiuhl |

ISliindiird l)c\i;ilion

(linnlie in |l.i'ii^lh/

ISliindiird l)c\ hilion

1 iviilmonl

( ii nils)

or Siiiiuliinl l!rror|

ISioniiissI

or Siiiiuliinl llrrorj

Control









[1]









[2]









[3]









[4]









[5]









[6]









./'









[ECx]





NOEC





LOEC





a Use superscript to identify the values repm ied in he simnlicanlly different from control.

R-9


-------
Reproductive: Briefly summarize reproduction endpoint results (if any). For multi-senerational studies, copy and paste Table
A.II.5 below for each generation with reproductive effects data.

General Notes: Comment on concentrations response relationship and slope of response ifprovided. Compare reproductive
endpoints in treatments with control group and/or the reference chemical.

Table A.II.5. Mean [Reproductive] Effect.

Mean [reproductive] effects for [generation] of [test organism] exposed to [test substance] for [test duration] under
[static/renewal/flow-through] conditions.









|Sliiildill'd



|Siiindiird

| M Oil n

| Si iin (hi rซl





|Sliiiul;ir(l



l)o\ iiilion



l)o\ iiilion

IlillCll

l)o\ iiilion



|\loiin

l)o\ iiilion or

| Moiin

OI*

|Moiin

or

Poroonl

or

Troiilmonl

Number of

Sliindiinl

Nil in hoi' of

Sliindiinl

Porconl

Sliindiird

Su r\ i\ill

Sliindiinl

(iinils)

S|):i\\ ns |

r.rrorl



r.rror|

ll;ik-h|

r.rrorl

Posll

r.rrorl

Control

















[11

















[21

















[31

















[41

















[51

















[61

















i

















[ECx]









NOEC









LOEC









a Use superscript to identify the values reported to he simnlicaiilK dilTerenl from control.

R-10


-------
Sublethal Toxicity Endpoints: Include other sublethal effect(s), including behavioral abnormalities or other signs of toxicity,
if any. Copy Table A.II.6 as needed to provide details for each sublethal effect observed.

General Notes: Briefly summarize observed sublethal effects otherwise not captured in the results table(s) below.

Table A.II.6. Mean [Sublethal] Effect.

Mean /"Sublethal effect (e.g., behavioral abnormalities, etc.)] in [test organism] during [test duration (acute/chronic)]
exposure to [test substance] under [static/renewal/flow-through] conditions.



|Mciin Suhlolhiil





Kespon se|

|Sliiii(liii'(l l)c\i;i(ion or

llViKllKIII

(units)

Siiindiird l.rmr|

Control





rn





[21





[31





[41





[51





[61





i





|ECxl



NOEC



LOEC



a Use superscript to identify the \ allies repni ied In be siginl ic;iiill> different from control

R-ll


-------
Reported Statistics: Copy and paste statistical section from publication.

R-12


-------
Part B: Detailed Review
I. Materials and Methods

Protocol/Guidance Followed: Indicate if provided by authors.

Deviations from Protocol: If authors report any deviations from the protocol noted above indicate here.

Study Design and Methods: Copy and paste methods section from publication.

TEST ORGANISM: Provide information under Details and any relevant or related information or clarifications in Remarks.

Parameler

Delails

Remarks

Species:

Common Name:
Scientific Name:

\nrih \incnc;i11 speaev'
SuiTnuak' fur Nui'ili Vmerican
Taxon"

(Place A' if applicable)

Strain/Source:

•	Wild caught from unpolluted areas [1]

o Quarantine for at least 14 days or until they are
disease free, before acclimation [1]

•	Must originate from same source and population [1]

•	Should not be used:

o If appeared stressed, such as discoloration or

unusual behavior [1]
o If more than 5% die during the 48 hours before

test initiation [1]
o If they were used in previous test treatments or
controls [2]

•	No treatments of diseases may be administered:
o Within 16 hour of field collection 111

o Within 10 days or testing or during testing 111





Age at Study Initiation:

Acute:

•	Juvenile stages preferred 111
Chronic:

•	Life-cycle test:

o Embryos or newly hatched voting 48 hours old

[2]

•	Partial life-cycle test:

o Immature juveniles at least 2 months prior to
active gonad development |2|

•	Early life-stage test:

o Shortly after fertilization |2|





Was body weight or length recorded al
test initiation?

Yes No



Was body weight or length recorded at
regular intervals?

Yes No

If yes, describe regular intervals:



R-13


-------
STUDY PARAMETERS: Provide information under Details and any relevant information of deficiencies in Remarks.
Complete for both Controlled Experiments and Field Studies/Observations.



Pummel or

Delails

Remarks

Number of Replicates per Treatment
Group:

•	At least 2 replicates/treatment recommended for
acute tests [1]

•	At least 2 replicates/treatment recommended for
chronic tests [3]

Control(s):



Treatment(s):



Number of Organisms per Replicate/
Treatment Group:

•	At least 10 organisms/treatment recommended [3]

•	At least 7 organisms/treatment acceptable [4]

Control(s):



Treatment(s):



Exposure Pathway:

(i.e., water, sediment, gavage, or diet).

Note: all other pathways (e.g., dermal, single dose via

gavage, and injection) are unacceptable.





Exposure Duration:

Acute

•	Should be 96 hours [2]

Chronic

•	Life-cycle tests:

o Ensure that all life stages and life processes are
exposed [2]

o Begin with embryos (or newly hatched young),
continue through maturation and reproduction, and
should end not less than 24 days (90 days for
salmonids) after the hatching of the next
generation [2]

•	Partial life-cycle tests:

o Allowed with species that require >1 year to reach
sexual maturity, so that all major life stages can be
exposed to the test material in <15 months [2]

o Begin with immature juveniles al least 2 months
prior to active gonad development, continue
through maturation and reproduction, and end not
less than 24 days (90 days for salmonids) alter the
hatching of the next generation 121

•	Early life-cycle tests:

o 28 to 32 day (60 day post hatch for salmonids)
exposures from shortly after fertilization through
embryonic, larval, and earlv juvenile development

m

Acute

I'ariial Life ( \ele
LarK Life Siaue
l ull Life ( \ele
Other (please remark):



Test Concentrations (remember units):

Recommended test concentrations include ai least three
concentrations other than the control; four or more will
provide a better statistical analysis [3]

Nummal:

Measured:

Media measured in:



Observation Intervals:

• Should be an appropriate number of observations
over the study to ensure water quality is being
properly maintained [4]





R-14


-------
CONTROLLED EXPERIMENT STUDY PARAMETERS: Provide information under Details and any relevant
information of deficiencies in Remarks. Complete for Controlled Experiments only.

-*

Paramcler

Details

Remarks

Acclimation/Holding:

•	Should be placed in a tank along with the water in
which they were transported

o Water should be changed gradually to 100%

dilution water (usually 2 or more days) [1]
o For wild-caught animals, test water temperature
should be within 5ฐC of collection water
temperature [1]
o Temperature change rate should not exceed 3ฐC
within 72 hours [1]

•	To avoid unnecessary stress and promote good
health:

o Organisms should not be crowded [1]
o Water temperature variation should be limited [1]
o Dissolved oxygen:

ฆ	Maintain between 60 - 100% saturation [1]

ฆ	Continuous gentle aeration if needed [1]

o Unionized ammonia concentration in holding and
acclimation waters should be < 35 jag/L [1]

Duration:

Feeding:

Water type:

Temperature (ฐC):

Dissolved Oxygen (mg/L):

Health (any mortality observed?):

Identify number of individuals excluded from testing and/or
analysis (if any):

Acclimation followed published guidance?

Describe, if any

Yes \n

If yes, indicate which guidance:



Test Vessel:

•	Test chambers should be loosely covered [1]

•	Test chamber material:

o Should minimize sorption of test chemical from
water [1]

o Should not contain substances that can be leached
or dissolved in solution and are free of substances
that could react with exposure chemical [1]
o Glass, No. 316 stainless steel, nylon screen and
perfluorocarbon (e.g. Teflon) are acceptable 111
o Rubber, copper, brass, galvanized mclal. epoxy
glues, lead and flexible tubing should not come
into contact with test solution, dil. water, or slock

[i]

•	Size/volume should maintain acceptable biomass
loading rates (see Biomass Loading Rate below) 111

Material:
Size:

l ill Volume

Briefly describe the test vessel:

Test Solution Delivery Systein/Melhod:

•	Flow-through preferred for some highly volatile,
hydrolysable or degradable materials [2|

o Concentrations should be measured often enough
using acceptable analytical methods [2]

•	Chronic exposures:

o Flow-through, measured tests required |2|

Tom ( niiceiilialmiis Measured
Yci No

Tom Solution Delivery System:
Static
Renewal

Indicate Interval:

Flow-through

Indicate Type of Diluter:



Source of Dilution Water:

•	Freshwater hardness range should be < 5 mg/L or <
10% of the average (whichever is greater) [1]

•	Saltwater salinity range should be < 2 g/kg or < 20%
of the average (whichever is greater) [1]

•	Dilution water must be characterized (natural surface
water, well water, etc.) [3]

o Distilled/deionized water without the addition of
appropriate salts should not be used [2]

•	Dilution water in which total organic carbon or
particulate matter >5 mg/L should not be used [2]

o Unless data show that organic carbon or particulate
matter do not affect toxicity [2]





Dilution Series (e.g., 0.5x, 0.6x, etc.):





R-15


-------


Paramcler

Details

Remarks





Dissolved Oxygen (mg/L):





Dilution Water Parameters:

Measured at the beginning of the experiment or
averaged over the duration of the experiment (details of

pH:





Temperature (ฐC):





Hardness (mg/L as CaCCh):





water quality parameters measured in test solutions
should be included under the results section)

Salinity (ppt):





Total Organic Carbon (mg/L):







Dissolved Organic Carbon (mg/L):





Aeration:

•	Acceptable to maintain dissolved oxygen at 60 -
100% saturation at all times [1]

•	Avoid aeration when testing highly oxidizable,
reducible and volatile materials [1]

•	Turbulence should be minimized to prevent stress on
test organisms and/or re-suspend fecal matter [1]

•	Aeration should be the same in all test chambers at all
times [1]

Yes No





Describe Preparation of Test
Concentrations (e.g., water exposure,
diet):







Test Chemical Solubility in Water:

List units and conditions (e.g., 0.01% at 20ฐC)





s>

Were concentrations in water or diet
verified by chemical analysis?

Measured test concentrations should be reported in
Table A.II.2 above.

	Yes 	No

Indicate media:



-

Were test concentrations verified by
chemical analysis in tissue?

Measured test concentrations can be verified in test
organism tissue (e.g., blood, liver, muscle) alone if a
dose-response relationship is observed.

Measured test concentrations should be reported in
Table A.II.2 above.

	Yes 	\n

Indicate tissue type:

If test concentrations were verified in test organism
tissue, was a dose-response relationship observed?



Were stability and homogeneity of test
material in water/diet determined?

	Yes 	No





Was test material regurgitated/avoided?

Yes \n





Solvent/Vehicle Type (Water or Dictan )

•	When used, a carrier solvent should be kept to a
minimum concentration [1]

•	Should not affect either survival or growth oTlcsl
organisms 111

•	Should be reagent grade or better [1J

•	Should not exceed 0.5 ml/L (static) or 0.1 nil I. (Mow
through) unless it was shown that higher
concentrations do not aJlecl toxicity [3]







Negative Control:

Yes No





Reference Toxicant Testing:

Yes No

If Yes, identify substance:



Other Control: If any (e.g. solvent control)





R-16


-------


Biomass Loading Rate:

•	Loading should be limited so as not to affect test
results. Loading will vary depending on temperature,
type of test (static vs. flow-through), species,
food/feeding regime, chamber size, test solution
volume, etc. [1]

•	This maximum number would have to be determined
for the species, test duration, temperature, flow rate,
test solution volume, chamber size, food, feeding
regime, etc.

•	Loading should be sufficiently low to ensure:

o Dissolved oxygen is at least 60% of saturation

(40% for warm-water species) [1,5]
o Unionized ammonia does not exceed 35 |ig/L [1]
o Uptake by test organisms does not lower test

material concentration by > 20% [1]
o Growth of organisms is not reduced by crowding

•	Generally, at the end of the test, the loading (grams of
organisms; wet weight; blotted dry) in each test
chamber should not exceed the following:

o Static tests: > 0.8 g/L (lower temperatures); > 0.5

g/L (higher temperatures) [1]
o Flow through tests: > 1 g/L/day or > 10 g/L at any
time (lower temperatures); >0.5 g/L/day or > 5
g/L at any time (higher temperatures) [1]

•	Lower temperatures are defined as the lower of 17ฐC
or the optimal test temperature for that species [ 11





R-17


-------


PiiniiiKMi'i-

IH'l.iils

Ki'iiiiirks

-*

l<"eediug:

• Unacceptable for acute tests [2]
o Exceptions:

ฆ	Data indicate that the food did not affect the
toxicity of the test material [2]

ฆ	Test organisms will be severely stressed if they
are unfed for 96 hours [2]

ฆ	Test material is very soluble and does not sorb
or complex readily (e.g., ammonia) [2]

Yes No





Lighting:

•	Depends on the type of test (acute or chronic) and
endpoint (e.g., reproduction) of interest.

o Embryos should be incubated under dim

incandescent lighting (< 20 fc) or total darkness
during early life-stage toxicity testing
o Embryos must not be subjected to prolonged
exposure to direct sunlight, fluorescent lighting, or
high intensity incandescent lighting

•	Generally, ambient laboratory levels (50-100 fc) or
natural lighting should be acceptable, as well as a
diurnal cycle consisting of 50% daylight or other
natural seasonal diurnal cycle.

•	Artificial light cycles should have a 15 - 30-minute
transition period to avoid stress due to rapid increases
in light intensity [1]





Study Design/Methods Classification: (Place Xby One Based on Overall Study Design/Methods Classification)
Provide details of Major or Minor Deficiencies/Concerns with Study Design in Associated Sections of Part A: Overview

This classification should be taken into consideration for the overall study classification for aquatic li/e criteria development in Part A.

	 Study Design Acceptable for Quantitative I se

	 Study Design Acceptable for Qualitative Use

	 Study Design Nol Acceptable for Use

Additional Notes: Provide additional considerations for the classification of study use based on the study design.

R-18


-------
OBSERVATIONS: Provide information under Details and any relevant information in Remarks. This information should be
consistent with the Results Section in Part A.

Parameter

Details

Remarks

Parameters measured including sublethal
effects/toxicity symptoms:

Common Apical Parameters Include:

Acute

•	EC50 based on percentage of organisms exhibiting
loss of equilibrium plus the percentage of organisms
immobilized plus percentage of organisms killed [2]
0 If not available, the 96-hr LC50 should be used [2]

Chronic

•	Life-cycle/Partial Life-cycle test:

0 Survival and growth of adults and young,

maturation of males and females, eggs spawned
per female, embryo viability (salmonids only), and
hatchability [2]

•	Early life-cycle test:

0 Survival and growth [2]

List parameters:



Was control survival acceptable?

Acute

•	> 90% control survival at test termination [2]
Chronic

•	> 80% control survival at test termination [2]

Yes \n
Control survival



Were individuals excluded from the
analysis?

Yes \n

If yes, describe justification provided:



Was water quality in test chambers
acceptable?

• If appropriate, describe any water quality issues
(e.g., dissolved oxygen level below 60% of
saturation)

Yes \n



Availability of concentration-response
data:

•	Were treatment level concentration-response
data included in study publication (can be from
tables, graphs, or supplemental materials)'?
specify endpoints in remarks

•	Were replicate level concentration-response
data included in study publication (can be from
tables, graphs, or supplemental materials)'?
specify endpoints in remarks

Yes \o
Yes No



• If treatment and/or replicate level
concentration-response data were included, how
was data presented? (check all thai applyj

Tables
Graphs

Supplemental Files



• Were concentration-response data estimated
from graphs study publication or supplemental
materials?

Yes No

If yes, indicate software used:

Yes No



• Should additional concentration-response data
be requested from study authors?

If concentration-response data are available, complete
Verification of Statistical Results (Part C)for sensitive
species.

Requested by:

Request date:

Date additional data received:



R-19


-------
Part C: Statistical Verification of Results

I.	Statistical Verification Information: Report the statistical methods (e.g., EPA TRAP, BMDS, R, other) used to verify the
reported study or test results for the five (5) most sensitive genera and sensitive apical endpoints (including for tests where such
estimates were not provided). If values for the LC50, LT50 and NOEC are greater than the highest test concentration, use the "> "
symbol.

Primary Reviewer: 	 Date: 	 	 EPA 	 Contractor {Place X by One)

Secondary Reviewer: 	 Date: 	 	 EPA 	 Contractor {Place X by One)

{At least one reviewer should be from EPA for sensitive taxa)

Endpoint(s) Verified:

Additional Calculated Endpoint(s):

Statistical Method (e.g., TRAP, BMDS, R, other):

II.	Toxicity Values: Include confidence intervals if applicable

NOEC:

LOEC:

MATC:

ECs:

EC10:

EC20:

ECso or LCso

Dose-Response Curve Classification: (Place X by One)

This classification should be taken into consideration for llie overall study classification for aquatic life criteria development in Part A

	Dose-Response Cim\ e AccqHaMe for Quantitative Use

	Dose-Response Cur\ e AcccplaMc lor Qualitative Use

	Dosc-Rcsponsc Cur\e \ol AccqMablc for Use

Summary of Statistical Veri Ileal ion: I'nividc .summary of methods used in statistical verification.

Additional Notes:

Attachments:

1.	Provide attachments to ensure all data used in Part C are captured, whether from study results reported in the publication
and/or from additional data requested from study authors

•	Data from study results of the publication should be reported in Results section of Part A

•	Additional data provided upon request from study authors should be reported in Table C.II.l below and original
correspondence with study authors should be included as attachments

2.	Model assessment output (including all model figures, tables, and fit metrics)

3.	Statistical code used for curve fitting

R-20


-------
III. Attachments: Include all attachments listed above after the table below.

Additional Data Used in Response-Curve: Provide all data used to fit dose-response curve not captured in Results section of PER above in Part A. Add rows as needed.
First row in italicized text is an example.

Table C.II.1 Additional Data Used in Dose-Response Curve.

( 11 I'M" II)

Spi-iii-s

r.iHipiiini

Tiv;iliiu-nl



| Si ;i n il:i I'll
IX'\ i;iiiun

HI"

Sl;i ihI;i nl
Krnir|

# of
Sun i\ urs

N'

k1

11 ฆ

kl'spilllsi-

kl'spilllsi-
I nil

ClIIH'

('mil' iinils

Alchronicl

Ceriodaphnia dubia

#of

young/female

0

6





10

10

I

IS

count

0.03

mg/L





























































































































































































































































aN = number of individuals per treatment; k= number of replicates per treatment le\ el. n number ol individuals per replicate

R-21


-------
Part D: References to Test Guidance

1.	ASTM Standard E 739, 1980. 2002. Standard guide for conducting acute toxicity tests on
test materials with fishes, macroinvertebrates, and amphibians. ASTM International,
West Conshohocken, PA.

2.	Stephan, C.E., D.I. Mount, DJ. Hansen, I.H. Gentile, G.A. Chapman and W.A. Brungs. 1985.
Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of Aquatic
Organisms and their Uses. PB85-227049. National Technical Information Service, Springfield,
VA.

3.	Stephan, C.E. 1995. Review of results of toxicity tests with aquatic organisms. Draft.
U.S. EPA, MED. Duluth, MN. 13 pp.

4.	OECD 203. 1992. Test No. 203: Fish, Acute Toxicity Test. OECD Guidelines for the
Testing of Chemicals, Section 2, OECD Publishing, Paris,

https://doi.org/10.1787/9789264069961-en.

5.	American Public Health Association (APHA). 2012. Standard methods lor the
examination of water and wastewater. Part 8000 - Toxicity. APHA. Washington, DC.

R-22


-------
R.2 Example Aquatic Invertebrate DER

Part A: Overview
I. Test Information

Chemical name:

CAS name:

Purity:

Solubility in Water (units):

Controlled Experiment

(,manipulated)

Primary Reviewer: 	

Secondary Reviewer: 	

(At least one reviewer should be from EPA for sensitive taxa)

CAS Number:
Storage conditions:

Field Study/Observation

(inot manipulated)

Date:

Date:

(I'/acc X by One)

IP A
LI'A

Con tractor (Place X by One)
Con tractor (Place X by One)

Citation: Indicate: author(s), year, study title, journal, volume, and/>ages.

(e.g., Keller, A.E and S.G. Zam. 1991. The acute toxicity of selected metals to the freshwater mussel. - Inoc/onta imbecilis. Environ. Toxicol. Chem. 10(4): 539-546.)

Companion Papers: Identify any companion papers associated with this paper using the citation format above.

(Ifyes, list file names of

Were other DERs completed lor Companion Papers?		 Yes 	 No DERs below)

Study Classification for Aqualic Life Criteria Development:

	 AcccplaMc lor Ouanlilalive Use

	 Acceptable for ()imlitati\e I se

	 Not Acceptable for I se/Unused

General Notes: Provide any necessary details regarding the study's use classification for all pertinent endpoints, including
non-apical endpoints within the study (e.g., note all study classifications for each endpoint if the use varies)

Major Deficiencies (note any slated exclusions): Check all that apply. Checking any of these items make the study "Not
Acceptable for Use"

, , ,,,	,, , . , . .	No Controls (for controlled experiments

Mixture (tor controlled experiments only)	only)

Excessive Control Mortality (> 10% for acute and > 20% for chronic)

. ,	. . .	.	Bioaccumulation: steady state not

Dilution water not adequately characterized	, ,

n J		reached

Dermal or Injection Exposure Pathway

Review paper or previously published without modification

R-23


-------
Other: (if any, list here)

POTENTIAL CHEMICAL MIXTURES : Describe any potential chemicals mixtures as characterized by study
authors (including any confirmation of chemical mixtures).

DESCRIPTION OF DILUTION WATER: Describe concerns with characterization of and/or major deficiencies
with dilution water.

General Notes:

Minor Deficiencies: List and describe any minor deficiencies or other concerns with lest, 't hese items may make the study
"Acceptablefor Qualitative Use" (exceptions may apply as noted )

For Field Studies/Observations: A field study/observation may be considered "Acceptable for Quantitative Use" if it

consisted of a range of exposure concentrations and the observed effects are justifiably contributed to a single chemical
exposure

	 Mixture (observed effects not justifiably contributed In single chemical exposure)

	 Uncharacterized Reference Sites ( ondilions

POTENTIAL CHEMICAL MIXTURES PRESENT AT SITE: Describe any potential chemicals mixtures present at
the site as characterized by study authors (including any confirmation of chemicals present at study site).

EXPOSURE VARI ABILITY ACROSS STUDY SITE(S). Describe any exposure variability across study site(s)
as characterized by study authors (i.e.. description of study design with reference and contaminated sites).

General Notes:

Reviewer's Comments: Provide additional comments that do not appear under other sections of the template.

R-24


-------
ABSTRACT: Copy and paste abstract from publication.

SUMMARY: Fill out and modify as needed.
Aaile

Species (lilesliiiie)

Melliod'1

Test
(lui'iilioii

( hemiciil
/ Piiriu

pll

Temp.

<ฐC)

Ihirdncss

(111li/l. ilS

CaiCO.o

or
S;ilini(\
(DDII

DOC
(lliu/l -)

tilled

Reported
tilled
( onceiil r;ition
(iiiiป/l.)

Verified
tilled
( onceiil r;i(ioii
(ill li/l.)

CI;issific;ilion























Quantitative / Qualitative /
Unused

3 S=static, R=renewal, F=flow-through, U=unmeasured, M=measured, T=total, D=dissolved, Diel diclars. M l maternal transfer

Chronic:

Species (lireslii^e)

Melliod-'

lesl
diiriilion

( hemiciil
/ Piiriu

pll

Temp.

<ฐC)

lliirdness

(lllfi/l. ilS

CaCO.o

or
Siilinilt
(ppl)

DOC

(iiiii/l.)

Chronic
l.imils

Reported
Chronic
\ iilue
(niii/l. or

Verified
Chronic
Value
dii^/l. or

( lironic
Value
tlndpoinl

Classification

























Quantitative /
Qualitative / Unused

a S=static, R=renewal, F=flow-through, U=unmeasuied. \1 measured. I U>i;il. I) dissnh cd. Dicl=dietary, MT=maternal transfer

R-25


-------
II. Results Provide results as reported in the publication (including supplemental materials). Include screen shots of tables and/or
figures reporting results from the article following tabulated data table in each associated results section for all studies. Complete
tabulated data tables for all studies for studies marked "Acceptable for Quantitative Use" and "Acceptable for Qualitative Use".

Water Quality Parameters: If only general summary data of water quality parameters is provided by study authors (i.e., no
specific details of water quality parameters on a treatment level is provided), summarize any information regarding water quality
parameters under General Notes below and include data not provided in Table A. II. 1.

General Notes: For aquatic life criteria development, measured water quality parameters in the treatments nearest the toxicity
test endpoint(s), e.g., LC50, EC20, etc., are most relevant.

Table A.II.1. Measured Water Quality Parameters in Test Solutions.

Dissolved oxygen, temperature, pH and [other parameters (hardness, saliniis. !)()( )| 111 lesi solutions during the /Ay-day
exposure of [test organism] to [concentration of treatments)] of [test substance / under I static renewal (flow-through]
conditions.

Parameler

Treatment

Mean

Range

Dissolved
oxygen

(% saturation
or mg/L)

[1]





[2]





j





j





Temperature
(O

[I]





[2]





j





j





pi"

HI





12/





J





J





Oilier (e.g..
hardness,
salinity, DOC)

11J





12]





j





J





R-26


-------
Chemical Concentrations: Summarize the concentration verification data from test solutions/media. Expand table to include
each measured concentration data for each media type (i.e., muscle, liver, blood, etc.).

General Notes: Provide any necessary detail regarding the measured concentrations, including any identified cause for
substantial differences between nominal and measured concentrations, if samples were collected on separate days (and if so provide
details), and any potential cross contamination.

Table A.II.2. Measured (and Nominal) Chemical Concentrations in Test Solutions/Media.

[Analytical Method] verification of test and control concentrations during an [X]-da> exposure of [test organism] to [test
substance] Milder [sialic renewal flou-ihioimh] conditions





|Mi'iin|





Nil in her of

ISiiiiuhinl





Nominal

Mc.isiiivd





Samples

l)e\ ialion or





Concentration

Concentration

Nil nihi l- olฐ

Noii-

Ik'low \
-------
Mortality: Briefly summarize mortality results (if any).

General Notes: Comment on concentrations response relations and slope of response ifprovided. Compare mortality with control
treatment and/or the reference chemical.

Table A.II.3. Mean Percent [Mortality or Survival].

Mean percent mortality [or number of immobilized] or survival of [test organism] exposed to [test substance] for [test
duration] under [static/renewal/flow-through] conditions.



|\k';in "A,

|M;iihI;ii'(I l)o\ iiilion

1 iviilmonl

Mor(;ilil\|

or Sliiiuliii'd I'Iitoi'I

Control





[1]





[2]





[3]





[4]





[5]





[6]





[LCX]



NOEC



LOEC



a Use superscript to identify the values reported U> he simnlicaiilly different from control.

R-28


-------
Growth: Briefly summarize growth results (if any).

General Notes: Comment on concentrations response relations and slope of response ifprovided. Compare growth endpoints with
control treatment and/or the reference chemical.

Table A.II.4. Mean [Growth].

Mean growth [length and/or weight] of [test organism] exposed to [test substance] for [test duration] under
[static/renewal/flow-through] conditions.



Menu (ii'owlh



Mosul IVrmil





| l.cMiiilh/W oiuhl |

ISliindiird l)c\i;ilion

(linnlie in |l.i-n^lh/

ISliindiird l)c\ hilion

1 iviilmonl

( ii nils)

or Siiiiuliinl l!rror|

ISioniiissI

or Siiiiuliinl llrrorj

Control









[1]









[2]









[3]









[4]









[5]









[6]









./'









[ECX]





NOEC





LOEC





a Use superscript to identify the values repm ied in he simnlicanlly different from control.

R-29


-------
Reproductive: Briefly summarize reproduction endpoint results (if any). For multi-senerational studies, copy and paste Table
A.II.5 below for each generation with reproductive effects data.

General Notes: Comment on concentrations response relations and slope of response if provided. Compare reproduction
endpoints with control treatment and/or the reference chemical.

Table A.II.5. Mean [Reproductive] Effect.

Mean [reproductive] effects for [generation] of [test organism] exposed to [test substance] for [test duration] under
[static/renewal/flow-through] conditions.









| Sl ;i il(l;i I'd



ISiiindiird



| Moiin

ISiiindiird



l)o\ iiilion



l)o\ iiilion



N u in her

l)o\ iiilion oi*

| M c;i n

or

| M Oil n

or

TrOilllllOlll

ol

Sliindiinl

Nil ill hoi' ol'

Siiiiidiird

Niiiiihoror

Siiiiidiird

(units)

Sp;m nsj

l-rror]



Ki'ioi'l

OITspriniil

I'.I'I'OI'I

Control













[11













[21













[31













[41













[51













[61













i













[ECxl







NOEC







LOEC







aUse superscript to identify I lie \ allies reported to he simnlicaiilK diN'crciil I'roiii control.

R-30


-------
Sublethal Toxicity Endpoints: Include other sublethal effect(s), including behavioral abnormalities or other signs of toxicity,
if any. Copy Table A.II.6 as needed to provide details for each sublethal effect observed.

General Notes: Briefly summarize observed sublethal effects otherwise not captured in the results table(s) below.

Table A.II.6. Mean [Sublethal] Effect.

Mean /"Sublethal effect (e.g., behavioral abnormalities, etc.)] in [test organism] during [test duration (acute/chronic)]
exposure to [test substance] under [static renewal flow-through] conditions.



|Mcan Sublethal





Kespon se|

|Siamlai'(l l)c\ia(ion or

Treadmill

(units)

Standard l.rror|

Control





[1]





[2]





[3]





[4]





[5]





[6]





./'





[ECx]



NOEC



LOEC



a Use superscript to identify the \ allies reported to be s>iginlicaiill> different from control

Reported Statistics: Copy and paste statistical section from publication.

R-31


-------
Part B: Detailed Review
I. Materials and Methods

PROTOCOL/GUIDANCE FOLLOWED: Indicate if provided by authors.

DEVIATIONS FROM PROTOCOL: If authors report any deviations from the protocol noted above indicate here.
Study Design and Methods: Copy and paste methods section from publication.

TEST ORGANISM: Provide information under Details and any relevant or related information or clarifications in Remarks.

Parameler

Delails

Remarks

Species:

Common Name:
Scientific Name:

Nui'lh \incnc;i11

SuiTnuak' I'm' \nrth American

Tavin"

(Place A' if applicable)

Strain/Source:

•	Wild caught from unpolluted areas [1]

o Quarantine for at least 7 days or until they are
disease free, before acclimation [1]

•	Must originate from same source and population [1]

•	Should not be used:

o If appeared stressed, such as discoloration or

unusual behavior [1]
o If more than 5% die during the 48 hours before

test initiation [1]
o If they were used in previous test treatments or
controls [2]

•	No treatments of diseases may be administered:
o Within 16 hours of field collection 111

o Within 10 days of testing or during testing 111





Age at Study Initiation:

Acute:

•	Larval stages preferred [1]

•	Mayflies and Stoneflies
o Early instar [ 11

•	Daphnids/cladocerans:
o < 24-hr old [11

•	Midges:

o 2ntl or 3ri instar larva 111

•	Hyalella azteca (chronic exposure)
o Generally, 7-8 days old |31

•	Freshwater mussels (chronic exposure)
o Generally, 2 month old juveniles |4|

•	Mysids (chronic exposure)
o < 24-hr old m





Was body weight or length recorded al
test initiation and/or at regular intervals?

Yes No



Was body weight or length recorded at
regular intervals?

Yes No

If yes, describe regular intervals:



R-32


-------
STUDY PARAMETERS: Provide information under Details and any relevant information of deficiencies in Remarks.
Complete for both Controlled Experiments and Field Studies Observations.		



Paramcler

Delails

Remarks



Number of Replicates per Treatment
Group:

•	At least 2 replicates/treatment recommended for
acute tests [1]

•	At least 2 replicates/treatment recommended for
chronic tests [5]

Control(s):





Treatment(s):





Number of Organisms per Replicate/
Treatment Group:

• At least 10 organisms/treatment recommended.

Control(s):





Treatment(s):



1

Exposure Pathway:

(i.e., water, sediment, or diet). Note: all other pathways
(e.g., dermal, injection) are unacceptable.







Exposure Duration:

Acute

•	Cladocerans and midges should be 48 hours [2]

o Longer durations acceptable if test species not fed
and had acceptable controls [2]

•	Freshwater mussel glochidia should be a maximum
of 24 hours [4]

o Shorter durations (6, 12, 18 hours) acceptable so
long as 90% survival of control animals achieved
(see below) [4]

•	Embryo/larva (bivalve mollusks, sea urchins,
lobsters, crabs, shrimp and abalones) should be 96
hours, but at least 48 hours [2]

Acute





( limine



<

• Other invertebrate species should be 96 hours

Other (please remark):



ฆi

Chronic

•	Daphnids/cladocerans should be 21 days (3-brood
test) [2]

o Exception 7 days acceptable for Ceriodaphnia
dubia [2]

•	Freshwater juvenile mussels should be al least 28

days [4]

•	Hyalella azteca should be at least 42 days
o Beginning with 7-8 day old animals |31

•	Mysids should continue until 7 days pasl the median
time of first brood release in the controls 14|







Test Concentrations (remember iniils):

Nmmiial





Recommended test concentrations include ai least three
concentrations other than the control: four or more will
provide a better statistical analysis.

Measured:





Media measured in:





Observation Intervals:

• Should be an appropriate number of observations
over the study to ensure waler quality is being
properly maintained f 11





R-33


-------
CONTROLLED EXPERIMENT STUDY PARAMETERS: Provide information under Details and any relevant
information of deficiencies in Remarks. Complete for Controlled Experiments only.



ParnimMcr

Details

Remarks



Acclimation/Holding:

• Should be placed in a tank along with the water in

Duration:

Identify number of individuals excluded from testing and/or
analysis (if any):



which they were transported [1]
o Water should be changed gradually to 100%

dilution water (usually 2 or more days) [1]
o For wild-caught animals, test water temperature
should be within 5ฐC of collection water

Feeding:





Water:





temperature [1]
o Temperature change rate should not exceed 3ฐC

Temperature (ฐC):





within 72 hours [1]

• To avoid unnecessary stress and promote good
health:

o Organisms should not be crowded [1]
o Water temperature variation should be limited
o Dissolved oxygen:

ฆ	Maintain between 60 - 100% saturation [1]

ฆ	Continuous gentle aeration if needed [1]

o Unionized ammonia concentration in holding and
acclimation waters should be < 35 jag/L [1]

Dissolved Oxygen (mg/L):





Health (any mortality observed?):



-

Acclimation followed published guidance?

Describe, if any

Yes \n



-*

If yes, indicate which guidance:



s

Test Vessel:

• Test chambers should be loosely covered [1]

Material:

Briefly describe the test vessel here



o Should minimize sorption of test chemical from
water [1]

o Should not contain substances that can be leached

Size:



ฃ

or dissolved in solution and free of substances that
could react with exposure chemical [1]
o Glass, No. 316 stainless steel, nylon screen and

Fill Volume'





perfluorocarbon (e.g. Teflon) arc acceptable 111
o Rubber, copper, brass, galvanized nielal. epoxv
glues, lead and flexible tubing should noi come
into contact with test solution, dilution water or
stock [1]

•	Size/volume should maintain acceptable bioniass
loading rates (see below) 111

•	Substrate:

o Required for some species (e.g.. Hvalclla azicca)

[3]

o Common types: stainless steel screen, nylon

screen, quail/, sand, cotton gauze and maple leaves

[3]

o More inert substances preferred over plant
material, since plants may break down during
testing and promote bacterial growth [3]
o Consideration should be given between substrate
and toxicant [3]

ฆ Hydrophobic organic compounds in particular
can bind strongly to Nitexฎ screen, reducing
exposure concentrations, especially for studies
using static or intermittent renewal exposure
methods [31







Sllllsll';||e I sod (ifapplicable)'.









R-34


-------


Parameter

Details

Remarks



Test Solution Delivery System/Method:

•	Flow-through preferred for some highly volatile,
hydrolyzable or degradable materials [2]

o Concentrations should be measured often enough
using acceptable analytical methods [2]

•	Chronic exposures:

o Flow-through, measured tests required [2]
o Exception: renewal is acceptable for daphnids [2]

Test Concentrations Measured
Yes No

Test Solution Delivery System:
Static
Renewal

Indicate Interval:

Flow-through

Indicate Type of Diluter



Source of Dilution Water:

•	Freshwater hardness range should be < 5 mg/L or <
10% of the average (whichever is greater) [1]

•	Saltwater salinity range should be < 2 g/kg or < 20%
of the average (whichever is greater) [1]

•	Dilution water must be characterized (natural surface
water, well water, etc.) [2]

o Distilled/deionized water without the addition of
appropriate salts should not be used [2]

•	Dilution water in which total organic carbon or
particulate matter exceed 5 mg/L should not be used
o Unless data show that organic carbon or particulate

matter do not affect toxicity [2]

•	Dilution water for tests with Hyalella azteca

o Reconstituted waters should have at least 0.02 mg
bromide/L; natural ground or surface water
presumed to have sufficient bromide [3]
o Recommended that control/dilution waters have
chloride concentrations at or above 15 mg/L [3]





Dilution Series (e.g., 0.5x, 0.6x, etc.):





Dilution Water Parameters:

Measured at the beginning of the experiment or
averaged over the duration of the experiment (details of
water quality parameters measured in test solutions
should be included under the results section)

Dissolved Oxygen ung/L):
pH:

Temperature (ฐC):

Hardness (mg/L as CaC03):
Salinity (ppt):

Total Organic Carbon (mg/L):
1 )issolved Organic Carbon (mg/L):



Aeration:

•	Acceptable to maintain dissolved oxygen al 60 -
100% saturation al all limes [1]

•	Avoid aeration when testing highly oxidizable.
reducible and volatile materials

•	Turbulence should be minimized to prevent stress on
test organisms and/or re-suspend fecal mailer 111

•	Aeration should be the same in all lesl chambers al all
times [1]

Yes No



Describe Preparation of Test
Concentrations (e.g., water exposure,
diet):





R-35


-------


Parameter

Details

Remarks



Test Chemical Solubility in Water:

• List units and conditions (e.g., 0.01% at 20ฐC)





Were concentrations in water or diet
verified by chemical analysis?

Measured test concentrations should be reported in
Table A.II.2 above.

Yes No



Indicate media:

Were test concentrations verified by
chemical analysis in tissue?

Measured test concentrations can be verified in test
organism tissue (e.g., blood, liver, muscle) alone if a
dose-response relationship is observed.

Measured test concentrations should be reported in
Table A.II.2 above.

Yes No

If test concentrations were verified in test organism
tissue, was a dose-response relationship observed?

Indicate tissue type:

Were stability and homogeneity of test
material in water/diet determined?

Yes \o





Was test material regurgitated/avoided?

Yes V-





Solvent/Vehicle Type:

•	When used, a carrier solvent should be kept to a
minimum concentration [1]

•	Should not affect either survival or growth of test
organisms [1]

•	Should be reagent grade or better [1]

•	Should not exceed 0.5 ml/L (static), or 0.1 ml/L (flow
through) unless it was shown that higher
concentrations do not affect toxicity [5]





Negative Control:

Yes No





Reference Toxicant Testing:

Yes No



If yes, identify substance:

Other Control: If any (e.g. solvent control I





Biomass Loading Rate:

•	Loading should be limited so as nol lo alfccl test
results. Loading will vary depending on icmpcrnlurc.
type of test (static vs. flow-through), species,
food/feeding regime, chamber size, test solution
volume, etc. [1]

•	This maximum number would have lo be determined
for the species, test duration, temperature. How rale,
test solution volume, chamber size. Ibod. leediim
regime, etc.

•	Loading should be sufficiently low to ensure:

o Dissolved oxygen is at least 60% of saturation

(40% for warm-water species) [1,6]
o Unionized ammonia does not exceed 35 |ig 1. 111
o Uptake by test organisms does not lower tesl

material concentration by 20% [1]
o Growth of organisms is nol reduced by crowding

•	Generally, at the end of the test, the loading (grams of
organisms; wet weight; blotted dry) in each test
chamber should not exceed the following:

o Static tests: > 0.8 g/L (lower temperatures); > 0.5

g/L (higher temperatures) [1]
o Flow through tests: > 1 g/L/day or > 10 g/L at any
time (lower temperatures); >0.5 g/L/day or > 5
g/L at any time (higher temperatures) [1]
o Lower temperatures are defined as the lower of
17ฐC or the optimal test temperature for that
species. [1]





R-36


-------
For Controlled Experiments Only

Feeding:

• Unacceptable for acute tests [2]
o Exceptions:

ฆ	Data indicate that the food did not affect the
toxicity of the test material [2]

ฆ	Test organisms will be severely stressed if they
are unfed for 96 hours [2]

ฆ	Test material is very soluble and does not sorb
or complex readily (e.g., ammonia) [21

Yes No





Lighting:

•	No specific requirements for lighting

•	Generally, ambient laboratory levels (50 - 100 fc) or
natural lighting should be acceptable, as well as a
diurnal cycle consisting of 50% daylight or other
natural seasonal diurnal cycle

•	Artificial light cycles should have a 15 - 30 minute
transition period to avoid stress due to rapid increases
in light intensity [ 1 ]

•	Depends on the type of test (acute or chronic) and
endpoint (e.g., reproduction) of interest.





Study Design/Methods Classification: (Place Xbv One Based on Overall Study Design/Methods Classification)
Provide details of Major or Minor Deficiencies/Concerns with Study Design in Associated Sections of Part A: Overview

This classification should be taken into consideration for the overall study classification for aquatic life criteria development in Part A.

	 Study Design Acceptable for Quantitative Use

	 Study Design Acceptable for Qualitative Use

	 Study Design Not Acceptable for Use

Additional Notes: Provide additional considerations for the classification of study use based on the study design.

R-37


-------
OBSERVATIONS: Provide information under Details and any relevant information in Remarks. This information should be
consistent with the Results Section in Part A.

Parameler

Details

Remarks

Parameters measured including sublethal
effects/toxicity symptoms:

Common Apical Parameters Include:

Acute

•	Daphnids/cladocerans:

o EC50 based on percentage of organisms

immobilized plus percentage of organisms killed

[2]

•	Embryo/larva (bivalve molluscs, sea urchins, lobsters,
crabs, shrimp, and abalones):

0 EC50 based on the percentage of organisms with
incompletely developed shells plus the percentage
of organisms killed [2]

ฆ	If not available, the lower of the 96 hour EC50
based on the percentage of organisms with
incompletely developed shells and the 96-hr
LC50 should be used [2]

•	Freshwater mussel (glochidia and juveniles):

0 Glochidia: EC50 based on 100 x number closed
glochidia after adding NaCl solution - number
closed glochidia before adding NaCl solution) /
Total number open and closed glochidia after
adding NaCl solution [4]

0 Juvenile: EC50 based on percentage exhibiting foot
movement within a 5-min observation period [4]

•	All other species and older life stages:

0 EC50 based on the percentage of organisms

exhibiting loss of equilibrium plus the percentage
of organisms immobilized plus the percentage of
organisms killed [2]

ฆ	If not available, the 96 hour LC50 should be
used [2]

Chronic

•	Daphnid:

0 Survival and young per female [2]

•	Mysids:

0 Survival, growth and young per female |2|

List parameters:



Was control survival acceptable?

Acute

•	> 90% control survival al lesi termination |2|

0 Glochidia 90% altar 24 hours, or. the nexl longest
duration less than 24 hours that had al leasi 90ฐ 0
survival [4]

Chronic

•	> 80% control survival at test termination 121

0 80% in 42 day test with Hyalella azteca. slightly
lower in tests substantially longer lhan 42 davs |3|

Yes No



Cuiiiiol survival (%):

R-38


-------
Paramcler

Details

Remarks

Were individuals excluded from the
analysis?

Yes No

If yes, describe justification provided:



Was water quality in test chambers
acceptable?

• If appropriate, describe any water quality issues
(e.g., dissolved oxygen level below 60% of
saturation)

Yes No



Availability of concentration-response
data:

• Were treatment level concentration-response
data included in study publication (can be from
tables, graphs, or supplemental materials)?
specify endpoints in remarks

Yes No



•	Were replicate level concentration-response
data included in study publication (can be from
tables, graphs, or supplemental materials)?
specify endpoints in remarks

•

Yes No



• If treatment and/or replicate level
concentration-response data were included, how
was data presented? (check all that apply)

Tables
Graphs

Supplemental Fi les



• Were concentration-response data estimated
from graphs study publication or supplemental
materials?

Yes No
If yes, indicate software used:

'les \n



Should additional concentration-response data be
requested from study authors?

If concentration-response data are available, complete
Verification of Statistical Results (Part C) tor sensitive

species.

Requested by:

Request date:

Dale additional data received:



R-39


-------
Part C: Statistical Verification of Results

I.	Statistical Verification Information: Report the statistical methods (e.g., EPA TRAP, BMDS, R, other) used to verify the
reported study or test results for the five (5) most sensitive genera and sensitive apical endpoints (including for tests where such
estimates were not provided). If values for the LC50, LT50 and NOEC are greater than the highest test concentration, use the "> "
symbol.

Primary Reviewer: 	 Date: 	 	 EPA 	 Contractor {Place X by One)

Secondary Reviewer: 	 Date: 	 	 EPA 	 Contractor {Place X by One)

{At least one reviewer should be from EPA for sensitive taxa)

Endpoint(s) Verified:

Additional Calculated Endpoint(s):

Statistical Method (e.g., TRAP, BMDS, R, other):

II.	Toxicity Values: Include confidence intervals if applicable

NOEC:

LOEC:

MATC:

ECs:

EC10:

EC20:

ECso or LCso

Dose-Response Curve Classification: (Place X by One)

This classification should be taken into consideration for llie overall study classification for aquatic life criteria development in Part A

	Dose-Response Cim \ e Acceptable for Quantitative Use

	Dose-Response Cin\ e Acceptable lor Qualitative Use

	Dose-Response Cui \e Not Acceptable for Use

Summary of Statistical Ver ideal ion: I'nividc .summary of methods used in statistical verification.

Additional Notes:

Attachments:

1.	Provide attachments to ensure all data used in Part C is captured, whether from study results reported in the publication
and/or from additional data requested from study authors

•	Data from study results of the publication should be reported in Results section of Part A

•	Additional data provided upon request from study authors should be reported in Table C.II.l below and original
correspondence with study authors should be included as attachments

2.	Model assessment output (including all model figures, tables, and fit metrics)

3.	Statistical code used for curve fitting

R-40


-------
III. Attachments: Include all attachments listed above after the table below.

Additional Data Used in Response-Curve: Provide all data used to fit dose-response curve not captured in Results section of PER above in Part A, rows as needed. First
row in italicized text is an example.

Table C.II.1 Additional Data Used in Dose-Response Curve.

( 11 I'M" II)

Spi-ik-s

r.iHipiiini

Tiv;ilnu-nl



| Si ;i n il:i I'll
l)i-\ i;iliun
hi*

Sl;iii(l;i|-(l
llrnirl

# of
Sun i\ urs

N'

k'

11 ฆ

Kl'S|>IIIISl-

ki'spimsi-
I nil

ClIIH'

Ciuu' iinils

Alchronicl

Ceriodaphnia dubia

#of

young/female

0

6





10

10

I

1S

count

0.03

mg/L





























































































































































































































































aN = number of individuals per treatment; k= number of replicates per treatment le\ el. n number ol individuals per replicate

R-41


-------
Part D: References to Test Guidance

6.	ASTM Standard E 739, 1980. 2002. Standard guide for conducting acute toxicity tests on
test materials with fishes, macroinvertebrates, and amphibians. ASTM International,
West Conshohocken, PA.

7.	Stephan, C.E., D.I. Mount, DJ. Hansen, J.H. Gentile, G.A. Chapman and W.A. Brungs. 1985.
Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of Aquatic
Organisms and their Uses. PB85-227049. National Technical Information Service, Springfield,
VA.

8.	Mount, D.R. and J.R. Hockett. 2015. Issue summary regarding test conditions and
methods for water only toxicity testing with Hyalella azleca. Memorandum to Kathryn
Gallagher, U.S. EPA Office of Water. U.S. EPA Office of Research and Development.
MED. Duluth, MN. 9 pp.

9.	Bringolf, R.B., M.C. Barnhart, and W.G. Cope. 2013. Determining the appropriate duration of
toxicity tests with glochidia of native freshwater mussels. Submitted to I idw aid I lammer. U.S.
EPA. Chicago, IL, May 8, 2013. 39 pp.

10.	Stephan, C.E. 1995. Review of results of toxicity tests with aquatic organisms. Draft.
U.S. EPA, MED. Duluth, MN. 13 pp.

11.	American Public Health Association (APHA). 2<) 12. Standard methods for the
examination of water and wastewater. Part 8000 - Toxicity APHA. Washington, DC.

R-42


-------