United States Science and, Technology EPA 823-R-97-0Q6
Environmental Protection (4305) September 1997
Agency					 ,		

*>EPA The Incidence And Severity

Of Sediment Contamination
In Surface Waters Of The
United States

Volume 1:

National Sediment
Quality Survey


-------
y^os%

S!	i

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460

JAN 7 19:

THE ADMINISTRATOR

The Honorable Albert Gore, Jr.

President of the Senate
Washington, D.C, 20510

Dear Mr. President:

As required by the Water Resources Development Act of 1992 (WRDA), I am pleased to
transmit the Environmental Protection Agency's (EPA) Report to Congress on the Incidence and
Severity of Sediment Contamination in Surface Waters of the United States. This report
describes the accumulation of chemical contaminants in river, lake, ocean, and estuary bottoms
and includes a screening assessment of the potential for associated adverse effects to human and
environmental health. It represents the first comprehensive EPA analysis of sediment chemistry
and related biological data to assess what is known about the national incidence and severity of
sediment contamination. As directed by WRDA, EPA consulted with the U.S. Army Corps of
Engineers and the National Oceanic and Atmospheric Administration in compiling data and
preparing the report.

EPA studied available data from sixty-five percent of the 2,111 watersheds in the
continental United States and identified ninety-six watersheds that contain "areas of probable
concern." In portions of these watersheds, environmental conditions may be unsuitable for
bottom dwelling creatures, and fish that live in these waters may contain chemicals at levels
unsafe for regular consumption. Areas of probable concern are located in regions affected by
urban and agricultural runoff, municipal and industrial waste discharge, and other pollution
sources. EPA recommends that resource managers fully examine the risks to human health and
the environment in these watersheds. Authorities should take steps to ensure that major pollution
sources are effectively controlled and that plans are in place to improve sediment conditions and
to support long-term health goals. EPA's goals for managing the problem of contaminated
sediment are provided as an enclosure to this letter.

Printed on Recycled Paper


-------
2

The process to produce EPA's Report to Congress on the Incidence and Severity of
Sediment Contamination in Surface Waters of the United States has been thorough and
extensive, meeting WRDA requirements for Federal agency consultation, as well as EPA's own
standards and policies regarding internal program and regional office review, external scientific
peer review, and external stakeholder review. I would be pleased to further discuss the contents
of this report at your convenience.

Sincerely,

Carol M, Browner

Enclosure


-------
,^gp8r%

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460

JAN ~7 199T

THE ADMINISTRATOR

The Honorable Newt Gingrich
Speaker of the House of Representatives
Washington, D.C. 20515

Dear Mr. Speaker:

As required by the Water Resources Development Act of 1992 (WRDA), I am pleased to
transmit the Environmental Protection Agency's (EPA) Report to Congress on the Incidence and
Severity of Sediment Contamination in Surface Waters of the United States. This report
describes the accumulation of chemical contaminants in river, lake, ocean, and estuary bottoms
and includes a screening assessment of the potential for associated adverse effects to human and
environmental health. It represents the first comprehensive EPA analysis of sediment chemistry
and related biological data to assess what is known about the national incidence and severity of
sediment contamination. As directed by WRDA, EPA consulted with the U.S. Army Corps of
Engineers and the National Oceanic and Atmospheric Administration in compiling data and
preparing the report.

EPA studied available data from sixty-five percent of the 2,111 watersheds in the
continental United States and identified ninety-six watersheds that contain "areas of probable
concern." In portions of these watersheds, environmental conditions may be unsuitable for
bottom dwelling creatures, and fish that live in these waters may contain chemicals at levels
unsafe for regular consumption. Areas of probable concern are located in regions affected by
urban and agricultural runoff, municipal and industrial waste discharge, and other pollution
sources. EPA recommends that resourcb managers fully examine the risks to human health and
the environment in these watersheds. Authorities should take steps to ensure that major pollution
sources are effectively controlled and that plans are in place to improve sediment conditions and
to support long-term health goals. EPA's goals for managing the problem of contaminated
sediment are provided as an enclosure to this letter.

Printed an Recycled Paper


-------
2

The process to produce EPA's Report to Congress on the Incidence and Severity of
Sediment Contamination in Surface Waters of the United States has been thorough and
extensive, meeting WRDA requirements for Federal agency consultation, as well as EPA's own
standards and policies regarding internal program and regional office review, external scientific
peer review, and external stakeholder review. I would be pleased to further discuss the contents
of this report at your convenience.

Sincerely,

Carol M. Browner

Enclosure


-------
Managing Contaminated Sediment in the United States

Issue Background

Many pollutants released to the environment settle and accumulate in the silt and mud
called sediment on the bottoms of rivers, lakes, estuaries, and oceans. Much of the
contaminated sediment in the U.S. was polluted years ago by such chemicals as DDT,
PCBs, and mercury, which have since been banned or restricted. These contaminants are
now found less frequently in overlying surface water than in the past. However, they can
persist for many years in the sediment, where they can cause adverse effects to aquatic
organisms and to human health. Some other chemicals released to surface waters from
industrial and municipal discharges, and polluted runoff from urban and agricultural
areas, continue to accumulate to environmentally harmful levels in sediment.

Costs of Sediment Contamination

Ecological and human health impairment due to contaminated sediment imposes costs
on society. Fish diseases causing tumors and fin rot and loss of species and communities
that cannot tolerate sediment contamination can severely damage aquatic ecosystems.
Contaminants in sediment can also poison the food chain. Fish and shellfish can become
unsafe for human or wildlife consumption. Potential costs to society include lost
recreational enjoyment and revenues or, worse, possible long-term adverse health effects
such as cancer or children's neurological and IQ impairment if fish consumption
warnings are not issued and heeded. The health and ecological risks posed by
contaminated sediment dredged from harbors can lead to increased cost of disposal and
lost opportunities for beneficial uses, such as habitat restoration.

Volume of Contaminated Sediments

The U.S. Environmental Protection Agency estimates that approximately 10 percent of
the sediment underlying our nation's surface water is sufficiently contaminated with toxic
pollutants to pose potential risks to fish and to humans and wildlife who eat fish. This
represents about 1.2 billion cubic yards of contaminated sediment out of the
approximately 12 billion cubic yards of total surface sediments (upper five centimeters)
where many bottom dwelling organisms live, and where the primary exchange processes
between the sediment and overlying surface water occur. Approximately 300 million


-------
2

cubic yards of sediments are dredged from harbors and shipping channels annually to
maintain commerce, and about 3-12 million cubic yards of those are sufficiently
contaminated to require special handling and disposal. These amounts are graphically
illustrated in the diagram below.

Volume of U.S.

Sediment by Category /

~ Sadlmant: approx 18 billion cubic yard»\

¦Contaminated 8adlmatrt: approx, 1,2 \

billion cubic yards	\

HODradgad Material; 300 million cublo yaida*

¦contaminated Dradgad matarial: 3-12
million cubic yard*'

•Annual ntlmilM •« raportad In EPA'a draft Contamlnatad Sadlmut Managanant Stiatagy

Where Is contaminated sediment a potential concern?

EPA has studied data from 1,372 of the 2,111 watersheds in the continental U.S. Of
these, EPA has identified 96 watersheds that contain "areas of probable concern" where
potential adverse effects of sediment contamination are more likely to be found. These
areas, identified in the figure below, are on the Atlantic, Gulf, Great Lakes, and Pacific
coasts, as well as in inland waterways, in regions affected by urban and agricultural
runoff, municipal and industrial
' waste discharges, and other
pollution sources. Some of
these areas have been studied
extensively, and now have
appropriate management
actions in place. However,
others may require further
evaluation to confirm that
environmental effects are
occurring.


-------
3

EPA's Contaminated Sediment Goals

EPA's Contaminated Sediment Management Strategy establishes four goals to
manage the problem of contaminated sediment, and describes actions the Agency intends
to take to accomplish those goals. The four goals are:

1.	Prevent the volume of contaminated sediment from increasing. To accomplish
this, EPA will employ its pollution prevention and source control programs. Both the
pesticides and toxic substances programs will use new and existing chemical registration
programs to reduce the potential for release of sediment contaminants to surface waters.
The water program will work with States and Tribes to identify waterbodies with
contaminated sediment as impaired and target them for Total Maximum Daily Load
evaluations. EPA will also work with the States and Tribes to enhance the
implementation of point and nonpoint source controls in these watersheds.

2.	Reduce the volume of existing contaminated sediment. EPA will consider a range
of risk management alternatives to reduce the volume and effects of existing
contaminated sediment, including in-situ containment and contaminated sediment
removal. In some cases, risk managers may select a combination of practicable
alternatives as the remedy. Where natural attenuation is part of the selected alternative,
EPA will accelerate pollution prevention and source control efforts, where appropriate, to
ensure that clean sediments will bury contaminated ones within an acceptable recovery
period. During the recovery period, EPA will work with the States to improve human
health protection by establishing and maintaining appropriate fish consumption
advisories. In all cases, environmental monitoring will be conducted to ensure that risk
management goals are achieved.

3.	Ensure that sediment dredging and dredged material disposal are managed in an
environmentally sound manner. EPA carefully evaluates the potential environmental
effects of proposed dredged material disposal. In addition, EPA is initiating a national
stakeholder review process to help the Agency review the ocean disposal testing
requirements and ensure that any future revisions reflect both sound policy and sound
science. EPA and the Army Corps of Engineers also will provide appropriate guidance to
further encourage and promote beneficial uses of dredged material.


-------
4

4. Develop scientifically sound sediment management tools for use in pollution
prevention, source control, remediation, and dredged material management. Such
tools include national inventories of sediment quality and environmental releases of
contaminants, numerical assessment guidelines to evaluate contaminant concentrations,
and standardized bioassay tests to evaluate the bioaccumulation and toxicity potential of
specific sediment samples.

Working with States and Tribes through existing statutory authorities, EPA can
identify impaired waterbodies and watersheds at risk from contaminated sediment,
implement appropriate actions to accomplish the goals described above, and monitor the
effectiveness of actions taken to accomplish the Agency's goals.


-------
The Incidence And Severity
Of Sediment Contamination
In Surface Waters Of The
United States:

Volume 1:

National Sediment
Quality Survey

September 1997

Office of Science and Technology
United States Environmental Protection Agency
401 M Street, SW
Washington, DC 20460


-------
The National Sediment Quality Survey is a screening-level assessment of sediment quality
that compiles and evaluates sediment chemistry data and related biological data taken from
existing databases. The data and information contained in this document could be used in
various EPA regulatory programs for priority setting or other purposes after further evaluation for
program-specific criteria. However, this document has no immediate or direct regulatory conse-
quence, It does not in' itself establish any legally binding requirements, establish or affect legal rights
or obligations, or represent a determination of any party's liability.


-------
Contents

;	Page

Tables

Figures	ix

Acknowledgments

Executive Summary	XV

Introduction	1 -1

What Is the National Sediment Quality Survey?							1-1

Why Is Contaminated Sediment an Important National Issue?..						 1-2

How Significant is the Problem?		...........			.							1-3

What Are the Potential Sources of Sediment Contamination?							1-4

Methodology	2—1

Background									2-2

Description of NSI Data					,					2-3

NSI Data Evaluation Approach								 2-4

Sediment Chemistry Data												2-12

Tissue Residue Data										2-15

Toxicity Data															2-15

Incorporation of Regional Comments on the Preliminary Evaluation of

Sediment Chemistry Data 											 2-16

Evaluation Using EPA Wildlife Criteria									2-16

.....a3—1

National Assessment									,												 3-1

Watershed Analysis												 3-12

Wildlife Assessment													 3-17

Regional and State Assessment							3-17

EPA Region 1												 3-22

EPA Region 2 															 3-27

EPA Region 3																	3-32

EPA Region 4															3-37

EPA Region 5							;								 3-43

EPA Region 6 													 3-49

EPA Region 7															3-54

EPA Region 8 																3-59

EPA Region 9																					3-63

EPA Region 10 																3-68

Potentially Highly Contaminated Sites Not Identified by the NSI Evaluation		3-73

in


-------
Contents (continued)

Pollutant Sources	4-1

Extent of Sediment Contamination by Chemical Class	.							4-2

Major Sediment Contaminant Source Categories „„						4-3

Land Use Patterns and Sediment Contamination					.					4-8

EPA's Point and Nonpoint Source Sediment Contaminant Inventories ...................... 4-14

Conclusions sind Discussion	5-1

Extent of Sediment Contamination			.					........... 5-2

Sources of Sediment Contamination											 5-4

Comparison of NSI Evaluation Results to Results of Previous Sediment

Contamination Studies																 5-4

Comparison of NSI Evaluation Results to Fish Consumption Advisories			5-5

Sensitivity of Selected PCB Evaluation Parameters					5-7

Strengths of the NSI Data Evaluation																5-8

Limitations of the NSI Data Evaluation									5-10

Limitations of Data 															 5-10

Limitations of Approach										 5-12

Recommendations	6-1

Recommendation 1: Further Investigate Conditions in the 96 Targeted Watersheds .. 6-1
Recommendation 2; Coordinate Efforts to Address Sediment Quality Through

Watershed Management Programs									6-2

Recommendation 3; Incorporate a Weight-of-Evidence Approach and

Measures of Chemical Bioavailability into Sediment Monitoring Programs...... 6-2

Recommendation 4; Evaluate the NSI's Coverage and Capabilities and Provide

Better Access to Information in the NSI							.... 6-3

Recommendation 5: Develop Better Monitoring and Assessment Tools		 6-4

Glossary	Glossary-1

Acronyms	Acronyms-1

References		References-1

6

Appendices

A.	Detailed Description of NSI Data									 A-l

B.	Description of Evaluation Parameters Used in the NSI Data Evluation		B-l

C.	Method for Selecting Biota-Sediment Accumulation Factors and Percent
Lipids in Fish Tissue Used for Deriving Theoretical Bioaccumulation
Potentials 																				 C-l

D.	Screening Values for Chemicals Evaluated							 D-l

E.	Cancer Slope Factors and Noncancer Reference Doses Used to Develop

EPA Risk Levels													E-l

F.	Species Characteristics Related to NSI Bi(accumulation Data				F-l

G.	Notes on the Methodology for Evaluating Sediment Toxicity Tests				 G-l

H.	Additional Analyses for PCBs and Mercury 								H-l

I.	NSI Data Evaluation Approach Recommended at the National Sediment
Inventory Workshop, April 26-27, 1994								 1-1

iv


-------
Tables

Table	Page

2-1 Number of Stations Evaluated in the NSI by State			2-5

2-2	NSI Data Evaluation Approach	.	2-9

3-1	National Assessment: Evalutaion of Results for Sampling Stations and River
Reaches by EPA Region				 3-3

3-2 Chemicals or Chemical Groups Most Often Associated With Tier 1 or Tier 2

Sampling Station Classifications					3-7

3-3 Number of Sampling Stations Classified as Tier 1 and Tier 2 Based on Each

Component of the Evaluation Approach					3-11

3-4 USGS Cataloging Unit Numbers and Names for Watersheds Containing APCs	3-14

3-5 River Reaches With 10 or More Tier 1 Sampling Stations Located in

Watershed Containing APCs 										3-17

3-6 Increased Number of Sampling Stations Classified as Her 1 and Her 2 by

Including Wildlife Criteria in the National Assessment			3-20

3-7 Region 1; Evaluation Results for Sampling Stations and River Reaches

by State									3-23

3-8 Region 1: Watersheds Containing Areas of Probable Concern for Sediment

Contamination	3-25

3-9 Region 1: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs					3-25

3-10 Region 1: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications												3-26

3-11 Region 2: Evaluation Results for Sampling Stations and River Reaches

by State			3-28

3-12 Region 2: Watersheds Containing Areas of Probable Concern for Sediment

Contamination			...3-30

3-13 Region 2: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs 								3-30

3-14 Region 2: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications							3-31

3-15 Region 3: Evaluation Results for Sampling Stations and River Reaches

by State					3-33

3-16 Region 3; Watersheds Containing Areas of Probable Concern for Sediment

Contamination					3-35

3-17 Region 3: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs					3-35

v


-------
Tables

I

Tables (continued)

3-18 Region 3: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications				 3-36

3-19 Region 4; Evaluation Results for Sampling Stations and River Reaches

by State			3-38

3-20 Region 4: Watersheds Containing Areas of Probable Concern for Sediment

Contamination							3-40

3-21 Region 4: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs	.....3-41

3-22 Region 4: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications			.	,,.3-42

3-23 Region 5: Evaluation Results for Sampling Stations and River Reaches

by State					3-44

3-24 Region 5: Watersheds Containing Areas of Probable Concern for Sediment

Contamination			3-46

3-25 Region 5: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs 								3-47

3-26 Region 5: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications			3-48

3-27 Region 6: Evaluation Results for Sampling Stations and River Reaches

by State								 3-50

3-28 Region 6: Watersheds Containing Areas of Probable Concern for Sediment

Contamination		 3-52

3-29 Region 6: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs					3-52

3-30 Region 6: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications						 3-53

3-31 Region 7: Evaluation Results for Sampling Stations and River Reaches

by State					3-55

3-32 Region 7: Watersheds Containing Areas of Probable Concern for Sediment

Contamination					3-57

3-33 Region 7; Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs			.3-57

3-34 Region 7: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications	..3-58

3-35 Region 8: Evaluation Results for Sampling Stations and River Reaches

by State					3-60

3-36 Region 8: Chemicals Most Often Associated With Her 1 or Tier 2 Sampling

Station Classifications 				3-62

vi


-------
Tables (continued)

3-37 Region 9: Evaluation Results for Sampling Stations and River Reaches

by State	.............											........	3-64

3-38 Region 9: Watersheds Containing Areas of Probable Concern for Sediment

Contamination									....3-66

3-39 Region 9: Water Bodies With Sampling Stations Classified as Her 1 Located in

Watersheds Containing APCs					»				 3-66

3-40 Region 9: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling

Station Classifications 																3-67

3-41 Region 10: Evaluation Results for Sampling Stations and River Reaches

by State																			....3-69

3-42 Region 10: Watersheds Containing Areas of Probable Concern for Sediment

Contamination									....3-71

3-43 Region 10: Water Bodies With Sampling Stations Classified as Tier 1 Located in

Watersheds Containing APCs													...3-71

3-44 Region 10: Chemicals Most Often Associated With Her 1 or Her 2 Sampling

Station Classifications 												3-72

3-45	Potentially Highly Contaminated Sites Not Identified in the NSI Evaluation		 3-73

4-1	Correlations of Sources to Chemical Classes of Sediment Contaminants	4-4

4-2 Tier 1 and Her 2 Station Classification by Chemical Class and Land Uses in

Watersheds Containing APCs																	4-9

4-3 Comparison of Percent Agricultural Land Use in Watersheds Containing APCs

to Percent of Tier 1 and Tier 2 Stations by Chemical Class					4-13

4-4	Comparison of Percent Urban Land Use in Watersheds Containing APCs

to Percent of Tier 1 and Tier 2 Stations by Chemical Class 						..4-14

5-1	Comparison of Contaminants Most Often Associated With Fish Consumption
Advisories and Those Which Most Often Cause Stations To Be Placed in Tier 1 or
Her 2 Based on the NSI Data Evaluation 											 5-5

5-2 National Sediment Inventory Database: Summary of QA/QC Information			5-11

vii


-------
viii


-------
*¦71 • -

Figures

Figure	Page

2-1 NSI Sediment Sampling Stations Evaluated	2-6

2-2 NSI Tissue Residue Sampling Stations Evaluated	2-7

2-3 NSI Toxicity Test Stations Evaluated			2-8

2-4 Aquatic Life Assessments: Sediment Chemistry Analysis for Organic Chemicals

and Metals Not Included in the AVS Analysis................		2-9

2-5 Aquatic Life Assessments: Sediment Chemistry Analysis for Divalent Metals	2-10

2-6 Aquatic Life Assessments: Sediment Toxicity Analysis						2-10

2-7 Human Health Assessments: Sediment Chemistry and Fish Tissue Residue

Analysis (excluding dioxins and PCBs)							2-11

2-8	Human Health Assessments: PCBs and Dioxin in Fish Tissue Analysis			2-11

3-1	Location of All NSI Sampling Stations			3-2

3-2 Sampling Stations Classified as Tier 1 (Associated Adverse Effects Probable)	3-4

3-3 National Assessment: Percent of River Reaches That Include Tier 1, Tier 2,

and Tier 3 Sampling, Stations						3-5

3-4 National Assessment: Percent of NSI Measurements That Indicate

Potential Risk											3-6

3-5 Sampling Stations Classified as Tier 1 or Tier 2 for Potential Risk to Aquatic Life .. 3-9

3-6 Sampling Stations Classified as Tier 1 or Tier 2 for Potential Risk to Human

Health	3-10

3-7 Watersheds Identified as Containing APCs			3-13

3-8 National Assessment: Watershed Classifications	:					...,3-16

3-9 Sampling Stations Classified as Tier 1 or Her 2 Based on Wildlife Criteria		 3-19

3-10 Region 1: Percent of River Reaches That Include Tier 1, Tier 2, and

Tier 3 Sampling Stations					3-22

3-11 Region 1: Watershed Classifications									3-22

3-12 Region 1: Location of Sampling Stations Classified as Tier 1 or Tier 2 and
Watersheds Containing Areas of Probable Concern for Sediment
Contamination						......3-24

3-13 Region 2: Percent of River Reaches That Include Tier 1, Tier 2, and

Tier 3 Sampling Stations	:			3-27

3-14 Region 2: Watershed Classifications			3-27

3-15 Region 2: Location of Sampling Stations Classified as Her 1 or Her 2 and

Watersheds Containing Areas of Probable Concern for Sediment Contamination... 3-29

3-16 Region 2: Percent of River Reaches That Include Tier 1, and Tier 3

Sampling Stations			3-32

ix


-------
Figures

Figures (continued)

3-17 Region 3: Watershed Classifications			3-32

3-18 Region 3: Location of Sampling Stations Classified as Tier 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination			3-34

3-19 Region 4: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations	.				 3-37

3-20 Region 4: Watershed Classifications	......3-37

3-21 Region 4; Location of Sampling Stations Classified as Tier 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination			3-39

3-22 Region 5: Percent of River Reaches That Include Tier 1, Tier 2, and

Tier 3 Sampling Stations			3-43

3-23 Region 5: Watershed Classifications					3-43

3-24 Region 5: Location of Sampling Stations Classified as Her 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination			3-45

3-2S Region 6: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations					3-49

3-26 Region 6: Watershed Classifications					3-49

3-27 Region 6: Location of Sampling Stations Classified as Tier 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination					...3-51

3-28 Region 7: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations 				3-54

3-29 Region 7: Watershed Classifications	3-54

3-30 Region 7: Location of Sampling Stations Classified as Tier 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination			3-56

3-31 Region 8: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations	3-59

3-32 Region 8: Watershed Classifications		3-59

3-33 Region 8: Location of Sampling Stations Classified as Tier 1 or Tier 2	3-61

3-34 Region 9: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations							3-63

3-35 Region 9: Watershed Classifications				 3-63

3-36 Region 9: Location of Sampling Stations Classified as Tier 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination 				3-65

3-37 Region 10: Percent of River Reaches That Include Tier 1, Tier 2, and

Her 3 Sampling Stations					3-68

x


-------
Nntiomil Sediment Qmilily Survey

Figures (continued)

3-38 Region 10: Watershed Classifications					3-68

3-39 Region 10: Location of Sampling Stations Classified as Her 1 or Tier 2
and Watersheds Containing Areas of Probable Concern for Sediment
Contamination							3-70

3-40	Location of Potentially Highly Contaminated Water Bodies Not Identified in the

NSI Evaluation					3-74

4-1	Average Percent Contamination in Watersheds Containing APCs by

Chemical Class			4-3

4-2 Percent Tier 1 and Tier 2 Stations vs. Agricultural Land Use in APCs	4-13

4-3	Percent Tier 1 and Her 2 Stations vs. Total Urban Land Use in APCs	4-14

5-1	Tier 1 and Tier 2 Sampling Stations for Potential Risk to Human Health Located
Within Water Bodies With Fish Consumption Advisories in Place for the Same
Chemical Responsible for the Tier 1 or Tier 2 Classfieation					5-6

5-2 Sampling Stations Classified as Her 1 or Tier 2 for Potential Risk to Human

Health Excluding PCBs					5-9

xi


-------
('inures


-------
Nafioiiiil St'dimciM Quality Siu \ < y

Acknowledgments

The United States Environmental Protection Agency's Office of Science and Technology (OST)
produced this report with technical support from Tetra Tech, Inc., under EPA Contract Num-
ber 68-C3-0374. Staff from EPA Regional Offices and other headquarters program offices
have participated in this project, provided technical guidance, and reviewed previous reports. We
greatly appreciate their efforts and helpful comments. In particular, we wish to acknowledge Catherine
Fox who served as the initial project officer and oversaw the development of the database and
evaluation methodology. We also wish to express our appreciation to the EPA Regional Office and
State personnel who participated in the review of the preliminary evaluation of sediment chemistry
data, as well as to the persons who participated in national workshops in April 1993 (data compila-
tion) and April 1994 (evaluation methodology).

We wish to offer a special expression of gratitude to several Agency scientists who provided
technical information, guidance, and expert counsel to OST. Gerald Ankley, Phil Cook, David
Hansen, Richard Swartz, and Nelson Thomas provided technical assistance in refining the evalua-
tion methodology. Charles Stephan assisted in the process of reviewing aquatic toxicity literature.
Samual Karickoff and J. Mac Arthur Long reviewed chemical property literature and recommended
organic carbon partitioning coefficients for specific chemicals.

We also wish to acknowledge the following persons who provided external peer review of the
evaluation methodology and its application to the data compiled for this report: William Adams of
Kennecott Utah Copper Corporation in Magna, Utah; Derek Muir of the Freshwater Institute in
Winnipeg, Manitoba; Spyros Pavlou of URS Grenier in Seattle, Washington; Anne Spacie of Purdue
University in Lafayette, Indiana; and William Stubblefield of ENSR Technology in Fort Collins,
Colorado. These persons reviewed the soundness of proposed evaluation methods for intended
purposes, and recommended appropriate and meaningful presentation and interpretation of results.
Chapter 4 and the Executive Summary of this document were not included in prior reviews because
they had not yet been completed. Participation in the review process does not imply concurrence by
these individuals with all observations and recommendations contained in this report.

We greatly appreciate the comments received from various stakeholders during the review of
the July 1996 of this report. Our thanks to all state government officials, trade association represen-
tatives, environmental advocacy professionals, and members of the scientific community who pro-
vided valuable insights. We also wish to acknowledge the consultation with the National Oceanic
and Atmospheric Administration, the United States Army Corps of Engineers, and the United
States Geological Survey.

Finally, we wish to recognize Andrew Zacherle, Jon Harcum, Alex Trounov, Esther Peters, and
all other participating staff and management at Tetra Tech, Inc. for their efforts and professionalism
in providing technical support and data management.

James Keating, Principal Investigator

Thomas Armitage, Acting Chief, Risk Assessment and Management Branch
Elizabeth Southerland, Acting Director, Standards and Applied Science Division

xiii


-------
Acknow lodgments


-------
National Sediment Quality Survey

Executive Summary

T|his report, The Incidence and Severity of Sedi-
ment Contamination in Surface Waters of the
United States, describes the accumulation of
chemical contaminants in river, lake, ocean, and estuary
bottoms and includes a screening assessment of the po-
tential for associated adverse effects on human and envi-
ronmental health. The United States Environmental
Protection Agency (EPA) prepared this report to Con-
gress in response to requirements set forth in the Water
Resources Development Act (WRDA) of 1992, which di-
rected EPA, in consultation with the National Oceanic and
Atmospheric Administration (NOAA) and the U.S. Army
Corps of Engineers (USAGE), to conduct a comprehen-
sive national survey of data regarding the quality of aquatic
sediments in the United States. The Act required EPA to
compile all existing information on the quantity, chemical
and physical composition, and geographic location of
pollutants in aquatic sediment, including the probable
source of such pollutants and identification of those
sediments which are contaminated. The Act further
required EPA to report to the Congress the findings,
conclusions, and recommendations of such survey,
including recommendations for actions necessary to
prevent contamination of aquatic sediments and to
control sources of contamination. The Act also re-
quires EPA to establish a comprehensive and continu-
ing program to assess aquatic sediment quality. As
part of this continuing program, EPA must submit a
national sediment quality report to Congress every 2
years.

To comply with the WRDA mandate, EPA's Office
of Science and Technology (OST) initiated the National
Sediment Inventory (NSI). The NSI is a compilation of
existing sediment quality data; protocols used to evaluate
the data; and various reports and analyses produced to
present the findings, conclusions, and recommendations
for action. EPA produced this first report to Congress in
four volumes:

• Volume 1: National Sediment Quality Sur-
vey—Screening analysis to qualitatively as-
sess the probability of associated adverse
human or ecological effects based on a weight-
of-evidence evaluation

•	Volume 2; Data Summaries for Areas of Prob-
able Concern (APCs)—Sampling station loca-
tion maps and chemical and biological summary
data for watersheds containing APCs

•	Volume 3: National Sediment Contaminant
Point Source Inventory—Screening analysis to
identify probable point source contributors of
sediment pollutants

•	Volume 4: National Sediment Contaminant
Nonpoint Source Inventory—Screening analy-
sis to identify probable nonpoint source con-
tributors of sediment pollutants (in preparation
for subsequent biennial reports)

EPA prepared Volume I, the National Sediment Qual-
ity Survey, to provide a national baseline screening-level
assessment of contaminated sediment over a time period
of the past 15 years. To accomplish this objective, EPA
applied assessment protocols to existing available data
in a uniform fashion. EPA intended to accurately depict
and characterize the incidence and severity of sediment
contamination based on the probability of adverse ef-
fects to human health and the environment. The process
has demonstrated the use of "weight-of-evidence" mea-
sures (including measures of the bioavailability of toxic
chemicals) in sediment quality assessment. Information
contained in this volume may be used to further investi-
gate sediment contamination on a national, regional, and
site-specific scale. Further studies may involve toxico-
logical investigations, risk assessment, analyses of tem-
poral and spatial trends, feasibility of natural recovery,
and source control.

The National Sediment Quality Survey is the first
comprehensive EPA analysis of sediment chemistry and
related biological data to assess what is known about the
national incidence and severity of sediment contamina-
tion. This volume presents a screening-level identifica-
tion of sampling stations in several areas across the
country where sediment is contaminated at levels sug-
gesting an increased probability of adverse effects on
aquatic life and human health. Based on the number and
percentage of sampling stations containing contaminated

xv


-------
I'Aix'utiu'Siitnnuirv

sediment within watershed boundaries, EPA identified a
number of watersheds containing areas of probable con-
cern where additional studies may be needed to draw con-
clusions regarding adverse effects and the need for actions
to reduce risks.

In addition to this and future reports to Congress,
EPA anticipates that products generated through the NSI
will provide managers at the federal, state, and local levels
with information. Many of the NSI data were obtained by
local watershed managers from monitoring programs tar-
geted toward areas of known or suspected contamina-
tion. NSI data and evaluation results can assist local
watershed managers by providing additional data that they
may not have, demonstrating the application of a weight-
of-evidence approach for identifying and screening con-
taminated sediment locations, and allowing researchers
to draw upon a large data set of information to conduct
new analyses that ultimately will be relevant for local as-
sessments.

Description of the NSI Database

The NSI is the largest set of sediment chemistry and
related biological data ever compiled by EPA. It includes
approximately two million records for more than 21,000
monitoring stations across the country. To efficiently
collect usable information for inclusion in the NSI, EPA
sought data that were available in electronic format, repre-
sented broad geographic coverage, and represented spe-
cific sampling locations identified by latitude and longitude
coordinates. The minimum data requirements for inclu-
sion of computerized data in the NSI were monitoring pro-
gram, sampling date, latitude and longitude coordinates,
and measured units. Additional data fields such as sam-
pling method and other quality assurance/quality control
information were retained in the NSI if available, but were
not required for a data set to be included in the NSI.

The NSI includes data from the following data stor-
age systems and monitoring programs:

• Selected data from EPA's Storage and Retrieval
System (STORET)

•	EPA Region 10/USACE Seattle District's Sediment
Inventory

•	EPA Region 9's Dredged Material Tracking Sys-
tem (DMATS)

•	EPA's Great Lakes Sediment Inventory

•	EPA's Environmental Monitoring and Assess-
ment Program (EMAP)

•	United States Geological Survey (Massachusetts
Bay) Data

In addition to sediment chemistry data, the NSI in-
cludes tissue residue, toxicity, benthic abundance, histo-
pathology, and fish abundance data. The sediment
chemistry, tissue residue, and toxicity data were evalu-
ated for this report to Congress. Data from 1980 to 1993
were used in the NSI data evaluation, but older data also
are maintained in the NSI.

Evaluation Approach

The WRDA defines contaminated sediment as
aquatic sediment that contains chemical substances in
excess of appropriate geochemical, toxicological, or sedi-
ment quality criteria or measures; or is otherwise consid-
ered to pose a threat to human health or the environment
The approach used to evaluate the NSI data focuses on
the risk to benthic organisms exposed directly to contami-
nated sediments, and the risk to human consumers of or-
ganisms exposed to sediment contaminants. EPA
evaluated sediment chemistry data, chemical residue lev-
els in edible tissue of aquatic organisms, and sediment
toxicity data taken at the same sampling station (where
available) using a variety of assessment methods.

The following measurement parameters and tech-
niques were used alone or in combination to evaluate the
probability of adverse effects:

Aquatic Life

(1) Comparison of sediment chemistry measurements
to sediment chemistry screening values

•	Draft sediment quality criteria (SQCs)

•	Sediment quality advisory levels (SQALs)

•	Effects range-median (ERM) and effects
range-low (ERL) values

•	NOAA's Coastal Sediment Inventory (COSED)

*	EPA's Ocean Data Evaluation System (ODES)

•	EPA Region 4's Sediment Quality Inventory

*	Gulf of Mexico Program's Contaminated Sedi-
ment Inventory


-------
•	Probable effects levels (PELs) and
threshold effects levels (TELs)

•	Apparent effects thresholds (AETs)

(2)	Comparison of the molar concentration of acid
volatile sulfides ([AVS]) in sediment to the molar
concentration of simultaneously extracted met-
als ([SEM]) in sediment (under equilibrium con-
ditions, sediment with [EVS] greater than [SEM]
will not demonstrate toxicity from metals)

(3)	Lethality based on sediment toxicity data

Human Health

(4)	Comparison of theoretical bioaccumulation
potential (TBP) of measured sediment contami-
nants to:

•	EPA cancer and noncancer risk levels

•	Food and Drug Administration (FDA)
tolerance, action, or guidance values

(5)	Comparison of fish tissue contaminant levels to

•	EPA cancer and noncancer risk levels

•	FDA tolerance, action, or guidance
values

The sediment chemistry screening values used in this
report are not regulatory criteria, site-specific cleanup stan-
dards; or remediation goals. Sediment chemistry screen-
ing values are reference values above which a sediment
ecotoxicological assessment might indicate a potential
threat to aquatic life. For example, independent analyses
of matching chemistry and bioassay data reveal that ERL/
ERMs and TEL/PELs frequently classify samples correctly
either as nontoxic when chemical concentrations are lower
than all these values or as toxic when concentrations ex-
ceed these values. (See Appendix B.) The sediment chem-
istry screening values include both theoretically and
empirically derived values. The theoretically derived
screening values (e.g., SQC, SQAL, [SEM]-[AVS]) rely on
the physical/chemical properties of sediment and chemi-
cals to predict the level of contamination that would not
cause an adverse effect on aquatic life under equilibrium
conditions in sediment. The empirically derived, or cor-
relative, screening values (e.g.,ERM/ERL, PEL/TEL, AET)
rely on paired field and laboratory data to relate incidence
of observed biological effects to the dry-weight sediment
concentration of a specific chemical. Correlative screen-
ing values can relate measured concentration to a prob-
ability of association with adverse effects, but do not

establish cause and effect for a specific chemical. Toxicity
data were used to classify sediment sampling stations
based on their demonstrated lethality to aquatic life in
laboratory bioassays.

Under an assumed exposure scenario, theoretical
bioaccumulation potential (TBP) and tissue residue data
can indicate potential adverse effects on humans from the
consumption of fish that become contaminated through
exposure to contaminated sediment. TBP is an estimate of
the equilibrium concentration (concentration that does
not change with time) of a contaminant in tissues of aquatic
organisms if the sediment in question were the only source
of contamination to the organism. At present, the TBP
calculation can be performed only for nonpolar organic
chemicals. The TBP is estimated from the concentration
of contaminant in the sediment, the organic carbon con-
tent of the sediment, the lipid content of the organism,
and the relative affinity of the chemical for sediment or-
ganic carbon and animal lipid content This relative affin-
ity is measured in the field and is called a biota-sediment
accumulation factor (BSAF, as discussed in detail in Ap-
pendix C). In practice, field measured BSAFs can vary by
an order of magnitude or greater for individual compounds
depending on location and time of measurement. For this
evaluation, EPA selected BSAFs that represents the cen-
tral tendency, suggesting an approximate 50 percent
chance that an associated tissue residue level would ex-
ceed a screening risk value.

Uncertainty is associated with site-specific measures,
assessment techniques, exposure scenarios, and default pa-
rameter selections. Many mitigating biological, chemical,
hydrological, and habitat factors may affect whether sedi-
ment poses a threat to aquatic life or human health. Because
of the limitations of the available sediment quality measures
and assessment methods, EPA characterizes this evaluation
as a screening-level analysis. Similar to a potential human
illness screen, a screening-level analysis should pick up
potential problems and note them for further study. A screen-
ing-level analysis will typically identify many potential prob-
lems that prove not to be significant upon further analysis.
Thus, classification of sampling stations in this analysis is
not meant to be definitive, but is intended to be inclusive of
potential problems arising from persistent metal and organic
chemical contaminants For this reason, EPA elected to evalu-
ate data collected from 1980 to 1993 and to evaluate each
chemical or biological measurement taken at a given sam-
pling station individually. A single measurement of a chemi-
cal at a sampling station, taken at any point in time over the
past 15 years, may have been sufficient to categorize the
sampling station as having an increased probability of asso-
ciation with adverse effects on aquatic life or human health.

xvii


-------
Executive Summary

In this report, EPA associates sampling stations with
their "probability of adverse effects," Each sampling sta-
tion falls into one of three categories, or tiers:

•	Tier 1: associated adverse effects are probable

•	Tier2: associated adverse effects are possible,
but expected infrequently

•	Tier 3: no indication of associated adverse
effects (any sampling station not classified as
Tier 1 or Tier 2; includes sampling stations for
which substantial data were available, as well
as sampling stations for which limited data
were available).

The potential risk of adverse effects on aquatic life and
human health is greatest in areas with a multitude of con-
taminated locations. The assessment of individual sam-
pling stations is useful for estimating the number and
distribution of contaminated spots and overall magnitude
of sediment contamination in monitored waterbodies of the
United States, However, a single "hot spot" might not pose
a great threat to either the benthic community at large or
consumers of resident fish because the spatial extent of
exposure could be small. On the other hand, if many con-
taminated spots are located in close proximity, the spatial
extent and probability of exposure are much greater. EPA
examined sampling station classifications within watersheds
to identify areas of probable concern for sediment contami-
nation (APCs), where the exposure of benthic organisms
and resident fish to contaminated sediment might be more
frequent. In this report, EPA defines watersheds by 8-digit
United States Geological Survey (USGS) hydrologic unit
codes, which are roughly the size of a county. Watersheds
containing APCs are those in which 10 or more sampling
stations were classified as Tier I, and in which at least 75
percent of all sampling stations were categorized as either
Tier lor Tier 2.

The definition of "area of probable concern" was de-
veloped for this report to identify watersheds for which fur-
ther study of the effects and sources of sediment
contamination, and possible risk reduction needs, would be
warranted. Where data have been generated through inten-
sive sampling in areas of known or suspected contamina-
tion within a watershed, the APC definition should identify
watersheds which contain even relatively small areas that
are considerably contaminated. However, this designation
does not imply that sediment throughout the entire water-
shed, which is typically very large compared to the extent of
available sampling data, is contaminated. On the other hand,
where data have been generated through comprehensive
sampling, or where sampling stations were selected randomly

or evenly distributed throughout a sampling grid, the APC
definition might not identify watersheds that contain small
or sporadically contaminated areas. A comprehensively
surveyed watershed of the size typically delineated by a
USGS cataloging unit might contain small but significant
areas that are considerably contaminated, but might be too
large in total area for 75 percent of all sampling stations to be
classified as Tier 1 or Tier 2. Limited random or evenly
distributed sampling within such a watershed also might
not yield 10 Tier 1 sampling stations. Thus, the process
used to identify watersheds containing APCs may both in-
clude some watersheds with limited areas of contamination
and omit some watersheds with significant contamination.
However, given available data EPA believes it represents a
reasonable screening analysis to identify watersheds where
further study is warranted.

Strengths and Limitations

For this report to Congress, EPA has compiled the most
extensive database of sediment quality information currently
available in electronic format. To evaluate these data, EPA
has applied sediment assessment techniques in a weight-
of-evidence approach recommended by national experts.
The process to produce this report to Congress has en-
gaged a broad array of government, industry, academic, and
professional experts and stakeholders in development and
review stages. The evaluation approach uses sediment chem-
istry, tissue residue, and toxicity test results. The assess-
ment tools employed in this analysis have been applied in
North America, with results published in peer-reviewed lit-
erature. Toxicity test data were generated using established
standard methods employed by multiple federal agencies.
The evaluation approach addresses potential impacts on
both aquatic life and human health. Some chemicals pose a
greater risk to human health than to aquatic life; for others,
the reverse is true. By evaluating both potential human
health and aquatic life impacts, EPA has ensured that the
most sensitive endpoint is used to assess environmental
impacts.

Two general types of limitations are associated with
this report to Congress—limitations of the compiled data
and limitations of the evaluation approach. Limitations of
the compiled data include the mixture of data sets derived
from different sampling strategies, incomplete sampling
coverage, the age and quality of data, and the lack of
measurements of important assessment parameters. Limi-
tations of the evaluation approach include uncertainties
in the interpretive tools to assess sediment quality, lack of
quantitative risk assessment that consideres exposure
potentials as well as contamination (e.g., fish consump-
tion rates within APCs for human health risk), and the
subsequent difficulties in interpreting assessment results.

xviii


-------
Nation;)! Sedlineijf Quality Survey

These limitations and uncertainties are discussed in detail
in Chapter 5 of this volume under "Limitations of the NSI
Data Evaluation."

Data compiled for this report were generated using a
number of different sampling strategies. Component sources
contain data derived from different spatial sampling plans,
sampling methods, and analytical methods. Most of the
NSI data were compiled from nonrandom monitoring pro-
grams. Such monitoring programs focus their sampling ef-
forts on areas where contamination is known or suspected
to occur. Reliance on these date is consistent with the stated
objective of this survey: to identify those sediments which
are contaminated. However, one cannot accurately make
inferences regarding the overall condition of the Nation's
sediment, or characterize the "percent contamination," us-
ing the data in die NSI because uncontaminated areas are
most likely substantially underrepresented.

Because this analysis is based only on readily avail-
able electronically formatted data, contamination prob-
lems exist at some locations where data are lacking.
Conversely, older data might not accurately represent cur-
rent sediment contamination conditions. The reliance on
readily available electronic data has undoubtedly excluded
a vast amount of information available from sources such
as local and state governments and published academic
studies. In addition, some data in the NSI were not evalu-
ated because of questions concerning data quality or be-
cause no locational information (latitude and longitude)
was available. NSI data do not evenly represent all geo-
graphic regions in the United States, nor do the data rep-
resent a consistent set of monitored chemicals.

EPA recognizes that sediment is dynamic and that
great temporal and spatial variability in sediment quality
exists. Movement of sediment is highly temporal, and
dependent upon the physical and biological processes at
work in the watershed. Some deposits will redistribute
while others will remain static unless disturbed by extreme
events. Because the data analyzed in this report were
collected over a relatively long period of time, conditions
might have improved or worsened since the sediment was
sampled. Consequently, this report does not definitively
assess the current condition of sediments, but serves as a
baseline for future assessments

The lack of data required to apply some important
assessment parameters hampered EPA's efforts to deter-
mine the incidence and severity of sediment contamina-
tion. For example, the component databases contain a
dearth of total organic carbon (TOC) and acid volatile
sulfide (AVS) measurements relative to the abundance of
contaminant concentration measurements in bulk sedi-

ment. TOC and AVS are essential pieces of information
for interpreting the bioavailability, and subsequent toxic-,
ity, of nonpolar organic and metal contaminants, respec-
tively. In addition, matched sediment chemistry with
toxicity tests, and matched sediment chemistry with tis-
sue residue data, were typically lacking.

It is important to understand both the strengths and
limitations of this analysis to appropriately interpret and
use the information contained in this report. The limita-
tions do not prevent intended uses, and future reports to
Congress on sediment quality will contain less uncertainty.
To ensure that future reports to Congress accurately re-
flect current knowledge concerning the conditions of the
Nation's sediment as our knowledge and application of
science evolve, the NSI will develop into a periodically
updated, centralized assemblage of sediment quality mea-
surements and state-of-the-art assessment techniques.

Findings

EPA evaluated more than 21,000 sampling stations
nationwide as part of the NSI data evaluation. Of the
sampling stations evaluated, 5,521 stations (26 percent)
were classified as Tier 1,10,401 (49 percent) were classi-
fied as Tier 2, and 5,174 (25 percent) were classified as Tier
3. This distribution suggests that state monitoring pro-
grams (accounting for the majority of NSI data) have been
efficient and successful in focusing their sampling efforts
on areas where contamination is known or suspected to
occur. The frequency of Tier 1 classification based on all
NSI data is greater than the frequency of Tier 1 classifica-
tion based on data sets derived from purely random sam-
pling.

The percentage of all NSI sampling stations where
associated effects are "probable" or "possible but expected
infrequently" (i.e„ 26 percent in Tier 1 and 49 percent in
Tier 2) does not represent the overall condition of sedi-
ment across the country: the overall extent of contami-
nated sediment is much less, as is the percentage of
sampling stations where contamination is expected to ac-
tually exert adverse effects. For example, a reasonable
estimate of the national extent of contamination leading to
adverse effects to aquatic life is between 6 and 12 percent
of sediment underlying surface waters (see Chapter 5 for
expanded discussion of "extent of contamination"). This
is primarily because most of the NSI data were obtained
from monitoring programs targeted toward areas of known
or suspected contamination (i.e., sampling stations were
not randomly selected).

The NSI sampling stations were located in 6,744 indi-
vidual river reaches (or water body segments) across the

xix


-------
I \ceuli\i*.Summitry

contiguous United States, or approximately 11 percent of
all river reaches in the country (based on EPA's River
Reach File 1). A river reach can be part of a coastal shore-
line, a lake, or a length of stream between two major tribu-
taries ranging from approximately 1 to 10 miles long. As
depicted in Figure 1, approximately 4 percent of all river
reaches in the contiguous United States had at least one
station categorized as Tier 1, approximately 5 percent of
reaches had at least one station categorized as Tier 2 (but
none as Tier 1), and all of the sampling stations were clas-
sified as Tier 3 in about 2 percent of reaches.

Watersheds containing areas of probable concern for
sediment contamination (APCs) are those that include at
least 10 Tier 1 sampling stations and in which at least 75
percent of all sampling stations were classified as either
Tier 1 or Tier 2. The NSI data evaluation identified 96
watersheds throughout the United States as containing
APCs (Figure 2 and Table 1). (The map numbers listed on
Table 1 correspond to the numbered watersheds identi-
fied in Figure 2.) These watersheds represent about 5
percent of all watersheds in the United States (96 of 2,111).
APC designation could result from extensive sampling
throughout a watershed, or from intensive sampling at a
single contaminated location or a few contaminated loca-
tions. In comparison to the overall results presented on
Figure 1, sampling stations are located on an average of
46 percent of reaches within watersheds containing APCs.
On the average, 30 percent of reaches in watersheds con-
taining APCs have at least one Tier 1 sampling station,
and 13 percent have no Tier 1 sampling
station but at least one Tier 2 sampling
station. In many of these watersheds, the
risk might be concentrated on certain wa-
ter bodies or river reaches. Within the 96
watersheds containing APCs, 57 river
reaches include 10 or more Tier 1 sam-
pling stations. For more detailed informa-
tion concerning individual watersheds
containing APCs, please consult Volume
2 of this report

The evaluation results indicate that
sediment contamination associated with
probable or possible but infrequent ad-
verse effects exists for both aquatic life
and human health. More sampling sta-
tions were categorized as either Tier 1 or
Tier 2 for aquatic life concerns than for
human health concerns. About 41 per-
cent more sampling stations were classi-
fied as Tier 1 for aquatic life (3,287 stations)
than for human health (2,327 stations).
About 60 percent more sampling stations

were categorized as Tier 2 for aquatic life (9,921 stations)
than for human health (6,196 stations).

Recognizing the imprecise nature of some assess-
ment parameters used in this report, Tier 1 sampling sta-
tions are distinguished from Tier 2 sampling stations based
on the magnitude of a contaminant concentration in sedi-
ment, or the degree of corroboration among the different
types of sediment quality measures. In response to un-
certainty in both biological and chemical measures of sedi-
ment contamination, environmental managers must balance
T^pe I errors (false positives: sediment classified as pos-
ing a threat that does not) with Type II errors (false nega-
tives: sediment that poses a threat but was not classified
as such). In screening analyses, the environmentally pro-
tective approach is to minimize Type n errors, which leave
toxic sediment unidentified. To achieve a balance and to
direct attention to areas most likely to be associated with
adverse effects, Tier 1 sampling stations are intended to
have a high rate of "correct" classification (e.g., sediment
definitely posing or definitely not posing a threat) and a
balance between T^pe I and Type II errors. On the other
hand, to retain a sufficient degree of environmental con-
servatism in screening, Tier 2 sampling stations are in-
tended to have a very low number of false negatives in
exchange for a large number of false positives.

To help judge the effectiveness of the evaluation ap-
proach described previously, EPA examined the agreement
between matched sediment chemistry and toxicity test re-

At Least One
Tier 1 Station
4%

At Least One
Tier 2 Station
5%

Tier 3 Stations
2%

No Data Available
89%

Although 77 percent of reaches with
sampling stations include m least one Tier
i or Tier 2 sampling station, if all reaches
in eluded sampling stations this proportion
would Ukely be much smaller because
most available data are from sampling
targeted toward contaminated areas.

Figure 1. National Assessment: Percent of River Reaches That Include
Tier 1, Tier 2, and Tier 3 Sampling Stations.

xx


-------

-------
IXmilh c Summary

I

liable 1. USGS Cataloging Unit Number and Name for Watersheds Containing APCs.

Map #

Cataloging Unit
Number

Cataloging Unit Name

1

1090001

Charles

2

1090002

Cape Cod

3

1090004

Narragansett

4

2030103

Hackensack-Passaic

5

2030104

Sandy Hook-Staten Island

6

2030105

Raritan

7

2030202

Southern Long Island

8

2040105

Middle Delaware-Musconetcong

9

2040202

Lower Delaware

10

2040203

Schuylkill

11

2040301

Mullica-Toms

12

2060003

Guhpowdsr-Patapseo

13

2070004

Conococheague-Opequon

14

3040201

Lower Pee Dee

15

3060101

Seneca

16

3060106

Middle Savannah

17

3080103

Lower St. Johns

18

3130002

Middle Chattahoochee-Lake Harding

19

3140102

Choctawhatchee Bay

20

3140107

Perdido Bay

21

3160205

Mobile Bay

22

4030102

Door-Kewaunee

23

4030108

Menominee

24

4030204

Lower Fox

25

4040001

Little Calumet-Galien

26

4040002

Pike-Root

27

4040003

Milwaukee

28

4050001

St Joseph

29

4060103

Manistee

30

4090002

Lake St. Clair

31

4090004

Detroit

32

4100001

Ottawa-Stony

33

4100002

Raisin

34

4100010

Cedar-Portage

35

4100012

Huron-Vermillion

36

4110001

Black-Rocky

37

4110003

Ashtabula-Chagrin

xxii


-------
Thblel. (Continued)

Map #

Cataloging Unit
Number

Cataloging Unit Name

38

4120101

Chautauqu a-Conneaut

39

4120103

Buffalo-Eighteenmile

40

4120104

Niagara

41

4130001

Oak Orchard-Twelvemile

42

4150301

Upper St. Lawrence

43

5030101

Upper Ohio

44

5030102

Shenango

45

5040001

Tuscarawas

46

5120109 .

Vermilion

47

512011!

Middle Wabash-Busseron

48

6010104

Holston

49

6010201

Watts Bar Lake

50

6010207

Lower Clinch

51

6020001

Middle Tennessee-Chickamauga

52

6020002

Hi was see

53

6030001

Guntersville Lake

54

6030005

Pickwick Lake

55

6040001

Lower Tennessee-Beech

56

6040005

Kentucky Lake

57

7010206

Twin Cities

58

7040001

Rush-Vermillion

59

7040003

Buffalo-Whitewater

60

7070003

Castle Rock

61

7080101

Copperas-Duck

62

7090006

Kishwaukee

63

7120003

Chicago

64

7120004

Des Plaines

65

7120006

Upper Fox

66

7130001

Lower llli nois-Senach wi ne Lake

67

71401001

Cahokia-Joachim

68

7140106

Big Muddy

69

7140201

Upper Kaskaskia

70

7140202

Middle Kaskaskia

71

8010100

Lower Mississippi-Memphis

72

8030209

Deer-Steele

73

8040207

Lower Ouachita


-------
I'Xmit i\ e Siiimiiarv

table 1. (Continued)

Map #

Cataloging Unit
Number

Cataloging Unit Name

74

8080206

Lower Calcasieu

75

8090100

Lower Mississippi-New Orleans

76

10270104

Lower Kansas

77

11070207

Spring

78

11070209

Lower Neosho

79

12040104

Buffalo-San Jacinto

80

17010303

Coeur D'Alene Lake

81

17030003

Lower Yakima

82

17090012

Lower Willamette

83

17110002

Strait of Georgia

84

17110013

Duwamish

85

17110014

Puyallup

86

17110019

Puget Sound

87

18030012

Tuiare-Buena Vista Lakes

88

18050003

Coyote

89

18050004

San Francisco Bay

90

18070104

Santa Monica Bay

91

18070105

Los Angeles

92

18070107

San Pedro Channel Islands

93

18070201

Seal Beach

94

18070204

Newport Bay

95

18070301

Aliso-San Onofre

96

18070304

San Diego

suits for the 805 sampling stations where both data types
are available. The toxicity test data indicate whether signifi-
cant lethality to indicator organisms occurs as a result of
exposure to sediment. Tier 1 classification for aquatic life
effects from sediment chemistry data correctly matched tox-
icity test results for about three-quarters of the sampling
stations, with the remainder balanced between false posi-
tives (12 percent) and false negatives (14 percent). In con-
trast, when Tier 2 classifications from sediment chemistry
data are added in, false negatives drop to less than 1 percent
at the expense of false positives (increases to 68 percent)
and correctly matched sampling stations (drops to 30 per-
cent). This result highlights the fact, already discussed
above, that classification in Tier 2 is very conservative, and
it does not indicate a high probability of adverse effects to
aquatic life. If bioassay test results for chronic toxicity end-
points were included in the NSI evaluation, the rate of false

positives would likely decrease and correctly matched sam-
pling stations would likely increase for both tiers.

Data related to more than 230 different chemicals or
chemical groups were included in the NSI evaluation.
Approximately 40 percent of these chemicals or chemical
groups (97) were present at levels that resulted in classifi-
cation of sampling stations as Tier 1 or Tier 2. The con-
taminants most frequently at levels in fish or sediment
where associated adverse effects are probable include
PCBs (58 percent of the 5,521 Tier 1 sampling stations)
and mercury (20 percent of Tier 1 sampling stations). Pes-
ticides, most notably DDT and metabolites at 15 percent
of Tier 1 sampling stations, and polynuclear aromatic hy-
drocarbons (PAHs) such as pyrene at 8 percent of Tier 1
sampling stations, also were frequently at levels where
associated adverse effects are probable.

xxiv


-------
National Sediment Quality Survey

Dry weight measures of divalent metals other than
mercury (e.g., copper, cadmium, lead, nickel, and zinc) in
sediment were not used to place a sampling station in Tier
1 without an associated measurement of acid volatile sul-
fide, a primary mediator of bioavailability for which data
are not often available in the database. As a result, metals
other than mercury (which also include arsenic, chromium,
and silver) are solely re^onsible for only 6 percent of Tier
1 sampling stations and overlap with mercury or organic
compounds at an additional 6 percent of Tier 1 sampling
stations. In contrast, metals other than mercury are solely
responsible for about 28 percent of the 15,922 Tier 1 and
Tier 2 sampling stations and overlap with mercury or or-
ganic compounds at an additional 28 percent of Tier 1 and
Tier 2 sampling stations. The remaining 44 percent of Tier

1	and Tier 2 sampling stations are classified solely for
mercury or organic compounds.

Two important issues in interpreting the results of
sampling station classification are naturally occurring
"background" levels of chemicals and the effect of chemi-
cal mixtures. Site-specific naturally occurring (or back-
ground) levels of chemicals may be an important risk
management consideration in examining sampling station
classification. This is most often an issue for naturally
occurring chemicals such as metals and PAHs. In addi-
tion, although the sediment chemistry screening levels
for individual chemicals are used as indicators of poten-
tial adverse biological effects, other co-occurring chemi-
cals (which may or may not be measured) can cause or
contribute to observed adverse effects at specific loca-
tions.

Because PCBs were the contaminants most often re-
sponsible for Tier 1 classifications in the NSI evaluation,
and because EPA took a precautionary approach (de-
scribed in Chapter 2) in evaluating the effects of PCB
exposure, the Agency conducted two separate analyses
of PCB data to determine the impact of the precautionary
approach on the overall classification of NSI sampling
stations. EPA first examined the effect of excluding PCBs
entirely from the NSI evaluation. If PCBs were excluded,
the number of Tier 1 stations would be reduced by 42
percent, from 5,521 to 3,209 stations. The number of Tier

2	stations would be increased by 18 percent, from 10,401
to 11,957 stations. This increase reflects the movement
of stations formerly classified as Tier 1 into Tier 2. In the
second PCB evaluation, EPA evaluated the effect on the
overall results of using a less precautionary noncancer
screening value (rather than the cancer screening value)
for predicting human health risk associated with PCB sedi-
ment contamination. When the noncancer screening value
was used, the number of Tier 1 stations decreased by 12

percent, from 5,521 to 4,844 stations, and the number of
Tier 2 stations increased by 4 percent, from 10,401 to
10,802 stations.

Conclusions and Recommendations

The characteristics of the NSI data, as well as the de-
gree of certainty afforded by available assessment tools,
allow neither an absolute determination of adverse effects
on human health or the environment at any location, nor a
determination of the areal extent of contamination on a na-
tional scale. However, the evaluation results strongly sug-
gest that sedi i nent contamination may be significant enough
to pose potential risks to aquatic life and human health in
some locations. The evaluation methodology was designed
for the purpose of a screening-level assessment of sediment
quality; further evaluation would be required to confirm that
sediment contamination poses actual risks to aquatic life or
human health for any given sampling station or watershed.

EPA's evaluation of the NSI data was the most geo-
graphically extensive investigation of sediment contami-
nation ever performed in the United States. The evaluation
was based on procedures to address the probability of
adverse effects on aquatic life and human health. Based
on the evaluation, sediment contamination exists at lev-
els where associated adverse effects are probable (Tier 1)
in some locations within each region and state of the
country. The water bodies affected include streams, lakes,
harbors, nearshore areas, and oceans. At the Tier 1 level,
PCBs, mere ry, organochlorine pesticides, and PAHs are
the most freijuent chemical indicators of sediment con-
tamination.

The results of the NSI data evaluation must be inter-
preted in the context of data availability. Many states and
EPA Regions appear to have a much greater incidence of
sediment contamination than others. To some degree,
this appearance reflects the relative abundance of readily
available electronic data, not necessarily the relative inci-
dence of sediment contamination.

Although the APCs were selected by means of a
screening exercise, EPA believes that they represent the
highest priority for further ecotoxicological assessments,
risk analysis, temporal and spatial trend assessment, con-
taminant source evaluation, and management action be-
cause of the preponderance of evidence in these areas.
Although the procedure for classifying APCs using mul-
tiple samplin g stations was intended to minimize the prob-
ability of iraking an erroneous classification, further
evaluation of conditions in watersheds containing APCs
is necessary because the same mitigating factors that might

xxv


-------
Executive Summary

reduce the probability of associated adverse effects at
one sampling station might also affect neighboring sam-
pling stations.

EPA chose the watershed as the unit of spatial analy-
sis because many state and federal water and sediment
quality management programs, as well as data acquisition
efforts, are centered around this unit. This choice reflects
the growing recognition that activities taking place in one
part of a watershed can greatly affect other parts of the
watershed, and that management efficiencies are achieved
when viewing the watershed holistically. At the same
time, the Agency recognizes that contamination in some
reaches in a watershed does not necessarily indicate that
the entire watershed is affected.

Watershed management is a vital component of
community-based environmental protection. The Agency
and its state and federal partners can address sediment
contamination problems through watershed management
approaches. Watershed management programs focus on
hydrologically defined drainage basins rather than areas
defined by political boundaries. Local management, stake-
holder involvement, and holistic assessments of water
quality are characteristics of the watershed approach. The
National Estuary Program is one example of the water-
shed approach that has led to specific actions to address
contaminated sediment problems. Specifically, the
Narragansett (Rhode Island) Bay, Long Island Sound,
New York/New Jersey Harbor, and San Francisco Bay
Estuary Programs have all recommended actions to re-
duce sources of toxic contaminants to sediment. Numer-
ous other examples of watershed management programs
are summarized in The Watershed Approach: 1993/94
Activity Report (USEPA, 1994g) and A Phase I Inventory
of Current EPA Efforts to Protect Ecosystems (USEPA,
1995b).

Available options for reducing health and environ-
mental risks from contaminated sediment include physical
removal and land disposal; subaqueous capping; in situ
or ex situ biological, physical/chemical, or thermal treat-
ment to destroy or remove contaminants; or natural re-
covery through continuing deposition of clean sediment.
Assuming further investigation reveals the need for man-
agement attention to reduce risks, the preferred means
depends on factors such as the degree and extent of con-
tamination, the value of the resource, the cost of available
options, likely human and ecological exposure, and the
acceptable time period for recovery. If risk managers an-
ticipate a lengthy period of time prior to recovery of the
system, state and local authorities can consider options
such as placing a fish consumption advisory on water

bodies or portions of water bodies where a significant
human health risk exists.

Some of the most significant sources of persistent
and toxic chemicals have been eliminated or reduced
as the result of environmental controls put into place
during the past 10 to 20 years. For example, the com-
mercial use of PCBs and the pesticides DDT and chlo-
rdane has been restricted or banned in the United
States. In addition, effluent controls on industrial and
municipal point source discharges and best manage-
ment practices for the control of nonpoint sources have
greatly reduced contaminant loadings to many of our
rivers and streams.

The feasibility of natural recovery, as well as the
long-term success of remediation projects, depends on the
effective control of pollutant sources. Although most ac-
tive sources of PCBs are controlled, past disposal and use
continue to result in evaporation from some landfills and
leaching from soils. The predominant continuing sources
of organochlorine pesticides are runoff and atmospheric
deposition from past applications on agricultural land. For
other classes of sediment contaminants, active sources con-
tinue to contribute substantial environmental releases. For
example, liberation of inorganic mercury from fuel burning
and other incineration operations continues, as do urban
runoff and atmospheric deposition of metals and PAHs. In
addition, discharge limits for municipal and industrial point
sources are based on either technology-based limits or
state-adopted standards for protection of the water column,
not necessarily for downstream protection of sediment qual-
ity. Determi ing the local and far-field effects of individual
point and nenpoint sources on sediment quality usually
requires site-specific in-depth study.

The primary recommendation of this report to Con-
gress is to encourage further investigation and assess-
ment of contaminated sediment. States, in cooperation
with EPA and other federal agencies, should proceed with
further evaluations of the 96 watersheds containing APCs.
In many cases, it is likely that much additional investiga-
tion and assessment has already occurred, especially in
well-known areas at risk for contamination, and some ar-
eas have been remediated. If active watershed management
programs are in place, these evaluations should be coordi-
nated within he context of current or planned actions. Fu-
ture assessment efforts should focus on areas such as the
57 water body segments located within the 96 watersheds
containing APCs that had 10 or more sampling stations
classified as Tier 1. The purpose of these efforts should be
to gather additional sediment chemistry and related biologi-
cal data, and to conduct further evaluation of data to deter-

xx vi


-------
National Sediment Quality Survey

mine human health and ecological risk, to determine tempo-
ral and spatial trends, to identify potential sources of sedi-
ment contamination and determine whether potential
sources are adequately controlled, and to determine
whether natural recovery is a feasible option for risk re-
duction.

Other recommendations resulting from the NSI evalu-
ation include the following:

•	Coordinate efforts to address sediment quality
through watershed management programs.
Federal, state, and local government agencies
should pool their resources and coordinate their
efforts to address their common sediment con-
tamination issues. These activities should sup-
port efforts such as the selection of future moni-
toring sites, the setting of priorities for
reissuance of National Pollutant Discharge Elimi-
nation System (NPDES) permits and permit syn-
chronization, pollutant trading between non-
point and point sources, and total maximum daily
load (TMDL) development.

•	Incorporate a weight-of-evidence approach
and measures of chemical bioavailability into
sediment monitoring programs. Future moni-
toring programs should specify collection of AVS
and SEM measurements where metals are a con-
cern and site-specific total organic carbon (TOC)
measurements where organic chemicals are a
concern. Future sediment monitoring programs

should also collect tissue residue, biological ef-
fects, and biological community measurements
as well as sediment chemistry measurements.

•	Evaluate the NSI's coverage and capabilities
and provide better access to information in the
NSI. EPA should consider whether to design
future evaluations of NSI data to determine the
temporal trends of contamination and to iden-
tify where and why conditions are improving or
worsening. EPA should consider whether to
expand the NSI to provide more complete na-
tional coverage of sediment quality data. EPA
should also consider increasing the number of
water bodies for evaluation and expanding the
suite of biological and chemical information avail-
able to evaluate each site. EPA should continue
its efforts to make the NSI data and evaluation
results more accessible to other agencies and to
the states.

•	Develop better monitoring and assessment
tools. EPA should continue to update the NSI
evaluation methodology as new assessment
tools become available and the state of the sci-
ence evolves. In the context of the budget pro-
cess, EPA and other federal agencies should
evaluate whether to request funding to support
the development of tools to better characterize
the sources, fate, and effects of sediment con-
taminants.

xxvii


-------
M\wul i\ eSummary

xxviii


-------
National Sediment Quality Surv<-y

Chapter 1

Introduction

What Is The National Sediment
Quality Survey?

The Water Resources Development Act (WRDA)
of 1992 directed the U.S. Environmental Protec-
tion Agency (EPA), in consultation with the
National Oceanic and Atmospheric Administration
(NOAA) and the U.S. Army Corps of Engineers, to con-
duct a comprehensive national survey of data regarding
the quality of sediments in the United States. The Act
required EPA to compile all existing information on the
quantity, chemical and physical composition, and geo-
graphic location of pollutants in aquatic sediment, includ-
ing the probable sources of such pollutants and
identification of those sediments which are contaminated.
The statute defines contaminated sediment as aquatic sedi-
ment that contains chemical substances in excess of ap-
propriate geochemical, toxicological, or sediment quality
criteria or measures, or is otherwise considered to pose a
threat to human health or the environment The Act fur-
ther required EPA to report to the Congress the find-
ings, conclusions, and recommendations of such
survey, including recommendations for actions neces-
sary to prevent contamination of aquatic sediments
and to control sources of contamination. In addition,
the Act requires EPA to establish a comprehensive and
continuing program to assess aquatic sediment quality.
As part of this continuing program, EPA must report to
Congress every 2 years on the assessment's findings.

To comply with the WRDA mandate, EPA's Office of
Science and Technology (OST) initiated the National Sedi-
ment Inventory (NSI). The goals of the NSI are to compile
sediment quality information from available electronic da-
tabases, gather information from available electronic da-
tabases and published reports on sediment contaminant
sources, develop screening-level assessment protocols
to identify potentially contaminated sediment, and pro-
duce biennial reports to Congress on the incidence and
severity of sediment contamination nationwide. The Inci-
dence and Severity of Sediment Contamination in Sur-
face Waters of the United States is the first of these reports
to Congress. To ensure that future reports to Congress
accurately reflect contemporary conditions of the Nation's
sediment as science evolves, the NSI will develop into a

regularly updated, centralized assemblage of sediment qual-
ity measurements and assessment techniques.

The Incidence and Severity of Sediment Contamina-
tion in Surface Waters of the United States is presented
as a four-volume series. This volume, Volume 1: The
National Sediment Quality Survey, presents a national
baseline screening-level assessment of contaminated sedi-
ment over a time period of the past 15 years using a weight-
of-evidence approach. The purpose of The National
Sediment Quality Survey is to depict and characterize the
incidence and severity of sediment contamination based
on the probability of adverse effects to human health and
the environment. Information contained in this volume
may be used to further investigate sediment contamina-
tion on a national, regional, and site-specific scale. Vol-
ume 2 of this series presents data summaries for
watersheds that have been identified in this volume as
containing areas of probable concern for sediment con-
tamination. Volume 3 presents a screening analysis to
identify probable point source contributors of sediment
pollutants. Volume 4 presents a screening analysis to
identify probable nonpoint contributors of sediment pol-
lutants (in preparation for subsequent biannual reports).

For The National Sediment Quality Survey, OST
compiled and analyzed historical data that were collected
from 1980 to 1993 from across the country and are cur-
rently stored in large electronic databases. This effort
required a substantial synthesis of multiple formats and
the coordinated efforts of many federal and state environ-
mental information programs that maintain relevant data.
Published data that have not been entered into databases,
or are not readily available to EPA, are not included in the
NSI at this time and thus were not evaluated for this report
to Congress. As data management systems and access
capabilities continue to improve, EPA anticipates that a
greater amount of data will be readily available in elec-
tronic form.

This report presents the results of the screening-level
assessment of the NSI data. For this assessment, OST
examined sediment chemistry data, associated fish tissue
residue levels, and sediment toxicity test results. The
purpose was to determine whether potential contamina-

1-1


-------
Introduction

tion problems either exist currently or existed over the
past 15 years at distinct monitoring locations. This report
identifies locations where available data indicate that di-
rect or indirect exposure to the sediment could be associ-
ated with adverse effects to aquatic life or human health.
However, because this analysis is based on readily avail-
able electronic data, contamination problems exist at some
locations where data are lacking. Furthermore, because
the data analyzed were collected over a relatively long
period of time, conditions might have improved or wors-
ened since the sediment was sampled. Consequently, this
report does not definitively assess the current overall con-
dition of all sediments across the country, but serves as a
baseline for future assessments, which will include addi-
tional sampling stations, incorporate contemporary data,
and examine trends.

In addition to this and future reports to Congress,
EPA anticipates that products generated through the NSI
will provide managers at the federal, state, and local levels
with information. Many of the NSI data were obtained by
local watershed managers from monitoring programs tar-
geted toward areas of known or suspected contamina-
tion. NSI data and evaluation results can assist local
watershed managers by providing additional data that they
might not have, demonstrating the application of a weight-
of-evidence approach for identifying and screening con-
taminated sediment locations, and allowing researchers
to draw upon a large data set of information to conduct
new analyses that ultimately will be relevant for local as-
sessments.

The National Sediment Quality Survey summarizes
national, regional, and state results from the evaluation of
NSI data. Chapter 1 provides background information
about sediment quality issues. Chapter 2 is an overview
of the assessment methods used to evaluate the NSI data.
Chapter 3 contains the evaluation results on a national,
regional, and state basis. Chapter 4 presents information
on probable sources of sediment contamination, includ-
ing point and nonpoint sources. A discussion of the
results is provided in Chapter 5. Chapter 6 presents rec-
ommendations for evaluating and managing contaminated
sediments. Several appendices present detailed descrip-
tions of both the NSI data and the approach used to
evaluate the data;

A: Detailed Description of NSI Data

B: Description of Evaluation Parameters Used in
the NSI Data Evaluation

C Method for Selecting Biota-Sediment Accumu-
lation Factors and Percent Lipids in Fish

Tissue Used for Deriving Theoretical
Bioaccumulation Potentials

D: Screening Values for Chemicals Evaluated

E: Cancer Slope Factors and Noncancer Refer-
ence Doses Used to Develop EPA Risk Levels

F: Species Characteristics Related to NSI
Bioaccumulation Data

G: Notes on the Methodology for Evaluating
Sediment Toxicity Tests

H: Additional Analyses for PCBs and Mercury

i NSI Data Evaluation Approach Recommended
by the National Sediment Inventory Work-
shop, April 26-27,1994

Why Is Contaminated Sediment An
Important National Issue?

Sediment provides habitat for many aquatic organ-
isms and functions as an important component of aquatic
ecosystems. Sediment also serves as a major repository
for persistent and toxic chemical pollutants released into
the environment. In the aquatic environment, chemical
waste products of anthropogenic (human) origin that do
not easily degrade can eventually accumulate in sedi-
ment. In feet, sediment has been described as the "ulti-
mate sink," or storage place, for pollutants (Salomons et
al., 1987). If that were entirely true, however, we would
not need to be concerned about potential adverse effects
from these "stored" pollutants. Unfortunately, sediment
can function as both a sink and a source for contami-
nants in the aquatic environment.

Adverse effects on organisms in or near sediment
can occur even when contaminant levels in the overlying
water are low. Benthic (bottom-dwelling) organisms can
be exposed to contaminants in sediment through direct
contact, ingestion of sediment particles, or uptake of dis-
solved contaminants present in the interstitial (pore) wa-
ter. In addition, natural and human disturbances can
release contaminants to the overlying water, where pe-
lagic (open-water) organisms can be exposed. Evidence
from laboratory tests shows that contaminated sediment
can cause both immediate lethality (acute toxicity) and
long-term deleterious effects (chronic toxicity) to benthic
organisms. Field studies have revealed other effects, such
as tumors and other lesions, on bottom-feeding fish.
These effects can reduce or eliminate species of recre-
ational, commercial, or ecological importance (such as
crabs, shrimp, and fish) in water bodies either directly or

1-2


-------
National Sediment Quality Survey

by affecting the food supply that sustainable popula-
tions require. Furthermore, sediment contaminants might
not kill the host organism, but might accumulate in edible
tissue to levels that cause health risks to wildlife and
human consumers.

In summary, environmental managers and others are
concerned about sediment contamination and the assess-
ment of sediment quality for the following reasons
(adapted from Power and Chapman in "Assessing Sedi-
ment Quality," 1992):

•	Various toxic contaminants found only in barely
detectable amounts in the water column can
accumulate in sediments to much higher levels.

•	Sediments serve as both a reservoir for contami-
nants and a source of contaminants to the water
column and organisms.

•	Sediments integrate contaminant concentrations
over time, whereas water column contaminant
concentrations are much more variable and dy-
namic.

•	Sediment contaminants (in addition to water
column contaminants) affect bottom-dwelling
organisms and other sediment-associated organ-
isms, as well as both the organisms that feed on
them and humans.

•	Sediments are an integral part of the aquatic en-
vironment that provide habitat, feeding, spawn-
ing, and rearing areas for many aquatic organ-
isms.

Contaminated sediments can affect aquatic life in a
number of ways. Areas with high sediment contaminant
levels can be devoid of sensitive species and, in some
cases, all species. For example, benthic amphipods were
absent from contaminated waterways in Commencement
Bay, Washington (Swartz et al., 1982). In Rhode Island,
the number of species of benthic molluscs was reduced
near an outfall where raw electroplating wastes and other
wastes containing high levels of toxic metals were dis-
charged into Narragansett Bay (Eisler, 1995). In Califor-
nia, pollution-tolerant oligochaete worms dominate the
sediment in the lower portion of Coyote Creek, which
receives urban runoff from San Jose (Pitt, 1995).

Sediment contamination can also adversely affect the
health of organisms and provide a source of contaminants
to the aquatic food chain (Lyman et al., 1987). For ex-
ample, fin rot and a variety of tumors have been found in

fish living above sediments contaminated by polycyclic
aromatic hydrocarbons (PAHs) located near a creosote
plant on the Elizabeth River in Virginia. These impacts
have been correlated with the extent of sediment contami-
nation in the liver (Van Veld et al., 1990). Liver tumors and
skin lesions have occurred in brown bullheads from the
Black River in Ohio, which is contaminated by PAHs from
a coke plant. The authors of the Black River study estab-
lished a cause-and-effect relationship between the pres-
ence of PAHs in sediment and the occurrence of liver
cancer in native fish populations (Baumann et al., 1987).
Examples of risks to fish-eating birds and mammals posed
by contaminated food chains include reproductive prob-
lems in Forster's terns on Lake Michigan near Green Bay
(Kubiak et al., 1989) and on mink farms where mink were
fed Great Lakes fish (Auerlich et al., 1973). In both cases,
high levels of polchlorinated biphenyls (PCBs) in fish were
identified as the cause of the reproductive failures. Con-
taminated sediments can also affect the food chain base
by eliminating food sources and, in some cases, altering
natural competition, which can impact the population dy-
namics of higher trophic levels (Burton et al., 1989; Landis
and Yu, 1995).

The accumulation of contaminants in fish tissue
(called bioaccumulation) and contamination of the food
chain are also important human health and wildlife con-
cerns because people and wildlife eat finfish and shell-
fish. In fact, the consumption of fish represents the most
significant route of aquatic exposure of humans to many
metals and organic compounds (USEPA, 1992a). Most
sediment-related human exposure to contaminants is
through indirect routes that involve the transfer of pollut-
ants out of the sediments and into the water column or
aquatic organisms. Many surface waters have fish con-
sumption advisories or fishing bans in place because of
the high concentrations of PCBs, mercury, dioxin, kepone,
and other contaminants. In 1995, over 1,500 water bodies
in the United States had fish consumption advisories in
place, affecting all but four states. Water supplies also
have been shut down because of contaminated sediments,
and in some places swimming is no longer allowed.

How Significant Is The Problem?

Puget Sound was one of the first areas in the country
to be studied extensively for sediment contamination.
Early studies from the 1980s demonstrated fairly exten-
sive sediment contamination, especially near major indus-
trial embayments (Dexter et al., 1981; Long, 1982; Malins
et al., 1980; Riley et al., 1981). These early assessments
demonstrated that Puget Sound sediments were contami-
nated by many organic and inorganic chemicals, includ-
ing PCBs, PAHs, and metals. Although contaminant

1-3


-------
Introduction

1

concentrations in sediment tended to decrease rapidly
with distance from the nearshore sources, researchers also
documented widespread low-level contamination in the
deepwater sediments of the main basin of Puget Sound
(Ginn and Pastorok, 1982). Also in the 1980s, several kinds
of biological effects, including cancerous tumors, were
reported in organisms from contaminated areas of Puget
Sound (Becker et al., 1987).

Several recent studies conducted in other parts of
the country further illustrate the significance of sediment
contamination and its potential widespread impact. For
example, Myers et al. (1994) investigated the relationships
between hepatic lesions (liver tumors) and stomach con-
tents, liver tissue, and bile in three species of bottom-
dwelling fish captured from 27 urban and nonurban sites
on the Pacific Coast from Alaska to southern California,
as well as the relationship of such lesions to associated
chemical concentrations in sediments. In general, the au-
thors found that lesions were more likely to occur in fish
from sites with higher concentrations of chemical con-
taminants in sediments. Certain lesions had a significantly
higher relative risk of occurrence at urban sites in Puget
Sound, San Francisco Bay, the vicinity of Los Angeles,
and San Diego Bay (Myers et al., 1994). The results of this
study provide strong evidence for the involvement of sedi-
ment contaminants in causing hepatic lesions in bottom
fish and clearly indicate the usefulness of these lesions as
indicators of contaminant-induced effects in fish (Myers
etal., 1994).

Several recent assessments of existing data on the
Nation's marine (saltwater) and freshwater sediments (e.g.,
NRC, 1989) indicate potentially widespread and serious
contamination problems. The NOAA National Status and
Trends Program has monitored coastal sediment contami-
nation since the mid-1980s and has linked elevated pol-
lutant concentrations to the potential for adverse
biological effects in many urban areas, including the
Hudson-Ran tan estuary, Boston Harbor, western Long
Island, and the Oakland estuary of San Francisco Bay
(Long and Morgan, 1990; Power and Chapman, 1992).
The U.S. and Canadian governments have also identified
widespread contaminated sediments in the Great Lakes
(DC, 1987; Fox andTuchman, 1996; Power and Chapman,
1992). The USEPA (1993a) summarizes other recent as-
sessment studies. However, there is still no national-
scale assessment of the incidence and severity of
sediment contamination, particularly in freshwater areas.
This report is the result of EPA's first assessment to de-
termine how significant the problem of sediment contami-
nation is on a national basis.

What Are The Potential Sources Of
Sediment Contamination?

Water bodies usually receive discharges of pollut-
ants as a result of the various human activities, past and
present, that take place nearby. The cumulative effect of
historical, nonpoint, and point sources can contribute to
sediment contamination. A point source is a single, iden-
tifiable source of pollution such as a pipe from a factory
or a wastewater treatment plant. Nonpoint source pollu-
tion is usually carried off the land by storm water runoff
and includes pollutants from agriculture, urban areas,
mining, marinas and boating, construction and other land
modifications, and atmospheric deposition. Many of the
current suspected and documented cases of sediment
contamination are caused by past industrial and agricul-
tural uses of highly persistent and toxic chemicals, such
as PCBs and chlordane. While the use of such chemicals
has since been banned or tightly restricted, monitoring
programs continue to study the extent and severity of
their accumulation in sediment, and subsequently in the
tissues of fish and shellfish. Other potential sediment
contaminants, including heavy metals, PAHs, some pes-
ticides, and existing and new industrial chemicals, con-
tinue to appear in point and nonpoint source releases.
However, significant progress over the past 10 to 15 years,
achieved through industry pollution prevention initia-
tives, National Pollutant Discharge Elimination System
(NPDES) permits, and national technology-based efflu-
ent guideline limitations, has substantially reduced the
discharge of toxic and persistent chemicals. Surficial sedi-
ments are often less contaminated than deeper sediments
indicating improved sediment conditions with reduced
discharges over the past 10 to 15 years.

The characteristics of local sediment contamination
are usually related to the types of land use activities that
take place or have taken place within the area that drains
into the water body (the watershed). For example, har-
bors, streams, and estuaries bordered by industrialized
or urbanized areas tend to have elevated levels of the
metals and organic compounds typically associated with
human activities in these land use areas. Sometimes the
contamination is localized beneath an outfall of industrial
or municipal waste; in other cases, natural mixing pro-
cesses and dredging disperse the pollutants. In addi-
tion, rivers and streams can carry pollutants from upstream
sources into larger downstream water bodies, where they
can contribute further to the problem of sediment con-
tamination. Drifting atmospheric pollutants that are even-
tually deposited in water bodies also contribute to
sediment contamination. For example, EPA estimates that

1-4


-------
76 to 89 percent of PCB loadings to Lake Superior have
come from air pollution (USEPA, 1994a),

Point source releases, including accidental or delib-
erate discharges, have resulted in elevated localized sedi-
ment contamination. Purposeful and accidental
contaminant additions include effluent discharges, spills,
dumping, and the addition of herbicides to lakes and res-
ervoirs, Both industrial mid municipal point sources have
contributed a wide variety of contaminants to sediments.
Municipal point sources include sewage treatment plants
and overflows from combined sewers (which mix the con-
tents of storm sewers and sanitary sewers). Industrial
point sources include manufacturing plants and power-
generating operations.

The pervasiveness of organic and metal compounds
in sediments near urban and agricultural areas and the
association of large inputs of these contaminants with
runoff events tend to support the importance of contami-
nant contributions from nonpoint sources like atmo-
spheric deposition and land drainage. For example, mining
is a significant source of sediment contamination in some
regions, as are runoff and seepage from landfills and
Superfund sites, and urban and agricultural runoff (Baudo
and Muntau, 1990; Canfieldetal., 1994; Hoffman, 1985;
Livingston and Cox, 1985; Ryan and Cox, 1985). Agricul-
tural runoff can contribute selenium, arsenic, and mer-
cury and a wide variety of pesticides. Urban runoff is a
frequently mentioned source of heavy metals and PAHs.

Atmospheric deposition can be one of the major sources
of lead, arsenic, cadmium, mercury, PAHs, DDT and other
organochlorine pesticides, and PCBs in many aquatic en-
vironments (USEPA, 1993c). However, it is often difficult
to determine the portion of these contaminants contrib-
uted by nonpoint versus point source discharges be-
cause the same contaminants can come from both (Baudo
and Muntau, 1990).

Kepone contamination in the James River in Virginia
is an example of historical sediment contamination.

Kepone is a very stable organic compound formerly used
in pesticides. Although active discharges of kepone at
the production site in Hopewell, Virginia, terminated in
1980, high levels of kepone can still be found in the sedi-
ment and finish and shellfish of the lames River down-
stream from the original discharge site (Huggett and
O'Conner, 1988; Nichols, 1990). In fact, a fish advisory
exists on portions of the James River because of high
levels of kepone in tissues of fish taken from the river.
Historical sediment contamination problems such as those
on the James River are often further complicated by on-
going discharge sources. Such historical sediment con-
tamination problems can also slow the natural recovery
of aquatic systems because of the stable nature of the
chemicals responsible for the contamination. Historical
sediment contamination can also cause new problems.
For example, during heavy storms contaminated sedi-
ments can be uncovered, resuspended, and carried down-
stream, where they cause problems in areas that were
previously uncontaminated.

1-5


-------
Introduction

1-6


-------
Chapter 2

Methodology

EPA faced two primary challenges to achieving
the short-term goals of the National Sediment
Inventory (NSI) and fulfilling the mandate of
the Water Resources Development Act (WRDA) of 1992,
as described in the introduction to this report. The first
challenge was to compile a database of consistent sedi-
ment quality measures suitable for all regions of the coun-
try. The second challenge was to identify scientifically
sound methods to determine whether a particular sedi-
ment is "contaminated," according to the definition set
forth in the statute.

In many known areas of contamination, visible and
relatively easy-to-recognize evidence of harmful effects
on resident biota is concurrent with elevated concentra-
tions of contaminants in sediment. In most cases, how-
ever, less obvious effects on biological communities and
ecosystems are much more difficult to identify and are
frequently associated with varying concentrations of sedi-
ment contaminants. In other words, bulk sediment chem-
istry measures arc not always indicative of toxic effect
levels. Similar concentrations of a chemical can pro-
duce widely different biological effects in different sedi-
ments. This discrepancy occurs because toxicity is
influenced by the extent to which chemical contaminants
bind to other constituents in sediment. These other sedi-
ment constituents, such as organic ligands and inorganic
oxides and sulfides, are said to control the bioavailability
of accumulated contaminants. Toxicant binding, or sorp-
tion, to sediment particles suspends the toxic mode of
action in biological systems. Because the binding ca-
pacity of sediment varies, the degree of toxicity exhib-
ited also varies for the same total quantity of toxicant.

The five general categories of sediment quality
measurements are sediment chemistry, sediment tox-
icity, community structure, tissue chemistry, and pa-
thology (Power and Chapman, 1992). Each of these
categories has strengths and limitations for a national-
scale sediment quality assessment. To be efficient in
collecting usable data of similar types, EPA sought
data that were available in electronic format, repre-
sented broad geographic coverage, and represented
specific sampling locations identified by latitude and
longitude coordinates. EPA found sediment chemis-

try and tissue chemistry to be the most widely avail-
able sediment quality measures.

As described above, sediment chemistry measures
might not accurately reflect risk to the environment.
However, EPA has recently developed assessment meth-
ods that combine contaminant concentration with mea-
sures of the primary binding phase to address
bioavailability for certain chemical classes, under as-
sumed conditions of thermodynamic equilibrium
(USEPA, 1993d). Other methods, which rely on statisti-
cal correlations of contaminant concentrations with in-
cidence of adverse biological effects, also exist (Barrick
et al„ 1988; FDEP, 1994; Long et al., 1995). In addi-
tion, fish tissue levels can be predicted using sediment
contaminant concentrations, along with independent field
measures of chemical partitioning behavior and other
known or assigned fish tissue and sediment characteris-
tics. EPA can evaluate risk to consumers from predicted
and field-measured tissue chemistry data using estab-
lished dose-response relationships and standard consump-
tion patterns. Evaluations based on tissue chemistry
circumvent the bioavailability issue while also account-
ing for other mitigating factors such as metabolism. The
primary difficulty in using field-measured tissue chem-
istry is relating chemical residue levels to a specific sedi-
ment, especially for those fish species which typically
forage across great distances.

Sediment toxicity, community structure, and pathol-
ogy measures are less widely available than sediment
chemistry and fish tissue data in the broad-scale elec-
tronic format EPA sought for the NSI. Sediment toxic-
ity data are typically in the form of percent survival,
compared to control mortality, for indicator organisms
exposed to the field-sampled sediment in laboratory bio-
assays (USEPA, 1994b, c). Although these measures
account for bioavailability and the antagonistic and syn-
ergistic effects of pollutant mixtures, they do not ad-
dress possible long-term reproductive or growth effects,
nor do they identify specific contaminants responsible
for observed lethal toxicity. Indicator organisms also
might not represent the most sensitive species. Com-
munity structure measures, such as fish abundance and
benthic diveisity, and pathology measures are potentially

2-1


-------
Mvlliodol»j>>				-	I

indicative of long-term adverse effects, yet there are a
multitude of mitigating physical, hydrologic, and bio-
logical factors that might not relate in any way to chemi-
cal contamination.

The ideal assessment methodology would be based
on matched data sets of all five types of sediment qual-
ity measures to take advantage of the strengths of each
measurement type and to minimize their collective
weaknesses. Unfortunately, such a database does not
exist on a national scale, nor is it typically available on
a smaller scale. Based on the statutory definition of
contaminated sediment in the WRDA, EPA can iden-
tify locations where sediment chemistry measures ex-
ceed "appropriate geochemical, toxicological, or
sediment quality criteria or measures." Again based
on the statutory definition, EPA can also use tissue chem-
istry and sediment toxicity measures to identify aquatic
sediments that "otherwise pose a threat to human health
or the environment" because there are either screening
values (e.g., EPA risk levels for fish tissue consump-
tion) or control samples for comparison. However, EPA
believes it cannot accurately evaluate community struc-
ture or pathology measures to identify contaminated
sediment, based on the statutory definition, without first
identifying appropriate reference conditions to which
measured conditions could be compared.

For this analysis, EPA evaluated sediment chemis-
try, tissue chemistry, and sediment toxicity data, taken
at the same sampling station, individually and in com-
bination using a variety of assessment methods. Be-
cause of the limitations of the available sediment quality
measures and assessment methods, EPA characterizes
this identification of contaminated sediment locations
as a screening-level analysis. Similar to a potential hu-
man illness screen, a screening-level analysis should
pick up potential problems and note them for further
study. A screening-level analysis will typically identify
many potential problems that prove not to be signifi-
cant upon further analysis. Thus, classification of sam-
pling stations in this analysis is not meant to be
definitive, but is intended to be inclusive of potential
problems arising from presistent metal and organic
chemical contaminants. For this reason, EPA elected
to evaluate data collected from 1980 to 1993 and to
evaluate each chemical or biological measurement taken
at a given sampling station individually. A single mea-
surement of a chemical at a sampling station, taken at
any point in time over the past 15 years, may have been
sufficient to classify the sampling station as having an
increased probability of association with adverse effects
to aquatic life or human health.

EPA recognizes that sediment is dynamic and that
great temporal and spatial variability in sediment qual-
ity exists. This variability can be a function of sam-
pling (e.g., a contaminated area might be sampled one
year, but not the next) or a function of natural events
(e.g., floods can move contaminated sediment from one
area to another, or can bury contaminated sediment).
Movement of sediment is highly temporal, and depen-
dent upon the physical and biological processes at work
in the watershed. Some deposits will redistribute while
others will remain static unless disturbed by extreme
events.

In this report, EPA associates sampling stations with
their "probability of adverse effects on aquatic life or
human health." Each sampling station falls into one of
three categories (tiers): associated adverse effects are
probable (Tier 1); associated adverse effects are possible,
but expected infrequently (Tier 2); or no indication of
associated adverse effects (Tier 3). A Tier 3 sampling
station classification does not neccesarily imply a zero
or minimal probability of adverse effects, only that avail-
able data (which may be substantial or limited) do not
indicate an increased probability of adverse effects. Rec-
ognizing the imprecise nature of the numerical assess-
ment parameters, Tier 1 sampling stations are
distinguished from Tier 2 sampling stations based on
the magnitude of a sediment chemistry measure or the
degree of corroboration among the different types of sedi-
ment quality measures.

The remainder of this chapter presents a short his-
tory of how EPA developed the NSI, a brief description
of the NSI data, and an explanation of the NSI data evalu-
ation approach.

Background

EPA initiated work several years ago on the devel-
opment of the NSI through pilot inventories in EPA Re-
gions 4 and 5 and the Gulf of Mexico Program. Based
on lessons learned from these three pilot inventories,
the Agency developed a document entitled Framework
for the Development of the National Sediment Inven-
tory (USEPA, 1993a), which describes the general for-
mat for compiling sediment-related data and provides a
brief summary of sediment quality evaluation techniques.
The format and overall approach were then presented,
modified slightly, and agreed upon at an interagency
workshop held in March 1993 in Washington, DC. Fol-
lowing the workshop, EPA began compiling and evalu-
ating data for the NSI. Data from several national and
regional databases were included as part of the effort.

2-2


-------
In the spring of 1994, EPA conducted a prelimi-
nary evaluation of NSI sediment chemistry data only.
The purpose of the assessment was to identify sampling
stations throughout the United States where measured
values of sediment pollutants exceeded sediment chem-
istry levels of concern. The results of that assessment
were then distributed to the EPA Regional offices for
their review. The Regional offices were asked to review
the preliminary evaluation and to:

•	Verify sampling stations targeted as areas of
concern.

•	Identify sampling stations that might be incor-
rectly targeted as areas of concern.

•	Identify potential areas of concern that were not
targeted, but should have been.

•	Inform EPA Headquarters of additional sedi-
ment quality data that should be included in
the NSI to make the inventory more accurate
and complete.

The EPA Regional offices completed their review of
the preliminary evaluation during the winter of 1994-
95. Regional comments on the results of the prelimi-
nary evaluation were incorporated into the NSI database.
EPA will add new data sets identified by the Regions to
the NSI and include them in the national assessment for
future reports to Congress.

In April 1994, EPA Headquarters held the Second
National Sediment Inventory Workshop (USEPA, 1994d).
The purpose of this workshop was to bring together ex-
perts in the field of sediment quality assessment to rec-
ommend an approach for integrating and evaluating the
sediment chemistry and biological data contained in the
NSI. The final approach recommended by workshop par-
ticipants provided the basis for the final approach adopted
to evaluate NSI data for this report to Congress. Appen-
dix I of this report provides a brief description of the
workshop approach and a list of attendees.

Description of NSI Data

The NSI includes data from the following data stor-
age systems and monitoring programs:

•	Selected data sets from EPA's Storage and Re-
trieval System (STORET) (69 percent of sam-
pling stations)

-	U.S. Army Corps of Engineers (USAGE)

-	U.S. Geological Survey (USGS)

-	EPA

-	States

•	NOAA's Coastal Sediment Inventory (COSED)
(5 percent of sampling stations)

•	EPA's Ocean Data Evaluation System (ODES)
(6 percent of sampling stations)

•	EPA Region 4's Sediment Quality Inventory (5
percent of sampling stations)

•	Gulf of Mexico Program's Contaminated Sedi-
ment Inventory (1 percent of sampling stations)

•	EPA Region 10/COE Seattle District's Sedi-
ment Inventory (8 percent of sampling stations)

•	EPA Region 9's Dredged Material Tracking
System (DMATS) (1 percent of sampling sta-
tions)

•	EPA's Great Lakes Sediment Inventory (less
than 1 percent of sampling stations)

•	EPA's Environmental Monitoring and Assess-
ment Program (EMAP) (2 percent of sampling
stations)

•	USGS (Massachusetts Bay) Data (3 percent of
sampling stations)

Although EPA elected to evaluate data collected since
1980 (i,e., 1980-93), data from before 1980 are still main-
tained in the NSI. At a minimum, EPA required that
electronically available data include monitoring program,
sampling date, latitude and longitude coordinates, and
measured units for inclusion in the NSI. Additional data
fields providing details such as sampling method or other
quality assurance/quality control information were re-
tained in the NSI if available. Additional information
about available data fields ami NSI component databases
is presented in Appendix A of this report.

The types of data contained in the NSI include the
following:

•	Sediment chemistry: Measurement of the
chemical composition of sediment-associated
contaminants.

•	Tissue residue: Measurement of chemical con-
taminants in the tissues of organisms.

2-3


-------
•	Benthic abundance: Measurement of the num-
ber and types of organisms living in or on sedi-
ments.

•	Toxicity: Measurement of the lethal or suble-
thal effects of contaminants in environmental
media on various test organisms.

•	Histopathology: Observation of abnormalities
or diseases in tissue (e.g., tumors).

•	Fish abundance: Measurement of the number
and types of fish found in a water body.

The NSI represents a compilation of environmental
monitoring data from a variety of sources. Most of the
component databases are maintained under known and
documented quality assurance and quality control proce-
dures. However, EPA's STORET database is intended to
be a broad-based repository of data. Consequently, the
quality of the data in STORET, both in terms of database
entry and analytical instrument error, is unknown and
probably varies a great deal depending on the quality
assurance management associated with specific data sub-
mittals.

Inherent in the diversity of data sources are contrast-
ing monitoring objectives and scope. Component sources
contain data derived from different spatial sampling
plans, sampling methods, and analytical methods. For
example, most data from EPA's EMAP program repre-
sent sampling stations that lie on a standardized grid
over a given geographic area, whereas data in EPA's
STORET most likely represent state monitoring data
sampled from locations near known discharges or thought
to have elevated contaminant levels. In contrast, many
of the National Status and Trends Program data in
NOAA's COSED database represent sampling stations
purposely selected because they are removed from known
discharges. However, many other sampling stations in
the COSED database were located within highly urban-
ized bays and estuaries where chemical contamination
was expected. These sampling stations include data from
regional bioeffects assesments in which NOAA exam-
ined sediment quality in several highly urbanized areas.
These surveys were region-wide assessments, not point
source or end-of-pipe studies.

From an assessment point of view, STORET data
might be useful for developing a list of contaminated
sediment locations, but might overstate the general ex-
tent of contaminated sediment in the Nation by focusing
largely on areas most likely to be problematic. On the
other hand, analysis of EMAP data might result in a more

balanced assessment in terms of the mix of contaminated
sampling stations and uncontaminated sampling stations.
Approximately two-thirds of sampling stations in the NSI
are from the STORET database. Reliance on these data
is consistent with the stated objective of this survey: to
identify those sediments which are contaminated. How-
ever, one cannot accurately make inferences regarding
the overall condition of the Nation's sediment, or char-
acterize the "percent contamination," using the data in
the NSI because uncontaminated areas are most likely
substantially underrepresented.

NSI data do not evenly represent all geographic re-
gions in the United State, nor do the data represent a
consistent set of monitored chemicals. For example, sev-
eral of the databases are targeted toward marine envi-
ronments or other geographically focused areas. Table
2-1 presents the number of stations evaluated per state.
More than 50 percent of all stations evaluated in the NSI
ate located in Washington, Florida, Illinois, California,
Virginia, Ohio, Massachusetts, and Wisconsin. Each of
these states has more than 700 monitoring stations. Other
states of similar or larger size (e.g., Georgia, Pennsylva-
nia) have far fewer sampling stations with data for evalu-
ation. Figures 2-1, 2-2, and 2-3 depict the location of
monitoring stations with sediment chemistry, tissue resi-
due, and toxicity data, respectively. Individual stations
may vary considerably in terms of the number of chemi-
cals monitored. Some stations have data that represent a
large number of organic and inorganic contaminants,
whereas others have measured values for only a few
chemicals. Thus, the inventory cannot be considered
comprehensive even for locations with sampling data.
The reliance on readily available electronic data has un-
doubtedly led to exclusions of a vast amount of informa-
tion available from sources such as local and state
governments and published reports. Other limitations,
including data quality issues, are discussed in Chapter 5
of this report.

NSI Data Evaluation Approach

The methodology developed for classifying sampling
stations according to the probability of adverse effects on
aquatic life and human health from sediment contami-
nation relies on measures of sediment chemistry, sedi-
ment toxicity, and contaminant residue in tissue.
Although the NSI also contains benthic abundance, his-
topathology, and fish abundance data, these types of data
were not used in the evaluation. Benthic and fish abun-
dance Cannot be directly associated with sediment con-
tamination based on the statutory definition and currently
available assessment tools, and available fish liver histo-
pathology data were very limited.

2-4


-------
Table 2-1. Number of Stations Evaluated in the NSI by State

Region 1

Connecticut

98

Region 6

Arkansas

107



Maine

55



Louisiana

460



Massachusetts

895



New Mexico

101



New Hampshire

7



Oklahoma

286



Rhode Island

42



Texas

662



Vermont

5







Region 2

New Jersey

448

Region 7

Iowa

228



New York

618



Kansas

203



Puerto Rico

30



Missouri
Nebraska

327
253

Region 3

Delaware

218

Region 8

Colorado

202



District of Columbia

4



Montana

38



Maryland

206



North Dakota

161



Pennsylvania

311



South Dakota

43



Virginia

1,051



Utah

47



West Virginia

120



Wyoming

44

Region 4

Alabama

477

Region 9

Arizona

124



Florida

1,776



California

1,443



Georgia

318



Hawaii

36



Kentucky

249



Nevada

96



Mississippi

318



,





North Carolina

612









South Carolina

563









Tennessee

646







Region 5

Illinois

1,669

Region 10

Alaska

267



Indiana

108

-

Idaho

95



Michigan

402



Oregon

291



Minnesota

438



Washington

2,225



Ohio

970









Wisconsin

703







The approach used to evaluate the NSI data focuses
on the protection of benthic organisms from exposure to
contaminated sediments and the protection of humans from
the consumption of fish that bioaccumulate contaminants
from sediment. In addition, potential effects on wildlife
from fish consumption were also evaluated. The wildlife
results were not included in the overall results of the NSI
data evaluation; however, they are presented separately.
Table 2-2 presents the classification scheme used in the
evaluation of the NSI data. Each component, or evalua-
tion parameter, of the classification scheme is numbered
on Table 2-2. Each evaluation parameter is discussed un-
der a section heading cross-referenced to these numbers.
Figures 2-4 through 2-8 depict the evaluation parameters
and sampling station classifications in flowchart format.

EPA analyzed the NSI data by evaluating each param-
eter in Table 2-2 on a measurement-by-measurement and
sampling station-by-sampling station basis. Each sampling
station was associated with a "probability of adverse ef-

fects" by combining parameters as shown
in Table 2-2 and Figures 2-4 through 2-8.
Because each individual measurement was
considered independently (except for diva-
lent metals, whose concentrations were
summed), a single observation of elevated
concentration could place a sampling sta-
tion into Tier I, (associated adverse effects
are probable). In general, the methodol-
ogy was constructed such that a sampling
station classified as Her 1 must be repre-
sented by a relatively large set of data or by
a highly elevated sediment concentration
of a chemical whose effects screening level
is well characterized based on multiple as-
sessment techniques. Fewer data were re-
quired to classify a sampling station as Tier
2. Any sampling station not meeting the
requirements to be classified as Tier 1 or
Tier 2 was classified as Tier 3. Sampling
stations in this category include those for
which substantial data were available with-
out evidence of adverse effects, as well as
sampling stations for which limited data
were available to determine the potential
for adverse effects.

Individual evaluation parameters, ap-
plied to various measurements indepen-
dently, could lead to different site
classifications. If one evaluation param-
eter indicated Tier 1, but other evaluation
parameters indicated Tier 2 or Her 3, a Her
1 classification was assigned to the sam-
pling station. For example, if a sampling station was cat-
egorized as Tier 2 based on all sediment chemistry data,
but was categorized as Her 1 based on toxicity data, the
station was placed in Tier 1. This principle also applies to
evaluating multiple contaminants within the same evalua-
tion parameter. For example, if the evaluation of sediment
chemistry data placed a sampling station in Her 1 for met-
als and in Tier 2 for PCBs, the station was placed in Her 1.

Recognizing the imprecise nature of some assessment
parameters used in this report. Tier 1 sampling stations are
distinguished from Tier 2 sampling stations based on the
magnitude of a contaminant concentration in sediment, or
the degree of corroboration among the different types of
sediment quality measures. In response to uncertainty in
both biological and chemical measures of sediment con-
tamination, environmental managers must balance Type I
errors (false positives: sediment classified as posing a threat
that does not) with Type II errors (false negatives: sedi-
ment that poses a threat but was not classified as such). In

2-5


-------

-------

-------

-------
Table 2-2. NSI Data Evaluation Approach (with numbered parameters)

Category of Sampling
Station
OmsHI cation*

Data U*ed to Determine OaaiifkatiQm

Sediment Otenjistiy



Tbtue Residue



Toxicity

Her 1:

Associated Adverse
Effects to Aquatic Life or
Human Heakh are
Probable

Sediment chemistry value*
exceed draft sediment quality
criteria far any one of the 6ve
cftetricais for which criteria have
beea developed by EPA (must
have measured TOO

1

OR

[SEM]*[AVS>5 for the sum of
molar ooacestratbns of Cd. Cu,
Ni Pb, and Zo*

2

OR

Sediment chemistry vsbes
exceed two or more of the
relevant qpper screening valuer
CERMj, AHTi (h£h), PELs,
SQALs, SQCs) for any one
chemical (other than Cd, Cu, Hi,
Pb, and Za) (can use de&uk
TOO

3

OR

Ussue lewis of dioxin or PCBs
m resklent species exceed EPA
risk levels
8

OR

Tbxfciy demonstrated by two or
more nosnicrobial scute toxicity
tests using two dM&reot specks
(one of which most be a sofld-
phase test)

11

OR

Sediment chemistry TDP
exceeds PDA levels or BPA rtik
leveb 4

AND

Tissue levels in reatient spec lot
exceed FDA toveb or EPA Hik
levels
9





Iter Z\

Associated Advene
Eflfecu to Aquatic LSe or
Hurron Heakh are
Poaifaie, but Expected

IreShsqtieraly

!SEM|.(AVS 1« 0 so 5 for the
cum of molar concenraUora of
Cd, Cu. Ni, Pb, and Zn

5

OR

Sedlrart chemistry values
exceed anyone of the relevant
lower screening vakxu OBRL),
AffI* (tow), H3L», SQALf,
SQCs) for any one chemical
(can use default TOC)

6

OR

Sediment chemistry IBP
exceed! FDA bveb or EPA
risk kveb ?

OR

Ibsuo levels in resident ipcctsi
exceed PDA levels or EPA risk
bvets
10

OR

Tbxfcly demonstrated by a
singfe««pocfes nonzntcrobiat
toxic ky test
12

Her 3:

No Indication of
Aucockled Adverse
Efiects

Any aanpfeag station not categorised as Her 1 or Tier 2. Available data (wftch imy be very limited or quite extensive) do
not indicate a bkefchood of adverse efiects to aquatic Bfe or hurmn heakh.

"Metals: Cd = cadmium. Oi = copper, Ni = nickel, Pb = lead, Za = zinc.

Does the dumieal
haveadnA SQCi

1

Wu TOC measured



tor the SBjiplinf imJon?

Wax TOC memir*d
for the sampling station!

Use measured TOC
value to determine TOC
rvocmaJ&ed chemtad
concentration for
comparijon with SQALs

Use measured TOC
value to deoermbe TOC

rtorma&ied chemical 	

concentration for
companion wfch draft SQC»

Utede6diTOCc(IX
to determirxs TOC
ramaiktd ch«nkal
concentration for
COmparocn with draft
SQCs and SQALt

Exceeded one or more
tower smwirig values

Unless catpjoraed bjr another parameter i

Figure 2-4. Aquatic Life Assessments: Sediment Chemistry Analysis for

Organic Chemicals and Metals Not Included in the AVS Analysis,


-------
Mi'thodology

I

Figure 2-5. Aquatic Life Assessments: Sediment Chemistry Analysis for
Divalent Metals.

Wu a toxicfty	1 f3

test performed? I

Wis toxicity demonstrated u*lng 2 or
mora nonmferobbl toxicity tens
u$fng 2 different spades (one of **hlch
wm *iotld-phue teit)?

		-v


-------
National Sediment Quality Survey

Unless categorized by another parameter

Figure 2-7. Human Health Assessments: Sediment Chemistry and Fish
Tissue Residue Analysis (excluding dioxins and PCBs).

Unless categorized by another parameter

Figure 2-8. Human Health Assessments: PCBs and Dioxin in Fish Tissue
Analysis.

2-11


-------
Mi'tlio(lnli><>\

screening analyses, the environmentally protective approach
is to minimize Type II errors, which leave toxic sediment
unidentified. To achieve a balance and to direct attention
to areas most likely to be associated with adverse effects,
Tier 1 sampling stations are intended to have a high rate of
"correct" classification (e.g., sediment definitely posing or
definitely not posing a threat) and a balance between Type
I and Type II errors. On the other hand, to retain a suffi-
cient degree of environmental conservatism in screening,
Tier 2 sampling stations are intended to have a very low
number of false negatives in exchange for a large number
of false positives.

The numbered evaluation parameters used in the NSI
data evaluation are briefly described below. A detailed de-
scription of the evaluation parameters is presented in Ap-
pendix B.

Sediment Chemistry Data

The sediment chemistry screening values used in this
report are not regulatory criteria, site-specific cleanup stan-
dards, or remediation goals. Sediment chemistry screen-
ing values are reference values above which a sediment
ecotoxicological assessment might indicate a potential
threat to aquatic life. The sediment chemistry screening
values used to evaluate the NSI data for potential adverse
effects of sediment contamination on aquatic life include
both theoretically and empirically based values. The theo-
retically based values rely on the physical/chemical prop-
erties of sediment and chemicals to predict the level of
contamination that would not cause an adverse effect on
aquatic life. The empirically based, or correlative, screen-
ing values rely on paired field and laboratory data to relate
incidence of observed biological effects to the dry-weight
sediment concentration of a specific chemical.

The theoretically based screening values used as pa-
rameters in the evaluation of NSI data include the sedi-
ment quality criteria, sediment quality advisory levels, and
comparison of simultaneously extracted metals to acid-vola-
tile sulfide concentrations. Empirically based, correlative
screening values used in the NSI evaluation include the
effects range-median/effects range-low values, probable ef-
fects levels/threshold effects levels, and apparent effects
thresholds. The use of each of these screening values in
the evaluation of the NSI data is described below. Another
theoretically based evaluation parameter, the theoretical
bioaccumulation potential (which was used for human
health assessments), is also described below. The limita-
tions associated with the use of these screening values are
discussed in Chapter 5.

Sediment Chemistry Values Exceed EPA Draft
Sediment Quality Criteria [1]

This evaluation parameter was used to assess the po-
tential effects of sediment contamination on benthic spe-
cies. EPA has developed draft sediment quality criteria
(SQCs) for the following five nonionic organic chemicals:

•	Acenaphthene (polynuclear aromatic
hydrocarbon, or PAH)

•	Dieldrin (pesticide)

•	Endrin (pesticide)

•	Fluoranthene (PAH)

•	Phenanthrene. (PAH)

EPA developed these draft criteria using the equi -
librium partitioning (EqP) approach (described in de-
tail in Appendix B) for linking bioavailability to toxicity.
The EqP approach involves predicting the dry-weight
concentration of a contaminant in sediment that is in
equilibrium with a pore water concentration that is pro -
tective of aquatic life. It combines the water-only ef-
fects concentration (the chronic water quality criteria)
and the organic carbon partitioning coefficient of the
chemical normalized to the organic carbon content of
the sediment. The draft criterion is compared to the
measured dry-weight sediment concentration of the
chemical normalized to sediment organic carbon con-
tent. If the organic-carbon-normalized concentration
of the contaminant does not exceed the draft sediment
quality criterion, adverse effects should not occur to at
least 95 percent of benthic organisms. The draft SQCs
are based on the highest quality data available, which
have been reviewed extensively.

For the NSI data evaluation, sediment chemistry mea-
surements with accompanying measured total organic car-
bon (TOC) values can place a site in Tier 1 based exclusively
on a comparison with a draft SQC. The amount of TOC in
sediment is one of the factors that determines the extent to
which a nonionic organic chemical is bound to the sedi-
ment and, thus, the availability for uptake by organisms
(bioavailability). If draft SQCs based on measured TOC
were not exceeded, or if none of the five nonpolar
organic chemicals that have been assigned draft SQC
values were measured, the sampling station was classi-
fied as Tier 3 unless otherwise categorized by another
parameter. Appendix B discusses the assumptions

2-12


-------
I

National Sediment Quality Survey

and limitations associated with the use of draft SQCs.
If a sample for any of the five contaminants for which
draft SQCs have been developed did not have accompa-
nying TOC data, the measured concentration was com-
pared to the draft SQC based on a default TOC value of
1 percent. In these instances, the draft SQC was treated
like other sediment quality screening values described
later in this section.

The assumption that the percent TOC for samples
without measured TOC is equal to 1 percent is based on
a review of values published in the literature. TOC can
range from 0.1 percent in sandy sediments to 1 to 4 per-
cent in silty harbor sediments and 10 to 20 percent in
navigation channel sediments (Clarke and McFarland,
1991). Long et al. (1995) reported an overall mean TOC
concentration of 1.2 percent from data compiled from
350 publications for their biological effects database for
marine and estuarine sediments. Ingersoll et al. (1996)
reported a mean TOC concentration of 2.7 percent for
inland freshwater samples. Based on this review of TOC
data, EPA selected a default TOC value of 1 percent for
the NSI evaluation. Consistent with the screening level
application, this value should not lead to an underesti-
mate of the bioavailability of associated contaminants
in most cases.

Comparison ofAVS to SEM Molar Concentrations
[2, 5]

The use of the total concentration of a trace metal
in sediment as a measure of its toxicity and its ability to
bioaccumulate is problematic because different sediments
exhibit different degrees of bioavailability for the same
total quantity of metal (Di Toro et al., 1990; Luoma,
1983). These differences have recently been reconciled
by relating organism toxic response (mortality) to the
metal concentration in the sediment interstitial water
(Adams et al., 1985; Di Toro et al., 1990). Acid-vola-
tile sulfide (AVS) is one of the major chemical compo-
nents that control the activities and availability of metals
in interstitial waters of anoxic (lacking oxygen) sedi-
ments (Meyer et al., 1994).

A large reservoir of sulfide exists as iron sulfide in
anoxic sediment. Sulfide will react with several diva-
lent transition metal cations (cadmium, copper, mercury,
nickel, lead, and zinc) to form highly insoluble com-
pounds that are not bioavailable (Allen et al., 1993). It
follows in theory, and with verification (Di TorO et al.,
1990), that divalent transition metals will not begin to
cause toxicity in anoxic sediment until the reservoir of
sulfide is used up (i.e., the molar concentration pf met-
als exceeds the molar concentration of sulfide), typically

at relatively high dry-weight metal concentrations. This
observation has led to a laboratory measurement tech-
nique of calculating the difference between simulta-
neously extracted metal (SEM) concentration and acid
volatile sulfide concentration from field samples to de-
termine potential toxicity.

To evaluate the potential effects of metals on benthic
species, the molar concentration of AVS ([AVS]) was
compared to the sum of SEM molar concentrations
([SEM]) for five metals: cadmium, copper, nickel, lead,
and zinc. Mercury was excluded from AVS comparison
because other important factors play a major role in de-
termining the bioaecumulation potential of mercury in
sediment. Specifically, under certain conditions mer-
cury binds to an organic methyl group and is readily
taken up by living organisms.

Sediment with measured [SEM] in excess of [AVS]
does not necessarily exhibit toxicity. This is because
other binding phases can tie up metals. However, re-
search indicates that sediment with [AVS] in excess of
[SEM] will not be toxic from metals, and the greater the
[SEM]-[AVS] difference, the greater the likelihood of
toxicity from metals. Analysis of toxicity data for fresh-
water and saltwater sediment amphipods (crustaceans)
from EPA's Environmental Research Laboratory in
Narragansett, Rhode Island, revealed that 80 to 90 per-
cent of the sediments were toxic at [SEM]-[AVS] > 5
(Hansen, 1995; see also Hansen et al., 1996). Thus,
EPA selected [SEM]-[AVS] = 5 as the demarcation line
between Tier 1 and Tier 2. For the purpose of this evalu-
ation, where [SEM]-[AVS] was greater than 5, the sam-
pling station was classified as Tier 1. If [SEM]-[AVS]
was between zero and 5, the sampling station was clas-
sified as Tier 2. If [SEM]-[AVS] was less than zero, or
if AVS or the five AVS metals were not measured at the
sampling station, the sampling station was classified as
Tier 3 unless otherwise classified by another parameter.
Appendix B discusses the assumptions and limitations
associated with the [SEM]-[AVS] approach.

Sediment Chemistry Values Exceed Screening
Values [3, 6]

Several sets of sediment contaminant screening val-
ues, developed using different methodologies, are avail-
able to assess potential adverse effects on benthic species.
The screening values selected for comparison with mea-
sured sediment levels are the draft SQCs using a default
TOC of 1 percent (for those samples which do not have
accompanying TOC data), sediment quality advisory lev-
els (SQALs) for freshwater aquatic life (developed using
the equilibrium partitioning approach discussed previ-

2-13


-------
Methodology

ously for the development of draft SQCs), the effects
range-median (ERM) and effects range-low (ERL) val-
ues developed by Long et al. (1995), the probable effects
levels (PELs) and threshold effects levels (TELs) devel-
oped for the Florida Department of Environmental Pro-
tection (FDEP, 1994), and the apparent effects thresholds
(AETs) developed by Barrick et al, (1988), The assump-
tions and approaches used to develop these screening
values are discussed in detail in Appendix B.

The draft SQCs and SQALs were both developed us-
ing the same EqP approach. However, the data used to
derive SQALs were not compiled from an exhaustive lit-
erature search, nor were the toxicity data requirements as
extensive as specified for draft SQCs. Toxicity values used
for SQAL development include final chronic values from
EPA ambient freshwater quality criteria and secondary
chronic values derived using EPA's Great Lakes Water Qual-
ity Initiative "Her II" water quality criteria methodology.
The data used to develop the latter values were taken pri-
marily from quality-screened studies in published litera-
ture. The development of SQALs is discussed in further
detail in Appendix B of this report EPA has also prepared
a document describing the derivation of the SQALs
(USEPA, 1996). The chemicals for which SQALs have
been developed are identified in Appendix D of this vol-
ume.

The ERLs/ERMs, PELs/TELs, and AETs relate the
incidence of adverse biological effects to the sediment
concentration of a specific chemical at a specific sam-
pling station using paired field and laboratory data. The
developers of the ERLs/ERMs define sediment concen-
trations below the ERL as being in the "minimal-effects
range," values between the ERL and ERM in the "pos-
sible-effects range," and values above the ERM in the
"probable-effects range." In the FDEP (1994) approach,
the lower of the two guidelines for each chemical (the
TEL) is assumed to represent the concentration below
which toxic effects rarely occur. In the range of concen-
trations between the TEL and PEL, effects occasionally
occur. Toxic effects usually or frequently occur at con-
centrations above the upper guideline (the PEL).

In independent analyses of the predictive abilities
of the ERL/ERMs and TEL/PELs, the precentages of
samples indicating high toxicity in laboratory bioassays
of amphipod survival were relatively low (10-12 per-
cent) when all chemical concentrations were in the mini-
mal effects range, intermediate (17-19 percent) in the
possible effects range, and higher (38-42 percent) in the
probable effects range. Furthermore, the percentages of
samples indicating high toxicity in any one of a battery
of 2-4 tests performed, including more sensitive bioas-

says with sublethal endpoints, were 5-28 percent, 59-64
percent, and 78-80 percent among samples within the
minimal, possible, and probable effects ranges (Long et
al,, in press).

The AET approach is not based on the probability
of incidence of adverse biological effects. The AET is
the highest concentration at which statistically signifi-
cant differences in observed adverse biological effects
from reference conditions do not occur, provided that
the concentration also is associated with observance of
a statisically significant difference in adverse biological
effects. Essentially, this identifies the concentration
above which an adverse biological effect always occurs
for a particular data set. Barrick et al. (1988) list spe-
cific AET values for several different species or biologi-
cal indicators. For the purposes of this assessment, EPA
defined the AET-low as the lowest AET among appli-
cable biological indicators, and the AET-high as the
highest AET among applicable biological indicators. By
the nature of how the AET is derived, less stringent val-
ues might evolve as more data sets become available.

For the NSI data evaluation, the upper screening val-
ues were considered to be the ERM, PEL, draft SQC
(when using default TOC value of 1 percent), SQAL,
and AET-high for a given chemical. The lower screen-
ing values were considered to be the ERL, TEL, draft
SQC (when using default TOC of 1 percent), SQAL, and
AET-low for a given chemical. Because they are not
based on ranges of effects, the single freshwater aquatic
life draft SQC and SQAL values for a given chemical
served as both the high and low screening values.

For a sampling station to be classified as Tier 1, a
chemical measurement must have exceeded at least two
of the upper screening values. If a sediment chemistry
measurement exceeded any one of the lower screening
values, the sampling station was classified as Tier 2. If
sediment concentrations at a sampling station did not
exceed any screening values or there were no data for
chemicals that have assigned screening values, the sam-
pling station was categorized as Her 3 unless otherwise
-categorized by another parameter.

Under this approach, a sampling station could be
classified as Tier 1 from elevated concentrations of cad-
mium, copper, lead, nickel, or zinc based only on a com-
parison of [SEM] to [AVS]; that is, sampling stations
could not be classified as Tier 1 based on an exceedance
of two upper screening values for any of the five metals.
However, sampling stations were classified as Tier 2 for
these five metals based on an exceedance of one of the
lower screening values if AVS data were not available.

2-14


-------
f	,	j	National Sediment Quality Survey

r	•	 ; i	;	i	t	

Sediment Chemistry TBPs Exceed Screening
Criteria [4, 7]

This evaluation parameter addresses the risk to hu-
man consumers of organisms exposed to sediment con-
taminants. The theoretical bioaccumulation potential
(TBP) is an estimate of the equilibrium concentration
(concentration that does not change with time) of a con-
taminant in tissues if the sediment in question were the
only source of contamination to the organism. At
present, the TBP calculation can be performed only for
nonpolar organic chemicals. The TBP is estimated from
the concentration of contaminant in the sediment, the
organic carbon content of the sediment, the lipid con-
tent of the organism, and the relative affinity of the
chemical for sediment organic carbon and animal lipid
content. This relative affinity is measured in the field
and is called a biota-sediment accumulation factor
(BSAF, as discussed in detail in Appendix C). In prac-
tice, field measured BSAFs can vary by an order of mag-
nitude or greater for individual compounds depending
on location and time of measurement. For this evalua-
tion, EPA selected BSAFs that represents the central
tendency, suggesting an approximate 50 percent chance
that an associated tissue residue level would exceed a
screening risk value.

In the evaluation of NSI data, if a calculated sedi-
ment chemistry TBP value exceeded a screening value
derived using standard EPA risk assessment methodol-
ogy or the Food and Drug Administration (FDA) toler-
ance/action or guidance level, and if a corresponding
tissue residue level for the same chemical for a resident
species at the same sampling station also exceeded one
of those screening values, the station was classified as
Tier 1. Individual chemical risk levels were considered
separately ; that is, risks from multiple contaminants were
not added. Both sediment chemistry and tissue residue
samples must have been taken from the same sampling
station. If tissue residue levels for the same chemical
for a resident species at the same sampling station did
not exceed EPA risk levels or FDA levels or there were
no corresponding tissue data, the sampling station was
classified as Tier 2. If neither TBP values nor fish tis-
sue residue levels exceeded EPA risk levels or FDA lev-
els, or if no chemicals with TBP values, EPA risk levels,
or FDA levels were measured, the sampling station was
classified as Tier 3 unless otherwise classified by an-
other parameter. A detailed description of the methods
used to develop TBP values and to determine the EPA
risk levels used in this comparison is presented in
Appendix B.

Tissue Residue Data [8, 9, 10]

Tissue residue data were used to assess potential
adverse effects on humans from the consumption of fish
that become contaminated through exposure to contami-
nated sediment. Only those species considered benthic,
non-migratory (resident), and edible by human popula-
tions were included in human health assessments. A
list of species included in the NSI and their characteris-
tics is presented in Appendix F.

Sampling stations at which human health screen-
ing values for dioxin and PCBs were exceeded in fish
tissues were classified as Tier 1. For these chemicals,
corroborating sediment chemistry data were not required.
If human health screening values for dioxin or PCBs in
fish tissue were not exceeded or if neither chemical was
measured, the sampling station was classified as Tier 3
unless otherwise classified by another parameter.

For other chemicals, both a tissue residue level ex-
ceeding an FDA tolerance/action or guidance level or
EPA risk level and a sediment chemistry TBP value ex-
ceeding that level for the same chemical were required
to classify a sampling station as Tier 1. If tissue residue
levels exceeded FDA levels or EPA risk levels but corre-
sponding TBP values were not exceeded at the same sta-
tion (or there were no sediment chemistry data from that
station), the sampling station was classified as Tier 2.
If neither fish tissue levels nor TBP values exceeded EPA
risk levels or FDA levels, or if no chemicals with TBP
values, EPA risk levels, or FDA levels were measured,
the sampling station was classified as Tier 3 unless oth-
erwise classified by another parameter.

Toxicity Data [11,12]

Toxicity data were used to classify sediment sam-
pling stations based on their demonstrated lethality to
aquatic life in laboratory bioassays. Nonmicrobial sedi-
ment toxicity tests with a mortality endpoint were evalu-
ated. Toxicity test results that lacked control data, or
had control data that indicated greater than 20 percent
mortality (less than 80 percent survival), were excluded
from further consideration. The EPA has standardized
testing protocols for marine and freshwater toxicity tests.
A review of several protocols for sediment toxicity tests
suggests that mortality in controls may range from 10 to
30 percent, depending on the species, to be considered
an acceptable test result (API, 1994). Current amphi-
pod test requirements indicate that controls should have
less than 10 percent mortality (API, 1994; USEPA,
1994b).

2-15


-------


For the NSI data evaluation, EPA considered sig-
nificant toxicity as a 20 percent difference in survival
from control survival. For example, significant toxicity
occurred if control survival was 80 percent and experi-
mental survival was 60 percent or less.

Fortius evaluation parameter, corroboration of mul-
tiple tests was considered more indicative of probable
associated adverse effects than the magnitude of the ef-
fect in a single test Lethality demonstrated by two or
more single-species tests using two different test spe-
cies (at least one of which had to be a solid-phase test)
placed a sampling station in Tier 1. A sampling station
was classified as Tier 2 if toxicity was demonstrated by
one single-species nonmicrobial toxicity test. If lethal-
ity was not demonstrated by a nonmicrobial toxicity test,
or if toxicity test data were not available, the sampling
station was classified as Tier 3 unless otherwise classi-
fied by another parameter.

Incorporation ofRegional Comments on the
Preliminary Evaluation ofSediment
Chemistry Data

Several reviewers from different EPA Regions and
states provided comments on the May 16, 1994,
preliminary evaluation of sediment chemistry data. The
comments included more than 150 specific comments
identifying additional locations with contaminated sedi-
ment that had not been identified in the preliminary
evaluation. Since the preliminary evaluation, the final
NSI methodology has been developed and implemented.
The updated methodology has been refined significantly
to include tissue residue and toxicity data as well as
revised screening values. Data corresponding to any
additional comments that required further review were
divided into two categories: (1) data that incorrectly
identified contaminated sediment and (2) additional wa-
ter bodies that contain areas of sediment contamination.
The first category primarily addressed sampling stations
identified in the preliminary assessment as exceeding
sediment chemistry screening values for specific con-
taminants that reviewers stated were located in water
bodies that are not contaminated from the chemical(s)
in question.

EPA examined all NSI sampling stations that had
been identified in the preliminary evaluation as exceed-
ing a sediment quality screening value, but were located
in water bodies that reviewers of the preliminary evalu-
ation identified as not being contaminated by that spe-
cific contaminant or contaminants. If the sampling
station in question was classified in this final evalua-
tion as Tier 1 based only on the specific contaminant(s)
identified by the reviewer as not being a problem, the
sampling station was removed from the Tier 1 category
and placed in the Tier 3 category. Only a few sampling
stations were moved from the Tier 1 category to the Her
3 category as a result of this procedure. Stations identi-
fied in the NSI evaluation as Tier 1 based on other chemi-
cals not identified by the reviewer or because of toxicity
data were not removed from Her 1.

Additional water bodies that reviewers identified as
potential areas of significant contamination were evalu-
ated to determine whether sampling stations along those
water bodies were classified as Tier 1 based on the final
NSI data evaluation. Locations or water bodies identi-
fied by reviewers as potential areas of significant con-
tamination are discussed separately in the results
(Chapter 3).

Evaluation Using EPA Wildlife Criteria

In addition to the evaluation parameters described
above and presented in Table 2-2, EPA conducted an
assessment of NSI data based on a comparison of sedi-
ment chemistry TBP values and fish tissue values to EPA
wildlife criteria developed for the Great Lakes. This
evaluation, however, was not included with the results
of evaluating the NSI data based on the other param-
eters. The results of evaluating NSI data based on wild-
life criteria are presented in a separate section of Chapter
3. Wildlife criteria based solely on fish tissue concen-
trations were derived for EPA wildlife criteria for water
that are presented in the Great Lakes Water Quality Ini-
tiative Criteria Documents for the Protection of Wild-
life (USEPA, 1995a). EPA has developed wildlife criteria
for four contaminants: DDT, mercury, 2,3,7,8-TCDD,
and PCBs. The method to adjust these wildlife criteria
for the NSI data evaluation is explained in detail in
Appendix B.

2-16


-------
Chapter 3

Findings

This chapter presents the results of the
evaluation of NSI data based on the
methodology described in Chapter 2. This dis-
cussion includes a summary of the results of national, re-
gional, and state assessments.

National Assessment

EPA evaluated a total of 21,096 sampling stations na-
tionwide as part of the NSI data evaluation (Figure 3-1).
Of the sampling stations evaluated, 5,521 stations (26 per-
cent) were classified as Tier 1, 10,401 (49 percent) were
classified as Tier 2, and 5,174 (25 percent) were classified
as Tier 3 (Table 3-1). This distribution suggests that state
monitoring programs (accounting for the majority of NSI
data) have been efficient and successful in focusing their
sampling efforts on areas where contamination is known
or suspected to occur. The frequency of Tier 1 classifica-
tion based on the evaluation of all NSI data is greater than
from data sets derived from purely random sampling.

The national distribution of Tier 1 sampling stations is
illustrated in Figure 3-2. The distribution of Tier 1 stations
depicted in Figure 3-2 must be viewed in the context of the
distribution of all sampling stations depicted in Figure 3-1.
Table 3-1 presents the number of sampling stations in each
tier by EPA Region. The greater number of Tier 1 and Her
2 sampling stations in some Regions is to some degree a
function of a larger set of available data. Although there
are 17 times more Her 1 stations in EPA Region 4 (south-
eastern states) than in EPA Region 8 (mountain state), there
are also 13 times more Tier 3 stations.

The NSI sampling stations were located in 6,744 in-
dividual river reaches throughout the contiguous United
States (based on EPA's River Reach File 1; Bondelid and
Hanson, 1990). A river reach can be part of a coastal
shoreline, a lake, or a length of stream between two ma-
jor tributaries ranging from approximately 1 to 10 miles
long. NSI sampling stations were located in approxi-
mately 11 percent of all river reaches identified in the
contiguous United States (Table 3-1 and Figure 3-3).
Four percent of all river reaches in the United States con-
tained at least one sampling station classified as Tier 1.

Five percent of all reaches contained at least one sam-
pling station classified as Tier 2 (but none as filer 1). In
2 percent of reaches in the contiguous United States, all
of the sampling stations were classified as Her 3. EPA
has not yet catalogued river reaches outside the contigu-
ous United States (e.g., Alaska, Hawaii, Puerto Rico),
and some sampling stations in the ocean were not linked
to a specific reach. Sampling bias toward areas of known
or suspected contamination may be more pronounced in
some Regions compared to others, aid may bfe related to
the relative extent of sampling. The results presented on
Table 3-1 appear to indicate that the smaller the percent-
age of reaches with available data, the greater the likeli-
hood those reaches will contain a Her 1 or Her;2 sampling
station.

Not all sampling programs target only sites of known
or suspected contamination. The NSI includes data from
the National Oceanic and Atmospheric Administration's
(NOAA's) National Status and Trends Program, which
is part of the COSED database, and EPA's Environmen-
tal Monitoring and Assessment Program (EM AP). These
are examples of sampling programs in whict^ most sam-
pling stations are not targeted at locations of known or
suspected contamination. Based on these data alone, the
percentage of sampling stations placed in each tier dif-
fers considerably from the percentage of sampling sta-
tions in each tier based on an evaluation of all the data in
the NSI. Smaller percentages of COSED and EMAP
sampling stations are categorized as Tier 1 (18 percent
for COSED and 14 percent for EMAP compared to 26
percent for all NSI sampling stations), greater percent-
ages are categorized as Tier 2 (75 percent for COSED
and 68 percent for EMAP compared to 49 percent for all
NSI stations), and smaller percentages are categorized
as Her 3 (7 percent for COSED and 18 percent for EMAP
compared to 25 percent for all NSI sampling stations).
This may reflect the lower detection limits of more sen-
sitive analytical chemistry techniques, the sensitivity of
Tier 2 evaluation parameters, and the nearly ubiquitous
presence of lower to intermediate levels of contamina-
tion in areas sampled by these programs.

The NSI contains over 1.5 million individual records
of contaminant measurements in sediment and fish

3-1


-------

-------
Table 3-1. National Assessment: Evaluation Results for Sampling Stations and River Reaches by EPA Region

Region (State)

Station Evaluation

River Reach Evatuatioif

Herl

Tier 2

Her 3

Number of
Stations

Not
Identified

fayanRFl
Reaclf

Reaches

It

Least 1
Station fai
Her 1

Reaches

w/at
Least 1
Station in
Her24

Reaches

wfaU
Stations
in Tier 3

Total#
Reaches

Wat
Least 1
Station
Evaluated

Total
Reaches
in Region

% of all
Reaches
in Region
w&t
Least 1
Station
Evaluated

%ot
Reaches

Wat
Least 1
Her 1 or
Her 2
Station

#

%"

#

%b

#

%-

Region 1

(CT, ME, MA, NH, RI, VT)

298

27

646

59

158

14

361

59

65

7

131

2,648

5

5

Region 2
(NY, NJ, PR)

355

32

559

51

182

17

173

116

147

29

292

1,753

17

15

Region 3

(DE, DC, MD, PA, VA, WV)

318

17

934

49

658

34

92

209

453

226

888

3,247

27

20

Region 4

(AL, FL, GA, KY, MS. NC, SC,
TO)

1,157

23

1,930

39

1,872

38

343

566

684

520

1,770

9,749

18

13

Region 5

OL, IN, MI, MN, OH, WI)

1,418

33

2,137

50

735

17

108

594

570

268

1,432

6,025

24

19

Region 6

(AR, LA, NM, OK, TX)

382

24

837

52

397

24

124

266

341

192

799

7,293

11

8

Region 7

(IA, KS, MO, NE)

330

33

393

39

288

28

N/A

246

182

88

516

4,857

11

9

Region 8

(CO, MT, ND, SD, UT, WY)

68

13

327

61

140

26

N/A

61

153

91

305

13,492

2

2

Region 9

(AZ, CA, HI, NV)

468

28

942

55

289

17

794

119

92

43

254

4,601

6

5

Regon10
(AK, ID, OR, WA)

727

25

1,696

59

455

16

497

147

174

72

393

10,178

4

3

¦ftilal for U.S.1

5,521

26

10,401

49

5,174

25

2,492

2,371

2,843

1,530

6,744

62,742

11

8

¦River reaches based on EPA River Resell File 1 (RF1).

'Percent of all stations evaluated in the NSI in the Region.

'Stations not identified by an RF1 reach were located in coastal or open water areas.

*No stations in these reaches were included in Her 1.

•Because some reaches occur in more than one Region, the total number of teaches in each caleogry for the country might not equal the sum of reaches in the Regions.


-------
Figure 3-2. Sampling Stations Classified as Tier 1 (Associated Adverse Effects are Probable),


-------
11' .	1

National Sediment Quality Survey

At Least One
Tier 1 Station
4%

At Least One
Tier 2 Station and
Zero Tier 1 Stations
5%

All Tier 3 Stations
2%

Figure 3-3. National Assessment: Percent of River Reaches That Include
Her 1, Tier 2, and Tier 3 Sampling Stations.

tissue (Figure 3-4). Slightly more than one-third of these
measurements represent concentrations recorded as above
a detection limit. Using available assessment parameters,

EPA could evaluate nearly two-thirds (approximately
380,000) of these measurements for the probability of
association with adverse effects. Approximately one-
quarter of the measurements above detection (nearly 40
percent of measurements that could be evaluated) reflect
either a Tier 1 or Her 2 level of contamination. Figure 3-4
also shows the distribution of measurements at the Tier I
and Tier 2 level of contamination by chemical class.

Chemicals that have been measured over the past 15 years,
can be evaluated using the NSI evaluation approach, and
accumulate to levels associated with an increased prob-
ability of adverse effects are predominantly persistent, hy-
drophobic organic compounds and metals.

Data related to more than 230 different chemicals or
chemical groups were included in the NSI evaluation.

Approximately 40 percent of these chemicals or chemi-
cal groups (97) were present at levels that resulted in
classification of sampling stations as Tier 1 or Tier 2.

Table 3-2 presents the chemicals or chemical groups that
resulted in classification of more than 1,000 Tier 1 or
Tier 2 sampling stations. Sampling stations are reported
more than once in Table 3-2 because it is common for a
station to have elevated concentration levels for multiple
chemicals.

The contaminants most frequently at levels in fish
or sediment where associated adverse effects are prob-

able include PCBs (58 percent
of the 5,521 Tier 1 sampling
stations) and mercury (20 per-
cent of Tier 1 sampling sta-
tions). Pesticides, most notably
DDT and metabolites at 15 per-
cent of Tier 1 sampling stations,
and polynuclear aromatic hy-
drocarbons (PAHs), such as
pyreen at 8 percent of Her 1
sampling stations, also were
frequently at levels where as-
sociated adverse effects are
probable.

Dry weight measures of
divalent metals other than mer-
cury (e.g., copper, cadmium,
lead, nickel, and zinc) were not
used to place a sampling sta-
tion in Tier 1 without an asso-
ciated measurement of acid
volatile sulfide, a primary me-
diator of bioavailabilty not of-
ten available in the data base. The [SEM]-[AVS]
methodology for sediment assessment is relatively new, and
AVS measurements have not commonly been made during
sediment analyses. As a result, metals otter than mercury
(which also include arsenic, chromium, and silver) are solely
responsible for only 6 percent of Her 1 sampling stations
and overlap with mercury or organic compounds at an ad-
ditional 6 percent of Tier 1 sampling stations. In contrast,
metals other than mercury are solely responsible for about
28 pa-cent of the 15,992 Her 1 and Tier 2 sampling sta-
tions, and overlap with mercury or organic compounds at
an additional 28 percent of Tier 1 and Tier 2 sampling sta-
tions. The remaining 44 percent of Tier 1 and Her 2 sam-
pling stations are classified solely for mercury or organic
compounds.

Two important issues in interpreting the results of
sampling station classification are naturally occurring
"background" levels of chemicals and the effect of chemi-
cal mixtures. Site-specific naturally occulting (or back-
ground) levels of chemicals may be an important risk
management consideration in examining sampling sta-
tion classification. This is most often an issue for natu-
rally occurring chemicals such as metals and PAHs. In
addition, although the sediment chemistry screening lev-
els for individual chemicals are used as indicators of po-
tential adverse biological effects, other co-occurring
chemicals (which may or may not be measured) can cause
or contribute to any observed adverse effect at specific
locations.

3-5


-------
Sediment and Fish Tissue
Measurements
(1,565,103)

Measurements Above
Detection Limit
(586,994)

Measurements Indicating
Potential Risk
(Tier land Tier 2*)
(142,004)

PCBs
15%

Tier 2

«% PmHcMm'
/	22%

NuVMytoEyahwfs
35%

'For Tier 1 alone: 27,358 measurements Indicate potential risk, distributed among PCBs (62 percent) PAH (13 percent), pesticides (S percent)
mercury (7 percent), other organlcs (5 percent), and other metals (4 percent)

Figure 3-4. National Assessment: Percent of NSI Measurements That Indicate Potential Risk.


-------
Table 3-2. Chemicals or Chemical Groups Most Often Associated With Tier 1 and Her 2 Sampling Station
Classifications

Chemical or
Chemical Group

Number of Stations

Total# of
Stations
Evaluated

Based on All Measurement Parameters

Based on
Aquatic Life
Parameters

Based oil
Human Health
Parameters

Combined
Tiers 1 & 2

Percent of
All Tier 1
and Tier 2
Stations

Tier!

Percent of
All Tier 1
Stations

Tier 2

Tier!

Tier 2

Tier 1

Tier 2

Copper

16,161

7,172

45

-

-

7,172

-

7,167

-

5

Nickel

12,447

6,284

39

-

-

6,284

-

6,284

-

-

Lead

16,791

5,681

36

-

-

5,681

-

5,415

-

328

Folychlorinated biphenyls

12,276

5,454

34

3,175

58

2,279

963

1,219

2,256

3,198

Arsenic

13,200

5,392

34

182

3

5,210

182

4,658

-

605

Cadmium

16,010

4t808

30

-

-

4,808

-

4,773

-

41

Mercury

15,649

4,333

27

1,122

20

3,211

1,122

3,127

-

103

Zinc

15,160

3,468

22

-

-

3,468

-

3,451

-

17

DDT (and metabolites)

11,462

3,422

21

803

15

2,619

798

2,203

21

1,402

Chromium

15,222

3,070

19

278

5

2,792

278

2,786

-

7

Dieldrin

10,284

2,597

16

58

1

2,539

49

1,006

9

2,456

Chlordane

10,697

2,169

14

11

<1

2,158

-

1,303

11

1,697

Benzo(a)pyrene

5,435

1,993

13

287

5

1,706

287

1,051

-

1,990

Pyrene

5,798

1,920

12

431

8

1,489

431

1,489

-

10

Chrysene

5,300

1,427

9

166

3

1,261

166

1,261

-

30

Dibetizo(a,h)anthracene

4,896

1,383

9

337

6

1,046

337

1,018

-

1,092

Benzo(a)anthracene

5,120

1,366

9

214

4

1,152

214

1,106

-

847

Bis(2-elhylhexyl)phthalate

3,559

1,190

7

347

6

843

347

823

-

406

Naphthalene

5,246

1,186

7

254

5

932

254

932

-

5

Fluoranthene

5,814

1,114

7

210

4

904

210

904

-

11

Fluorene

5,175

1,107

7

201

4

906

201

906

-

5

Silver

8,022

1,096

7

302

5

794

302

794

-

-

Total for all chemicals in
the NSI database

21,096

15,922

-

5,521

-

10,401

3,287

9,921

2,327

6,196

The total number of sampling stations classified as
Tier 1 or Tier 2 for a given chemical as presented in
Table 3-2 may not be representative of the potential risk
posed by that chemical. Although there may be few over-
all observations for some chemicals, the frequency of
detection in sediment and tissue and the frequency with
which those chemicals result in Tier 1 or Tier 2 risk may
be high. (See Appendix D, Table D-2.)

The results of the analysis for three chemicals (arsenic,
silver, and phthalate esthers) might be misleading. Arsenic is
typically analyzed in biota as "total arsenic", which includes
all forms of arsenic. The EPA risk level for comparison with
measured values was derived for the highly toxic effects of
inorganic arsenic. However, arsenic in the edible portions of
fish and shellfish is predominantly found in a nontoxic or-
ganic form (USEPA, 1995c). For this analysis, a precautionary

3-7


-------
approach was taken to account for the human health risk from
the small amount of inorganic arsenic included in total ar-
senic measures and for measures that, in fact, represent only
inorganic arsenic. Silver, like copper, cadmium, lead, nickel,
and zinc, binds to sulfide in sediment However, silver can-
not be evaluated like these other metals in the [SEMHAVS]
assessment for a number of reasons, including that one mol-
ecule of sulfide binds two molecules of silver rather than just
one as is the case for the other metals. Recent research sug-
gests that if any AVS is measured, silver will not be bioavail-
able or toxic to exposed aquatic organisms (Berry et al., 1996).
In the NSI data evaluation, silver is not evaluated on the basis
of AVS measurement, and exceedance of two upper thresh-
olds for aquatic life protection can classify a sampling station
as Tier 1. In the case of phthalate esthers, high concentra-
tions in samples might be an indication of contamination dur-
ing sample handling and not necessarily an indication of
sediment contamination at the sampling station.

Table 3-2 also separately identifies the number of
sampling stations categorized as Tier 1 or Tier 2 for aquatic
life effects and for human health effects. Evaluation pa-
rameters indicative of aquatic life effects include:

•	Comparison of sediment chemistry measure-
ments to EPA draft sediment quality criteria
(SQCs).

•	Comparison of sediment chemistry measure-
ments to other screening values (SQCs when
percent organic carbon is not reported, SQALs,
ERL/ERMs, PEL/TELs, and AETs).

•	Comparison of [SEM] to [AVS].

•	Results of toxicity tests.

Human health evaluation parameters included:

•	Comparison of sediment chemistry TOP to EPA
risk levels or FDA tolerance/action or guide-
line levels.

•	Comparison of fish tissue levels of PCBs and di-
oxin to EPA ride levels. (A sampling station can
be classified as Her 1 without corroborating sedi-
ment chemistry data.)

•	Comparison of fish tissue levels to EPA risk lev-
els and FDA tolerance/action or guideline levels.

The evaluation results indicate that sediment contami-
nation associated with probable or possible but infrequent
adverse effects exists for both aquatic life and human
health. More sampling stations were classified as either

Her 1 or Tier 2 for aquatic life concerns than for human
health concerns. About 41 percent more sampling sta-
tions were classified as Tier 1 for aquatic life (3,287 sta-
tions) than for human health (2,327 stations). About 60
percent more sampling stations were classified as Tier 2
for aquatic life (9,921 stations) than were classified as
Tier 2 for human health (6,196 stations). The locations
of sampling stations classified as Tier 1 or Tier 2 for
aquatic life concerns are illustrated in Figure 3-5, and the
locations of those classified as Tier 1 or Tier 2 for human
health concerns are illustrated in Figure 3-6.

EPA analyzed the results to determine which evalua-
tion parameters most often caused sampling stations to
be classified as either Tier 1 or Tier 2 (see Table 3-3).
Most of the sampling stations classified as Tier 1 (3,283
stations) or Tier 2 (9,882 stations) were placed in those
categories because measured sediment contaminant lev-
els exceeded screening values. The comparison of fish
tissue levels of PCBs and dioxins to EPA risk levels trig-
gered placement of the second highest number of sam-
pling stations in Tier 1 (2,313 stations). The comparison
of sediment chemistry TBP values to FDA levels and EPA
risk levels triggered placement of the second highest num-
ber of sampling stations in Tier 2 (5,671 stations). The
AVS and toxicity parameters triggered placement of the
fewest sampling stations in Her 1 (8 stations each) and
Tier 2 (146 stations for AVS and 183 stations for toxic-
ity). These results reflect both data availability and evalu-
ation parameter sensitivity.

The lack of data required to apply some important
assessment parameters hampered EPA's efforts to deter-
mine the incidence and severity of sediment contamina-
tion. For example, a Tier 1 classification based on divalent
metal concentrations in sediment required an associated
acid-volatile sulfide (AVS) measurement. Also, a Tier I
classification for potential human health effects required
both sediment chemistry and fish tissue residue data for
all chemicals except PCBs and dioxins. These data com-
binations frequently were not available. Table A-2 in Ap-
pendix A presents the total number of NSI stations where
sediment chemistry data, related biological data, and
matched data (i,e., sediment chemistry and biological data
taken at the same sampling station) were collected. AVS
measurements were available at only 1 percent of the
evaluated stations. Likewise, matched sediment chemis-
try and fish tissue data were available at only 8 percent of
the evaluated stations. Toxicity data were also limited:
bioassay results were available at only 6 percent of the
evaluated stations.

To help judge the effectiveness of the NSI data evalu-
ation approach, EPA examined the agreement between

3-8


-------

-------

-------
Njilioiijil .SVcliinciH Qualify Sm^'cv

Table 3-3. Number of Sampling Stations Classified as Tier 1 and Tier 2 Based on Each Component of
the Evaluation Approach (see Table 2-2)

Measurement Parameter

Number of
Sampling
Stations In
Tier 1

Number of
Sampling
Stations in
Tier 2

Sediment chemistry values exceed draft sediment quality criteria

97

NA

ESEMHAVS] comparison

8

146

Sediment chemistry values exceed threshold values

3,283

9,882

Sediment chemistry TBP and fish tissue levels exceed risk levels or action levels

126

NA

Sediment chemistry TBP exceeds risk levels or action levels

NA

5,671

Fish tissue levels exceed risk levels or action levels

NA

2,789

Tissue levels of PCBb or dioxins exceed risk levels

2,313

NA

Toxicity test results

8

183

matched sediment chemistry and toxicity test results for
the 805 NSI sampling stations where both data types were
available and could be evaluated. The toxicity test data
indicate whether significant lethality to indicator organ-
isms occurs as a result of exposure to sediment. Tier 1
classifications for aquatic life effects from sediment chem-
istry data correctly matched toxicity test results for about
three-quarters of the sampling stations, with the remain-
der balanced between false positives (12 percent) and false
negatives (14 percent). In contrast, when Her 2 classifi-
cations from sediment chemistry data are added in, false
negatives drop to less than 1 percent at the expense of
false positives (which increase to 68 percent) and cor-
rectly matched sampling stations (which drop to 30 per-
cent). This result highlights the fact that classification in
Tier 2 is very conservative, and it does not indicate a high
probability of adverse effects to aquatic Mfe. If bioassay
test results for sublethal (chronic) endpoints such as re-
productive effects were included in the NSI evaluation,
the rate of false positives would likely decrease and cor-
rectly matched sampling stations would likely increase
for both tiers.

EPA also conducted a separate analysis of the corre-
lation of toxicity data and exceedances of SQCs and
SQALs (exclusive of other threshold values). From the
results of this study, there are 2,037 observations of a
SQC or SQAL exceedance at 916 sampling stations.
These 916 sampling stations are located in 405 distinct
RF1 reaches, which are in turn located in 218 distinct
watersheds. Matching toxicity test data are available at

39 of these 916 sampling stations. Toxicity test results
indicate that one or more SQC or SQAL exceedances are
associated with significant lethality (acute effects) to in-
dicator organisms slightly more than half of the time (22
of 39 sampling stations). SQCs and SQALs are levels set
to be protective of acute and chronic effects, such as ef-
fects on reproduction or growth, for 95 percent of benthic
species. The NSI currently does not contain matching
chronic toxicity test data to compare with sediment chem-
istry measures.

For a number of reasons, known contaminated sedi-
ment locations in the United States might not have been
classified as Her 1 or Tier 2 based on the evaluation of
NSI data. The NSI does not presently include data de-
scribing every sampled location in the Nation. There-
fore, numerous sampling stations were not evaluated for
this first report to Congress. However, additional data-
bases will be added to the NSI and more sampling sta-
tions will be evaluated for future reports to Congress.

During an initial screening of the NSI data, EPA
noted data quality problems that might have affected
all or many of the data reported in a given database
(e.g., the Virginia State Water Control Board organic
chemical data reported in STORET). Databases with
obvious quality problems were not included in the NSI
data evaluation. Also, if a database included in the
NSI did not have associated locational information
(latitude/longitude), data in that database were not in-
cluded in the NSI data evaluation (e.g., EPA's Great

3-11


-------
l'imlin

Lakes Sediment Quality Database). To reduce the
chances of overlooking sampling locations that have
obvious sediment contamination problems, EPA sent
a preliminary evaluation of sediment chemistry data
to each EPA Region so knowledgeable staff would
have an opportunity to list additional contaminated
sediment locations not identified in the NSI evalua-
tion. These locations are presented at the end of this
chapter. Despite such efforts, some sediment sampling
locations known to have contamination problems still
have not been listed in this first report to Congress.

Watershed Analysis

The potential risk of adverse effects to aquatic life
and human health is greatest in areas with a multitude of
contaminated locations. The assessment of individual
sampling stations is useful for estimating the number and
distribution of contaminated spots and the overall mag-
nitude of sediment contamination in monitored
waterbodies of the United States. However, a single "hot
spot" might not pose a great threat to either the benthic
community at large or consumers of resident fish because
the spatial extent of exposure could be small. On the
other hand, if many contaminated spots are located in
close proximity, the spatial extent and probability of ex-
posure are much greater. EPA examined sampling sta-
tion classifications within watersheds to identify areas
of probable concern for sediment contamination (APCs),
where the exposure of benthic organisms and resident
fish to contaminated sediment may be more frequent. In
this report, EPA defines watersheds by 8-digit United
States Geological Survey (USGS) hydrologic unit codes
(the cataloging unit), which are roughly the size of a
county.

Watersheds containing APCs are those that include
at least 10 Tier 1 sampling stations, and in which at least
75 percent of all sampling stations were classified as ei-
ther Tier 1 or Her 2. These dual criteria are based on
empirical observation of the data. NSI Sampling sta-
tions are located within 1,367 watersheds, or approxi-
mately 65 percent of the total number of watersheds in
the continental United States. To identify APCs, EPA first
examined the frequency distribution of the number of
Tier 1 sampling stations within these watersheds. The
upper 10 percent of watersheds with sampling stations
had 10 or more sampling stations classified as Tier 1.
Because approximately three-quarters of all sampling
stations in the nation are classified as Tier 1 or Tier 2,
EPA determined that APCs should also reflect at least
this distribution. This second requirement slightly re-
duced the number watersheds containing APCs.

The definition of "area of probable concern" v/as
developed for this report to identify watersheds for which
further study of the effects and sources of sediment con-
tamination, and possible risk reduction needs, would be
warranted. Where data have been generated through in-
tensive sampling in areas of known or suspected con-
tamination within a watershed, the APC definition should
identify watersheds which contain even relatively small
areas that are considerably contaminated. However, this
designation does not imply that sediment throughout the
entire watershed, which is typically very large compared
to the extent of available sampling data, is contaminated.
On the other hand, where data have been generated
through comprehensive sampling, or where sampling sta-
tions were selected randomly or evenly distributed
throughout a sampling grid, the APC definition might
not identify watersheds that contain small or sporadically
contaminated areas. A comprehensively surveyed wa-
tershed of the size typically delineated by a USGS cata-
loging unit might contain small but significant areas that
are considerably contaminated, but might be too large in
total area for 75 percent of all sampling stations to be
classified as Tier 1 or Tier 2. Limited random or evenly
distributed sampling within such a watershed also might
not yield 10 Tier 1 sampling stations. Thus, the process
used to identify watersheds containing APCs may both
include some watersheds with limited areas of contami-
nation and omit some watersheds with significant con-
tamination. However, given available data, EPA believes
it represents a reasonable screening analysis to identify
watersheds where further study is warranted.

The application of this procedure identified 96 wa-
tersheds that contain APCs. The location of these water-
sheds is depicted on Figure 3-7. The name and cataloging
unit number on Table 3-4 correspond to the labels on
Figure 3-7. These watersheds represent about 5 percent
of all watersheds in the continental United States (96 of
2,111). The watershed analysis also indicated that 39
percent of all watersheds in the country contain at least
one Tier 1 sampling station, 15 percent contain at least
one Tier 2 sampling station but no Tier 1 stations, and 6
percent contain all Tier 3 sampling stations (Figure 3-8).
Thirty-five percent of all watersheds in the country did
not include a sampling station.

The definition of an APC requires that a watershed
include at least 10 sampling stations, because at least 10
must be classified as Tier 1. About one-quarter of the
watersheds in the country (488 of 2,111) met this require-
ment, and thus were eligible to contain an APC: approxi-
mately 20 percent (96 of 488) of these contain APCs.
Although a minimum amount of sampling was required

3-12


-------

-------
I'inclines

Ibble 3-4. USGS Cataloging Unit Numbers and Names for Watersheds Containing APCs

Map#

Cataloging Unit Number

Cataloging Unit Name

1

1090001

Charles

2

1090002

Cape Cod

3

1090004

Nairagansett

4

2030103

Hackensack-Passaic

5

2030104

Sandy Hook-Staten Island

6

2030105

Raritan

7

2030202

Southern Long Island

8

2040105

Middle Delaware-Musconetcong

9

2040202

Lower Delaware

10

2040203

Schuylkill

11

2040301

Muilica-Totns

12

2060003

Gunpowder-Patapsco

13

2070004

Conococheague-Opequon

14

3040201

Lower Pee Dee

15

3060101

Seneca

16

3060106

Middle Savannah

17

3080103

Lower St. Johns

18

3130002

Middle Chattahoochee-Lake Harding

19

3140102

Choctawhatehee Bay

20

3140107

Peidido Bay

21

3160205

Mobile Bay

22

4030102

Door-Kewaunee

23

4030108

Menominee

24

4030204

Lower Fox

25

4040001

Little Calumet-Galien

26

4040002

Pike-Root

27

4040003

Milwaukee

28

4050001

St. Joseph

29

4060103

Manistee

30

4090002

Lake St Clair

31

4090004

Detroit

32

4100001

Ottawa-Stony

33

4100002

Raisin

34

4100010

Cedar-Portage

35

4100012

Huron-Vermillion

36

4110001

Black-Rocky

37

4110003

Ashtabula-Chagrin

3-14


-------
National Sedinii'iil Qu^il'ilx Stir\c\

Table 3-4. (continued)

Map#

Cataloging Unit Number

Cataloging Unit Name

38

4120101

Chautauqua-Conneaut

39

4120103

Buffalo-Eighteenmile

40

4120104

Niagara

41

4130001

Oak Orchard-Twelvemile

42

4150301

Upper St. Lawrence

43

5030101

Upper Ohio

44

5030102

Shenango

45

5040001

Tuscarawas

46

5120109

Vermilion

47

5120111

Middle Wabash-Busseron

48

6010104

Holston

49

6010201

Watts Bar Lake

50

6010207

Lower Clinch

51

6020001

Middle Tennessee-Chickamauga

52

6020002

Hiwassee

53

6030001

Guntersville Lake

54

6030005

Pickwick Lake

55

6040001

Lower Tennessee-Beech

56

6040005

Kentucky Lake

57

7010206

Twin Cities

58

7040001

Rush-Vermillion

59

7040003

Buffalo-Whitewater

60

7070003

Castle Rock

61

7080101

Copperas-Duck

62

7090006

Kishwaukee

63

7120003

Chicago

64

7120004

Des Plaines

65

7120006

Upper Fox

66

7130001

Lower Illinois-Senachwine Lake

67

71401001

Cahokia-Joachim

68

7140106

Big Muddy

69

7140201

Upper Kaskaskia

70

7140202

Middle Kaskaskia

71

8010100

Lower Mississippi-Memphis

72

8030209

Deer-Steele

73

8040207

Lower Ouachita

3-15


-------
Finding

Table 3-4. (continued)

Map#

Cataloging Unit Number

Cataloging Unit Name

74

8080206

Lower Calcasieu

75

8090100

Lower Mississippi-New Orleans

76

10270104

Lower Kansas

77

11070207

Spring

78

11070209

Lower Neosho

79

12040104

Buffalo-San Jacinto

80

17010303

Coeur D'Alene Lake

81

17030003

Lower Yakima

82

17090012

Lower Willamette

83

17110002

Strait of Georgia

84

17110013

Duwamish

85

17110014

Puyallup

86

17110019

PugetSound

87

18030012

Tulare-Buena Vista Lakes

88

18050003

Coyote

89

18050004

San Francisco Bay

90

18070104

Santa Monica Bay

91

18070105

Los Angeles

92

18070107

San Pedro Channel Islands

93

18070201

Seal Beach

94

18070204

Newport Bay

95

18070301

Aiiso-San Oiiofre

96

18070304

San Diego

Figure 3-8. National Assessment: Watershed Classifications.

3-16


-------
for consideration as an APC, sampling effort alone did
not determine APC identification. In fact, other than
defining a ceiling, the total number of sampling stations
in a watershed is not indicative of the number of Tier 1
sampling stations. A simple statistical regression analy-
sis of total number of sampling stations versus number
of Tier 1 sampling stations for the nearly 500 watersheds
eligible to contain an APC (including at least 10 and up
to 200 sampling stations) resulted in a correlation coef-
ficient (R-square) of 0.44, a value which indicates a large
amount of variation.

APC designation could result from extensive sam-
pling throughout a watershed, or from intensive sampling
at a single or few contaminated locations. In compari-
son to the overall results presented in Figure 1, sampling
stations are located on an average of 46 percent of reaches
within watersheds containing APCs. On the average, 30
percent of reaches in watersheds containing APCs have
at least one Tier 1 sampling station, and 13 percent have
no Tier 1 sampling station but at least one Her 2 sam-
pling station. In many of these watersheds, contaminated
areas may be concentrated in specific river reaches in a
watershed. Within the 96 watersheds containing APCs
across the country, 57 individual river reaches or water
body segments have 10 or more Tier 1 sampling stations
(Table 3-5). These are localized areas within the water-
shed for which an abundance of evidence indicates po-
tentially severe contamination. Because EPA's Reach File
1 was used to index the location of NSI sampling sta-
tions, some sampling stations might not actually occur
on the identified Reach File 1 stream, but on a smaller
stream that is hydrologically linked or is relatively close
to the Reach File 1 stream.

Volume 2 of this report contains more detailed in-
formation for each watershed containing an APC. This
information includes maps showing watershed bound-
aries, major waterways (RF1), and the location and clas-
sification of sampling stations. In addition, Volume 2
provides tables summarizing the sediment chemistry, fish
tissue, and toxicity test data collected within those wa-
tershed that were used for this evaluation.

Wildlife Assessment

As described in Chapter 2, EPA conducted a sepa-
rate analysis of the NSI data to determine the number of
sampling stations where chemical concentrations of DDT,
mercury, dioxin, and PCBs exceeded levels set to be pro-
tective of wildlife (i.e., EPA wildlife criteria). The wild-
life criteria used in this evaluation were derived from
those presented in the Great Lakes Water Quality Initia-
tive Criteria Documents for the Protection of Wildlife

(USEPA, 1995a) subtracting out exposure from direct
water consumption. The only assumed route of expo-
sure for this evaluation was the consumption of contami-
nated fish tissue by wildlife.

Data were available to evaluate a total of 13,691 NSI
sampling stations using the wildlife criteria. Based on
wildlife criteria alone, 162 sampling stations would be
classified as Tier 1 (matched sediment chemistry and fish
tissue data), and 7,634 sampling stations would be clas-
sified as Tier 2 (sediment chemistry TBP or fish tissue
data). Figure 3-9 shows the location of Her 1 and Tier 2
sampling stations based on exceedance of wildlife crite-
ria. Table 3-6 presents a comparison of the sampling
stations classified as Tier 1 or Tier 2 with and without
the use of wildlife criteria. If wildlife criteria had been
used to complete the national assessment, 619 sampling
stations classified as Tier 3 would have been classified
as Tier 2 and 16 sampling stations classified as Tier 2
would have been classified as Tier 1. Most of the change
is from an increase in Tier 2 sampling stations classified
for DDT (from 2,619 to 4,276) and mercury (from 3,211
to 5,199).

Additional sampling stations would be classified as
Tier 1 or Tier 2 using wildlife criteria for two reasons:
(1) the wildlife criteria for DDT and mercury are signifi-
candy lower (8 and 19 times lower, respectively) than
the EPA risk levels used in the corresponding human
health evaluations; (2) the lipid content used in the wild-
life TBP analysis (10.31 percent for whole body) ex-
ceeded the lipid content used in the human health TBP
analysis (3.0 percent for fillet).

No additional sampling stations would be classified
as Tier 1 based on mercury or dioxins wildlife criteria.
For a sampling station to be classified as Tier 1, both sedi-
ment chemistry TBP and measured fish tissue concentra-
tions taken from that sampling station had to exceed the
wildlife criteria. At very few sampling stations in the NSI
were both sediment chemistry and fish tissue levels for
dioxin measured. In those few cases where contaminants
in both media were measured, there were no additional
sampling stations (stations not already classified as Tier
1) where both the sediment chemistry TBP and fish tissue
levels exceeded the wildlife dioxin criteria. No additional
sampling stations were classified as Her 1 for exceedance
of the wildlife criteria for mercury because sediment chem-
istry TBPs cannot be calculated for metals.

Regional and State Assessment

The remainder of this chapter presents more de-
tailed results from the evaluation of NSI data for sam-

3-17


-------
Table 3-5. River Reaches With 10 or More Tier 1 Sampling Stations Located in Watersheds Containing
APCs

EPA Region

Cataloging

Unit
Number

Cataloging Unit Name

RF1 Reach ID

RF1 Reach Name

Number of
Tierl
Stations

Total Number
of Stations in
Reach

1

01090001

Charts

01090001022

Boston Bay

72

146

01090001015

Boston Bay

42

149

01090001013

Atlantic Ocean

37

58

01090001024

Boston Bay

16

45

1

01090004

Narragansett

01090004023

Seekonk River

16

17

2

02030103

Hackensack-Passaic

02030103023

Rockaway River

26

56

2

02030104

Sandy Hook-Staten Island

02030104003

Arthur Ki

10

10

2

04120103

Biiffkb-Eghteermnile

04120103007

Buflab Creek

26

42

04120103001

Lake Brie, U.S. Store

17

22

2

04120104

Niagara

04120104007

Niagara River

12

20

2

04130001

Oak Orchard-lWelverrile

04130001001

Lake Ontario, U.S. Shore

14

27

4

03060106

Middle Savannah

03060106047

Horse Creek

10

11

4

03080103

Lower St Johns

03080103017

St Johns River

10

27

4

06010201

Walls Bar Lake

06010201026

Little River

15

23

06010201035

Tennessee River

10

12

4

06010207

Lower Clinch

06010207022

Poplar Creek

19

25

06010207021

Poplar Creek, Brushy
Fork

17

23

06010207003

Clinch River

16

20

4

06020001

Middle Temessee-Chrekamauga

06020001003

Lookout Creek

29

41

4

06030005

Pickwick Lake

06030005046

Wilson Lake

22

25.

5

04030108

Menominee

04030108001

Menominee River

10

12

5

04030204

Lower Fox

04030204001

Fox River

13

13

04030204010

Fox River

12

13

04030204004

Fox River

10

10

5

04040001

Little CahinEt-Galfcn

04040001010

Indiana Harbor

15

15

04040001006

Calumet River

12

20

5

04040002

Pike-Root

04040002002

Lake Michigan

15

33

5

04040003

Milwaukee

04040003001

Milwaukee River

48

64

5

04090004

Detroit

04090004006

Detroit River

27

38

04090004014

River Rouge

12

12

04090004011

Detroit River

11

11

04090004004

Detroit River

10

12

5

04100002

Raisin

04100002001

River Raisin

16

32

3-18


-------
Sediment (Juulity Survey

Thble 3-5. (Continued)

EPA Region

Cataloging

Unit
Number

Cataloging Unit Name

RF1 Reach ID

RF1 Reach Name

Number of
Tier 1
Stations

Total Number
of Stations in
Reach

5

07010206

Twin Cities

7010206001

Mississippi River

10

15

5

07120003

Chicago

7120003001

Chicago Sanitary Ship
Canal

35

36







7120003006

Little Calumet River

13

42

5

07120004

Des Plaines

7120004011

Des Plains River

11

20

6

08040207

Lower Ouachita

8040207005

Bayou De Siard

11

11

6

08080206

Lower Cakaseu

8080206033

Calcaseu River

13

40







8080206034

Bayou D'Inde

11

30

6

08090100

Lower Mississippi-New Orleans

8090100004

Mississippi River

13

23

9

18030012

Tblare-Buena Vista Lakes

18030012014

Kings River

10

12

9

18050004

San Francisco Bay

18050004001

San Francisco Bay

11

27

9

18070104

Santa Monica Bay

18070104003

Pacific Ocean

20

37

9

18070105

Los Angeles

18070105001

Los Angefcs River

12

31

9

18070201

Seal Beach

18070201001

Pacific Ocean

18

47

9

18070204

Newport Bay

18070204002

San Dego Creek

11

22

9

18070304

San Diego

18070304014

San Diego Bay

30

46

10

17110002

Strait of Georgia

17110002019

BeHingham Bay

13

26

10

17110013

Duwanrish

17110013003

EDiott Bay

41

100

10

17110019

Puget Sound

17110019086

Puget Sound

119

232







17110019085

Puget Sound

105

264







17110019068

Budd Inlet

41

112







17110019084

Puget Sound

32

57







17110019087

Puget Sound

32

164







17110019020

Bainbridge Island

31

88







17110019022

Sinclair Inlet

25

44

3-19


-------
LO

ro
o

Figure 3-9. Sampling Stations Classified as Tier 1 or Tier 2 Based on Wildlife criteria.


-------
Table 3-6. Increased Number of Sampling Stations Classified as Her 1 and Tier 2 by Including Wildlife

Criteria in the National Assessment"

Chemical or Chemical
Group

Number of Stations Excluding
Wildlife Assessment

Number of Stations Including
Wildlife Assessment

Her 1

Tier 2

Tier 1

Tier 2

DDT (and metabolites)

803

2,619

868

4,276

Dioxin

311

33

311

60

Mercury

1,122

3,211

1,122

5,199

PCBs

3,175

2,279

3,181

2,289

All Data

5,521

10,401

5,537

11,004

'The wildlife assessment used a default lipid content of 10.31 percent to compute the sediment chemistry TBP,

pling stations located in each of the EPA Regions and
each state. The sections that follow present the num-
ber of Tier 1» Tier 2, and Tier 3 sampling stations in
each Region and state and lists of the chemicals most
often responsible for Tier 1 and Tier 2 classifications.
Tables and figures similar to those presented in the
national assessment of sampling station evaluation re-
sults and river reach evaluation results are included.
Regional maps display the location of Tier 1 and Tier
2 sampling stations and APCs. The presentation for-
mat is identical for each Region.

These summary results are not inclusive of locations
with contaminated sediment not identified in this sur-
vey. The data compiled for the NSI are primarily from
large national electronic databases. Data from many sam-
pling and testing studies have not yet been incorporated
into the NSI. Thus, there might be additional locations

with sediment contamination that do not appear in this
summary. On the other hand, data in the inventory were
collected between 1980 and 1993 and any single mea-
surement of chemical at a sampling station, taken any
point in time during that period, could result in the clas-
sification of the sampling station in Tier 1 or Tier 2.
Because the evaluation is a screening level analysis, sam-
pling stations appearing in Tier 1 or Her 2 might not
cause unacceptable impacts. In addition, management
programs to address identified sediment contamination
might already exist.

It is important to emphasize here that some Re-
gions, such as Region 4 and Region 5, have signifi-
cantly more data in the NSI than do most other
Regions. This would, to some degree, account for the
relatively large number of sampling stations classified
as Tier 1 in these Regions.

3-21


-------
Findings

II

EPA Region. 1

Connecticut, Maine, Massachusetts, New Hampshire,
Rhode Island, Vermont

EPA evaluated 1,102 sampling stations in Region 1
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 254 of these sampling sta-
tions, and possible but infrequent (Tier 2) at 613 of these
sampling stations. For human health, data for 44 sam-
pling stations indicated probable association with adverse
effects (Her 1), and 246 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2). Overall,
this evaluation resulted in the classification of 298 sam-
pling stations (27 percent) as Tier 1,646 (59 percent) as
Tier 2, and 158 (14 percent) as Tier 3. The NSI sam-
pling stations in Region 1 were located in 131 separate
river reaches, or 5 percent of all reaches in the Region.
Two percent of all river reaches in Region 1 included at
least one Tier 1 station, 3 percent included at least one
Tier 2 station but no Tier 1 stations, and less than one
percent had only Her 3 stations (Figure 3-10), Table 3-
7 (on the following page) presents a summary of sam-
pling station classification and evaluation of river reaches
for each state and for the Region as a whole.

Figure 3-10. Region 1: Percent of River Reaches
That Include Tier 1, Tier 2, and Iter
3 Sampling Stations.

This evaluation identified 3 watersheds containing
areas of probable concern for sediment contamination
(APCs) out of the 61 watersheds (5 percent) in Region 1
(Figure 3-11). In addition, 39 percent of all watersheds
in the Region had at least one Tier 1 sampling station but
were not identified as containing APCs, 11 percent had
at least one Tier 2 station but no Tier 1 stations, and 2
percent had only Tier 3 stations. Forty-three percent of
the watersheds in Region 1 did not include a sampling
station. The locations of the watersheds containing APCs
and the Tier 1 and Tier 2 sampling stations in Region 1
are illustrated in Figure 3-12.

Within the three watersheds in Region 1 identified
as containing APCs (Table 3-8), 14 water bodies have at
least 1 Tier 1 sampling station; 3 water bodies have 10 or
more Tier 1 sampling stations (Table 3-9). The Massa-
chusetts Bay area appears to have the most significant
sediment contamination in Region 1. The water bodies
listed on Table 3-9 are not inclusive of all locations con-
taining a Tier 1 sampling station because only water bod-
ies within watersheds containing APCs are listed.

The chemicals most often associated with Her 1 and ,
Tier 2 sampling station classifications in Region 1 over-
all and in each state in Region 1 are presented in
Table 3-10.



At Least One
Tier 1 Station
39%





.y. / '"X



At Least One
Tier 2 Station and P
Zero Tier 1 Stations^"
11% £



11 APCs

B 5%

All Tier 3 Stations
2%



'No Data
43%

Total number of watersheds = 61



Figure 3-11. Region 1: Watershed Classifications.

At Least One
Tier 1 Station

9%

At Least One
Tier 2 Station and
Zero Tier 1 Stations

All Tier 3
Stations
<1%

3%

No Data
95%

Total number of river reaches = 2,648

3-22


-------
Table 3-7. Region 1: Evaluation Results for Sampling Stations and River Reaches by State



Station Evaluation

River Reach Evaluation^



Herl

Tier 2

Tier 3

Number of
Stations

Not
Identified
by an RF1
Reachb







Total#
Reaches

w/at
Least 1
Station
Evaluated



% of An
Reaches
inState

w/at
Least 1
Station
Evaluated

% of
Reaches

Wat
Least 1
Tierl or
Tier 2
Station

State

No.

%

No.

%

No.

%

Reaches

w/at
Least 1
Station in
Her 1

Reaches

w/at
Least 1
Station in
Tier2c

Reaches

w/All
Stations
in Tier 3

Total
Reaches
inState

Connecticut

20

20

67

68

11

11

8

16

24

4

44

215

21

19

Maine

13

24

37

67

5

9

28

9

7

2

18

1,583

1

1

Massachusetts

242

27

516

58

137

15

316

25

27

-

52

270

19

19

New Hampshire

4

57

1

14

2

29

-

2

-

2

4

279

1

1

Rhode Island

16

38

24

57

2

5

9

6

7

-

13

56

23

23

Vermont

3

60

1

20

1

20

-

3

-

-

3

355

1

1

REGION ld

298

27

646

59

158

14

361

59

65

7

131

2,648

5

5

•River reaches based on EPA River Reach File 1 (RF1).

"Stations not identified by an RF1 reach were located in coastal or open water areas.
cNo stations in these reaches were included in Tier 1.

'Because some reaches occur in more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------

-------
Table 3-8. Region 1: Watersheds Containing Areas of Probable Concern for Sediment Contaminat

Cataloging
Unit Number





Number of Sampling
Stations

Percent of
Sampling
Stations in Tier t
or Tier 2

Name

State(s)"

Her 1

Her 2

Tier 3

01090001

Charfes

MA

195

402

111

84

01090004

Narragansett

MA, RI

28

20

0

100

01090002

Cape Cod

MA, (RI)

15

73

20

81

on

* No dati were available for states listed in parenthesis

Table 3-9. Region 1: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs



# of Tier 1



# of Tier 1

Wiiter Body

Stations

Water Body

Stations

Boston Bay

141

Bass River

3

Atlantic Ocean

46

Potowomut River

3

Seckonk Elver

16

Conaricut Island

2

Boston Harbor and Mystic River Area

9

Pawtuxet River

2

Buzzards Bay

5

Acushnet River

1

Martha's Vineyard*

4

Charles River

1

Narragansett Bay

4

Taunton River

1

•Subsequent data review indicates these sampling stations may, in fact, be located in Buzzards Bay.

3


-------
I'ill
-------
EPA Region 2

New Jersey, New York, Puerto Rico

EPA evaluated 1,096 sampling stations in Region 2
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 319 of these sampling sta-
tions, and possible but infrequent (Tier 2) at 523 of these
sampling stations. For human health, data for 37 sam-
pling stations indicated probable association with adverse
effects (Tier 1), and 533 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2). Overall,
this evaluation resulted in the classification of 355 sam-
pling stations (32 percent) as Tier 1, 559 (51 percent) as
Tier 2, and 182 (17 percent) as Her 3. The NSI sam-
pling stations in Region 2 were located in 292 separate
river reaches, or 17 percent of all reaches in the Region.
Seven percent of all river reaches in Region 2 included
at least one Tier 1 station, 8 percent included at least one
Tier 2 station but no Tier 1 stations, and 2 percent had
only Tier 3 stations (Figure 3-13). Table 3-11 (on the
following page) presents a summary of sampling station
classification and evaluation of river reaches for each
state and for the Region as a whole.

This evaluation identified 12 watersheds containing
areas of probable concern for sediment contamination
(APCs) out of the 63 watersheds (19 percent) in Region
2 (Figure 3-14). In addition, 41 percent of all water-

2%

Total number of river reaches = 1,753

Figure 3-13. Region 2: Percent of River Reaches

That Include Iter 1, Tier 2, and Tier 3
Sampling Stations.

sheds in the Region had at least one Tier 1 sampling sta-
tion but were not identified as containing APCs, 30 per-
cent had at least one Tier 2 station but no Her 1 stations,
and none of the watersheds evaluated had only Her 3
stations. Ten percent percent of the watersheds in Re-
gion 2 did not include a sampling station. The locations
of the watersheds containing APCs and the Tier 1 and
Tier 2 sampling stations in Region 2 are illustrated in
Figure 3-15.

Within the 12 watersheds in Region 2 identified as
containing APCs (Table 3-12), 52 water bodies have at
least 1 Tier 1 sampling station; 9 water bodies have 10
or more Tier 1 sampling stations (Table 3-13). Several
areas in Region 2 appear to have significant sediment
contamination. They include the Niagara River, Buffalo
Creek, and Lake Erie near Buffalo, New York; Lake
Ontario between Rochester, New York, and the Niagara
River; the St. Lawrence River in the northern part of New
York; Arthur Kill in New York and New Jersey; the
Hackensack/Passaic watershed in New York and New
Jersey; the Atlantic Ocean beyond Staten Island; and oth-
ers. The water bodies listed on Table 3-13 are not inclu-
sive of all locations containing a Tier 1 sampling station
because only water bodies within watersheds containing
APCs are listed.

The chemicals most often associated with Tier 1 and
Tier 2 sampling station classifications in Region 2 over-
all and in each state in Region 2 are presented in
Table 3-14.

At Least One Tier 2 Station
and Zero Tier 1 Stations
30%

Total number of watersheds = 63
Figure 3-14, Region 2: Watershed Classifications.

3-27


-------
Table 3-11, Region 2: Evaluation Results for Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation*

Ueri

Tier 2

Her 3

Number of
Stations

Not
Identified
by an RFI
Reachk

Reaches

Wat
Least 1
Station in
Tier!

Reaches

Wat
Least 1
Station in
Her 2C

Reaches

Wall
Stations
in Tier 3

Total#
Reaches

Wat
Least 1
Station
Evaluated

Total
Reaches
in State

% of all

Reaches
inState

Wat
Least 1
Station
Evaluated

%of
Reaches

Wat
Least 1
Tier! or
Tier 2
Station

No.

%

No.

%

No.

%

New Jersey

142

32

228

51

78

17

62

59

56

14

129

285

45

40

New York

208

34

310

50

100

16

81

58

93

15

166

1,488

11

10

Puerto Rico

5

17

21

70

4

13

30

-

-

-

-

-

-

-

REGION 2d

355

32

559

51

182

17

173

116

147

29

292

1,753

17

15

•Rivet teaches based on EPA River Reach File 1 (RFI).

'Stations not identified by an RFI reach were located in coastal or open water areas,

•No stations ill these reaches were included in Her 1.

¦¦Because some reaches occur in more than one state, the total number of reaches in each category for the Region might tot equal the sum of reaches in the states.


-------
w Figure 3-15, Region 2: Location of Sampling Stations Classified as Tier 1 or Tier 2 and Watersheds Containing Areas of Probable Concern for
w	Sediment Contamination (APCs).

?


-------
I'indinjis

Table 3-12. Region 2: Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

Statc(s)*

Number of Sampling
Stations

Percent of
Sampling
Stations in Her 1
or Tier 2

Tier 1

Her 2

"Her 3

02030104

Sandy Hook-Staten Island

NY, NJ

60

21

19

81

04120103

Bufftio-Eghteenmile

NY

59

33

9

91

02030103

Hackers ack-Passaic

NY, NJ

43

58

2

98

04130001

Oak Orchard-TVvelvsnile

NY

39

46

1

99

04120104

Niagara

NY

24

16

I

98

04120101

Chautauqua-Conneaur

NY, PA, OH

21

.. 86

3

97

04150301

Upper St Lawrence

NY

21

5

5

84

02040202

Lower Delaware

PA, NJ

18

29

10

82

02030105

Raritan

NJ

13

37

15

77

02030202

Southern Long Island

NY

11

24

8

81

02040105

Middfe Delaware-Musconetcong

PA, NJ

11

26

11

?7

02040301

Muffica-Toms

NJ

10

22

10

76

TMble 3-13. Region 2: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs



# of Her 1



# of Tier 1

Water Body

Stations

Water Body

Stations

Lake Ontario, U.S. Shorn

31

Shrewsbury River

2

Buflhb Creek

30

Stony Bk.

2

Rockaway Riwr

26

Bass River

1

Lake Ere, U.S. Shore

24

Be den Brook

1

Atlantis Ocean

22

Big Timber Creek

1

Niagara River

21

Caasnovia Creek

1

St. Lawrence River

21

Cooper River

1

Arthur Kffl

10

Cranbury Bk.

I

Staten Island

10

Great South Bay

1

Sandy Hook Bay

8

Green Bk.

1

Delaware River

8

Hammonton C reek

1

Newark Bay

6

Matchaponix Bk.

1

Smoke Creek

6

Mistone River

1

Passak River

6

MuDica River

1

Hackensack River

5

Rahway River

1

Manasquan River

4

Rareiocas Creek, N. Br.

1

Musconetcong River

3

Raritan Bay

1

Tbnawanda Creek

3

Raritan River, N. Br.

1

Bamegat Bay

2

Raritan River, S. Br.

1

EighteereiiSe Creek

2

SB Rockaway Creek

1

Lower Bay

2

Sliinnecock Bay

1

Manalapan Bk.

2

South River

I

Moriches Bay

2

Toms River

1

Porrpton Creek

2

Wanaque Reservoir

1

Rancocas Creek, S. Br.

2

Whippany River

1

Saddb River

2

Yellow Brook

1

3-30


-------
jN;i(ion;iI Scdlfiiciif OuaJilvSur\< \

I	¦		_	I		• i	:	¦	;			V

Table 3-14. Region 2; Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling Station
Classifications*



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station

Region 2
Overall

Copper
Lead

546

467



546
467

New Jersey
(continued)

Cadmium
Chromium

128
119

22

128
97



Nickel

443

-

443

New York

Copper

332

-

332



Polychiorinated biphenyls

442

151

291



Nickel

321

-

321



Mercury

388

144

244



Lead

268

-

268



Cadmium

360

--

360



Polychiorinated biphenyls

261

108

153



Zinc

358

--

358



Cadmium

230

..

230



DDT

351

114

237



Mercury

224

70

154



Arsenic

282

6

276



Zinc

210

_

210



Chromium

247

26

221



DDT

155

66

89



Chlordane

229

--

229



Pyrene

147

52

95



Pyrene

214

64

150



Chromium

126

4

122



Benzo(a)pyrene

180

36

144

Puerto Rico

Copper

22

-

22



Naphthalene

155

30

125



Nickel

10

-

10



Fluoranthene

151

41

110



Arsenic

9

-

9

New Jersey

DDT

195

48

147



Lead

8

-

8



Copper

192

-

192



Mercury

6

4

2



Lead

191

-

191



Zinc

5

--

5



Polychiorinated biphenyls

181

43

138



Silver

4

1

3



Mercury

158

70

88



Bis(2-ethylhexyl)phthalat

2

1

1



Arsenic

151

6

145



Diethyl phthalate

2

1

!



Zinc

143

-

143



Cadmium

2

_

2



Chlordane

139



139











'Stations may be listed for more ihao one chemical.

3-31


-------
E

imlrnjjs

EPA Region 3

Delaware, District of Columbia, Maryland, Pennsylva-
nia, Virginia, West Virginia

EPA evaluated 1,910 sampling stations in Region 3
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able CTier 1) was found at 86 of these sampling stations,
and possible but infrequent (Tier 2) at 915 of these sam-
pling stations. For human health, data for 239 sampling
stations indicated probable association with adverse ef-
fects (Her 1), and 222 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2). Overall,
this evaluation resulted in the classification of 318 sam-
pling stations (17 percent) as Tier 1,934 (49 percent) as
Her 2, and 658 (34 percent) as Tier 3. He NSI sam-
pling stations in Region 3 were located in 888 separate
river reaches, or 27 percent of all reaches in the Region.
Six percent of all river reaches in Region 3 included at
least one Her 1 station, 14 percent included at least one
Tier 2 station but no Tier 1 stations, and 7 percent had
only Tier 3 stations (Figure 3-16). Table 3-15 (on the
following page) presents a summary of sampling station
classification and evaluation of river reaches for each
state and for the Region as a whole.

This evaluation identified 8 watersheds containing
areas of probable concern for sediment contamination
(APCs) out of the 128 watersheds (6 percent) in Region
3 (Figure 3-17). In addition, 63 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not identified as containing APCs, 22
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 5 percent had only Tier 3 stations. Four per-
cent of the watersheds in Region 3 did not include a
sampling station. The locations of the watersheds con-
taining APCs and the Tier 1 and Tier 2 sampling stations
in Region 3 are illustrated in Figure 3-18.

Within the 8 watersheds in Region 3 identified as
containing APCs (Table 3-16), 27 water bodies have at
least 1 Tier 1 sampling station; 4 water bodies have 10 or
more Tier 1 sampling stations (Table 3-17). The Dela-
ware River; the Schuykill River in Pennsylvania (near
Philadelphia); coastal areas of Lake Erie near Erie, Penn-
sylvania; and the Ohio River near Pittsburgh appear to
have some of the most significant sediment contamina-
tion in Region 3. The water bodies listed on Table 3-17
are not inclusive of all locations containing a Tier 1 sta-
tion because only water bodies within watersheds con-
taining APCs are listed.

The chemicals most often associated with Her 1 and
Tier 2 sampling station classifications in Region 3 over-
all and in each state in Region 3 are presented in
Table 3-18.

No Data/
73% /

% ;. J At Least One
ffcjgtJaTier 1 Station

HSpjf 6%

All Tier 3

Stations

7%

"yAt Least One
Tier 2 Station
and Zero Tier 1
Stations
14%

Total number of river reaches = 3,247



At Least One Her 2 Station
and Zero Tier 1 Stations

22%

Total number of watersheds = 128

Figure 3-16. Region 3: Percent of River Reaches Figure 3-17. Region 3: Watershed Classifications.
That Include Tier 1, Tier 2 and Tier 3
Sampling Stations.

3-32


-------
Table 3-15, Region 3: Evaluation Results for Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation*

Her 1

Her 2

Her 3

Number of
Stations

Not
Identified
by an RF1
Reachh

Reaches

Wat
Least 1
Station in
Herl

Reaches

Wat
Least 1
Station in
Her2«

Reaches

WaD
Stations
in Her 3

Total P
Reaches

Wat
Least 1
Station
Evaluated

Total
Reaches
inState

% of all
Reaches
inState

Wat
Least 1
Station
Evaluated

% of
Reaches

Wat
Least 1
Tier 1 or
Tier 2
Station

No.

%

No.

%

No.

%

Delaware

21

10

35

16

162

74

13

10

7

22

39

77

51

22

District of Columbia

3

75

1

25

¦

-

-

3

-

-

3

11

27

27

Maryland

50

24

68

33

88

43

29

31

36

30

97

400

24

17

Pennsylvania

127

41

106

34

78

25

4

78

27

34

139

677

21

16

Virginia .

73

7

691

66

287

27

46

61

362

112

535

1279

42

33

West Virginia

44

37

33

27

43

36

-

30

23

31

84

993

9

5

REGION 3d

318

17

934

49

658

34

92

209

453

226

888

3247

27

20

'River reaches based on EPA River Reach File 1 (RI'l).

'Stations not identified by an RF1 reach were located in coastal or open water areas.

'No stations in these reaches were included in Tier 1,

'Because some reaches occur in more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------
I?

Figure 3-18, Region 3: Location of Sampling Stations Classified as Her 1 of Tier 2 and Watersheds Containing Areas of Probable Concern for
Sediment Contamination (APCs).


-------
Tfabie 3-16, Region 3; Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State (s)*

Number of Sampling
Stations

Percent of
Sampling
Stations lit Tlerl
or Tier 2

Tlerl

Tlerl

Tier 3

04120101

Chaiteuqua-Conneaut

NY.PA.OH

21

86

3

97

02040202

Lower Delaware

PA.NJ

18

29

10

82

02060003

Gunpowder-Patapseo

MD,(PA)

17

7

5

83

02040203

Sctaytkfl

PA

12

23

9

80

05030101

Upper Oho

WyPA,OH

12

29

12

77

02040105

Middle Delaware-Musconetcong ¦

PAJSfJ

11

26

11

77

02070(X)4

Conoco che ague-Opequon

WV,VA,MD,(P-
A)

11

12

6

79

05030102

Shenango

OH,PA

11

1

3

80

•No dati were available for states listed in parentheses.

Table 3-17, Region 3: Water Bodies With Sampling Stations Classified as Her 1 Located in Watersheds
Containing APCs



# of Tier 1



# of Tier 1

W»ferBo
-------
Finding

Table 3-18, Regjon 3: Chemicals Most Often Associated With Tier 1 or Her 2 Sampling Station
Classifications*





# Tier 1









# Tier I









& Tier 2

# Tier 1

# Tier 2





& Tier 2

# Tier 1

# Tier 2



Chemical

Stations

Station

Station



Chemical

Stations

Station

Station

Region 3

Nickel

634

„

634

Maryland

Nickel

50

-

50

Overall

Copper

626

-

626

(continued)

Copper

42

--

42



Lead

626

-

626



Chromium

41

4

37



Arsenic

529

1

528



DDT

35

-

35



Zinc

371

-

371



Chlordane

33

-

33



Polychtorinated biphenyls

353

243

no



Zinc

32

--

32



Cadmium

346

-

346



Benzo(a)pyrcne

31

_

31



Mercury

320

42

278

Pennsylvania

Polychlorinated biphenyls

141

112

29



Chromium

249

12

237



Lead

87

--

87



Chlordane

161

_

161



Chlordane

81

-

81



DDT

135

9

126



Nickel

63

-

63



Dieldrin

116

-

116



Cadmium

56

_

56



Benzo(a)pyrene

106

6

100



Dieldrin

55

_

55



BHC

69

2

67



Copper

46

_

46



Dibenzo(aJi)anthraccne

64

4

60



Zinc

44

--

44

Delaware

Polychiorinated biphenyls

33

14

19



DDT

38

6

32



DDT

27

3

24



Mercury

25

3

22



Lead

24

_

24

Virginia

Copper

520

-

520



Chromium

19

2

17



Nickel

497

--

497



Arsenic

18

_

18



Arsenic

412

_

412



Nickel

15

-

15



Lead

411

-

411



BHC

13

_

13



Zinc

279

-

279



Mercury

12

3

9



Mercury

260

34

226



Benzotajpyiene

12

_

12



Cadmium

255

-

255



Copper

8

-

S



Chromium

167

3

164

District of

PolychJorinated biphenyls

4

2

2



Polychlorinated biphenyls

62

30

32

Columbia

Dioxlns

2

2

-



Benzo(a)pyrene

48

4

44



Benzo(a)pyrene

2

-

2

West Virginia

Polychlorinated biphenyls

42

41

-



Chlordane

2

-

2



Lead

35

¦: --

35



Copper

2

-

2



Chlordane

29

-

29



Dieldrin

2

_

2



Dieldrin

16

.

16



Nickel

2

..

2



Cadmium

12

-

12



Sliver

1

1

-



Copper

8

-

8



Arsenic

i

-

I



Zinc

8

..

8



Benzo(a)anthracene

1



1



HeptachJor epoxide

7

_

7

Maryland

Polychlorinated biphenyls

71

44

27



Nickel

7

-

7



Arsenic

70

_

70



Aldrin

6

-

6



Lead

68

-

68











'Stations may fee lifted for more than one chemical.

3-36


-------
EPA Region 4

Alabama, Florida, Georgia, Kentucky, Mississippi, North
Carolina, South Carolina, Tennessee

EPA evaluated 4,959 sampling stations in Region 4
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 637 of these sampling sta-
tions, and possible but infrequent (Tier 2) at 1,888 of
these sampling stations. For human health, data for 561
sampling stations indicated probable association with ad-
verse effects (Tier 1), and 1,006 sampling stations indi-
cated possible but infrequent adverse effects (Tier 2).
Overall, this evaluation resulted in the classification of
1,157 sampling stations (23 percent) as Tier 1, 1,930 (39
percent) as Tier 2, and 1,872 (38 percent) as Tier 3. The
NSI sampling stations in Region 4 were located in 1,770
separate river reaches, or 18 percent of all reaches in the
Region. Six percent of all river reaches in Region 4 In-
cluded at least one Tier 1 station, 7 percent included at
least one Tier 2 station but no Tier 1 stations, and 5 per-
cent had only Tier 3 stations (Figure 3-19). Table 3-19
(on the following page) presents a summary of sampling
station classification and evaluation of river reaches for
each state and for the Region as a whole.

This evaluation identified 19 watersheds containing
areas of probable concern for sediment contamination

(APCs) out of the 308 watersheds (6 percent) in Region
4 (Figure 3-20). In addition, 59 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not identified as containing APCs, 17
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 8 percent had only Tier 3 stations. Ten per-
cent of the watersheds in Region 4 did not include a
sampling station. The locations of the watersheds con-
taining APCs and the Tier 1 and Tier 2 sampling stations
in Region 4 are illustrated in Figure 3-21.

Within the 19 watersheds in Region 4-identified as
containing APCs (Table 3-20), 65 water bodies have at
least 1 Tier 1 sampling station; 15 water bodies have 10 or
more Tier 1 sampling stations (Table 3-21). Several areas
in Region 4 appear to have potential sediment contamina-
tion. They include the Tennessee River and Lookout Creek
in Tennessee and Georgia, Wilson Lake and Mobile Bay
in Alabama, the St Johns River in Florida, and other loca-
tions. The water bodies listed on Table 3-21 are not inclu-
sive of all locations containing a Her 1 sampling station
because only water bodies within watersheds containing
APCs are listed.

The chemicals most often associated with Tier 1 and
Tier 2 sampling station classifications in Region 4 overall
and in each state in Region 4 are presented in Table 3-22.

No Data
82%

At Least One
Tier 1 Station
6%

At Least One
Tier 2 Station
and Zero Tier 1
Stations
All Her 3 7%
Stations
5%

Total number of river reaches - 9,749

Figure 3-19. Region 4; Percent of River Reaches

That Include Tier 1, Tier 2, and Tier 3
Sampling Stations.

At Least One
Tier 1 Station.

5i%

No Data
10%

Jl Tier 3 Stations

8%

At Least One Tier 2 Station
and Zero Tier 1 Stations
17%

Total number of watersheds = 308

Figure 3-20. Region 4i Watershed Classifications.

3-37


-------
Table 3-19. Region 4: Evaluation Results for Sampling Stations and River Reaches by State

State

State Evaluation

River Reach Evaluation*

Tier I

Ber2

Her 3

NimVberof
Stations

Not
Identified
by anRFl
Reach*

Reaches

Wat
Least 1
Station in
Her I

Reaches

Wat
Least I
Station in

Iter 2*

Reaches

Wall
Stations

to Her 3

Total#

Reaches

Wat
Least 1
Station
Evaluated

Total

Readies
instate

% of all

Reaches
inState

Wat
Least 1
Station
Evaluated

%of
Reaches

Wat
Least 1
Her 1 or
Her 2
Statioa

No.

%

No,

%

No.

%

Alabama

160

34

178

37

139

29

65

68

57

57

182

1,531

12

8

Florida

211

12

672

38

893

50

190

70

115

126

311

855

36

22

Georpa

115

36

100

32

103

32

3

75

57

54

186

1,658

11

8

Kentucky

69

28

131

52

49

20

-

49

60

26

135

1,247

11

9

Mississippi

54

17

142

45

122

38

61

21

47

35

103

984

11

7

Noah Carolina

71

12

294

48

247

40

22

50

156

107

313

1,415

22

15

South CatoBna

161

29

254

45

148

26

2

105

138

28

271

1,055

26

23

Tennessee

316

49

159

25

171

26

-

132

63

97

292

1,417

21

14

REGION ¥

1,157

23

1,930

39

1,872

38

343

566

684

520

1 7"»ti



. 18

13

•River reaches based on EPA River Reach FUe 1 (RF!>.

'Stations not identified by an RF1 reset were located to coastal or open water areas.

•No stations in these reaches were included in Tier 1.

^Because some reaches occur to more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------


Figure 3-21. Region 4; Location of Sampling Stations Classified as Tier 1 or Tier 2 and Watersheds Containing Areas of Probable Concern for
Sediment Contamination (APCs).


-------
Findings.

Tbble 3-20. Region 4; Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State(s)"

Number of Sampling
Stations

Percent of
Sampling
Stations in Tier 1
or Her 2

Tier I

Tier 2

Tier 3

06010201

Watts Bar Lake

TN

63

7

19

79

06010207

Lower Cftich

TN

61

14

4

95

06030005

Pickwick Lake

TN, AL, (MS)

49

9

11

84

06020001

Miridfc Tennessee- ChickanHuga

GA, TN, (AL)

47

29

18

81

03080103

Lower St Johns

FL

32

111

45

76

03160205

MobJb Bay

AL

31

43

7

91

06030001

GunlersvBte Lake

TN, AL, (CA)

25

46

21

77

03130002

Miridfe Chattahoochee-Lake
Harding

GA, (AL)

21

4

2

93

03060106

Middle Savannah

GA, SC

20

11

5

86

03140102

Cboctawhatchee Bay

CT

19

23

9

82

06040001

Lower Temessee-Beech

TN, (MS)

15

6

4'

84

06040005

Kentucky Lake

KY, TN'

15

14 .

1

97

08010100

Lower Mississippi-Mentha

AR, MS, KY,

MO, TN

14

3

3

85

06020002

Hlwassee

GA. NC. TN

13

17

3

91

06010104

Hofeton

TN

12

2

1

93

03040201

Lower Pee Dee

NC, SC

11

20

3

91

08030209

Deer-Steels

MS, (LA)

11

10

0

100

03060101

Seneca

NC, SC

10

3

3

81

03140107

Peidklo Bay

FL, AL

10

24

4

89

*Ko dsta were iviitiWe for stttet listed la pareothescs.

3-40


-------
Table 3-21. Region 4: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs



# of Tier 1



# of Tier 1

Water Body

Stations

Water Body

Stations

Tennessee River

80

Cypress Creek

2

St. Johns River

30

Deer River

2

Lookout Creek

29

Long Cane Creek

2

Mobile Bay

29

Seneca River

2

Wilson Lake

27

Shoal Creek

2

Poplar Creek

21

Spring Creek

2

Clinch River

18

Twelvemile Creek

2

Choctawhatchee Bay

17

West Pont Lake

2

Guntersvilte Lake

17

Beech Creek

1

Poplar Creek, Bmsfcy Fork

17

Big Black Creek

1

Little River

1,6

Big Sandy Creek

1

Chattahoochee River

14

Chatugue Lake

1

Walts Bar Lake

14

Conecross Creek

I

Mississippi River

12

Coon Creek

1

Horse Creek

10

Hlevenmile Creek

1

Black Bayou

9

Golden Creek

1

Holston River

9

Hiwassee Lake

1

Kentucky Lake

9

Jeffries Creek

1

Savannah River

9

Lake Harding

1

Hiwassee River

8

Lake Keowee

1

Perdido Bay

7

Lake Washington

1

Melton Hill Lake

5

Lafayette Creek

1

Cherokee Lake

3

Little Horse Creek

1

Fort Loudoun Lake

3

Mountain Creek

1

Gulf Of Mexico

•3

Mud Creek

1

Reservoir

3

Notlely Lake

1

Lake Chickamauga

3

Oostanaula Creek

1

Pee Dee River

3

Pottsburg Creek

i

Pickwick Lake

3

Rogers Creek

1

Big Nance Creek

2

Sinking Creek

1

Black Creek

2

Steele Bayou

1

Catfish Creek

2

Sweetwater Creek

1

Crooked Creek

2





3-41


-------
Findings

Table 3-22. Region 4: Chemicals Most Often Associated With Her 1 or Tier 2 Sampling Station
Classifications*





# Tier 1







# Tier 1









& Tier 2

# Tier 1

# Tier 21



& Tier 2

#Tierl

# Tier 2



Chemical

Stations

Station

Station 1

Chemical

Stations

Station

Station

Region 4

Polychlorinated biphenyls

1034

669

365

Kentucky

Arsenic

65

3

62

Overall

Lead

989

_

989

(continued)

Copper

55

-

55



Copper

935

„

935



Polychlorinated biphenyls

50

48

2



Mercury

923

235

688



Zinc

43

-

43



Nickel

820

--

820



Chlordane

41

3

38



DDT

751

157

594



Dieldrin

40

' 3

37



Cadmium

751

..

751



Mercury

35

5

30



Arsenic

734

37

697

Mississippi

DDT

99

31

68



Chromium

459

26

433



Nickel

66

-

66



Zinc

438

_

438



Arsenic

63

1

62



Chlordane

374

7

367



Polychlorinated biphenyls

44

15

29



Benzo(a;pyrene

289

28

261



Cadmium

33

_

33



Pyrene

279

62

217



Chromium

32

_

32



Dieldrin

252

9

243



Lead

28

-

28



Fluoranther.e

207

34

173



Dieldrin

24

-

24

Alabama

Mercury

125

42

83



Copper

22

_

22



Arsenic

118

4

114



Senzo{a)pyren6

13

-

13



Polychlorinated biphenyls

114

98

16

North

Copper

150



150



Cadmium

103

-

103

Carolina

Mercury

133

30

103



Nickel

97



97



Lead

128

-

128



Copper

94

_

94



Nickel

99

-

99



Lead

85

_

85



Arsenic

75

_

75



DDT

76

8

68



Chromium

72

2

70



Zinc

76

_

76



Cadmium

62



62



Chromium

69

I

68



Polychlorinated biphenyls

60

28

32

Florida

Mercury

302

52

250



Zinc

45

-

45



Polychlorinated biphenyls

293

82

211



DDT

27

1

26



Lead

291

_

291

South

Lead

198

_

198



Copper

283

_

283

Carolina

DDT

188

48

140



DDT

242

48

194



Mercury

144

19

125



Cadmium

208



208



Copper

141

_

141



Benzo(a)pyrene

193

19

174



Polychlorinated biphenyls

132

93

39



Pyrene

176

30

146



Nickel

131

_

131



Arsenic

171

7

164



Cadmium

129

_

129



Chlordane

169

_

169



Chromium

63

12

51

Georgia

Polychlorinated biphenyls

111

82

29



Arsenic

62

18

44



Aisenic

62



62



Zinc

58

-

58



Cadmium

60

_

60

Tennessee

Polychlorinated biphenyls

230

223

7



Copper

60

_

60



Nickel

164

_

164



Lead

46

_

46



Lead

137



137



Chlordane

45

4

41



Mercury

134

75

59



Mercury

43

12

31



Copper

130

-

130



Nickel

38

„

38



Arsenic

118

4

114



DDT

36

11

25



Cadmium

87

-

87



Chromium

33

2

31



Zinc

83



83

Kentucky

Nickel

105

_

105



DDT

57

6

51



Lead

76

..

76



Dieldrin

52

3

49



Cadmium

69

-

69











'SuUoei may fee listed for matt than cm chemical.

3-42


-------
EPA Region. 6

Illinois, Indiana, Michigan, Minnesota, Ohio, Wisconsin

EPA evaluated 4,290 sampling stations in Region
5 as part of the NSI evaluation. Sediment contamina-
tion where associated adverse effects to aquatic life are
probable (Tier 1) was found at 642 of these sampling
stations, and possible but infrequent (Tier 2) at 2,011 of
these sampling stations. For human health, data for 777
sampling stations indicated probable association with ad-
verse effects (Tier 1), and 1,469 sampling stations indi-
cated possible but infrequent adverse effects (Tier 2).
Overall, this evaluation resulted in the classification
of 1,418 sampling stations (33 percent) as Tier 1,2,137
(50 percent) as Tier 2, and 735 (17 percent) as Tier 3.
(It should be noted that the NSI includes sampling data
from the Great Lakes Sediment Inventory that, because
of a lack of latitude and longitude data, were not in-
cluded in the NSI evaluation. Had those data been
included in the NSI evaluation, an additional 221 sta-
tions would have been categorized as Tier 1, 392 as
Tier 2, and 84 as Tier 3.) The NSI sampling stations
in Region 5 were located in 1,432 separate river
reaches, or 24 percent of all reaches in the Region.
Ten percent of all river reaches in Region 5 included
at least one Tier 1 station, 10 percent included at least
one Tier 2 station but no Tier 1 stations, and 4 percent
had only Tier 3 stations (Figure 3-22). Table 3-23 (on
the following page) presents a summary of sampling sta-
tion classification and evaluation of river reaches for each
state and for the Region as a whole.

This evaluation identified 36 watersheds containing
arras of probable concern for sediment contamination
(APCs) out of the 278 watersheds (13 percent) in Re-
gion 5 (Figure 3-23). In addition, 59 percent of all wa-
tersheds in the Region had at least one Tier 1 sampling
station but were not categorized as containing APCs,
7 percent had at least one Tier 2 station but no Tier 1
stations, and 3 percent had only Tier 3 stations. Eigh-
teen percent of the watersheds in Region 5 did not in-
clude a sampling station. The locations of the watersheds
containing APCs and the Tier 1 and Tier 2 sampling
stations in Region 5 are illustrated in Figure 3-24.

Within the 36 watersheds in Region 5 identified
as containing APCs (Table 3-24), 102 water bodies
have at least 1 Tier 1 sampling station; 18 water bod-
ies have 10 or more Tier 1 sampling stations (Table 3-
25). The Detroit River, Fox River, Milwaukee River,
Mississippi River, Chicago Ship Canal, and several
coastal areas of Lake Michigan and Lake Erie appear
to have the most significant sediment contamination
in Region 5. The water bodies listed on Table 3-25
are not inclusive of all locations containing a Tier 1
sampling station because only water bodies within
watersheds containing APCs are listed.

The chemicals most often associated with Tier 1
and Tier 2 sampling station classifications in Region 5
overall and in each state in Region 5 are presented in
Table 3-26.

No Data /
76% /

c-'.'j At Least One
WS Tier 1 Station
V-7 10%

All Tier 3
Stations
4%

' At Least One
Tier 2 Station
and Zero Tier 1
Stations
10%

Total number of river reaches = 6,025



7%

Total number of watersheds = 278

Figure 3-22. Region 5: Percent of River Reaches Figure 3-23. Region S: Watershed Classlflcations.
That Include Tier 1, Tier 2, and Tier 3
Sampling Stations.

3-43


-------
Table 3-23, Region 5: Evaluation Results for Sampling Stations and River Reaches by State

Slate

Station Evaluation

River Reach Evaluation*

•Berl

Tier 2

Her 3

Number of
Stations

Not
Identified
by anRFl
( Reach1'

Reaches

wfat
Least 1
Station in
Tier 1

Reaches

wfot
Least 1
Station in

Her 2"

Reaches

Wall
Stations

in Her 3

Total #

Readies

Wat
Least 1
Station

Evaluated

Tbtal
Reaches
inStale

%oraH

Reaches
in State

Wat
Least 1
Station
Evaluated

%Qf
Readies

Wat
Least 1
Tier lor
Tier 2
Station

No,

%

No.

%

No.

%

Hinois

428

26

1,075

64

166

W

8

182

255

30

467

920

51

48

Indiana

67

62

23

21

18

17

3

35

8

1

44

559

8

8

Michigan

219

54

144

36

39

10

20

64

41

11

116

1,145

10

9

Minnesota

220

50

65

15

153

35

-

140

34

90

264

1,355

20

13

Oto

130

13

704

73

136

14

71

56

191

57

304

1,054

29

23

Wisconsin

354

50

126

18

223

32

6

130

47

82

259

1,174

22

15

REGION 5'

1,418

33

2,137

50

735

17

108

594

570

268

1,432

6,025

24

19

*River roaches based on EPA River Reach Hie 1 (RF1).

^Stations not identified by an RF1 reach were located in coasial or open water areas.
cNo stations in these reaches were included in Her I,

^Because some reaches occur in more than oik slate, the total number of reaches in each category for the Region might not equal the sum of reaches in the slates.


-------

-------
Thble 3-24. Region 5: Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number





Number of Sanding
Stations

Percent of
Sampling

Name

State (s)*

TSep 1

Tier 2

Tier 3

or Her 2

04090004

Detroit

MI

85

29

1

99

07120003

Chicago

IN, EL.

64

36

3

97

07120004

DesPIaines

WI.IL

61

43

6

95

04040003

Milwaukee

WI

60

16

14

84

04030204

Lower Fox

WI

49

2

0

100

04040001

Little Calumst-Gafien

IL, IN, (MI) •

45

26

18

80

04040002

Pike-Root

WI, EL

34

30

8

89

07140201,

Upper Kaskaskia

IL

31

24

0

100

07010206

TWin Cities

WI, MN

26

2

7

80

04110001

Black-Rocky

OH

24

31

4

93

07140106

Big Muddy

IL

23

65

6

94

04120101

Chautauqua-Cormeaut

NY, PA, OH

21

86

3

97

07070003

Castk Rock

WI

20

0

2

91

04100002

Kaisil

MI, (OH)

18

19

1

97

07140101

Cabokia-Joachim

MO, IL

18

34

4

93

04050001

St. Joseph

IN, MI

17

9

6

81

07040003

Buffab-Whitewater

WI, MN

17

3

6

77

07080101

Copperas-Duck

IL, IA

17

5

5

81

05120111

Middle Wabash-Busseron

IN, IL

15

17

1

97

07120006

Upper Fox

WI, IL

15

40

5

92

04090002

Lake St Clair

MI

13

5

1

95

04100001

Ottawa-Stony

OH, MI

13

15

1

97

04100010

Cedar-Portage

MI, OH

13

39

4

93

07040001

Rush-Venriffion

WI, MN

13

1

0

100

07140202

Middle Kaskaskia •

IL

13

22

3

92

04030102

Door-Kewaunee

WI

12

5

3

85

04030108

Menominee

MI, WI

12

6

3

86

05030101

Upper Ohio

WV, PA, OH

12

29

12

77

05120109

\fermtBon

IL, (IN)

12

16

0

100

04060103

Manistee

MI

11

3

0

100

05030102

Shenango

OH, PA

11

1

3

80

07130001

Lower linois-Senachwine Lake

IL

11

10

0

100

04100012

Huron-\fentilkm

OH

10

35

0

100

04110003

Ashtabula-Chagrin

OH

10

18

3

90

05040001

Tbscarawas

OH

10

53

15

81

07090006

Kishwaukee

IL, (WI)

10

24

0

100

•No diu were available for states listed in parentheses.

3-46


-------
Table 3-25. Region 5: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs

Water Body

# of Tier 1
Stations

Water Body

# of Tier 1
Stations

Detroit River

64

Becks Creek

2

Lake Erie, U.S. Shore

60

Castle Rock Flowage

2

Fox River

58

Coldwater River

2

Mississippi River

56

Crab Orchard Creek

2

Milwaukee River

55

Crooked Creek

2

Lake Michigan

45

Hickory Creek

2

Chicago Sanitary Ship Canal

41

Kaskaskia Creek, E. Fork

2

Des Plains River

27

Kaskaskia River, Lake Fork

2

Kaskaskia River

21

Lake Shelbyville

2

Calumet River

19

Little Creek

2

River Raisin

16

Portage River, E. Br.

2

Indiana Harbor

15

Ramsey Creek

2

Wisconsin River

15

Saline River

2

Wabash River

14

Vermilion River

2

Lake St. Clair

13

Barton Lake

1

Little Calumet River

13

Beaucoup Creek

1

River Rouge

13

Big Bureau Creek

1

Menominee River

12

Big Muddy River, M. Fork

1

Du Page River

9

Buffalo Creek

1

Illinois River

9

Bums Ditch

1

Cahokia Canal

8

Clark Lake

1

Manistee Lake

8

Coon River

1

Big Muddy River, Casey Fork

7

Deep River

1

Black River

7

East River

1

Crab Orchard Lake

7

Eliza Creek

1

Du Page River, E. Br.

7

Garvin Brook

1

Du Page River, W. Br.

7

Gilmore Creek

1

Grosse Isle

7

Grosse Isle

1

Lake Minnetonka

7

Hog Creek

1

St. Joseph River

7

Kaskaskia Creek, N. Fork

1

Tuscarawas River

7

Kilbourn Ditch

1

Lake Calumet

6

Killbuck Creek

1

Ashtabula River

5

Lake Creek

1

Cedar Creek

5

Lemon weir River

I

Fox Lake

5

Little Crooked Creek

1

Kishwaukee River, S. Br.

5

Little Roche A Cri Creek

1 ,

Lake Michigan,•Green Bay

5

Mill Creek

1

Chicago Ship Canal

4

Ottawa Creek

1

Root River

4

Petenwell Flowage

1

Salt Creek

4

Pigeon River

1

Vermilion River, Salt Fork

4

Piscasaw River

1

Big Muddy River

3

Rend Lake

1

Chicago River, N. Br.

3

Rocky River

1

Huron River

3

Sturgeon Bay

1

Kishwaukee River

3

Sugar Creek

1

3-


-------
I

'hidings

Table 3-25. (continued)



# of Tier 1



# of Tier 1

Water Body

Stations

Water Body

Stations

Manistee River

3

Swan Creek

I

Nimishillen Creek

3

Upper Salt Fork Drainage Ditch

I

Ohnathan Creek

3

Vermilion River, M. Fork

1

Paw Paw River

3

W Bureau Creek

1

Vermilion River, N. Fork

3

Wall Town Drainage Ditch

I

W Okaw River

3

Whitewater River

1

Table 3-26. Region 5: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling Station Classifications*



Chemical

# Tier J
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station



Chemical

# Tier 1
& Tier 2
Stations

# Tier I
Station

# Tier 2
Station

Region 5
Overall

Copper

Polychlorinated biphenyls

1,625
1,460

1,113

1,625
347

Michigan
(continued)

Nickel
DDT

198
182

97

198
85



Lead

1,326

~

1,326



Zinc

170

-

170



Dieldrin

1,318

36

1,282



Mercury

140

53

87



Nickel

1,260

-

1,260



Pyrene

140

50

90



Cadmium

1,203

-

1,203



Cadmium

140

-

140



Arsenic

1,019

32

987



Fluoranthene

133

20

113



Zinc

915

-

915

Minnesota

Polychlorinated biphenyls

225

216

¦ 9



Mercury

761

197

564



Dieldrin

88

-

88



Chlordane

723

-

723



Cadmium

66

_

66



DDT

668

Ill

491



DDT

30

-

30



Chromium

414

81

333



Copper

24

-

24



Heptachlor epoxide

338

-

338



Lead

21

-

21



Pyrene

300

103

197



Mercury

17

-

17



Fluoranthene

290

59

231



Dioxins

10

10

-

Illinois

Dieldrin

1019

33

986



Chromium

9

-

9



Copper

616

_

616



Aldrin

5

_

5



Chlordane

518

_

518

Ohio

Nickel

644

-

644



Polychlorinated biphenyls

503

318

185



Copper

577

-

577



Lead

464

-

464



Lead

472

_

472



Cadmium

460

-

460



Arsenic

459

2

457



Arsenic

380

18

362



Cadmium

420

-

420



Nickel

342

_

342



Zinc

381

_

381



Mercury

330

72

258



Mercury

125

16

109



DDT

275

36

239



Chromium

123

19

104

Indiana

Polychlorinated biphenyls

66

59

7



Fluoranthene

108

17

91



Arsenic

53

3

50



Polychlorinated biphenyls

97

65

32



Dieldrin

51

3

48

Wisconsin

Polychlorinated biphenyls

319

304

15



Chlordane

48

-

48



Copper

159

-

159



Heptachlor epoxide

42

-

42



Mercury

127

42

85



Copper

36

--

36



Lead

120

-

120



Lead

36

-

36



DDT

100

15

85



BHC

33

7

26



Cadmium

88

_

88



DDT

33

6

27



Dieldrin

76

-

76



Cadmium

29

-

29



Pyrene

62

21

41

Michigan

Polychlorinated biphenyls

250

151

99



Zinc

60

-

60



Copper

213

-

213



Nickei

54

-

54



Lead

213

-

213











'Suilosu m»y be listed for mote Ihaa one chemical.

3-48


-------
EPA Region 6

Arkansas, Louisiana, New Mexico, Oklahoma, Texas

EPA evaluated 1,616 sampling stations in Region 6
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 222 of these sampling sta-
tions, and possible but infrequent (Tier 2) at 852 of these
sampling stations. For human health, data for 189 sam-
pling stations indicated probable association with adverse
effects (Tier 1), and 421 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2). Overall,
this evaluation resulted in the classification of 382 sam-
pling stations (24 percent) as Tier 1, 837 (52 percent) as
Tier 2, and 397 (24 percent) as Tier 3. The NSI sam-
pling stations in Region 6 were located in 799 separate
river reaches, or 11 percent of all reaches in the Region.
Three percent of all river reaches in Region 6 included
at least one Tier 1 station, 5 percent included at least one
Tier 2 station but no Tier 1 stations, and 3 percent had
only Tier 3 stations (Figure 3-25). Table 3-27 (on the
following page) presents a summary of sampling station
classification and evaluation of river reaches for each
state and for the Region as a whole.

This evaluation identified 8 watersheds containing
areas of probable concern for sediment contamination

(APCs) out of the 403 watersheds (2 percent) in Region
6 (Figure 3-26). In addition, 36 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not identified as containing APCs, 21
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 10 percent had only Tier 3 stations. Thirty-
one percent of the watersheds in Region 6 did not include
a sampling station. The locations of the watersheds con-
taining APCs and the Tier 1 and Her 2 sampling stations
in Region 6 are illustrated in Figure 3-27.

Within the 8 watersheds in Region 6 identified as
containing APCs (Table 3-28), 17 water bodies have at
least 1 Tier 1 sampling station; 4 water bodies have 10 or
more Tier 1 sampling stations (Table 3-29). The
Calcasieu River and Mississippi River in Louisiana ap-
pear to have some of the most significant sediment con-
tamination in Region 6. The water bodies listed on Table
3-29 are not inclusive of all locations containing a Tier 1
sampling station because only water bodies within wa-
tersheds containing APCs are listed.

The chemicals most often associated with Her 1 or Tier
2 sampling station classifications in Region 6 overall and
in each state in Region 6 are presented in Table 3-30.

Total number of river reaches = 7.293



At Least One



Fe 1 Station



36%

At Least One -Cvv V



Tier 2 Station and iVVv



Zero Tier 1 StationsfOOvV

J APCs

21% fXVvvy^?—

2%

\ /



v J

Mo Data

All Tier 3 Stations^/

10% 	

31%

Total number of watersheds = 403



Figure 3-25. Region 6: Percent of River Reaches Figure 3-26. Region 6: Watershed Classifications.
That Include Tier 1, Tier 2, and
Tier 3 Sampling Stations.

3-49


-------
Table 3-27. Region 6: Evaluation Results for Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation"

Tierl

Her 2

Her 3

Number of
Stations

Not
Identified
by ail RFI
Reach"1

Reaches

Wat
Least 1
Station in
Tierl

Reaches

Wat
Least 1
Station in
Tier 2C

Reaches

Wall
Stations
in Her 3

Total#
Reaches

Wat
Least 1
Station
Evaluated

Total
Reaches
in State

% of all
Reaches
inState

Wat
Least 1
Station
Evaluated

%of
Reaches

Wat
Least 1
Tier lor
Her 2
Station

No.

%

No.

%

No.

%

Arkansas

18

17

39

36

50

47

-

17

31

40

88

855

10

6

Louisiana

111

24

270

59

79

17

57

45

68

29

142

840

17

13

New Mexico

4

4

40

40

57

56

-

4

28

28

60

919

7

3

Oklahoma

122

43

95

33

69

24

-

97

59

41

197

1,308

15

12

Texas

127

19

393

59

142

22

67

104

160

56

320

3,588

9

7

REGION

382

24

837

52

397

24

124

266

341

192

799

7,293

11

8

•River reaches based on EPA River Reach File 1 (RF1).

'Stations not identified by an RF1 reach were located in coastal or open water areas.

'No stations in these reaches were included in Her 1.

'Because some reaches occur in more than one state, the toutl number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------
Figure 3-27. Region 6: Location of Sampling Stations Classified as Tier 1 or Tier 2 and Watersheds Containing Areas of Probable Concern for
Sediment Contamination (APCs).


-------
I

[•'inclines

Table 3-28. Region 6: Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State(s)"

Number of Sampling

Stations

Percent of
Sampling Stations
in Tier 1 or Tier 2

Tier 1

Tier 2

Tier 3

08080206

Lower Calcasieu

LA

26

52

22

78

08090100

Lower Mississippi-New Orleans

LA

16

34

1

98

08010100

Lower Mississippi-Memphis

AR, MS, KY,
MO, TO

14

3

3

85

11070209

Lower Neosho

OK, (AR)

13

3

4

80

08040207

Lower Ouachita

LA

12

0

0

100

08030209

Deer-Steele

MS, (LA)

11

10

0

100

11070207

Spring

OK, MO, KS

10

25

6

85

12040104

BuSUo-San Jaciuo

TX

10

23

3

92

*Mo dais were available for states listed In parentheses.

Table 3-29. Region 6: Water Bodies With Sampling Stations Classified as Her 1 Located in Watersheds
	Containing APCs	

Water Body

# of Tier 1
Stations

Water Body

# of Tier 1
Stations

Calcasieu River

15

Neosho River

2

Mississippi River

15

Pryor Creek

2

Bayou D'lnde

11

Greens Bayou



Bayou De Siard

11

LakeEucha

1

Bufiab Bayou

5

Mississippi River, Grand Pass

1

Fort Gfcson Lake

4

Mississippi River, Pass Loutre

1

Lake Hudson

3 .

Ouachita River

1

Bosch Island

2

Spavinaw Lake

1

Galveston Bay

2





3-52


-------
National Si'diim iM Qimlilv Siir\ v\

Table 3-30. Region 6: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling Station
Classifications*



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station

Region 6
Overall

Nickel

Polychlorinated biphenyls

460
434

216

460
218

Louisiana
(continued)

Dibenzo(a,h)anthracene
Lead

59
57

i

58
57



Arsenic

429

3

426

New Mexico

Copper

24

-

24



Copper

350

-

350



Cadmium

23

-

23



DDT

327

70

257



Arsenic

17

-

17



Cadmium

325

-

325



Nickel

12

-

12



Lead

297

--

297



Lead

8

-

8



Chromium

290

9

281



Zinc

6

-

6



Mercury

235

47

188



Mercury

5

3

2



Chlordane

189

4

185



Chromium

4

-

4



Silver

144

32

112



Polychlorinated biphenyls

2

2

-



Zinc

133

--

133



Chlordane

2

-

2



Dieldrin

132

10

122

Oklahoma

Polychlorinated biphenyls

135

118

17



BHC

123

16

107



Arsenic

78

1

77



Dibenzo(a,h)anthracene

122

2

120



Chlordane

73

3

70

Arkansas

Arsenic

25

-

25



Cadmium

60

-

60



DDT

23

6

17



DDT

58

7

51



Mercury

15

, 3

12



Lead

43

-

43



PolycMorinated biphenyls

14

7

7



Dieldrin

35

1

34



Lead

13

, --

13



Copper

27

-

27



Dieldrin

7

-

7



Mercury

26

3

23



Dioxins

6

6

--



Toxaphene

20

-

20



Chlordane

6

..

6

Texas

Nickel

259

-

259



Cadmium

4

--

4



Copper

185

-

185



Copper

3

--

3



Cadmium

182

-

182

Louisiana

Nickel

178

-

178



Lead

176

-

176



Arsenic

141

1

140



Arsenic

168

1

167



Chromium

132

3

129



Polychlorinated biphenyls

164

45

119



PolycMorinated biphenyls

119

44

75



Chromium

152

6

146



Copper

111

' -

111



DDT

135

31

104



DDT

110

26

84



Silver

135

30

105



SEM (est)'

75

-

75



Mercury

118

17

101



Mercury

71

21

50











'Stations may be listed for more than one chemical,
^Simultaneously extracted roeuds.

3-53


-------
rindiii^s

EPA Region 7

Iowa, Kansas, Missouri, Nebraska

EPA evaluated 1,011 sampling stations in Region 7 as
part of the NSI evaluation. Sediment contamination where
associated adverse effects to aquatic life are probable (Tier
1) was found at 32 of these sampling stations, and possible
but infrequent (Tier 2) at 242 of these sampling stations.
For human health, data for 299 sampling stations indicated
probable association with adverse effects (Tier 1), and 230
sampling stations indicated possible but infrequent adverse
effects (Tier 2). Overall, this evaluation resulted in the clas-
sification of 330 sampling stations (33 percent) as Tier 1,
393 (39 percent) as Her 2, and 288 (28 percent) as Tier 3.
The NSI sampling stations in Region 7 were located in 516
separate river reaches, or. 11 percent of all reaches in the
Region. Five percent of all river reaches in Region 7 in-
cluded at least one Tier 1 station, 4 percent included at least
one Her 2 station but no Tier 1 stations, and 2 percent had
only Tier 3 stations (Figure 3-28). Table 3-31 (on the fol-
lowing page) presents a summary of sampling station clas-
sification and evaluation of river reaches for each state and
for the Region as a whole.

This evaluation identified 5 watersheds containing
areas of probable concern for sediment contamination

Figure 3-28. Region 7: Percent of River Reaches
That Include Tier 1, Tier 2, and Tier 3
Sampling Stations.

(APCs) out of the 239 watersheds (2 percent) in Region
7 (Figure 3-29). In addition, 49 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not identified as containing APCs, 16
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 5 percent had only Tier 3 stations. Twenty-
eight percent of the watersheds in Region 7 did not
include a sampling station. The locations of the water-
sheds containing APCs and the Tier 1 and Tier 2 sam-
pling stations in Region 7 are illustrated in Figure 3-30.

Within the 5 watersheds in Region 7 identified as
containing APCs (Table 3-32), 12 water bodies have at
least 1 Tier 1 sampling station; 1 water body has 10 or
more Tier 1 sampling stations (Table 3-33). The water
bodies listed on Table 3-33 are not inclusive of all loca-
tions containing a Tier 1 sampling station because only
water bodies within watersheds containing APCs are
listed.

The chemicals most often associated with Tier 1 or
Tier 2 sampling station classifications in Region 7 over-
all and in each state in Region 7 are presented in
Table 3-34.

At Least One Tier 1 Station
49%

A

At Least One VvOvW	/

Tier 2 Station and \^S-Syj	/

Zero Tier 1 Stations yvV—j	/

16 ... -i- o 01/^^3 .-^NoData
All Tier 3 Stations^1	— 28%

5%

Total number of watersheds = 239

Figure 3-29. Region 7: Watershed Classifications.

No Data 	



89%/^





\ At Least 1



(Tier One Station



8 5%

\

y"\7 At Least One Tier 2

^ Station and Zero



—/ Tier 1 Stations



All Tier 3 4%



Stations



2%

Total number of river reaches

= 4,857

3-54


-------
Table 3-31. Region 7: Evaluation Results for Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation"

Tier!

Tier 2

Her 3

Number of
Stations

Not
Identified
byanRFl
Reach1*

Reaches

Wat
Least 1
Station in
Her I

Reaches

Wat
Least 1
Station in
Her2c

Reaches

Wall
Stations
in Tier 3

Total#
Reaches

Wat
Least 1
Station
Evaluated

Ttrtal
Reaches
instate

% of an
Reaches
instate

Wat
Least 1
Station
Evaluated

% of
Reaches

Wat
Least 1
Her 1 or
Her 2
Station

No.

%

No.

%

No.

%

Iowa

75

33

104

46

49

21

-

61

50

19

130

1,198

11

9

Kansas

76

38

98

48

29

14

-

64

48

13

125

1,184

11

9

Missouri

124

38

98

30

105

32

-

76

32

18

126

1,364

9

8

Nebraska

55

22

93

37

105

41

-

45

62

39

146

1,265

12

8

REGION I4

330

33

393

39

288

28

-

246

182

88

516

4,857

11

9

'River reaches based on EPA River Reach File 1 (RF1).

'Stations not identified by an RFl reach were located in coastal or open water areas.

"No stations in these reaches were included in Tier I.

'Because some reaches occur in more than one state, the total number of reaches in each categoiy for the Region might not equal the sum of reaches in the states.


-------
Ul
OS

Figure 3-30. Region 7: Locations ©f Sampling Stations Classified as Tier 1 or Tier 2 and Watersheds Containing Areas of Probable Concern for
Sediment Contamination (APCs).


-------
Table 3-32. Region 7: Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State(s)

Number of Sampling
Stations

Percent of
Sampling
Stations in Tier 1
or Tier 2

Tier 1

Tier 2

Her 3

07140101

Cahokia-Joachim

MO, IL

18

34

4

93

07080101

Copperas-Duck

II, IA

17

5

5

81

08010100

Lower Mississippi-Memphis

AR, MS, SCY,
MO, TN

14

3

3

85

10270104

Lower Kansas

MO, KS

12

15

2

93

11070207

Spring

OK, MO, KS

10

25

6

85

Table 3-33. Region 7: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs



# of Tier 1



# of T!er 1

Water Body

Stations

Water Body

Stations

Mississi>pi River

17

Duck Creek

1

Kansas River

7

Joachim Creek

1

Spring River

5

Kin Creek

1

Center Creek

3

Stranger Creek

1

Cedar Creek

2

Hirkey Creek

1

Cow Creek

1

Wakarusa River

1

3-


-------
I incliiiiis

Table 3-34. Region 7: Chemicals Most Often Associated With Tier 1 or Tier 2 Sampling Station
Classifications*





# Tier 1









# Tier 1









& Tier 2

# Tier 1

# Tier 2





& Tier 2

# Tier 1

# Tier 2



Chemical

Stations

Station

Station



Chemical

Stations

Station

Station

Region?

Dieldrin

336

2

334

Kansas

Arsenic

52

—

52

Overall

Chlordane

329

_

329

(continued)

Nickel

49

-

49



Polychlorinaied biphenyls

305

291

14



Cadmium

36

-

36



Arsenic

171

_

171



Lead

34

_

34



Heptachlor epoxide

138

-

138



Chromium

27

1

26



Nickel

121

-

121



Zinc

23

_

23



Cadmium

115

-

115



Copper

20

-

20





84

-

84

Missouri

Chlordane

119

-

119



Copper

74

-

74



Polychlorinaied biphenyls

116

102

14



Chromium

50

5

45



Dieldrin

76

-

76



Dioxins

44

42

2



Heptachlor epoxide

53

-

53



Zinc

43

-

43



Arsenic

43

-

43



Bis(2-ethylhexyl)phthalal

37

9

28



Cadmium

36

-

36



DDT

33

_

33



Lead

33

-

33



AJdrin

31

-

31



Dioxins

31

29

2

Iowa

Dieldrin

126

2

124



Nickel

29

~

29



Chlordane

91

-

91



Copper

27

_

27



Polychlortoated biphenyls

71

71

-

Nebraska

Dieldrin

72

-

72



Heptachlor epoxide

54

-

54



Chlordane

52

-

52



Arsenic

34

_

34



Polychlorinaied biphenyls

50

50

-



Copper

17

-

17



Arsenic

42

-

42



Cadmium

14

-

14



Cadmium

29

-

29



Nickel

14

-

14



Nickel

29

-

29



EOT

12

-

12



Chromium

17

2

15



Lead

10

-

10



Aldrin

13

-

13

Kansas

Polychlorinaied biphenyls

68

68

-



Heptachlor epoxide

12

-

12



Chlordane

67

-

67



Bis(2-ethylhexyi)phthalat

10

4

6



Dieldrin

62

-

62











*5Ut!oet be lilted for more thus one chemical.

3-58


-------
EPA Region 8

Colorado, Montana, North Dakota, South Dakota, Utah,
Wyoming

EPA evaluated 535 sampling stations in Region 8 as
part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 39 of these sampling stations,
and possible but infrequent (Tier 2) at 325 of these sam-
pling stations. For human health, data for 29 sampling
stations indicated probable association with adverse ef-
fects (Tier 1), and 19 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2). Overall,
this evaluation resulted in the classification of 68 sam-
pling stations (13 percent) as Tier 1,327 (61 percent) as
Tier 2, and 140 (26 percent) as Tier 3. The NSI sam-
pling stations in Region 8 were located in 305 separate
river reaches, or 2 percent of all reaches in the Region.
Less than 1 percent of all river reaches evaluated in Re-
gion 8 included at least one Tier 1 station, 1 percent in-
cluded at least one "Her 2 station but no Tier 1 stations,
and less than 1 percent had only Tier 3 stations (Figure
3-31). Table 3-35 (on the following page) presents a
summary of sampling station classification and evalua-

Figure 3-31. Region 8: Percent of River Reaches
That Include Her 1, Tier 2, and Tier 3
Sampling Stations.

tion of river reaches for each state and for the Region as
a whole.

None of the 385 watersheds in Region 8 were iden-
tified as watersheds containing areas of probable con-
cern for sediment contamination. Fourteen percent of
all watersheds in the Region had at least one Tier 1 sam-
pling station, 12 percent had at least one Tier 2 station
but no Tier 1 stations, and 9 percent had only Her 3 sta-
tions (Figure 3-32). Sixty-five percent of the watersheds
in Region 8 did not include a sampling station. The lo-
cations of the Tier 1 and Tier 2 sampling stations in Re-
gion 8 are illustrated in Figure 3-33.

Lack of multiple sampling site data did not allow
identification of any watersheds in Region 8 as contain-
ing APCs. Therefore, specific water bodies with Tier 1
sampling stations are not listed in a separate table, as for
other Regional summaries.

The chemicals most often associated with Tier 1 or
Tier 2 sampling station classifications in Region 8 over-
all and in each state in Region 8 are presented in
Table 3-36.

Total number of watersheds = 385

Figure 3-32. Region 8: Watershed Classifications.

No Data



98% \





\ At Least One



\T»r 1 Station



\ <1%

\ /



j /At Least One Her 2



— Station and Zero



j\ Tierl Stations



/ \ 1%



/All Tier 3



/ Stations



<1%

Total number of river reaches =

13,492

3-59


-------
Table 3-35. Region 8: Evaluation Results of NSI Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation^

Tier 1

Her 2

Tier 3

Number of
Stations

Not
Identified
by anRFl
Reactf*

Reaches

Wat
Least 1
Station in
TIerl

Reaches

Wat
Least 1
Station in
Tier?

Reaches

Wall
Stations
in Tier 3

Tbtal#
Reaches

Wat
Least 1
Station
Evaluated

Total
Reaches
instate

% of all
Reaches
instate
Wat Least
1 Station
Evaluated

%of
Reaches

Wat
Least 1
Her lor
Her 2
Station

No.

%

No.

%

No.

%

Colorado

11

6

140

69

51

25

-

8

73

34

115

2,178

5

4

Montana

9

24

18

47

11

29

-

9

10

8

27

5,490

1

<1

North Dakota

24

15

112

70

25

15

-

22

36

9

67

992

7

6

South Dakota

13

30

21

49

9

21

-

11

6

7

24

1,611

2

1

Utah

7

15

24

51

16

34

-

7

16

10

33

1,034

3

2

Wyoming

4

9

12

27

28

64

-

4

12

25

41

2,421

2

1

REGION 8d

68

13

327

61

140

26

-

61

153

91

305

13,492

2

2

'River reaches based on EPA River Reach File 1 (RF1).

^Stations not identified by an RF1 reach were located in coastal or open water areas.

®No stations in these reaches were included in Tier 1.

^Because some reaches occur in more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------

-------
I'in
-------
Naliomil Sediment Qiiiilily Survey

EPA Region 9

Arizona, California, Hawaii, Nevada

EPA evaluated 1,699 sampling stations in Region 9
as part of the NSI evaluation. Sediment contamination
where associated adverse effects to aquatic life are prob-
able (Tier 1) was found at 433 of these sampling sta-
tions, and possible but infrequent (Tier 2) at 894 of these
sampling stations. For human health, data for 40 sam-
pling stations indicated probable association with adverse
effects (Tier 1), and 765 sampling stations indicated pos-
sible but infrequent adverse effects (Tier 2), Overall,
this evaluation resulted in the classification of 468 sam-
pling stations (28 percent) as Tier 1,942 (55 percent) as
Her 2, and 289 (17 percent) as Tier 3. The NSI sam-
pling stations in Region 9 were located in 254 separate
river reaches, or 6 percent of all reaches in the Region.
Three percent of all river reaches in Region 9 included
at least one Tier 1 station, 2 percent included at least one
Her 2 station but no Tier 1 stations, and 1 percent had
only Tier 3 stations (Figure 3-34). Table 3-37 (on the
following page) presents a summary of sampling station
classification and evaluation of river reaches for each
state and for the Region as a whole.

This evaluation identified 10 watersheds containing
areas of probable concern for sediment contamination

Figure 3-34. Region 9: Percent of River Reaches
That Include Tier 1» Tier 2, and Tier 3
Sampling Stations.

(APCs) out of the 279 watersheds (4 percent) in Region
9 (Figure 3-35). In addition, 22 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not classified as containing APCs, 10
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 5 percent had only Tier 3 stations. Fifty-nine
percent of the watersheds in Region 9 did not include a
sampling station. The locations of the watersheds con-
taining APCs and the Tier 1 and Her 2 sampling stations
in Region 9 are illustrated in Figure 3-36.

Within the 10 watersheds in Region 9 identified as
containing APCs (Table 3-38), 19 water bodies have at
least 1 Tier 1 sampling station; 7 water bodies have 10 or
more Tier 1 sampling stations (Table 3-39). San Diego
Bay, San Francisco Bay, and offshore areas around San
Diego and Los Angeles appear to have the most signifi-
cant sediment contamination in Region 9. The water bod-
ies listed on Table 3-39 are not inclusive of all locations
containing a Tier 1 sampling station because only water
bodies within watersheds containing APCs are listed.

The chemicals most often associated with Tier 1 or
Tier 2 sampling station classifications in Region 9 over-
all and in each state in Region 9 are presented in
Table 3-40.

At Least One Tier 2 Station
and Zero Tier 1 Stations

All Tier 3 Stationsx^^vfP^ ?

At Least One
feaJier 1 Station
^X22%



_^^tdAPCs

4%

No Data"-^,^
59%



Total number of watersheds = 279



Figure 3-35. Region 9: Watershed Classifications.

No Data

84%

At Least 1 Tier One
Station
3%

At Least One Tier 2
- Station and Zero
Tier 1 Stations

All Tier 3 2%

Stations

1%

Total number of river reaches = 4,601

3-63


-------
Table 3-37, Region 9: Evaluation Results for NSI Sampling Stations and River Reaches by State



Station Evaluation

River Reach Evaluation'



Herl

Her 2

Her 3

Number of
Stations

Not
Identified
by an RFI
Reach*







Total#
Reaches

Wat
Least 1
Station
Evaluated



% of all
Reaches
Instate

Wat
Least 1
Station
Evaluated

%or
Reaches

Wat
Least 1
Herl or
Her 2
Station

State

No.

%

No.

%

No.

%

Reaches

Wat
Least 1
Station In
Herl

Readies
Wat
Least 1
Station In
Her 2s

Reaches

Wall
Stations
In Tier 3

Total
Readies
IttState

Arizona

44

35

58

47

22

18

-

30

33

11

74

1,146

7

5

Califomia

392

27

822

57

229

16

758

75

44

26

145

2,606

6

5

Hawaii

8

22

23

64

5

14

36

-

-

-

-

-

-

-

Nevada

24

25

39

41

33

34

-

16

15

6

37

916

4

3

REGION 9"

468

28

942

55

289

17

794

119

92

43

254

4,601

6

5

•River reaches based on EPA River Reach File 1 (RFI),

'Stations not identified by an RFI reach were located in coastal or open water areas,

'No stations in these reaches were included in Tier 1.

'Because some reaches occur in more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------
w Figure 3-36. Region 9: Location of Sampling Stations Classified as Tier 1 or Tier 2 and Watersheds Containing Areas of Probable Concern
gj	for Sediment Contamination (APCs).


-------
Dable 3-38. Region 9; Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State(s)

Number of Sampling
Stations

Percent of
Sampling
Stations In Tier 1
or Her 2

Tier 1

Tier 2

Her 3

18070104

Sana Monfca Bay

CA

79

31

22

83

18070201

Seal Beach

CA

63

339

40

91

18070304

San Diego

CA

53

51

3

97

18070204

Newport Bay

CA

24

68

16

85

18050004

SanFransfeco Bay

CA

19

37

8

88

18050003

Coyote

CA

18

6

0

100

18070105

Los Angebs

CA

14

19

4

89

18070107

Sail Pedro Channel Islands

CA

14

10

1

96

18030012

TUIare-Buena Vista Lakes

CA

10

5

5

75

18070301

AlisoSan Onofre

CA

10

22

0

100

liable 3-39, Region 9: Water Bodies With Sampling Stations Classified as Tier 1 Located in Watersheds
Containing APCs



# of Tier 1



# of Tier 1

Water Body

Stations

Water Body

Stations

Pacifis Ocean

178

Corte Madera Creek

2

San Diego Bay

32

Los Gates Creek

2

San Francisco Bay

19

Coyote Creek

1

Los Angeles River

14

Lcxitgton Reservor

1

Santa Catalna Island

14

Oso Creek

1

San Diego Creek

12

Peters Canyon Wash

1

Kiigs River

10

San Diego River

1

AJanAos Creek

8

San Juan Creek

1

Cakto Reservoir

4

Sweetwater River

1

Also Creek

2





3-66


-------
N;i(iun;il Si'diimnl Oulililv Sur\o\

Table 3-40. Region 9: Chemicals Most Often Associated with Her 1 or Her 2 Sampling Station
Classifications*



Chemical

# Tier I
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station

Region 9
Overall

Copper
DDT

678

675

179

678
496

California
(continued)

Cadmium
Nickel

406

373

—

406
373



Arsenic

455

12

443



Arsenic

357

3

354



Nickel

454

-

454



Mercury

336

103

233



Cadmium

446

-

446



Bis(2-ethylhexyl)phthalat

264

48

216



Polychlorinated biphenyls

445

100

345



Lead

253

-

253



Mercury

403

134

269



Chromium

239

40

199



Lead

314

-

314

Hawaii

Nickel

20

-

20



Bis(2-ethylhexyl)phthalat

302

69

233



Copper

19

-

19



Chromium

265

42

223



Mercury

16

4

12



Zinc

238

-

238



Arsenic

16

1

15



Silver

209

23

186



Lead

14

_

14



BHC

164

9

155



Zinc

13

-

13



Benzo(a)pyrene

158

6

152



DDT

10

2

8



Dieldrin

125

--

125



Chromium

10

1

9

Arizona

Copper

72

¦-

72



Polychlorinated biphenyls

8

3

5



Arsenic

55

8

47



Cadmium

8

-

8



Nickel

50



50

Nevada

Mercury

29

15

14



Lead

37

--

37



Arsenic

27

-

27



Zinc

28



28



Copper

14

-

14



Bis(2-ethylhexyl)phthalat

26

15

11



Nickel

11

-

11



Cadmium

24

-

24



Zinc

11

_

11



DDT

23

9

14



Lead

10

-

10



Mercury

22

12

10



Polychlorinated biphenyls

9

4

5



Silver

15

7

8



B i s(2-ethy lhexy l)phthalat

8

4

4

California

DDT

640

168

472



Cadmium

8

-

8



Copper

573

-

573



Chlordane

8

-

8



Polychlorinated biphenyls

418

87

331











'Stations may be listed for more than one chemical.

3-67


-------
Findings

EPA Region 10

Alaska, Idaho, Oregon, Washington

EPA evaluated 2,878 sampling stations in Region
10 as part of the NSI evaluation. Sediment contamina-
tion where associated adverse effects to aquatic life are
probable (Tier 1) was found at 623 of these sampling
stations, and possible but infrequent (Her 2) at 1,658 of
these sampling stations. For human health, data for 112
sampling stations indicated probable association with ad-
verse effects (Her 1), and 1,285 sampling stations indi-
cated possible but infrequent adverse effects (Tier 2).
Overall, this evaluation resulted in the classification of
727 sampling stations (25 percent) in Region 10 as Tier
1, 1,696 (59 percent) as Tier 2, and 455 (16 percent) as
Tier 3. The NSI sampling stations in Region 10 were
located in 393 separate river reaches, or 4 percent of all
reaches in the Region. One percent of all river reaches
in Region 10 included at least one Tier 1 station, 2 per-
cent included at least one Her 2 station but no Tier 1
stations, and 1 percent had only Tier 3 stations (Figure
3-37). Table 3-41 (on the following page) presents a
summary of sampling station classification and evalua-
tion of river reaches for each state and for the Region as
a whole.

No Data
86%/

\ At Least One
\ Her 1 Station

\/ 2%

At Least One Her 2
s3r~ Station and Zero
A Tier 1 Stations

TUfTler 3 1%

Stations
1%

Total number of river reaches = 10,178

Figure 3-37. Region 10: Percent of River Reaches
That Include Tier 1, Tier 2, and Tier 3
Sampling Stations.

This evaluation identified 7 watersheds containing
areas of probable concern for sediment contamination
(APCs) out of the 219 watersheds (3 percent) in Region
10 (Figure 3-38). In addition, 28 percent of all water-
sheds in the Region had at least one Tier 1 sampling sta-
tion but were not categorized as containing APCs, 14
percent had at least one Tier 2 station but no Tier 1 sta-
tions, and 6 percent had only Tier 3 stations. Forty-nine
percent of the watersheds in Region 10 did not include a
sampling station. The locations of the watersheds con-
taining APCs and the Tier 1 and Tier 2 sampling stations
in Region 10 are illustrated in Figure 3-39.

Within the 7 watersheds in Region 10 identified as
containing APCs (Table 3-42), 34 water bodies have at
least 1 Tier 1 sampling station; 8 water bodies have 10 or
more Tier 1 sampling stations (Table 3-43). Puget Sound
appears to have the most significant sediment contami-
nation in Region 10. The water bodies listed on Table 3-
43 are not inclusive of all locations containing a Tier 1
sampling station because only water bodies within wa-
tersheds containing APCs are listed.

The chemicals most often associated with Tier 1 or
Tier 2 sampling station classifications in Region 10 over-
all and in each state in Region 10 are presented in
Table 3-44.

At Least One
Tier 2 Station and ^
Zero Tier 1 Stations, ¦/2v'

14% AJv\

At Least One .
Tier 1 Station
"^28%

\

All Tmr 3 Stations

fi% |	—

3%

No Data
49%

Total number of watersheds = 219



Figure 3-38. Region 10: Watershed Classifications.

3-68


-------
Table 3-41, Region 10: Evaluation Results for NSI Sampling Stations and River Reaches by State

State

Station Evaluation

River Reach Evaluation*

Her 1

Tier 2

Tier 3

Number of
Stations

Not
Identified
by an RF1
Reachb

Reaches

Wat
Least 1
Station in
Tier 1

Reaches

Wat
Least 1
Station in
Her 2*

Reaches

Wall
Stations
teller 3

Total#
Reaches
Wat Least
1 Station
Evaluated

Total
Reaches
instate

% of all
Reaches
in State
Wat Least
1 Station
Evaluated

% of
Reaches

Wat
Least 1
Tkri or
Tier 2
Station

No.

%

No.

%

No.

%

Alaska

21

8

191

71

55

21

267

-

-

-

-

-

-

-

Idaho

43

•45

36

38

16

17

-

30

16

7

53

3,227

2

1

Oregon

81

28

158

54

52

18

2

45

43

25

113

4,203

3

2

Washington

582

26

1,311

59

332

15

228

75

115

40

230

2,924

8

6

REGION 10"

727

25

1,696

59

455

16

497

147

174

72

393

10,178

4

3

¦River reaches based on EPA River Reach File 1 (RF1).

•Stations not identified by 211RF1 reach were located in coastal or open water areas,
cNo stations in these reaches were included to Tier 1.

'Because some reaches occur in more than one state, the total number of reaches in each category for the Region might not equal the sum of reaches in the states.


-------
Figure 3-39, Region 10; Location of Sampling Stations Classified as Tier 1 or Her 2 and Watersheds Containing Arras of Probable Concern for
Sediment Contamination (APCs),


-------
Table 3-42. Region 10: Watersheds Containing Areas of Probable Concern for Sediment Contamination

Cataloging
Unit Number

Name

State(s)*

Number of Sampling
Stations

Percent of
Sampling
Stations in Tier 1
or Iter 2

fieri

Tier 2

Tier 3

17110019

Puget Sound

WA

418

851

114

92

17110013

Duwarrish

WA

48

69

10

92

17110002

Strait Of Georgia

WA

32

168

63

76

17030003

Lower Yakima

WA

23

19

5

89

17090012

Lower Willamette

OR

21

51

4

95

17110014

PuyallH>

WA

12

6

1

95

17010303

Coeur D'Afcne Lake

ID, (WA)

10

13

0

100

*No data were available for states listed in parentheses.

Table 3-43. Region 10: Water Bodies With Sampling Stations Classified as Tier 1 Located in Areas of
Probable Concern for Sediment Contamination

Water Body

# of Tier 1
Stations

Water Body

# of Tier 1
Stations

Puget Sound

306

Lake Whatcom

2

Budd Inlet

41

Samnish Bay

2

Elliot Bay

41

Samnish River

2

Bainbridg: Island

31

Whidbey Island

2

Sinclair Inlet

28

Spring Creek

2

Bellingham Bay

,22

Thonpson Lake

2

Yakima River

19

Ahtanum Creek

1

Willamette River

10

C amino Island

I

Carbon River

8

Duwanish Waterway



Columbia Slough

S

Fidalgo Island

1

Green River

6

Padden Lake

1

Coeur D'aiene Lake

4

Port Orchard

1

Dyes Inlet

4

Pott Susan

1

Puyallup River

4

Spanaway Lake

1

Coeur D'aiene River

3

Toppenish Creek

1

Johnson Creek

3

White Hall Creek

1

Chambers Creek

2

Wolf Lodge Creek

1

3-71


-------
Finding's

Ihble 3-44. Region 10: Chemicals Most Often Associated with Tier 1 or Tier 2 Sampling Station
Classifications*



Chemical

#Tier 1
& Tier 2
Stations

# "fieri
Station

# Tier 2
Station



Chemical

# Tier 1
& Tier 2
Stations

# Tier 1
Station

# Tier 2
Station

Region 10
Overall

Copper
Nickel

1,518
1,409

:

1,518
1,409

Idaho

{continued)

Cadmium
Copper

29
28

1IIU

29
28



Arsenic

1,231

55

1,176



Zinc

28

--

28



Lead

881

_

881



DDT

25

-

25



Benzo(a)pyrene

803

103

700



Dieldrin

21

-

21



Pyrcne

770

160

610



Toxaphene

14

--

14



Mercury

760

133

627



Silver

11

8

3



Cadmium

754

-

754

Oregon

Copper

125

_

125



Polychlorinated biphenyls

710

289

421



Nickel

107

-

107



Dibcnzo(a,h)anthracene

709

245

464



Arsenic

86

1

85



Chiysene

704

86

618



Polychlorinated biphenyls

84

46

38



Benzo(a)anthracene

669

107

562



DDT

73

19

54



Naphthalene

589

104

485



Zinc

59

--

59



Fluorcne

547

77

470



Mercury

53

7

46



Chromium

546

17

529



Cadmium

51

--

51

Alaska

Chromium

135

12

123



Chromium

46

3

43



Arsenic

89

-

89



Lead

44

-

44



Copper

50

-

50

Washington

Copper

1,315

--

1,315



Nickel

41

-

41



Nickel

1,256

--

1,256



Cadmium

35

_

35



Arsenic

1,017

41

^ 976



Naphthalene

31

2

29



Lead

788

-

788



Polychlorinated biphenyls

29

2

27



Benzo(a)pyrene

754

101

653



Zinc

29

-

29



Pyrene

735

156

579



Phenanthrene

26

-

26



Mercury

683

121

562



Fluorcne

22

-

22



Chrysene

682

83

599

Idaho

Arsenic

39

13

26



Dibenzo(a,h)anthracene

681

240

441



Polychlorinated biphenyls

32

28

4



Benzo(a)anthracene

646

104

542



Lead

32

-

32











*Sutiofti rtuy be listed for more than one chemical.

3-72


-------
Potentially Highly Contaminated
Sites Not Identified by the NSI
Evaluation

Several Regions and states provided comments on
the May 16, 1994, preliminary evaluation of sediment
chemistry data contained in the NSI. They identified
receiving streams that should have been but were not
identified as locations of potential adverse effects, based

on the NSI data evaluation. The specific water bodies
that reviewers of the preliminary evaluation identified
as potentially contaminated, but which are not presently
included in the NSI because data are inadequate to cat-
egorize sampling stations as Her 1, are presented in Table
3-45 and Figure 3-40. If a water body had previously
been identified as having at least one Tier 1 sampling
station using the NSI evaluation methodology, it was not
included in Table 3-45 or Figure 3-40,

Water Body

EPA Region

State

Chemicals Potentially Present

Onandaga Lake

2

NY

pesticides, metals, PAHs, PCBs

Ley Creek

2

NY

mercury

Kill van Kull

2

NY

metals, dioxin

Newtown Creek

2

NY

PAHs

Sc^jaquada Creek

2

NY

metals, PCBs

Skaneateles Creek

2

NY

PCBs

Hudson River

2

NY

PCBs

Southern reaches of the Maurice River

2

NJ

arsenic

Elizabeth River

3

VA

PAHs

James River

3

¥A

kepone

Anacostia River

' 3

DC

chlordane, PCBs

Lake 0' the Pines

6

TX

lead, zinc

Linneville Bayou

6

TX

lead, chromium

Humboldt River Basin

9

NV

selenium

Dry Lake

9

AZ

dioxin

Table 3-45. Potentially Highly Contaminated Sites Not Identified in the NSI Evaluation

3-73


-------
Figure 3-40, Location of Potentially Highly Contaminated Water Bodies Not Identified in the NSI Evaluation.


-------
Chapter 4

Pollutant Sources

Toxic chemicals that accumulate in sediment and
are associated with contamination problems
enter the environment from a variety of sources.
These sources can be broadly differentiated as point sources
and nonpoint sources. The term "point source" is defined
in the Clean Water Act (CWA) and generally refers to any
specific conveyance, such as a pipe or ditch, from which
pollutants are discharged. In contrast, nonpoint sources
do not have a single point of origin and generally include
diffuse sources, such as urban areas or agricultural fields,
that tend to deliver pollutants to surface water during and
after rainfall events. Some sources, such as landfills and
mining sites, ate difficult to categorize as either a point or
nonpoint source. Although these land areas represent dis-
crete sources, pollution from such areas tends to result from
rainfall runoff and leaching. Likewise, atmospheric depo-
sition of pollutants, generally considered to be a nonpoint
source of water pollution, arises from the emission of chemi-
cals from discrete stationary and mobile source points of
origin. The CWA specifies water vessels and other float-
ing craft as point sources although, taken as a whole, they
function as a diffuse source.

Many point and nonpoint pollutant sources have
been the subject of federal and other action over the past
25 years. The direct discharge of pollutants to water-
ways from municipal sewage treatment and industrial
facilities requires a permit under the CWA. Many states
have been authorized to issue permits in lieu of EPA.
These permits contain technology-based and water qual-
ity-based pollutant discharge limits and monitoring re-
quirements. More recently, replacement of aging
combined sewer systems and other storm water control
measures has addressed the discharge of pollutants from
urban areas through municipal facilities. The disposal
of sediment dredged to maintain navigation channels is
managed under both the CWA and the Marine Protec-
tion, Research, and Sanctuaries Act (MPRSA) to ensure
that unacceptable degradation from chemical pollutants
in the dredged material does not occur at the disposal
location. Emission standards and controls on station-
ary and mobile sources of air pollutants have also been
established in federal regulations promulgated under the
authority of the Clean Air Act (CAA). These actions
have reduced emissions of gaseous compounds such as
inorganic oxides, as well as pollutants that eventually

enter water bodies and accumulate in sediment. The
Toxic Substances Control Act (TSCA) and Federal In-
secticide, Fungicide, and Rodenticide Act (FIFRA) have
greatly reduced the toxic pollutant input to the environ-
ment through bans and use restrictions on many pesti-
cides and industrial-use chemicals.

Federal, state, and local laws have also addressed
land-based pollutant sources. Under the Resource Con-
servation and Recovery Act (RCRA), the transport, stor-
age, and disposal of pollutants in landfills and other
repositories of hazardous waste are tracked and con-
trolled. At sites where past disposal practices, either
purposeful or accidental, have resulted in severe con-
tamination, remediation has been undertaken under the
federal Superfund laws. Where applicable, land devel-
opment projects may be subject to an assessment of the
environmental impact conducted under National Envi-
ronmental Policy Act (NEPA) authority. Under the au-
thority of the Coastal Zone Management Act (CZMA),
EPA has developed nonregulatory management measures
to reduce pollutant delivery via nonpoint sources, such
as runoff from urban and agricultural areas.

The combined impact of these actions has yielded
improvements in water quality. In at least some docu-
mented cases, pollutant levels in sediment are also de-
creasing. (For example, see the discussion of the Palos
Verdes case study presented in Chapter 5.) However,
improvement in sediment quality might lag behind im-
provement in overlying water because of the persistent
nature of many pollutants, as well as the storage and
sink functions of sediment, and because the most toxic
bioaccumulative pollutants are difficult to monitor and
regulate. It is beyond the scope of this baseline assess-
ment to determine the temporal trends of pollutant con-
centrations in sediment on a national scale. Future
reports to Congress will address that issue.

Natural recovery of contaminated sediment can oc-
cur through source reduction, contaminant degradation,
and continuing deposition of clean sediment. The fea-
sibility of natural recovery, as well as the long-term suc-
cess of remediation projects, depends on the effective
control of pollutant sources. For some classes of sedi-
ment contaminants, such as PCBs and organochlorine

4-1


-------
I'ottnUmt .Sources

pesticides, use and manufacture bans or severe restric-
tions have been in place for many years. Past disposal
and use of PCBs continue to result in evaporation of
these contaminants from some landfills and leaching
from soils, but most active PCB sources have been con-
trolled. The predominant sources of organochlorine pes-
ticides are runoff and atmospheric deposition from past
applications on agricultural land, and occasional dis-
charge from municipal treatment facilities. For other
classes of sediment contaminants, active sources con-
tinue to contribute substantial environmental releases.
For example, liberation of inorganic mercury from fuel
burning and other incineration operations continues,
as do urban runoff and atmospheric deposition of met-
als and PAHs. In addition, discharge limits for munici-
pal and industrial point sources are based on
technology-based limits and state-adopted standards for
protection of the water column, not necessarily for down-
stream protection of sediment quality. Determining the
local and far-field effects of individual point and
nonpoint sources on sediment quality usually requires
site-specific study.

The purposes of this chapter are to:

•	Present the extent of sediment contamination
by chemical class in the 96 watersheds identi-
fied as areas of probable concern for sediment
contamination (APCs).

•	Identify the major source categories of these
chemical classes and summarize key studies
that link these source categories to sediment
contamination.

• • Analyze land use patterns and the extent of
sediment contamination by chemical class in
the 96 APCs.

•	Briefly describe current EPA efforts to further
characterize point and nonpoint sources of sedi-
ment contaminants.

Extent of Sediment Contamination
by Chemical Class

The individual chemicals evaluated for this re-
port can be grouped into six chemical classes: met-
als, PCBs, pesticides, mercury, PAHs, and other
organic chemicals. Pesticides include the organochlo-
rine pesticide compounds assessed in this report, such
as DDT and metabolites, dicldrin, and chlordane.
PAHs include both low- and high-molecular-weight
polynuclear aromatic hydrocarbons, and other organ-

ics include all organics not otherwise classified. Mer-
cury is grouped separately from other metals because
of its unique behavior in the environment (e.g., me-
thylation and bioaccumulation potential) and because
of recent attention focused on its impact as a primary
sediment and fish contaminant of concern.

Figure 4-1 presents, by chemical class, the average
percent of stations that are contaminated in the 96 APCs.
For this analysis, the percent contamination is derived by
taking the number of stations where an individual chemi-
cal constituent of a particular chemical class places a sta-
tion into Tier 1 or Tier 2 and dividing by the total number
of stations in the watershed. Each constituent, or any con-
stituent representative of a chemical class, might not have
been measured at all stations in the watershed. In addi-
tion, the total number of stations in each watershed varies
extensively, as does the spatial extent of sampling within
the watershed. The resulting percent contamination by
chemical class varies a great deal—from 0 percent to 100
percent for each class—among the watersheds. Figure 4-1
presents the average value at both Tier 1 and combined
Tier 1 and Tier 2 contamination levels.

Figure 4-1 indicates that at the Tier 1 level of con-
tamination, PCBs are the dominant chemical class with
an average extent of contamination of 29 percent. Among
Tier 1 stations, all other classes of contaminants account
for contamination at a lower percent of the stations on
the average (6 to 10 percent). The relative importance
of PCBs reflects, in part, the fact that a station can be
designated Tier 1 for human health effects based on el-
evated fish tissue concentrations alone for this chemical
class, whereas elevated levels in fish tissue and corre-
sponding elevated levels in sediment are required for
all other classes. At the combined Tier 1 and Tier 2
level of contamination, metals are the dominant chemi-
cal class measured by average extent of contamination
(59 percent), followed by PCBs and pesticides (both at
43 percent), mercury (29 percent), and PAHs and other
organics (19 and 14 percent, respectively). The very
large increase in the relative importance of metals from
Tier 1 to combined Tier 1 and Tier 2 also reflects the
evaluation methodology because a divalent transition
metal concentration cannot place a station into Tier 1
without an accompanying acid-volatile sulfide concen-
tration ([AVS]) measurement, which is typically not
available.

Figure 4-1 graphically displays the relative differ-
ences in certainty of assessing the probable effects of
metals versus assessing the effects of PCBs. More con-
fidence can be placed in the assertion that PCBs exhibit
"probable association with adverse effects" than in mak-

4-2


-------
National Sediment Quality Survey

60%-

50%-

40%

30%-

20%-

10%-

TIERS 1 & 2

Metals

Pesticides

TIER 1

Mercury

PAHs

Other
Organics

Figure 4-1. Average Percent Contamination in Watersheds Containing
APCs by Chemical Class.

ing this assertion for metals. The relatively high per-
cent of PCB contamination at the Tier 1 level reflects
the relative certainty that elevated PCB levels in fish
are associated with elevated levels in sediment. The
relatively low percent of metal contamination at the Tier

1	level primarily reflects the lack of confirming data
(i.e., AVS) regarding important binding phases and
bioavailability, not necessarily the lack of significance
of metal contamination. In fact, the very high percent
contamination indicated at the combined Tier 1 and Her

2	level demonstrates the potential importance of this
chemical class. It should also be noted, however, that
correlative screening values such as ERMs do not indi-
cate causality, rather they are concentrations associated
with effects.

This analysis does not imply that certain chemical
classes are always dominant, nor that other chemical
classes can be dismissed altogether. In fact, contamina-
tion from constituents in any class may be of paramount
importance in a given watershed or location. The dif-
ferences in extent of chemical class contamination on
the average in the 96 APCs is intended to provide some
perspective to the ensuing sections of this chapter.

Major Sediment Contaminant
Source Categories

To identify the important sources of sediment con-
taminants, EPA searched the scientific and technical lit-
erature for studies that link specific pollutant sources to

evidence of sediment contamina-
tion. EPA focused this review on
studies appearing in peer-re-
viewed journals and government
reports published after 1980.
The majority of studies related
sediment contamination to a
source through qualitative
means, including associations of
land use or specific activity with
the types of contaminants de-
tected, and spatial analyses. For
example, organochlorine pesti-
cide contamination is associated
with agricultural land use where
past application practices and hy-
drologic routes of rainfall runoff
are known. Some researchers
made the association with con-
tamination source by more quan-
titative means such as loadings
measurements, runoff or deposi-
tion estimates, or mass balance
models of contaminant inputs. Most research has fo-
cused on the chemicals or chemical classes listed above.
The studies reviewed attributed sediment contamination
from the six classes of chemicals to four general nonpoint
source categories and two general point source catego-
ries. Table 4-1 summarizes the correlations of source
category to chemical class documented in literature.

Table 4-1 does not specifically list some important
sources that are difficult to categorize as a point or
nonpoint source. These sources include leachate from
landfills, direct inputs from recreational and commer-
cial boating, and disposal of contaminated dredged ma-
terial. As mentioned at the beginning of this chapter,
landfills are not easily classified as a point or nonpoint
source. Evaporation and subsequent deposition of mod-
erately volat le contaminants from landfills represent an
atmospheric source, yet leachate is typically considered
as neither "urban runoff' nor a controlled point source.
Nonetheless, leachate from landfills is an important
documented source of sediment contaminants. For ex-
ample, landfill leachate and past effluent discharges from
electronics manufacturers have contaminated New
Bedford Harbor in Massachusetts with PCBs and heavy
metals (Garton et al., 1996). Boaring and shipping ac-
tivities can be important sources of a variety of contami-
nants, including PAHs and antifouling paint additives
such as tributyl tin and copper. As for dredged material
disposal, past dredging operations to maintain naviga-
tion channels could be responsible for contaminated sedi-
ment at specifically designated dump sites. Dredging

4-3


-------
Pollutant Sources

Table 4-1. Correlations of Sources to Chemical Classes of Sediment
Contaminants

Source/Chemical Class

Mercury

PCBs

PAHs

Metals

Pesticides

Other
Qrganics

Harvested Croplands









~



Inactive and Abandoned Mine Sites

•





•





Atmospheric Deposition

•

•

•

•

•

•

Urban Sources

•



•

•

~

•

Industrial Discharges

•

~

•

•

~

•

Municipal Discharges

•

•

•

•

•

•

A" Sour-x from put activities
• Onjolnj source

practices are currently managed under federal, state, and
local authority to ensure that appropriate testing and safe
disposal occur. In addition to these sources, uncontrol-
lable and accidental point source releases, such as im-
proper disposal practices and spills, have occurred and
continue to occur.

A notable feature of Table 4-1 is the extent to which
multiple sources can be associated with each chemical
class. This is the primary factor in making source as-
sessment and effective source control such difficult tasks.
The table does not provide any indication of which
sources are the most significant. The significance of
any given source depends on the areal extent of the source
and intensity of the activity in the watershed. Because a
variety of sources are present (or were present in the
past) in most watersheds, and the extent and intensity
of each source vary, the most important source of a par-
ticular chemical or class of chemical contaminants at a
given location also varies. In addition, there is typically
overlap among source categories. The most obvious
overlap is between atmospheric deposition and urban
sources. For example, fuel combustion in urban areas
releases PAHs to the atmosphere, which are subsequently
deposited in various parts of the watershed or transported
to other areas.

Despite these cautions, the results of EPA's litera-
ture review allow some broad assertions regarding source
associations. For harvested croplands, organochlorine
pesticides are the major contaminants of concern. Inac-
tive and abandoned mine sites contribute mercury and
other heavy metals to sediment. Atmospheric deposi-
tion is a primary contributor of mercury, PCBs, and
PAHs. Urban sources are most closely associated with
metals and PAHs. Although permit monitoring records

and industry-supplied release es-
timates, as well as specific spa-
tial analysis studies, indicate that
municipal and industrial dis-
charges of sediment contaminants
(particularly metals and other or-
gan ics) continue, the relative con-
tribution compared to nonpoint
sources is an open question and
undoubtedly varies substantially
by watershed. A brief summary
of the literature review for major
source categories follows.

At many sites, elevated lev-
els of pesticides in the Nation's
sediment can be attributed to past
agricultural practices. Crop growers deliberately apply
pesticides tc protect their yield from insects, fungus, and
weeds. In the past, organochlorine compounds such as
DDT and chlordane were used without restriction to rid
harvested croplands of a broad range of unwanted spe-
cies. These compounds tend to be persistent in the en-
vironment, adsorptive to soil and sediment particles,
highly bioaccumulative in living tissue, and lethal to
many non-target organisms. As these effects became
apparent and regulatory authorities began restricting or
banning the use of persistent pesticides in the United
States, chemical manufacturers developed newer orga-
nophosphate pesticides that might be more easily de-
gradable and, in some cases, more narrowly targeted to
specific organisms. In addition, modern pesticides must
undergo fed; ral registration procedures designed to pro-
tect human health and the environment before they can
be approved for intended new uses.

Although the current-use pesticides are applied
throughout the country in large amounts, they are not
frequently analyzed in routine sediment monitoring, nor
are they frequently detected in sediment when included
in monitoring studies (Pereira et al„ 1994). Because of
the lack of monitoring data, and the absence of avail-
able levels of concern in sediment, current-use pesti-
cides were not included in this evaluation of sediment
quality. However, these compounds exhibit toxicity to
non-taiget organisms. Furthermore, although these com-
pounds have shorter half-lives and greater water solu-
bility than organochlorines in general, the chemical and
physical properties of some of these compounds indi-
cate significant bioconcentration potential (Willis and
McDowell, 1983). Thus, further assessment of the pres-
ence of current-use pesticides in fish and sediment is
warranted.

4-4


-------
National Sediment Quality Survey

The discharge of pollutants from agricultural lands
to surface water is largely driven by precipitation. Con-
taminants also reach the aquatic ecosystem via irriga-
tion return flows through interflow or ground water
seepage. Most of the literature reviewed identifies agri-
culture as the source of pesticides in sediment because
of upstream land use, chemical use, and the nature of
the chemicals detected in sediments. Contamination of
sediment associated with major agricultural areas of the
United States has been reported in numerous studies.
For example, the San Joaquin River, in the highly agri-
cultural central valley of California, has bed-sediment
concentrations of the pesticides DDT and dieldrin among
the highest of all major rivers in the United States
(Gilliom and Clifton, 1990). Researchers have also
found continued elevated levels of highly persistent or-
ganochlorines in bottom-feeding fish, a condition that
is often a consequence of sediment contamination. In
the Yakima River in Washington, which drains a largely
agricultural region, concentrations of DDT in fish for
the years 1989-90 were found to be similar to concen-
trations for the years 1970-76 (USGS, 1993).

Contaminant contributions from past mining activi-
ties are so significant that several former mining sites
in the United States have been included on the EPA
Superfund Program's National Priorities List of sites for
remediation, including the Clark Fork River Basin in
Montana, the Bunker Hill Complex in Idaho, White-
wood Creek and the Belle Fourche River in South Da-
kota, Tar Creek in Oklahoma, Iron Mountain in
California, and the Arkansas River and tributaries near
Leadville, Colorado. The persistence and mobility of
heavy metals have resulted in concentrations in sedi-
ments up to 65 miles downstream of discharge similar
to the elevated concentrations found in the mine tail-
ings themselves (Henny et al., 1994). Based on infor-
mation provided by the states, the Bureau of Mines
estimated that abandoned coal and metal mines and their
associated wastes adversely affect more than 12,000 miles
of rivers and streams and more than 180,000 acres of
lakes and reservoirs (Kleinman, 1989).

The primary sediment contaminants of concern as-
sociated with mining are heavy metals such as lead, mer-
cury, zinc, cadmium, copper, manganese, and silver.
These metals are primarily associated with historical
mining of silver, gold, lead, and zinc. A literature re-
view of studies related to mining pollution provided pub-
lications describing the effects of mining on water
quality; however, few researchers have directly addressed
the effects of mining on sediments. A monitoring study
performed on Idaho's Lake Coeur d'Alene surface sedi-
ment found that ores and wastes from a mining district

were the source of elevated sediment concentrations of
several heavy metals via transport down the Coeur
d'Alene River (Horowitz et al., 1993). Moore et al.
(1991) performed an integrated sediment-water-biota
monitoring study on the effects of acid mine effluent on
the Blackfoot River in Montana. These researchers found
elevated levels of heavy metals in sediment from tribu-
taries with known historical mine effluent input that were
higher than levels in nonaffected tributaries. In another
study from the gold mining region of northern Georgia,
elevated mercury concentrations decreased as distance
of the sampling sites from the mining district increased
(Leigh, 1994). The author further suggests that similar
occurrences of mercury contamination could exist
throughout the gold mining region of the Southern Pied-
mont because of the historical amalgamation processes
used by gold miners.

Atmospheric deposition is often identified as a ma-
jor source of mercury, PCBs, and PAHs to aquatic sys-
tems. Studies have also implicated atmospheric sources
as an important contributor of metals. Sources that emit
large amounts of many toxic chemicals to the atmosphere
include industrial point sources, fuel combustion in mo-
tor vehicles, volatilization of compounds from landfills
and open water, combustion of wood and other fuels to
produce heat, and waste incineration. In addition, long-
range atmospheric transport of organochlorine pesticides
from countries where their use is still permitted contrib-
utes these ci impounds to aquatic environments in this
country (Keeler et al., 1993).

Atmospheric sources of mercury include coal com-
bustion, wriste incineration, and paint application.
Sorensen et al. (1990) compared mercury levels in sedi-
ment cores from lakes in northern Minnesota with pre-
cipitation loadings from monitoring and concluded that,
on the average, direct wet atmospheric deposition ac-
counts for 60 percent of the mercury in lake sediment.
A 1994 EPA report to Congress entitled Deposition of
Air Pollutants to the Great Waters also describes mass
balance studies from Wisconsin and Sweden indicating
that atmospheric deposition is responsible for most of
the mercury in lakes (USEPA, 1994a). The Swedish
study also points out that mercury deposited onto forest
soils is stored, for potentially long periods of time, be-
fore it enters the lake through storm water runoff. This
further illustrates the relationship between atmospheric
deposition and runoff.

Sources of PCBs to the atmosphere include munici-
pal and haziirdous waste landfills, refuse and sewage
sludge incinerators, and occasional leakage from elec-
trical transfcrmers and capacitors (Keeler et al., 1993).

4-5


-------
Pollutant Sources

Researchers have developed a mass balance for PCBs in
Lake Superior that indicates that approximately 77 to
89 percent of the annual PCB input to the lake is from
atmospheric deposition (Baker et al., 1993, cited in
USEPA, 1994a). These researchers have also estimated
the percent contribution of PCBs from atmospheric depo-
sition for other Great Lakes, keeping track of the frac-
tion contributed from atmospheric deposition to upstream
lakes. For example, about 63 percent of PCB input to
Lake Huron is from direct atmospheric deposition, an
additional 15 percent is from atmospheric deposition to
the upstream Lakes Superior and Michigan, and the re-
maining 22 percent is from other sources. Lakes Erie
and Ontario receive only about 13 percent and 7 per-
cent, respectively, of their annual PCB load from atmo-
spheric sources.

Sources of atmospheric PAHs include stationary fuel
combustion, industrial production facilities, transporta-
tion, solid waste incineration, and forest and prairie fires.
Routine installation of catalytic converters in motor ve-
hicles, as well as other combustion emission controls,
have decreased PAH releases to the atmosphere. Atmo-
spheric transport of PAHs generated during fuel com-
bustion has often been inferred to account for the
appearance of PAHs in soils and sediments in regions
distant from known combustion sources, but quantifica-
tion of this process is scarce in the literature (Prahl et
al., 1984). Researchers typically state that the types of
PAHs detected in sediments at a particular study site are
indicative of combustion sources, thereby implying that
atmospheric deposition is probably the primary source
to the aquatic environment (Helfrich and Armstrong,
1986; Rice et al., 1993). In a rare attempt to quantify
this contribution, Prahl et al. (1984) studied atmospheric
particulate matter and surface sediment in Washington
State coastal sediments and estimated that atmospheric
transport accounted for about 10 percent of the PAHs in
sediment. However, unlike the examination of PCBs in
the Great Lakes described above, the authors did not
account for the atmospheric contribution to upstream
watcrborne inputs.

Metals are released to the atmosphere from sources
such as primary and secondary metal production and, in
the past, use of leaded gasoline. Mass balance studies
of metal inputs to the aquatic environment have identi-
fied atmospheric deposition as an important contribu-
tor, but less significant than riverine and upstream
sources. As was the case with the PAH mass balance in
Washington, these studies do not identify the atmospheric
portion of riverine or upstream sources. In one study,
estimates of loadings to Narragansett Bay, Rhode Island,
indicated that atmospheric deposition contributes 2 per-

cent of copper and zinc and 33 percent of lead in sedi-
ment (Bricker, 1993). Based on a mass balance study
on Delaware Bay, direct atmospheric deposition accounts
for 7 percent of the cadmium loading to the bay; rivers
(72 percent) and salt marshes (21 percent) account for
the remaining cadmium input. Some portion of the riv-
erine input originates from the air (USEPA, 1994a).

Atmospheric deposition is a significant source of
dioxins and furans found in sediment. These highly
persistent compounds are grouped with "other organ-
ics" in Figure 4-1. Municipal and industrial waste in-
cineration and residential and industrial wood
combustion were both listed as important sources of di-
oxins and furans to the environment in two recent re-
views (Voldner and Smith, 1989 and Johnson et al., 1992,
cited in Keeler et al., 1993).

The category "urban sources" refers broadly to run-
off from roadways, residential and commercial areas,
construction sites, and marinas and shipyards. Accord-
ing to EPA's National Urban Runoff Program (NURP)
studies, the principal toxic pollutants found in urban
runoff are metals, oil and grease, PAHs, and petroleum
hydrocarbons (USEPA, 1992b). Much of the pollution
in urban runoff is associated with atmospheric deposi-
tion, particularly for mercury and PAHs. Other classes
of chemicals, such as metals and petroleum hydrocar-
bons, have many land-based sources. Lead was formerly
contributed by car exhaust, but most contributions now
come from exterior paints and industrial runoff. Cad-
mium is also associated with paints. Zinc is associated
with weathering and abrasion of galvanized iron and
steel. Car brake linings and leaching and abrasion of
copper pipes and brass fittings contribute copper to run-
off. Chromium is contributed to runoff through car and
machinery corrosion (Cohn-Lee and Cameron, 1991).
Sources of petroleum hydrocarbons include disposal of
automobile and industrial lubricants, spillage from oil
storage facilities, and leakage from motor vehicles
(Brown et al., 1985). In addition to agricultural uses,
organochlorine pesticides were also used extensively in
urban and residential areas for a variety of pest control
purposes.

The association of urban sources and metal enrich-
ment of sediment is well documented in the literature.
For example, a study of storm water detention ponds in
Florida, Virginia, Maryland, and Minnesota found that
metal concentrations in surface sediments were typically
5 to 30 times higher than those in the parent soils
(Schueler, 1994). This study also reported the highest
metal concentrations in ponds associated with indus-
trial land use, followed by those associated with roads

4-6


-------
National Sediment Quality Survey

and commercial land use, then those associated with resi-
dential land use. In contrast to atmospheric transport,
which can carry pollutants far from their original source,
runoff of metals tends to affect areas in close proximity
to the source. For example, Yousef et al. (1985) sampled
water and sediments in detention ponds in Florida and
found that metals from highway runoff are retained by
bottom sediments close to the point of entry to the water-
way.

Hydrocarbons, PAHs, and mercury are also fre-
quently associated with urban sources. Using analyti-
cal chemistry techniques, Brown et al. (1985) discovered
that crankcase oil was a primary contributor to sediment
hydrocarbon contamination in Tampa, Florida. Gas
chromatograms of used crankcase oil, storm water run-
off, and sediment samples all showed similar peaks, in-
dicating that the type of petroleum found in sediment
very closely resembled that found in storm water runoff.
Sources of PAHs that are concentrated in urban areas
include emissions from commercial and residential fuel-
burning furnaces and vehicular emissions. An inven-
tory of sediment contamination in Casco Bay, Maine,
showed that the highest PAH concentrations occurred at
locations closest to the city of Portland (Kennicutt et al.,
1994). Mastran et al. (1994) found that sediments from
urban areas tend to have lower fluoranthene/pyrene ra-
tios than those from remote areas. These ratios are in-
dicative of pollution caused by gas exhaust residues in
urban runoff. A study of ambient air in the southern
Lake Michigan basin revealed that concentrations of
mercury, both gaseous and particulate, are significantly
higher (approximately 5 times higher) in the Chicago
urban/industrial area than levels measured at the same
time in surrounding areas (Keeler, 1994, as reported in
USEPA, 1994a).

In addition to the nonpoint source categories dis-
cussed above, municipal and industrial point sources
have been associated with sediment contaminated by
each of the chemical classes examined in this report.
Much of this contamination has been caused by past in-
dustrial and municipal discharges. For example, sedi-
ment core samples from southwestern Long Island, New
York, revealed levels of metals that increased to several
times the preindustrial concentrations, then decreased
approximately 50 percent between the mid-1960s and
late 1980s. PCBs, chlordane, and other chlorinated or-
ganics in sediment also decreased between the late 1960s
and the late 1980s. Local improvements in wastewater
treatment and national efforts to restrict the use of spe-
cific chemicals are cited as explanations for the declines
(Bopp et al., 1993). As previously mentioned, past ef-

fluent discharges from electronics manufacturers are
linked to PCB contamination in New Bedford Harbor,
Massachuselts (Garton et al., 1996; Lake et al., 1992).
Perhaps the best example of pesticide contamination in
sediment from past industrial activity is kepone in the
James River, Virginia. Kepone escaped undetected from
a manufacturing site for over 9 years and contaminated
miles of the James (Nichols, 1990).

A well-documented case of the effects of point
sources on sediment quality is the Newark Bay estuary
in New Jersey, which encompasses the Passaic River,
Hackensack River, Kill van Kull, and Arthur Kill.
Wenning et al. (1994) examined sediment core samples
from the lower Passaic River in New Jersey and con-
cluded that the sediment is heavily contaminated with
PCBs, PAHs, and metals from recent and historical mu-
nicipal and industrial discharges from local and upstream
sources. The authors identify industrial effluent, either
directly discharged or released through combined sewer
overflows, as the most likely primary source. Research-
ers have also measured high levels of dioxin in sedi-
ment in the estuary adjacent to an industrial site in
Newark where chlorinated phenols had been produced
(Bopp et al., 1991). In a recent study, researchers deter-
mined that the magnitude of current loading estimates
for metals and organics from major sources, such as in-
dustrial and municipal discharges and combined sewer
overflows, likely exceeds the capacity of the Newark Bay
estuary to absorb and dilute the various waste streams
(Crawford et al., 1995).

EPA has conducted an inventory and analysis of
point source releases of sediment contaminants in the
United States. This inventory includes examination of
data from effluent monitoring required by discharge per-
mits and chemical release estimates provided by indus-
try under the community right-to-know provision of the
Superfund Amendments and Reauthorization Act of
1986 (SARA). Permit monitoring data indicate that mu-
nicipal sewage treatment plants and major industrial fa-
cilities discharge all chemical classes of sediment
contaminants. Metals are monitored at the greatest num-
ber of facilities and released in the largest amounts.
Mercury, PAHs, and other organics are also released from
many facilities. PCBs and pesticides are less frequently
monitored, and a relatively small number of records in-
dicate positive detections. Industry-supplied release es-
timates provided under SARA indicate that
manufacturing facilities transfer the majority of their
sediment contaminants, primarily metals and other or-
ganics, to municipal sewage treatment plants. The analy-
sis of these data addresses the potential to adversley affect

4-7


-------
Pollutant Sources

sediment quality, but does not indicate whether these
discharges actively contribute to documented cases of
sediment contamination.

Land Use Patterns and Sediment
Contamination

The characteristics of local sediment contamination
are usually related to the types of land use activities that
take place or have taken place within the area that drains
into the water body (the watershed). The previous sec-
tion of this chapter provided numerous examples of these
relationships from published studies. For this report,
EPA examined the relationship between the extent of
sediment contamination by chemical class and patterns
of land use in the 96 APCs. 'EPA identified individual
watersheds where land use appears to provide impor-
tant information concerning the types of contaminants
present, and summarized general trends that emerge by
looking at the percent of urban and agricultural land
areas in watersheds.

This analysis was based on a comparison of the ex-
tent of contamination by chemical class (described ear-
lier in this chapter) within each watershed to the percent
of land area developed for certain uses within the water-
shed. EPA used the Agency's modeling tool, Better As-
sessment Science Integrating Point and Nonpoint
Sources (BASINS), for spatial analysis to quickly ob-
tain land use data originally compiled by the U.S. Geo-
logical Survey (USGS) on a watershed basis. Although
these land use data might be as much as 20 years old,
the data compiled for the NSI have also been collected
over the past 15 years. The original land use data are
divided into 10 categories. EPA combined residential,
commercial/industrial, and other urban land uses in the
"total urban" land use category for this analysis. EPA
also combined cropland and other agricultural land/
rangeland in a "total agricultural" land use category.
This allowed comparison of attributes such as the per-
cent of stations with pesticide contamination and the
percent total agricultural land use.

Several difficulties are associated with this approach
to comparing land use to the evaluation of NSI sam-
pling stations. First, the frequency and spatial extent of
sampling data in the NSI vary by watershed. Second,
the acreage of a land use activity is not indicative of the
intensity of that use. For example, a small amount of
land in a watershed might be devoted to an industrial
activity that contributes a large amount of pollution.
Most watersheds contain at least a small fraction of each

land use activity. There are also problems of scale.
Localized problems in specific reaches might be caused
by land use activity in the immediate vicinity of the reach
rather than the overall land use in the watershed. Lastly,
many individual pollutants and chemical classes are as-
sociated with multiple types of sources. Some classes of
pollutants, like the highly persistent PCBs, have been
cycled in the environment for many years and trans-
ported far from their original source. These chemicals
would not be expected to be associated with any general
land use category.

Table 4-2 lists each of the 96 APCs with the num-
ber of Tier 1 and Tier 2 stations by chemical class and
the percent land use information. In general, EPA found
that a diversified set of land uses yields a diversified set
of pollutants. However, in some cases a preponderance
of one land use type is associated with expected chemi-
cal classes of sediment contaminants. For example, the
Lower Yakima watershed in Washington, an intensive
fruit and vegetable growing region, is approximately
81 percent agricultural and only 2 percent urban. In
this watershed, nearly 90 percent of the sampling sta-
tions were contaminated with pesticides, whereas no sta-
tions exhibited mercury contamination and less than 10
percent exhibited contamination from metals or PAHs.
These percentages were substantially different from the
average values presented in Figure 4-1. Similar find-
ings were evident in other highly agricultural watersheds,
such as the Tulare-Buena Vista Lakes in California.

In some cases, the absence of a particular land use
in a watershed can provide clues about the source of in-
place contaminants. Some watersheds, such as the Lower
Mississippi-New Orleans in Louisiana and the
Hackensack-Passaic in New Jersey, have very low agri-
cultural land usage, yet a high percentage of contami-
nation from pesticides. High levels of contaminants in
recent sediment deposition may indicate upstream de-
livery of contaminants, whereas high levels in buried
sediment may be indicative of pesticide manufacture/
formulation or urban applications in the past. In the
Coeur D'Alene watershed in Idaho, there is very little
agricultural land use and almost no urban land use. In
this watershed, where mining is a known source of con-
tamination, over 90 percent of the stations exhibited
metal contamination, whereas none indicated PAH or
pesticide contamination. In other watersheds with very
low percent urbanization, there was substantial contami-
nation from all chemical classes except PAHs. This phe-
nomenon was evident in several nonurbanized
watersheds in the Southeast and upper Midwest, such
as Pickwick Lake and Guntersville Lake. Further ex-

4-8


-------
Table 4-2. Tier 1 and Tier 2 Station Classification by Chemical Class and Land Uses in Watersheds Containing Areas of Probable Concern for
Sediment Contamination (APCs)







Number of Stations With a Probability of Adverse Effects

Percent of Total Area is Each Watershed

EPA
Rat.

Cataloging
Ural#

Name

Her

Mercury

Other
Metals

PCBs

Pesticides

PAHs

Otter

All
Chemicals'

Total
#of
Stations

Residential

Commercial/
Industrial

Other
Urban

Cropland

Other
Agricultural

fisrestland

Bays &
Estuaries

Other
Water

Other

Missing/
Unknown

1

01090001

Charles

1

2

146

216

63
486

35
54

8

50

11

50

1

0

195

402

708

25.43%

5.95%

4.56%

3.06%

0.04%

39.57%

7,82%

5.86%

1.47%

6.23%

i

01090004

Narragaesetf

1

2

8
20

18
27

4

17

3
18

2
22

0
0

28
20

48

13.74%

3,58%

4.61%

7.41%

0.86%

51.56%

9.96%

6.27%

1.14%

0.88%

1

01090002

Cape Cod

1

2

6
27

3

60

8
33

1

33

5
34

0
0

15

73

108

5.90%

0.81%

1.77%

1.84%

4,12%

22.90%

35.05%

4.26%

13?%

21.98%

2

04120103

Buffalo-Bgbteerowie

1

2

20
45

7

79

29
31

29

31

43
17

29
15

59
33

101

8,27%

3.54%

3.20%

42.85%

0.10%

30.94%

10.31%

0.35%

©.43%

0.02%

2

02030103

Hackensack-Pa$$aic

1

2

21

39

12
75

13
34

23

42

10
13

4

19

43

58

103

33.33%

7,24%

5.65%

2.62%

0.26%

38.99%

0.00%

6.94%

1.33%

3.64%

2

04130001

Oak Orchard-TVelvemilc

1

2

10
30

20
61

4

15

8

20

4

12

2
13

39
46

¦ §6

2.25%

44.43%

1.25%

10.48%

3.29%

8.42%

26.77%

178%

0.29%

QM%

2

02030104

Sandy Hook-Siaieo Island

1

2

53
11

40
30

19
9

17
19

12
29

20
5

60

21

100

30,58%

10.23%

7,70%

6.99%

0.49%

7.83%

13.66%

7.27%

122%

13.03%

2

04120104

Niagara

1

2

5
16

0
29

11
9

13
U

19
9

16
M

24
16

41

9.35%

32.02%

3.91%

31.59%

0.24%

17.47%

0.02%

3.61%

0.92%

0,87%

2

04150301

Upper St. Lawrence

1

2

5
8

0

17

21
5

3

a

3
6

9
5

21
5

31

1.51%

0,85%

1.29%

36.31%

0.75%

28.47%

0.06%

26.73%

0,21%

3.82%

2

02030105

Ran tan

1

2

1

11

1

39

4

25

5

27

I
4

1

3

13

37

65

15,15%

4,87%

2,99%

25,86%

0,49%

26.55%

0.00%

165%

1.01%

20.43%

2

02040301

Mullica-Toms

1

2

2
10

0
24

2
10

2
11

I

15

5
4

10
22

42

8,54%

1.71%

1.18%

6.04%

0.52%

43.11%

7.97%

20.75%

132%

7.86%

2

02040105

Middle Delaware-Musconetcong

1

2

1
3

1

19

8
13

1

20

1

2

0
0

U
26

48

5,49%

1.53%

1.26%

38.02%

0.16%

33.98%

0.00%

168%

0.67%

16.22%

2

02030202

Southern Long Island

1

2

7
12

4

25

I
8

4

8

1

14

2

2

U
24

43

23.38%

5.03%

5,06%

4.29%

0.74%

10.73%

19,75%

3-26%

1.88%

25.88%

3

02060003

Gunpowder-Fatapsco

1

2

2
6

3

19

15
4

0
21

1

1

0
4

17
7

29

13-47%

5.10%

4,32%

40,80%

0,11%

26.70%

4.62%

4.11%

0.76%

0.01%

3

02040203

Schuylkill

1

2

0
5

1

16

U
6

0
14

0
0

2
0

12

23

44

9.17%

168%

2.78%

41.37%

0.26%

25.81%

0.00%

0.65%

146%

14.82%

3

05030101

Upper Ohio

1

2

0
0

0
29

12
0

0
9

0

0

0

1

12
29

53

13.08%

2.52%

118%

3526%

0.34%

43.13%

0.00%

1.07%

142%

0.00%

3

02070004

Cotiocochtague-Opcquoa

1

2

0

2

0
17

11

0
13

0
0

1
0

11

12

29

1.88%

0.98%

0,89%

50.58%

1.55%

43.24%

0.00%

0.51%

0.34%

0.02%

3

02040202

Lower Delaware

1

2

1
7

I

23

12

20

5
33

1

2

5
0

IS
29

57

26.68%

13.51%

6.47%

21.76%

1.90%

18.45%

0.18%

9.61%

1.17%

0.27%

3

05030102

Shenango

1

2

0
0

0
2

U
0

0
8

0
0

0
0

11
1

15

3.93%

0.76%

2.20%

74.41%

0.02%

12.85%

0.00%

5.36%

0.44%

0.02%

3

04120101

Chaitaaqua-Conaeaut

1

2

1

22

0
101

18
15

0

20

3

29

4
13

21
86

110

4.07%

1.13%

2.05%

38,07%

0.21%

21.58%

31.10%

0.18%

011%

1.40%

4

06010201

Waas Bar Lake

1

2

5

5

0

10

58
2

0
14

0

0

1
1

63
7

89

9.71%

1.84%

1.29%

27.72%

0.06%

52.32%

0.00%

5.20%

1.87%

0.01%

4

06010207

Lower Clinch

1

2

46
11

19
33

24

0

0
7

4
14

3

20

61
14

79

11.76%

1.74%

1.24%

24.98%

0,04%

56.28%

0.00%

116%

1.63%

0.16%

4

0IO3OOO5

Pickwick Lake

1

2

8
11

1

24

45

2

1

23

0
0

0

2

49
9

69

1.93%

0.60%

0.33%

40.73%

0.07%

44.51%

0,00%

4.07%

1.35%

6.41%

4

06020001

Middle Tertt)csscc-Chkkamauga

2

14

15

57

16
1

1

12

26
0

7
9

47

29

94

8.14%

1.58%

1.19%

19.50%

0.04%

64.76%

0.00%

3.54%

1.44%

0.00%

4

03080103

Lower St Johns

1

2

7

35

0

76

5
18

3

48

22
57

2
1

32
111

188

6.99%

1.71%

1.57%

9,03%

1.72%

51.60%

0.00%

25.04%

1.98%

0.36%


-------
Table 4-2. (Continued)









Number of Swk*tt Willi * Probability of Advene

Effects



Percent of TouU Area !a Bad) Watershed























Total





















EPA

Catalog^;







Other









All

§ of



CcHBmerciaJ/

Other



Other



BaysJt

OAer



Minfogf

R*e,

Unit#

Name

Tier

Mercury

Meul?

PC?s

tadddes



Other

CbetakaU*

Stations

Residential

fcfktfUtal

Vtbm

Crooland

Arricvtturai

Rjrestfind

Esluwfes

Water

Other

Uninowo

4

06030001

OuctertvUltf Lai*

1

7

1

15

3

0

e

25

92

0.97 £

0J3%

0.23%

40.41%

0.05%

5234%

0,00%

5.18%

0.55%

0.05%







2

36

60

0

11

0

0

46























4

03130002

Middle Chauatheocfaee-Lake Harding

I

0

1

19

4

0

7

21

27

4.86%

0,77%

©.95%

1141%

0.12%

75.59%

0.00%

0.98%

1.27%

0.05%







2

3

8

3

14

2

2

4























4

03060106

Middle Savaoaah

I

II

1!

19

3

2

6

20

36

3.75$

1.7856

0,81%

16.90%

0.18%

62.67%

0.00%

1210%

1.80%

aoo%







2

6

10

3

13

2

2

11























4

03140102

Qoctawhatchee Bay

1

0

7

2

9

2

0

19

51

3.04%

4.94%

1.10%

3.03%

0.01%

61.80%

17.57%

3.14%

1.25%

4.13%







2

H

32

9

11

15

0

23























4

06040005

Kentucky Lake

1

0

0

14

0

0

1

15

30

125%

0.33%

0.26%

25.78%

0,00%

58,59%

0.00%

13.00%

0.76%

0.03%







2

9

25

Q

2

0

2

14























4

06040001

Lower Tennessee-Boedj

1

1

0

14

0

0

I

15

25

0.38%

0.12%

020%

28.06%

0.01%

65.47%

0:00%

3.01%

1.82%

0,94%







2

I

11

0

13

0

0

6























4

06020002

Hiwwsce

1

1

0

12

0

0

2

13

33

265*

0.51%

0.58%

18,99%

0.11%

58.13%

0.00%

1.63%

1.77%

15.63%







2

6

ES

0

6

0

0

17























4

08010100

Lower .Miisisrippl-Mesaphii

1

I

r

12

0

0

4

14

20

0.57%

0.81%

0.35%

49.87%

0.06%

21.07%

0,00%

25.08%

209%

0.03%







2

0

3

2

15

0

0

3























4

06010104

Holflon

1

3

1

10

0

0

2

12

15

4.73%

1.14%

0,45%

4435%

0.01%

43,72%

0.00%

529%

0.30%

0.00%







2

3

6

1

4

0

0

2























4

03040201

Lower Pee Dee

1

1

0

7

5

0

2

11

34

202%

0.55%

0.47%

3203%

0.20%

54.90%

0.01%

9,43%

0.38%

0.01%







2

16

16

1

16

I

0

20























4

03160205

Mobile Bay

1

11

13

2

1

4

0

31

81

4.22%

0,91%

0.97%

2.68%

0.43%

9,60%

18.20%

1.97%

(X33%

60.70%







2

14

38

6

16

21

0

43























4

08030209

Dier-Stee!*

1

0

0

0

11

0

0

11

21

1.23%

0.57%

0.77%

74.35%

©.91%

18.66%

0.00%

3.34%

0.03%

0.08%







2

0

7

0

10

0

0

10























4

Q3I4O107

PodtdoBay

1

s

0

1

0

1

1

10

38

8,04%

2.35%

1.12%

159%

0.16%

14.87%

8.08%

4.77%

1.61%

56.39%







2

8

IS

3

0

9

0

24























4

03060101

Seoeca

1

1

1

9

3

0

0

10

16

0,54%

0.02%

0.02%

0.12%

0.00%

13.24%

0.00%

0.58%

0.36%

85.13%







2

1

8

2

1

0

0

3























5

04090004

Detroit

I

42

21

74

42

53

38

85

ns

42.87%

1265%

8.99%

24.55%

0.18%

5.95%

0.78%

2.29%

1.74%

0.00%







2

27

90

31

1

19

17

29























5

07120003

Chicago

1

21

23

34

IS

0

0

64

103

36,16%

19.12%

8.10%

20,63%

0.00%

4.45%

8.76%

1.14%

1.63%

0.00%







2

27

52

Ifi

37

0

O

36























5

07120004

Des Plaines

1

12

4

54

a

0

1

61

110

21.71%

9.97%

6.61%

43.40%

0.31%

7.47%

0.00%

2.04%

3.48%

0.00%







2

18

53

24

76

0

0

43























5

&WM0003

Milwaukee

i

5

6

43

6

20

14

60

90

11.83%

5.78%

4.20%

65.30%

0.08%

6.64%

0.10%

4.68%

0,41%

0.00%







2

22

38

3

32

6

15

16























5

04030204

Lower Fox

1

21

3

41

8

5

5

49

51

8.94%

5.28%

2.88%

76.15%

0.04%

3.43%

0.11%

2,19%

0.98%

0-00%







2

5

27

1

16

14

19

2























5

04040001

Little Cdumet-GaUeo

1

10

14

40

9

7

10

45

89

7.34%

6.16%

2.59%

37.11%

0.22%

- 1287%

30.51%

2.12%

1.08%

©,00%







2

24

48

6

12

0

3

26























5

04040002

Pike-Root



5

4

2S

3

I

1

34

72

12.02%

5.19%

4.10%

33.68%

0.04%

0,93%

43.58%

0.18%

0.29%

0.00%







2

16

40

n

16

3

3

30























5

07140201

Upper Kaskaskia



0

0

23

14

0

0

31

55

1,19%

0.39%

0.69%

90.79%

0.02%

5.83%

0.00%

1.05%

0.04%

0.00%







2

4

8

6

38

0

0

24























5

07010206

Twin Cities



0

0

26

0

0

0

26

35

21.99%

5,24%

5.12%

48.03%

0.03%

4.39%

0.00%

14.24%

0.95%

0.00%







2

1

2

0

5

0

1

2























5

07140106

Big Muddy



2

2

20

0

0

0

23

94

1.96%

0.91%

0.66%

70.37%

0.51%

20.43%

0.00%

3.60%

1.56%

0.00%







2

14

61

13

39

0

0

65























5

07070003

Castle Rock



0

0

20

0

0

2

20

22

1.05%

0.53%

0.55%

40.77%

0.05%

37.43%

0.00%

18.97%

0,64%

0.00%







2

2

1

0

5

0

0

0
























-------
Table 4-2. (Continued)







Number of Stations With a Probability of Adverse Effects

Percent of Total Area in Each Watershed























Total





















EPA

Ctiaiogmg







Other









All

#of



Commercial/

Other



Other



Bays &

Other



Missing/

Reft*

Unit#

Name

Tier

Mercury

Metals

PCBi

Pesticides

PAHs

Other

Chemicals*

Stations

Residential

Industrial

Urban

Ciooland

Agricultural

Forestland

Ea futile*

Water

Other

Unknown

5

04100002

Raisin

1

1

0

17

7

1

©

18

39

2.25%

1.00%

0.74%

87.13%

0.15%

5,46%

0.01%

2.90%

0.35%

0.00%







2

2

7

17

13

2

6

19























5

04050001

St. Joseph

1

0

1

3

7

7

3

17

32

3.08%

1,42%

1,02%

79.21%

1.25%

9.23%

0.03%

4,45%

0.31%

0.00%







2

0

18

0

5

2

6

9























5

07040003

Buffalo-WhlEtwa^r

1

0

0

17

0

0

©

1?

26

0.74%

0,29%

0,40%

54.93%

0.05%

37,00%

0.0©%

6.50%

0.08%

0.00%







2

i

2

o

6

0

0

3























5

04U0001

Black-Rocky

1

2

0

12

7

21

9

24

53

11.18%

2.79%

4,40%

66.45%

0,20%

U.U%

3.20%

0.38%

0.29%

0.(30%







2

n

U

7

4

7

i

•*?























5

0?I20006

Upper Pox

1

0

©

15

0

0

0

15

60

10,36%

2.44%

2.38%

63.18%

0.61%

10.84%

0.00%

7.42%

2.77%

0.00%







2

12

37

H

27

9

9

40























5

0512011!

Middle Wabaih-Bussecon

1

7

0

9

0

0

0

15

33

2,49%

0.92%

1.02%

79.64%

0,09%

13.31%

0.00%

1.50%

1.03%

0.00%







2

9

23

ff

30

0

ft

17























5

07140202

Middle KaikaskUs

1

1

©

5

8

0

0

13

38

1,21%

0.40%

0.60%

78.52%

0.09%

16.06%

0.00%

3,01%

0.10%

0.00%







2

4

16

ft



fl

ft

77























5

07040001

Rush-Vermillion

1

0

0

13

0

0



13

14

1.38%

0.59%

0.44%

80.68%

0,06%

9.43%

0.00%

7.07%

0.34%

0.00%







2

2

3

9

3

9

p

1























5

05120109

Vermilion

1

8

0

4

0

0

0

12

28

- 3.92%

1.00%

0,73%

90.08%

0.10%

3.51%

. 0.00%

0,15%

0,50%

0,00%







2

2

19

1

26

0

0

16























5

04030108

Menominee

1

5

4

5

0

2

1

12

21

0.55%

0.17%

0.29%

10.13%

0.01%

<>738%

0.01%

20,94%

0.31%

0.01%







2

8

7

1

2

1

0

6























5

04090002

Lake St, Clair

1

1

2

10

S

5

9

13

19

18.44%

3.81%

2,35%

28.70%

0.06%

3,60%

38,06%

4,87%

an%

0.00%







2

10

13

6

S

8

5

5























5

07140101

Cahokia-Joachira

1

4

1

11

2

0

5

1$

56

10.64%

4.50%

4.32%

42,42%


-------
>3 Tiable 4-2. (Continued)







Nwa&er oC Swieos With x Pnbb&ty of A4vmc Meets

Pereeaj of Toul Area, ia Each Waiertbed

EPA
Ref»

Caltkfieg
Unilf

Nunc

Tier

Mercury

Other
Metals

PCBs

Pesticides

PAH*

Other

Aft
Chejaicals*

TotH
iof
Stations

Residential

CoMBKSCtll/

liKfau&ial

OUter
Utfaaa

Oopiitid

Other
Agricultural

Beralliad

B*y*&
Estuaries

Other
Water

Other

Mhttaj/
Uakccwi

6

12040104

BuMnSib Jacioto

1

2

0
14

i
26

9
15

3
14

1

11

3
3

10
23

36

2331%



632%

45.96%

0.06%

1338%

0.04%

197%

0.80%

0.08%

7

10210 lM

Lower Kansas

1

2

©
1

I

14

11
0

0
22

€
1

1

3

12
15

29

3.70%

1.82%

1.83%

82.75%

0.91%

7.67%

0.00%

0.92%

0.40%

0.00%

7

I10IO2S7

Spring

1

2

0

1

0

29

8

1

0
7

1
O

2

1

10

25

41

1,84%

0.67*

0.79%

80.42%

0.12%

14.27%

0.00%

0.19%

1.70%

0.01%

?

070S0101

Copperas-Duck

1

2

1

1
7

1?
0

0
18

0

1

1

2

17
5

2?

5.40%

2^3%

1.58%

68.60%

0.18%

9.58%

0.00%

9.04%

0.54%

2.55%

9

18070304

Sen Diego

1

2

IS
26

4

93

33
45

13
47

7

39

2
4

53
51

107

11.02%

4.09%

2.72%

6.92%

54.85%

9.62%

1.36%

0.86%

1.98%

6.60%

9

18070104

Santa Moaica Bay

2

15

33

6
94

22
34

66

21

4
18

1
3

79
31

132

17.03%

7,90%

2.86%

1.18%

20.81%

0.68%

0.41%

0.20%

0.96%

47.95%

9

18070201

Seal Beach

1

2

5
38

0
21!

8

142

23
288

2
30

32

182

63
339

442

41.18%

22.80%

4.68%

4.98%

0.12%

0.00%

0.75%

1.15%

157%

23.05%

9

18050003

Coyote

1

2

14
8

8

12

0

1

0
0

0

1

0
0

18

6

24

20,29%

9.69%

9.13%

6,07%

23.27%

27.93%

1.58%

138%

0.66%

0.01%

9

18070204

Newport Bay

1

2

10
13

0

62

1

19

11
48

0
8

2
25

24
68

108

19.51%

13.49%

6.60%

18.96%

28.16%

0.25%

1.09%

0.91%

333%

7.69%

9

18050004

San Francisco Bay

1

2

10
33

9
41

1

18

0
19

5
21

0

0

19
37

64

12.06%

7.21%

3.48%

4.43%

27.36%

28.64%

1430%

1.98%

0.65%

0.00%

9

18070105

Los Angeles

1

2

4

16

0

33

2
4

8
10

3
5

0

1

14
19

37

3836%

13.78%

6.51%

1.31%

31.59%

6.65%

0.02%

0.30%

1.46%

0.01%

9

18030012

Tulare-Buena Vista Lakes

1

2

0

1

0

5

1
4

10

S

1

0

I
0

10
5

20

1.76%

U3%

0.70%

55,36%

38.72%

0.90%

0.00%

0.74%

0.26%

0.03%

9

18070107

San Pedro Channel Islands

1

2

7
3

2
22

2
6

10
3

0
4

0

3

14
10

25

0.00%

0.08%

0.01%

0.00%

2.59%

0.00%

0.02%

0.00%

0.18%

97.12%

9

18070301

Aliso-Sac Ouofre

1

2

5
7

2

29

0
9

5
7

0

2

0
0

10
22

32

3.18%

1.26%

1.22%

4.37%

60.80%

5.39%

0.03%

0.26%

1.49%

22.01%

10

17110019

Puget Sound

1

2

98
449

52
1116

146

317

37
106

296
490

32
317

418
8SI

1383

12.36%

2.12%

Z05%

3.75%

0.32%

4135%

34.95%

2.62%

0.48%

0.00%

10

17110013

Duwamish

1

2

0

27

3

107

34

10

3
!7

12
SB

6
23

48

69

127

1199%

2.97%

4.23%

6.82%

035%

70J5%

0.00%

0.96%

0.63%

0.00%

10

17110002

Strait of Georgia

1

2

16
51

1

180

1

15

4

34

12
73

4

28

32
168

263

4.22%

0.75%

1.22%

10.95%

0.46%

28.13%

5138%

2.61%

0.20%

0,07%

10

17030003

Lower Yakima

1

2

0
0

0
4

5
0

19
23

0

1

1

10

23
19

47

1.13%

0.52%

0.26%

25.97%

55.06%

15.65%

0.00%

1.23%

0.17%

0.01%

10

17090012

Lower Willamette

1

2

1

12

0
51

13

24

10

18

5
11

4

IS

21

51

76

31.21%

6.41%

4.69%

13,32%

0.97%

19.01%

0.00%

3.77%

0.61%

0.00%

10

17110014

Puyallap

1

2

0
0

3
8

1
6

0

1

8

9

1
6

12
6

19

5.85%

0.55%

0.79%

3,78%

4.44%

81.43%

0.00%

0.68%

2-47%

0.01%

10

17010303

Cocur D'AJeoe Lake

1

2

1
t

8
13

2
0

0
0

0
0

0

0

10
13

23

0.73%

0.13%

0.42%

1268%

0.65%

75.10%

0.00%

10.14%

0.14%

0.00%

'Because of the numerous chemicals moaiiored at each station, the tola] |q this column is not equal to die sum of the numbers in tbe columns for the different chemical classes.
"Adapted from USGS land use and land cover classiHeatioo system for use with remote sensor data.


-------
Natimuil Sediment Quality Survey

Ave. Agri. Use
Ave. Urban Use
Ave. Forest Use

10%
20%
36%

amination of percent agricultural
and urban land use revealed some
general trends that are illustrated
by these examples.

A high percentage of agricul-
tural land use in a watershed
tended to correspond with a mark-
edly higher percent contamina-
tion from pesticides and lower
percent contamination from met-
als, mercury, and PAHs. This
phenomenon is presented graphi-
cally in Figure 4-2 and in tabular
form on Table 4-3. For this analy-
sis, EPA grouped watersheds into
quartiles based on percent total
agricultural land use and calcu-
lated the average percent of sam-
pling stations with contamination
by chemical class. Some general
trends that would be expected
were clearly evident. In water-
sheds with greater than 75 per-
cent of the land devoted to
agriculture, pesticide contamina-
tion jumped from under 40 per-
cent of all stations to 64 percent.

In contrast, metal, mercury, and
PAH contamination all steadily
decreased, with all three classes
exhibiting a percent contamina-
tion in the over 75 percent agri-
culture group at least 10
percentage points under the over-
all average for each class. PCBs
and other organics did not exhibit any trend and never
varied more than 5 percentage points from the overall
average.

In contrast, increasingly higher percentages of ur-
ban land use in watersheds correlated with steadily in-
creasing contamination from most chemical classes.
Figure 4-3 and Table 4-4 present the results of a trend
analysis for total urban land use. For this analysis; EPA
placed watersheds into groups of under 5 percent urban
area, 5 to 10 percent urban area, 10 to 20 percent urban
area, and greater than 20 percent urban area to best il-
lustrate trends. The percent PAH and metal contamina-
tion were both 10 percentage points under the overall
average for the least urbanized watershed group, then
rose sharply as the proportion of urban area crossed the
5 percent threshold. The extent of metal contamination
rose to an average of 71 percent, more than 10 percent-
age points above the overall average of 59 percent, in



Metal*

>

PCBs



Pesticides

X

Mercury

X

PAHs



Others

100*

Percent Agricultural Land Use

36%
19%
28%

63%
10%
17%

83%
5%
8%

Figure 4-2. Percent Tier 1 and Tier 2 Stations vs. Agricultural Land Use in
APCs.

Table 4-3. Comparison of Percent Agricultural Land Use in Watersheds
Containing APCs to Percent of Tier 1 and Tier 2 Stations by
Chemical Class



Percent Total Agricultural Land Area

<25% .

25-50%

50-75%

>75%

Overall
Average

Average Percent Agricultural Land Area in Group

10%

36%

63%

83%

39%

Number of Watersheds in Group

32

34

13

17



Metals

66%

60%

58%

47%

59%

PCBs

38%

48%

45%

42%

43%

Pesticides

37%

39%

40%

64%

43%

Mercury

32%

34%

20%

18%

29%

PAHs

30%

17%

12%

9%

19%

Others

13%

16%

9%

12%

14%

watersheds with more than 20 percent total urban land
use. Mercury contamination rose steadily and reached
a peak of 40 percent in the most heavily urbanized wa-
tersheds. The mercury and PAH trends perhaps illus-
trate the effect of atmospheric deposition from local
urban sources. Contamination from other organics also
rose steadily, but never varied more than 6 percentage
points from the overall average. Pesticide contamina-
tion initially decreased as percent urbanization increased,
but it rose more than 10 percentage points from the 10
to 20 percent urban group to the over 20 percent urban
group. As mentioned previously, this may reflect up-
stream delivery of contaminants, pesticide manufacture
or formulation, or urban applications in the past. As
was the case with the agriculture analysis, the average
percent PCB contamination for the urban groups showed
no trend and never varied substantially from the overall
average.

4-13


-------
Pollutant Sources



Metals

¦

PCBs

—A—

Pesticides

X

Mercury

X

PAHs



Others

15% 20%

Percent Total Urban Land Use

Ave. Urban Use
Ave. Agrl. Use
Ave. Forest Use

2% 7%
51% 38%
29% 27%

14%
40%
29%

35% 40%

38%
26%
18%

Figure 4-3. Percent Tier 1 and Tier 2 Stations vs. Urban Land Use in APCs.

Table 4-4. Comparison of Percent Urban Land Use in Watersheds

Containing APCs to Percent of Tier 1 and Tier 2 Stations by
Chemical Class



Percent Total Urban Land Area

<5%

5-10%

10-20%

>20%

Overall
Average

Average Percent Urban Land Area in Group

2%

7%

14%

38%

16%

Number of Watersheds in Group

32

18

19

27



Metals

49%

61%

59%

71%

59%

PCBs

47%

37%

40%

45%

43%

Pesticides

50%

39%'

32%

44%

43%

Mercury

21%

24%

30%

40%

29%

PAHs

9%

25%

23%

25%

19%

Others

8%

12%.

15%

20%

14%

based on 1994 permit monitor-
ing records in EPA's Permit
Compliance System (PCS) and
chemical release estimates in
the 1993 Toxic Release Inven-
tory (TRI). The report presents
a screening analysis to identify
probable point source contribu-
tors of sediment pollutants
based on release amount,
chemical toxicity, and inherent
physical/chemical properties of
the contaminant. The report
serves as Volume 3 of the com-
plete report to Congress on the
incidence and severity of sedi-
ment contamination in surface
waters of the United States. As
previously stated, discharge
limits for point sources are not
necessarily protective of down-
stream sediment quality. The
Agency believes an effective
source control strategy should
focus on areas at greatest risk
on a watershed scale. The re-
port identifies 29 watersheds
among the 96 APCs where the
potential for point source con-
tribution to sediment contami-
nation is the greatest.

EPA's Point and Nonpoint Source
Sediment Contaminant Inventories

As part of the National Sediment Inventory (NSI)
and mandate under the Water Resources Development
Act (WRDA) of 1992, EPA is conducting inventories of
point and nonpoint sources of sediment contaminants.

The objective of the point source assessment com-
ponent of the NSI is to compile available data regard-
ing the purposeful discharge of sediment contaminants
from industrial facilities and municipal sewage treat-
ment plants and to determine the potential to adversely
affect sediment quality by chemical class, watershed,
and industrial category. EPA has produced the Na-
tional Sediment Contaminant Point Source Inventory

The objective of the non-
point source assessment com-
ponent of the NSI is to prepare
a nationwide assessment of an-
nual nonpoint source contribu-
tions of selected sediment
contaminants on a watershed basis. Given the num-
ber and diversity of nonpoint sources, the Agency is
focusing its initial efforts on four major categories:
harvested croplands, urban areas, atmospheric dep-
osition, and inactive and abandoned mine sites (where
information is available). Although these nonpoint
sources do not constitute the full range of sediment
contaminant sources, they are frequently cited in the
scientific literature as significant sources of mercury,
PCBs, PAHs, metals, pesticides, and other organic
compounds.

The nonpoint source assessment is intended to be a
screening-level study that begins to correlate contami-
nated sediment locations with suspected sources of these
contaminants. As part of this assessment, EPA is com-
piling data from the Bureau of the Census, the U.S.

4-14


-------
Department of Agriculture, the U.S. Department of the
Interior's U.S. Geological Survey and Bureau of Mines,
and others. EPA will compile information and data con-
cerning these nonpoint source activities to identify wa-
tersheds for further investigation and assessment.

Given the breadth of nonpoint sources, EPA antici-
pates that the process of conducting future assessments

will be iterative. Additional nonpoint sources will be
added to the inventory to discriminate more fully be-
tween contaminant types and known sources and to char-
acterize their proximity to known or suspected
contaminated sediment sites. This iterative process will
allow EPA to identify regions of the country where
nonpoint sources are known to exist, but data on sedi-
ment quality are either limited or lacking.

4-15


-------
4-16


-------
Chapter 5

Conclusions and Discussion

T[he National Sediment Inventory (NSI) is EPA's
largest compilation of sediment chemistry data
and related biological data. It includes approxi-
mately 2 million records for more than 21,000 monitoring
stations across the country. EPA's evaluation of the NSI
data was the most geographically extensive investiga-
tion of sediment contamination ever performed in the
United States. The evaluation was based on procedures
to address the probability of adverse effects to aquatic
life and human health.

The characteristics of the NSI data, as well as the
degree of certainty afforded by available assessment
tools, allow neither an absolute determination of adverse
effects on human health or the environment at any loca-
tion, nor a determination of the areal extent of contamina-
tion on a national scale. However, the evaluation results
strongly suggest that sediment contamination may be
significant enough to pose potential risks to aquatic life
and human health in some locations. The evaluation meth-
odology was designed for the purpose of a screening-
level assessment of sediment quality; further evaluation
would be required to confirm that sediment contamina-
tion poses actual risks to aquatic life or human health for
any given site or watershed.

Based on the number and percentage of sampling
stations containing contaminated sediment within water-
shed boundaries, EPA identified a number of watersheds
containing areas of probable concern for sediment con-
tamination (APCs) where additional studies may be
needed to draw conclusions regarding adverse effects
and the need for actions to reduce risks. Although the
APCs were selected by means of a screening exercise,
EPA believes that they represent the highest priority for
further ecotoxicological assessments, risk analysis, tem-
poral and spatial trend assessment, contaminant source
evaluation, and management action because of the pre-
ponderance of evidence in these areas. Although the
procedure for classifying APCs using multiple sampling
stations was intended to minimize the probability of mak-
ing an erroneous classification, further evaluation of con-
ditions in watersheds containing APCs is necessary
because the same mitigating factors that might reduce
the probability of associated adverse effects at one sam-

pling station may also affect neighboring sampling sta-
tions.

EPA chose the watershed as the unit of spatial analy-
sis because many state and federal water and sediment
quality management programs, as well as data acquisi-
tion efforts, are centered around this unit. This choice
reflects the growing recognition that activities taking place
in one part of a watershed can greatly affect other parts
of the watershed, and that management efficiencies are
achieved when viewing the watershed holistically. At
the same time, the Agency recognizes that contamina-
tion in some reaches in a watershed does not necessarily
indicate that the entire watershed is affected.

Watershed management is a vital component of com-
munity-based environmental protection. The Agency and
its state and federal partners can address sediment con-
tamination problems through watershed management ap-
proaches. Watershed management programs focus on
hydrologically defined drainage basins rather than areas
defined by political boundaries. These programs recog-
nize that conditions of land areas and activities within
the watershed affect the water resource. Local manage-
ment, stakeholder involvement, and holistic assessments
of water quality are characteristics of the watershed ap-
proach. The National Estuary Program is one example of
the watershed approach that has led to specific actions
to address contaminated sediment problems. Specifically,
the Narragansett (RI) Bay, Long Island Sound, New York/
New Jersey Harbor, and San Francisco Bay Estuary Pro-
grams have all recommended actions to reduce sources
of toxic contaminants to sediment. Numerous other ex-
amples of watershed management programs are summa-
rized in The Watershed Approach: 1993/94 Activity
Report (USEPA, 1994g) and A Phase I Inventory of Cur-
rent EPA Efforts to Protect Ecosystems (USEPA, 1995b).

This chapter presents some general conclusions
about the extent of sediment contamination in the United
States and sources of sediment contaminants. It. also
includes comparisons to other national studies that ad-
dress the extent of sediment contamination and to a na-
tional survey of state-issued fish consumption advisories.
In addition, this chapter presents the results of an analy-

5-1


-------
sis of the sensitivity of parameters used to evaluate po-
tential human health effects from exposure to PCBs and
mercury, which was performed to show how the use of
different screening values affect the results. The chap-
ter concludes with a discussion of the strengths and
limitations of the NSI data and evaluation method.

It'is important to understand both the strengths and
limitations of this analysis to appropriately interpret and
use the information contained in this report. The limita-
tions do not prevent intended uses, and future reports
to Congress on sediment quality will contain less uncer-
tainty. To ensure that future reports to Congress accu-
rately reflect current knowledge concerning the
conditions of the Nation's sediment as our knowledge
and application of science evolves, the NSI will develop
into a perodically updated, centralized assemblage of sedi-
ment quality measurements and assessment techniques.

Extent of Sediment Contamination

Based on the evaluation, sediment contamination
exists at levels where associated adverse effects are prob-
able CTier 1) in some locations within each region and
state of the country. The water bodies affected include
streams, lakes, harbors, nearshore areas, and oceans. A
number of specific areas in the United States had large
numbers of sampling stations where associated adverse
effects are probable. Puget Sound, Boston Harbor, the
Detroit River, San Diego Bay, and portions of the Ten-
nessee River were among those locations. Several U.S.
harbors (e.g., Boston Harbor, Puget Sound, Los Ange-
les, Chicago, Detroit) appear to have some of the most
severely contaminated sediments in the country. This
finding is not surprising since major U.S. harbors have
been affected throughout the years by large volumes of
boat traffic, contaminant loadings from upstream sources,
and many local point and nonpoint sources.

Thousands of other water bodies in hundreds of
watersheds throughout the country contain sampling
stations classified as Tier 1. Many of these sampling
stations may represent isolated "hot spots" rather than
widespread sediment contamination, although insuffi-
cient data were available in the NSI to make such a deter-
mination. EPA's River Reach File 1 (RF1) delineates the
Nation's rivers and waterways into segments, or reaches,
of approximately 1 to 10 miles in length. Based on RF1,
approximately 11 percent of all river reaches in the United
States contained NSI sampling stations. More than 5,000
sampling stations in approximately 2,400 river reaches
across the country (4 percent of all reaches) were classi-
fied as Tier 1. Another 10,000 sampling stations were

classified as Tier 2. In total, over 5,000 river reaches in
the United States—approximately 8 percent of all river
reaches—include at least one Tier 1 or Tier 2 station.

EPA cannot determine the areal extent or number of
river miles of contaminated sediment in the United States
because the NSI does not provide complete coverage for
the entire nation, sampling locations are largely based on
a nonrandom sampling design, and sediment quality can
vary greatly within very short distances.

Most of the NSI data were compiled from nonran-
dom monitoring programs. Such monitoring programs
focus sampling efforts on areas where contamination is
known or suspected to occur. As a result, assuming all
other factors are the same, the frequency of Tier 1 or Tier
2 classification based on the NSI data evaluation is prob-
ably greater than that which would result from purely
random sampling. Swartz et al. (1995) demonstrated the
effects of nonrandom sampling design on the frequency
of detecting contaminated sampling stations. They com-
pared the percent of sediment sampling stations that ex-
ceeded PAH screening effects levels (ERL, SQC, AET)
based on random sampling station selection (Virginian
Province EMAP stations) to the percent of sampling sta-
tions that exceeded those levels based on sampling sta-
tion selection on the basis of known PAH contamination
(such as creosote-contaminated Eagle Harbor, Washing-
ton). They found that the frequency of exceeding a sedi-
ment chemistry screening value in sampling stations
known to be contaminated was 5 to 10 times greater than
that for randomly selected sampling stations.

The percentage of all NSI sampling stations where
associated adverse effects are "probable" or "possible
but expected infrequently" (i.e., 26 percent in Tier 1 and
49 percent in Tier 2) does not represent the overall condi-
tion of sediment across the country: the overall extent of
contaminated sediment is much less, as is the percentage
of sampling stations where contamination is expected to
actually exert adverse effects. For example, a reasonable
estimate of the national extent of contamination leading
to adverse effects to aquatic life is between 6 and 12
percent of sediment underlying surface waters. This is
primarily because the majority of sampling stations in the
NSI are located in known or suspected areas of sediment
contamination (i.e., sampling stations were not randomly
selected). However, some individual data sets that are
included in the NSI, as well as the results of independent
investigations conducted by other researchers, can be
applied to represent the areal extent of sediment contami-
nation in their respective study areas. EPA's EMAP data
collection effort featured a probabilistic, or random, sam-
pling design. In the Virginian and Louisianian EMAP

5-2


-------
National Sediment Quality Survey

Provinces, located on the Mid-Atlantic and Gulf coasts
respectively, 104 of678 (15.3 percent) of sediment samples
were toxic to amphipods. With a 5 percent false positive
rate (statistical alpha=0.05), EMAP toxicity data suggest
that about 10 percent of marine and estuarine sites are
sufficiently contaminated to cause lethality to benthic
organisms (Richard Swartz, personal communication,
December 27,1996). In another recent study, Long et al.
(1996) examined amphipod survival in test sediment col-
lected from 1,176 locations in 22 estuarine areas through-
out the nation. These authors concluded that the areal
extent of toxic sediment comprised approximately 11 per-
cent of the combined study area.

To apply the NSI evaluation to estimate the areal
extent of toxic sediment in the United States, three fac-
tors must be accounted for: (1) most of the NSI data were
generated from sampling targeted toward areas of known
or suspected contamination, (2) sediment chemistry
screening values only identify sediment associated with
a probability of toxicity, and (3) toxicity is demonstrated
at some sampling stations where sediment chemistry
screening values are not exceeded. The latter condition
could be a result of false positives (i.e., laboratory toxic-
ity that would not be present in the field), toxic chemicals
present in the field but not measured or evaluated, or
toxicity that correlative screening values do not predict
(e.g., by definition 10 percent of toxic samples in the "ef-
fects distribution" lie blow the ERL).

Using information from available data and published
studies, the effects of each of the above factors can be
quantified. Swartz et al. (1995) suggest that exceeding a
sediment chemistry screening value at sites of known or
suspected contamination is 5 to 10 times more likely than
at sites where sediment is randomly sampled. However,
comparison of Tier 1 classification for Virginian and Loui-
sianan EMAP data to the entire NSI data base suggests
that the mix of sampling strategies in the NSI data base as
a whole results in screening value exceedance at 2 to 4
times as many sampling stations than purely random sam-
pling. Long et al., (in press), as well as a comparison of
matched sediment chemistry and toxicity data within the
NSI, suggest that approximately 40 percent of Tier 1 sam-
pling stations, and 20 percent of Tier 2 sampling stations,
would exhibit significant lethality to bottom dwelling
aquatic organism. Both data sets also suggest that sig-
nificant lethality occurs at approximately 10 percent of
Tier 3 stations, where no screening value is exceeded.
Alternatively, one could assume that significant labora-
tory toxicity at randomly sampled locations classified as
Tier 3 only represents "false positives", and therefore
that no toxicity occurs at Tier 3 sampling stations classi-
fied from random sampling.

In the NSI evaluation, 3,283 and 9,688 of the 17,884
sampling stations with sediment chemistry data available
were classified as Her 1 and Tier 2, respectively, for risk
to bottom dwelling aquatic organisms. Using a 40 per-
cent probability of lethality at Tier 1 and a 20 percent
probability of lethality at Tier 2, and further assuming 10
times less frequent Tier 1 and Tier 2 classification (upper
end of range from Swartz et al., 1995) in a random sample
and no lethality at Tier 3 sampling stations, the estimated
extent of sediment contamination in the United States
associated with lethality to bottom dwelling aquatic or-
ganisms is 2 percent. At the other extreme, assuming 2
times less frequent Tier 1 and Tier 2 classification (lower
end of range from EMAP/NSI comparisons) in a random
sample and a 10 percent probability of lethality at all re-
sulting Tier 3 sampling stations (11,399; including the
additional sampling stations previously classified as Tier
1 and Tier 2 before adjusting for random sampling), the
estimated extent of sediment contamination associated
with lethality to bottom dwelling aquatic organisms is 15
percent. Avoiding either extreme, assuming 2 to 5 times
less frequent Tier 1 and Tier 2 classification in a random
sample and a 10 percent probability of lethality for only
the original Tier 3 sampling stations (4,913; prior to ad-
justing for random sampling), the range narrows to 6 to
12 percent—about 1,000 to 2,000 toxic sampling stations
out of approximately 18,000. This range encompasses
the areal extent point estimates from EMAP toxicity data
and Long et al. (1996). EPA believes these are reasonable
estimates of the extent of sediment contamination across
the United .States.

The results of the NSI data evaluation must be inter-
preted in the context of data availability. Many states
and EPA Regions appear to have a much greater inci-
dence of sediment contamination than others. To some
degree, this appearance reflects the relative abundance
of readily available electronic data, not necessarily the
relative incidence of sediment contamination. For example,
182 of the 920 river reaches in Illinois contain a Tier 1
sampling station, whereas only 9 of the 5,490 reaches in
Montana contain a Tier 1 sampling station. However, the
NSI includes sampling station data for over 50 percent of
the river reaches in Illinois but less than 1 percent of the
river reaches in Montana. Therefore, although the abso-
lute number of Tier 1 and Tier 2 stations in each state is
important, relative comparisons of the incidence of sedi-
ment contamination between states is not possible be-
cause the extent of sampling and data availability vary
widely.

For a number of reasons, some potentially contami-
nated sediment sites were missed in this evaluation. The
most obvious reason is that the NSI does not include all

5-3


-------
Conclusions and Discussion

sediment quality data that have ever been collected. For
example, the NSI does not include many EPA Superfund
Program data and therefore sampling stations in the vi-
cinity of hazardous waste sites might not have been in-
cluded in the NSI evaluation. Additional data sets will be
added to the NSI for future evaluations to provide better
national coverage. In addition, some data in the NSI were
not evaluated because of questions concerning data qual-
ity or because no locational information (latitude and lon-
gitude) was available.

Sources of Sediment
Contamination

Some of the most significant sources of persistent
and toxic chemicals have been eliminated or reduced as
the result of environmental controls put into place during
the past 10 to 20 years. For example, the commercial use
of PCBs and the pesticides DDT and chlordane has been
restricted or banned in the United States. In addition,
effluent controls on industrial and municipal point source
discharges and best management practices for the con-
trol of nonpoint sources have greatly reduced contami-
nant loadings to many of our rivers and streams.

The results of better controls over releases of sedi-
ment contaminants are evident from studies such as that
conducted by Swartz et al. (1991) on the Palos Verdes
Shelf. These researchers examined sediment cores col-
lected at two sites on the Palos Verdes Shelf near the Los
Angeles County Sanitation District's municipal waste-
water outfalls, and at two reference sites in Santa Monica.
They found that the vertical distribution of sediment tox-
icity near the outfalls was significantly correlated with
profiles of total organic carbon and sediment chemical
contamination. Dating of core horizons showed that sedi-
ment toxicity also was significantly correlated with his-
torical records of the mass emission rate of suspended
solids from the outfalls. The vertical profiles showed
that the toxicity of surficial sediments increased after the
initiation of the discharge in the 1950s, remained rela-
tively high until the early 1970s, and then decreased after
the implementation of source controls and improved ef-
fluent treatment (Swartz et al., 1991).

Based on the NSI data evaluation, metals and persis-
tent organic chemicals are the contaminants most often
associated with sediment contamination. Despite recent
progress in controlling sediment contaminant releases to
the environment, active sources of these contaminants
still exist. These include nonpoint source loadings such
as surface water runoff and atmospheric deposition, point
source loadings, and resuspension of in:place sediment
contaminants from historical sources.

Some correlations between land use and sediment
contamination caused by specific classes of chemicals
were identified in Chapter 4. Agricultural land use was
correlated with the extent of sediment contaminated with
organochlorine pesticides in APC watersheds, especially
those with more than 75 percent of land area devoted to
crop production or rangeland. In contrast, the extent of
sediment contaminated with PAHs, mercury, and other
metals in APC watersheds correlated with the extent of
urban land use. Land use did not appear to be associated
with the extent of PCB contamination.

Comparison of NSI Evaluation
Results to Results of Previous
Sediment Contamination Studies

The results of this study are consistent with the find-
ings of other national assessments of sediment contami-
nation. For example, in EPA's 1992 National Water
Quality Inventory report, 27 states identified 770 known
contaminated sediment sites (USEPA, 1994e). The iden-
tified "sites" probably best correlate to river reaches from
this analysis in terms of areal extent. The NSI evaluation
identified approximately 2,400 river reaches in 50 states
that contain a Tier 1 sampling station. In the National
Water Quality Inventory report, the states frequently listed
metals (e.g., mercury, cadmium, and zinc), PCBs, DDT (and
its by-products), chlordane, and priority organic chemi-
cals as the cause of sediment contamination. They iden-
tified industrial and municipal discharges (past and
present), landfills, resource extraction, abandoned haz-
ardous waste disposal sites, and combined sewer over-
flows as the most important sources of sediment
contamination.

In a 1987 overview of sediment contamination (which
was based on a limited amount of national data), EPA
estimated that hundreds of sites located in all regions of
the United States have in-place sediment contaminants
at concentrations of concern (USEPA, 1987). The study
identified harbor areas, both freshwater and marine, as
some of the most severely impacted areas in the country.
The study identified municipal and industrial point source
discharges, urban and agricultural runoff, combined sewer
overflows, spills, mine drainage, and atmospheric depo-
sition as frequently cited sources of sediment contami-
nation.

In 1994, the National Oceanic and Atmospheric Ad-
ministration (NOAA) released its Inventory of Chemical
Concentrations in Coastal and Estuarine Sediments
(NOAA, 1994). This study categorized 2,800 coastal sites
as either "high" or "hot" based on the contaminant con-
centrations found at the sampling locations. NOAA did

5-4


-------
National Sediment Quality Survey

not use risk-based screening values for its analysis. Us-
ing the National Status and Trends Mussel Watch data
set, "high" values were defined as the mean concentra-
tion for a specific chemical plus one standard deviation.
High values corresponded to about the 85th percentile of
contaminant concentration. "Hot" concentrations were
defined as those exceeding five times the "high" values.
Most of the "hot" sites were in locations with high ship
traffic, industrial activity, and relatively poor flushing,
such as harbors, canals, and intracoastal waterways
(NOAA, 1994). Mercury and cadmium exceeded the
NOAA "hot" thresholds at a greater percentage of sites
where they were measured (about 7 percent each) than
other sediment contaminants.

Comparison of NSI Evaluation
Results to Fish
Consumption
Advisories

Table 5-1.

EPA recently published a Na-
tional Listing of Fish Consumption
Advisories issued by state gov-
ernments. As of 1994, 1,532 fish
consumption advisories were in
place in 46 states. (Each advisory
might apply to several water body
segments, or reaches, as defined
in this study.) Mercury was the
contaminant most often associ-
ated with fish consumption advi-
sories; 1,119 water bodies had
advisories that included mercury.
States also issued a large number
of advisories because of high lev-
els of chlordane, PCBs, and diox-
ins in fish tissue.

A direct comparison of the
fish advisory contaminants and
NSI contaminants is not possible
because states often issue advi-
sories for groups of chemicals.
Nevertheless, five of the top six
contaminants associated with fish
advisories (PCBs, DDT, dieldrin,
chlordane, and dioxins) are also
among the contaminants most of-
ten responsible for the Tier 1 clas-
sification of water bodies based on
potential human health effects
(Table 5-1). As illustrated in Fig-
ure 5-1, many sampling stations
categorized as Tier 1 or Tier 2 for

human health effects are located in water bodies for which
fish consumption advisories have been issued for the
chemical(s) responsible for the Tier 1 or Tier 2 categoriza-
tion. Tier 1 and Tier 2 stations are located predominantly
where data have been collected and compiled for the NSI,
whereas fish consumption advisories are located in states
with active fish advisory programs. Unlike the NSI data
evaluation, which is applied consistently to available data,
risk assessment methods used by states may vary.

Although there is good agreement for other chemi-
cals, mercury is notably absent from the Tier 1 category
in Table 5-1. Using the NSI evaluation methodology, mer-
cury cannot place a sampling stations in Tier 1 for poten-
tial human health effects. For chemicals other than PCBs
and dioxins, sediment chemistry and fish tissue data must
both indicate human health risk for Tier 1 assignment.

Comparison of Contaminants Most Often Associated With Fish
Consumption Advisories and Those Which Most Often Cause
Stations to Be Placed in Tier 1 or Tier 2 Based on the NSI Data
Evaluation





Number of River Reaches That Include





at Least One Tier 1 or Tier 2 Station





Based oil the NSI Data Evaluation of





Human Health Fish Consumption





Advisories Parameters"





# of Water Bodies with







Chemical*

Fish Advisories"

Tier 1

Tier 2'

Total

Mercury

1,119

0

89

89

PCBs

387

1,498

732

2,230

Chlordane

114

11

1,026

1,037

Dioxins

53

242

8

250

DDT and metabolites

28

19

656

675

Dieldrin

15

9

1,296

1,305

Selenium

12

0

4

4

Mirex

10

0

15

15

FAHs

5

0

529

529

Toxaphene

4

0

183

183

Hexachlorobenzene

3

0

53

53

Lead

2

0

259

259

Hexachlorobutadiene

2

0

6

6

Creosote*

2

-

-

.

Chromium

1

0

6

6

Copper

1

0

4

4

Zinc

1

0

14

14

'Other chemical groups responsible for fish consumption advisories (i.e., pesticides (24 water bodies], "multiple" [4
water bodies], "not specified" f4 water bodies), and meials [6 water bodies]) could not be directly compared to NSI
chemicals.

No reference values were available for creosote; therefore, it was noi evaluated in the NSI data evaluation,
cDocs not include statewide advisories

Mercury: New York, New Jersey, Maine, Massachusetts, Michigan, coastal Florida
Chlordane: Missouri
PCBs: New York
Dloxin: coastal Maine
dA water body can be composed of numerous river reaches.

*River reaches that include at least one Her 2 sampling station but no Tier i sampling stations.

5-5


-------
Figure 5-1. Tier 1 and Tier 2 Sampling Stations for Potential Risk to Human Health Located Within Water Bodies with Fish Consumption Advisories
in Place for the Same Chemical Responsible for the Tier 1 or Tier 2 Classification.


-------
National Sediment Quality Survey

Unfortunately, the bioaccumulation potential of mercury
based on concentrations in sediment cannot be assessed
because the biota sediment accumulation factors (BSAFs)
used for this study apply only to nonionic organic com-
pounds. In addition, available fish tissue data for mer-
cury did not place a large number of sampling stations in
Tier 2 for potential human health effects, compared to the
number of fish consumption advisories issued.

There are three possible explanations for the rela-
tively small number of sampling stations categorized as
Tier 2 for mercury in comparison to the number of fish
consumption advisories in place for mercury. The first
explanation is that the NSI evaluation was limited to data
from resident demersal species, whereas data used in sup-
port of issuing state fish advisories probably included
pelagic and migratory species. The second possible ex-
planation is that the evaluation parameters used in the
analysis were not as stringent as the ones used to sup-
port fish consumption advisory issuance. The third ex-
planation is that the NSI does not include all of the data
used by the states to issue fish advisories.

To examine these possible explanations, EPA per-
formed additional analyses of mercury fish tissue data
included in the NSI. The current evaluation, using a fish
tissue screening value of 1 part per million (ppm), yields
103 Tier 2 sampling stations (4 percent of all stations with
detectable levels). If data from all edible pelagic and mi-
gratory species are included in the analysis, there are 374
Tier 2 sampling stations (9 percent of all stations with
detectable levels). A fish tissue threshold of 0.6 ppm,
derived using the more stringent reference dose (0.00006
mg/kg-day) recommended to states for issuing fishing
advisories to protect against developmental effects
among infants (USEPA, 1994f), yields 821 Tier 2 sampling
stations (20 percent of all stations with detectable levels)
when applied to all edible species using the consumption
rate for an average consumer of 6.5 grams per day. How-
ever, fish consumption advisories are often issued for
more highly exposed populations, such as recreational or
subsistence fishers. The 0.2 ppm Canadian guideline limit
for mercury in fish that are part of a subsistence diet
yields 2,308 Tier 2 sampling stations (56 percent of all
stations with detectable levels) when applied to all edible
species in the NSI database. Further details of the addi-
tional mercury analyses are provided in Appendix H.

The conclusion resulting from these additional analy-
ses is that all three explanations for the discrepancy in
numbers of fish advisories and Tier 1 and Tier 2 sampling
stations for mercury probably have an effect. Most fish
consumption advisories are issued to protect infants from
developmental effects for populations where exposure is

greater than 6.5 grams of fish per day. It is also likely that
many of the data used to develop state fish consumption
advisories are not included in the NSI, or are not evalu-
ated for sediment contamination because they are mea-
surements in pelagic or migratory fish.

Sensitivity of Selected PCB
Evaluation Parameters

Because PCBs and dioxin are extremely hydrophobic
chemicals commonly associated with sediment, and be-
cause of their toxicity to humans, EPA believes that el-
evated levels of PCBs and dioxins in fish tissue of
resident, demersal species are sufficient evidence to indi-
cate a higher probability of adverse human health effects
and to place a sampling station in Tier 1. Based on the
NSI data evaluation, PCBs were responsible for the Tier 1
classification of more sampling stations than any other
chemical. Therefore, EPA conducted a sensitivity analy-
sis of some PCB evaluation parameters to determine the
effect on the number of sampling stations classified as
Tier 1 or Tier 2.

In the NSI evaluation, EPA selected a precautionary
approach for the analysis of PCBs. The approach is pre-
cautionary because it does not require matching sedi-
ment chemistry and tissue residue data for PCB, and it is
based on the risk of cancer for all PCBs congeners or
total PCB measurements. However, some PCB congeners
are considered a greater threat for noncancer effects than
for cancer. The evaluation currently places 2,256 tissue
sampling stations in Tier 1 based on human health cancer
risk. Only 542 of these sampling stations included match-
ing sediment and tissue data for PCBs. Therefore, the
number of sampling stations classified as Tier 1 would
have decreased significantly if this match had been re-
quired.

EPA performed additional evaluations to determine
the number of sampling stations that exceed other screen-
ing values which are less precautionary than those se-
lected for the PCB evaluation in this study. The complete
results are presented in Appendix H, which includes a
comparison of the number of sediment and fish tissue
sampling stations with detectable levels of PCBs that ex-
ceed various evaluation parameters for both aquatic life
and human health.

Sampling station evaluation based on PCB contami-
nation is quite sensitive to the selection of evaluation
parameters. For protection of fish consumers, there are
essentially three distinct levels of protection. Using an
EPA cancer risk of 10"s (i.e., a 1 in 100,000 extra chance of
cancer over a lifetime of 70 years) or greater, 85 percent or

5-7


-------
<'(inclusions stud Discussion

more of the sampling stations with detectable PCB levels
are classified as Tier 1. About one-half to two-thirds of
the sampling stations are classified as Tier 1 for
exceedances of PCB levels protective of noncancer
health effects, cancer risk at a 10"4 risk level, or levels
exceeding the wildlife criterion. Less than one-third of
the stations are classified as Tier 1 using the FDA level
of protection. As documented in Appendix H, these per-
centages vary depending on use of a BSAF safety fac-
tor, and whether one is examining the set of fish tissue
data or sediment chemistry data. These three levels of
protection vary within two orders of magnitude, a range
that covers most of the distribution of PCB measure-
ments.

Although sampling station classification for PCB
contamination is quite sensitive to selection of evalua-
tion parameters, overall station classification using the
complete NSI evaluation for all chemicals is more robust
Using the selected PCB evaluation parameters, there are
15,922 total Tier 1 and Tier 2 sampling stations. If PCBs
are dropped from the analysis entirely, the total number
of Tier 1 and Tier 2 sampling stations remains about the
same (less than a 5 percent decrease), but the number of
Tier 1 sampling stations decreases by approximately 40
percent. If PCBs are evaluated using a noncancer hu-
man health threshold, the total number of Tier 1 and Tier
2 sampling stations decreases by less than 2 percent and
the number of Tier 1 sampling stations decreases by ap-
proximately 12 percent Figure 5-2 shows the location of
Tier 1 and Tier 2 sampling stations that exhibit potential
human health risks for all chemicals other than PCBs for
comparison to Figure 3-6 in the results section. Approxi-
mately 78 percent (6,670 of 8,523) of the total number of
Tier 1 and Tier 2 sampling stations indicating human health
riskremainafterexcludingPCBs from the evaluation.

Strengths of the NSI Data
Evaluation

For this report to Congress, EPA has compiled the
most extensive data base of sediment quality informa-
tion currently available in electronic format. To evaluate
these data, EPA has applied sediment assessment tech-
niques in a weight-of-evidence approach recommended
by national experts. The process to produce this report
to Congress has engaged a broad array of government,
industry, academic, and professional experts and stake-
holders in development and review stages. The evalua-
tion approach utilizes sediment chemistry, tissue residue,
and toxicity test results. The assessment tools employed
in this analysis have been applied in North America with
results published in peer reviewed literature. Toxicity
test data were generated using established standard

methods employed by multiple Federal agencies. The
evaluation approach addresses potential impacts to both
aquatic life and human health.

Because of the complex nature of the reactions among
different chemicals in different sediment types, in water,
and in tissues, no single sediment assessment technique
can be used to adequately evaluate potential adverse ef-
fects from exposure to all contaminants. Uncertainties
and limitations are associated with all sediment quality
evaluation techniques. To compensate for those limita-
tions, EPA has used multiple assessment techniques, alone
and in combination, to evaluate the NSI data. For example,
EPA developed draft SQCs based on the best scientific
data available and extensive peer review. Therefore, EPA
believes that the draft SQCs are reliable benchmarks for
protecting sediment quality, and with measured TOC can
indicate a higher probability for adverse effects to aquatic
life. In addition, EPA believes that other sediment chemis-
try screening values (ERMs/ERLs, PELs/TELs, AETs, and
SQALs) are also useful indicators of probability for aquatic
life impacts. The Agency applied a weight-of-evidence
approach for evaluating contaminant levels using these
screening values, requiring the exceedance of multiple
upper sediment chemistry screening values (i.e., ERM,
PEL, AET-high, or SQAL) for classification ofTier 1 sam-
pling stations.

The screening values used to evaluate the NSI data
include both theoretical and correlative approaches. The
theoretical approaches (e.g., draft SQCs, SQALs, and
TBPs) are based on the best information available con-
cerning how chemicals react in sediments and organisms
and how organisms react to those chemicals. The correla-
tive approaches (i.e., ERMs/ERLs, PELs/TELs, and AETs)
are based on matched sediment and biological data gath-
ered in the field and in the laboratory, and they provide
substantial evidence of actual biological effects from sedi-
ments contaminated with specific concentrations of the
chemicals.

The NSI evaluation approach includes assessments
of potential impacts to both human health and aquatic
life. Some chemicals pose a greater risk to human health
than to aquatic life; for others, the reverse is true. By
evaluating both potential human health and aquatic life
impacts, EPA has ensured that the most sensitive end-
point is used to assess environmental impacts.

Because sediment chemistry data are not the only
indicators of potential environmental degradation due to
sediment contamination, the NSI data evaluation approach
also includes evaluations of fish tissue residue and toxic-
ity data. If high levels ofPCBs or dioxins (which are highly

5-8


-------

-------
hydrophobic organic chemicals commonly found associ-
ated with sediments) were measured in fish tissue at a
given sampling station, the station could be categorized
as Tier 1 with no corroborating sediment chemistry data.
For other chemicals, high concentrations in tissues alone
were not sufficient to categorize a sampling station as
Tier 1; corroborating sediment chemistry data were also
required. For a sampling stations to be categorized as
Tier 1 based on toxicity data alone, multiple toxicity tests
with positive results using two different test species were
required. One of the tests had to be a solid-phase test.

Although EPA has developed draft SQCs for only
five nonionic organic chemicals, the Agency has devel-
oped similar values, the SQALs, for an additional 35 chemi-
cals as part of the NSI data evaluation. The SQALs have
allowed EPA to evaluate more chemicals using multiple
assessment techniques, thereby adding more weight of
evidence to the results of this evaluation.

limitations of the NSI Data
Evaluation

This methodology was designed for the purpose of
a screening-level assessment of sediment quality, A con-
siderable amount of uncertainty is associated with the
site-specific measures, assessment techniques, exposure
scenarios, and default parameter selections. Therefore,
the results of evaluating particular sampling stations
based on this methodology should be followed up with
more intensive assessment efforts, when appropriate (e.g.,
for water bodies with multiple Tier 1 sampling stations
located in APCs). Two types of limitations are associ-
ated with the evaluation of the NSI data: limitations asso-
ciated with the data themselves and limitations associated
with the evaluation of the data.

Limitations of Data

The NSI is a multimedia compilation of environmen-
tal monitoring data obtained from a variety of sources,
including state and federal government offices. Inherent
in the diversity of data sources are contrasting monitor-
ing objectives and scopes, which make comparison of
data from different data sets difficult. For example, sev-
eral of the databases contain only information from ma-
rine environments or other geographically focused areas.
The potential for inconsistencies in measured concentra-
tions of contaminants at different stations exists for
samples taken from different monitoring programs. For
example, sampling different age profiles in sediments,
applying different sampling and analysis methods, and
sampling for different objectives can affect the results of

the NSI evaluation. Although numerous data sets identi-
fied sampling and laboratory methods, most data did not
have this information. In addition, some data sets included
in the NSI were not peer-reviewed (i.e.. Region 4's Sedi-
ment Quality Inventory, the Gulf of Mexico Program's
Contaminated Sediment Inventory, and some data sets
from EPA's STORET). Furthermore, each monitoring pro-
gram used unique sampling and analysis protocols. For
example, PCBs, the chemical group most often respon-
sible for placing sites in Tier 1, were measured by nearly
all of the programs but were analyzed and reported as
aroclor-specific data, congener-specific data, total PCBs,
or a combination of these.

The only quality assurance/quality control (QA/QC)
information required for data to be included in the NSI
was information on the sour® of the data and the loca-
tion of the sampling station. Available information on
several types of QA/QC procedures that can influence
the quality of the data and can be used to check the
quality of data was included In the NSI. None of this
information, however, was required before a data set could
be included in the NSI. Evaluation of such information
can provide an indication of the quality of the data used
to target a specific site. Table 5-2 presents a summary of
the known QA/QC information associated with each of
the data sets included in the NSI.

Data reporting was also inconsistent among the dif-
ferent data sources. Inconsistencies that required reso-
lution included the lack or inconsistent use of Chemical
Abstract Service (CAS) numbers, analyte names, species
names, and other coding conventions, as well as the lack
of detection limits and associated data qualifiers (remark
codes). The evaluation of toxicity data required the pres-
ence of control data. Control data were not often initially
reported with the data, and significant follow-up work
was required to acquire such data. In addition, 4 of the 11
sources of toxicity test data used in the NSI evaluation
did not report the use of laboratory replicates.

Some of the data included in the NSI were compiled
as early as 1980 (the data cover the period of 1980-93) and
might not reflect current conditions. The analysis did
not include a temporal assessment of trends in sediment
contaminant levels. Emissions of many prominent
contaminants declined during the 1980s, and significant
remediation efforts have taken place at many locations
since that time. In addition, dredging, burial, and scour-
ing might have removed contaminants from some sam-
pling stations. The lack of a trend analysis in sediment
contamination over time is an important limitation of this
study and will be investigated in future NSI evaluations.

5-10


-------
Table 5-2. National Sediment Inventory Database: Summary of QA/QC Information

Database

Are There
QA/QC Reports
lo Accompany
the Data?

Were the Data
Peer-Reviewed?

Are the Sampling

and Analytical
Methods Identified
In the Database?

Are the Detection

Limits for the
Analytes Included
in the Database?

Comments

ODES

Yes

Yes, 301(h) data

Yes

Yes

Data Qualifiers

EMAP (VA and LA Provinces)

Yes

Yes

Yes

Yes

Data Qualifiers

Seattle; U.S. Army Corps of
Engineers

Yes

Yes

Yes

Yes

Data Qualifiers

Region 4

Some

No

Some

Yes

Data Qualifiers

Gulf of Mexico

Some

No

Some

Yes

Data Qualifiers

COSED

Yes

Yes

Yes

Some



Great Lakes

Yes

Yes

Yes

Yes



DMATS

Some

Yes

Yes

Yes

Data Qualifiers

STORET

Unknown

Unknown

No

Yes

Data Qualifiers

Massachusetts Bay (USGS)

Some

Yes

Yes

Yes



Some data parameters are consistently absent
throughout the NSI database. (Refer to Appendix A, Tables
A-1 and A-2, for information on the number of NSI sta-
tions at which the various types of data were collected.)
For example, very few site-specific TOC or AVS data are
available, and toxicity data or matched sediment chemis-
try and biological data were available at relatively few
sampling stations. For many of the fish tissue data in-
cluded in the NSI, the species was not identified.

The lack of AVS data in the NSI was a significant
limitation for the evaluation of metals data. The NSI in-
cludes a relatively large amount of metals data, and the
data indicate that metals concentrations in sediment are
elevated in many areas. At some stations the elevated
metals concentrations might indicate a potential prob-
lem; however, no sampling stations in the NSI could be
placed in Tier 1 solely from measured concentrations of
cadmium, copper, nickel, lead, or zinc. This reflects in
large part the absence of AVS data, which are required to
place sampling stations contaminated with those metals
in Tier I.

The unavailability of matching sediment chemistry
and tissue residue data also limited the NSI data evalua-
tion. In several instances, fish tissue was not analyzed
for the same suite of chemicals for which sediment was
analyzed. Spatial and temporal limitations of the data might
have directly affected the analysis. Although some sedi-
ment chemistry and tissue residue data might have been

collected in the same or very similar sampling stations, if
the station names were not identical, the data could not
be treated as if they were collected from the same loca-
tion. This very likely resulted in an underestimate of the
number of Tier 1 stations identified based on potential
human health effects. The underestimate occurred be-
cause exceedances of sediment TBP and tissue levels
(EPA risk levels and FDA levels) at the same sampling
station were required to categorize stations as Tier 1.

The lack of consistency among the different moni-
toring programs in the suite of chemicals analyzed also
represents an area of uncertainty in the NSI data evalua-
tion. Certain databases contain primarily information de-
scribing concentrations of metals or pesticides, whereas
others (e.g., STORET and ODES) contain data describing
concentrations of nearly every chemical monitored in all
of the NSI data. Many monitoring programs use a screen-
ing list of chemicals that are indicator pollutants for
contaminated sediments. Thus, many of the specific
chemicals assessed in the NSI data evaluation are not
always measured in samples. In addition, certain classes
of in-place sediment contaminants might not be
recognized as causing significant impacts and thus are
not routinely measured.

Information describing local background levels of
sediment contaminants was usually not presented with
the data included in the NSI and thus was not considered
when the significance of elevated contaminant concen-

5-11


-------
Conclusions sind Discussion

trations in sediment was evaluated. Background condi-
tions can be important in an evaluation of potential ad-
verse effects on aquatic life because ecosystems can
adapt to their ambient environmental conditions. For ex-
ample, high metals concentrations in samples collected
from a particular station might occur from natural geo-
logical conditions at that location, as opposed to the ef-
fects of human activities.

Most data are associated with a specific location. As
a result, establishing the extent of contaminated sedi-
ment within a water body is not possible because it is
difficult to assess the extent to which a monitoring sta-
tion represents a larger segment of a water body. Fur-
thermore, the NSI data are geographically biased. More
than 50 percent of all sampling stations evaluated in the
NSI are located in 8 states (Washington, Florida, Illinois,
California, Virginia, Ohio, Massachusetts, and Wiscon-
sin), which have more than 700 monitoring stations each.
Finally, EPA did not verify reported latitude and longi-
tude coordinates for each sampling station.

Limitations of Approach

Sediment Chemistry Screening Values

There are significant gaps in our knowledge con-
cerning sediment-pollutant chemistry (especially bioavail-
ability) and direct and indirect effects on aquatic biota.
The certainty with which sediment toxicity can be pre-
dicted for each chemical using the various screening val-
ues included in the NSI evaluation can vary significantly
based on the quality of the available data and the appro-
priateness of exposure assumptions. For example, draft
SQCs and SQALs are not equivalent, even though they
were developed using the same methodology. EPA has
proposed SQCs for five chemicals based on the highest
quality toxicity and octanol/water partitioning data, which
have been reviewed extensively. The draft SQCs have
also undergone extensive field validation experiments.
However, SQALs for additional chemicals are in many
cases based on a less extensive toxicity data set and have
not been field validated. The AET values used in this
evaluation were based on empirical data from Puget Sound.
Direct application of values from Puget Sound to a spe-
cific location or region in another part of the country
might be overprotective or underprotective of the re-
sources in that area. Extensive collection of data and ad-
ditional analyses would be required to develop AETs for
other locations.

The bioavailability of metals in sediment is addressed
by the comparison of the molar concentration of sulfide
anions (i.e., acid-volatile sulfide [AYS]) to the molar con-

centration of metals (i.e., simultaneously extracted metals
[SEM]). The [SEM]-[AVS] difference is most applicable
as an indicator of when metals are not bioavailable. If
[AVS] exceeds [SEM], there is sufficient binding capacity
in the sediment to preclude metal bioavailability. How-
ever, if [SEM] exceeds [AVS], metals might be bioavail-
able or other nonmeasured phases might bind up the
excess metals. To apply the [SEM]-[AVS] difference to
indicate positive bioavailability and toxicity for this evalu-
ation, EPA used laboratory data that indicated the prob-
ability of observed toxic effects at various [SEM]-[AVS]
levels. Based on these data, EPA defined the Tier 1 level
as [SEM]-[AVS]>5, Thus, this use of [SEM]-[AVS] repre-
sents a hybrid of a theoretical approach and a correlative
approach.

Only those chemicals for which sediment chemistry
screening values (i.e., draft SQCs, SQALs, ERLs/ERMs,
PELs/TELs, and AETs) are available were evaluated in
the analysis of NSI data. Therefore, the methodology
could not identify contamination associated with chemi-
cal classes such as ionic organic compounds (e.g„ alkyl
phenols) and organometallic complexes (e.g., tributyl tin).

Biological effects correlation approaches such as
ERMs or PELs are based on the evaluation of paired field
and laboratory data to relate incidence of adverse bio-
logical effects to the dry-weight sediment concentration
of a specific chemical at a particular sampling station.
Researchers use these data sets to identify level-of-con-
cem chemical concentrations based on the probability of
observing adverse effects. Exceedance of the identified
level-of-concern concentration is associated with a likeli-
hood of adverse organism response, but it does not dem-
onstrate that a particular chemical is solely responsible.
In fact, a given sample typically contains a mixture of
chemicals that contribute to observed adverse effects to
some degree. Therefore, these correlative approaches
tend to result in screening values that are lower than the
theoretical draft SQCs and SQALs, which address the
effects of a single contaminant. However, these correla-
tive approaches are better at predicting toxicity in com-
plex mixtures of contaminants in sediment. The effects
range approaches to assessing sediment quality also do
not account for such factors as organic matter content
and AVS, which can mitigate the bioavailability and, there-
fore, the toxicity of contaminants in sediment.

Another concern is the application of screening val-
ues based on freshwater data (draft SQCs and SQALs)
and those based on saltwater data alone (ERLs/ERMs,
PELs/TELs, and AETs) to evaluate sediment contaminant
concentrations in the NSI from both freshwater and salt-
water habitats. Freshwater organisms exhibit tolerance to

5-12


-------
N:i(iiin;il Si'dimcnl 0";>'¦'> Survey

toxic chemicals similar to that of saltwater species when
tested in their respective water; however, estuarine or-
ganisms might be less tolerant if osmotically stressed
(Rand and Petrocelli, 1985). Thus, the relative toxicity of a
chemical in water (i.e., its chronic threshold water con-
centration) is usually within an order of magnitude for
saltwater and freshwater species, although final chronic
values and proposed sediment quality criteria values are
usually slightly higher for saltwater species. Ingersoll et
al., (1996) reported similar reliability and predictive ability
between marine and freshwater guidelines. In addition
Long et al., (1995) compared the ERLs and ERMs with
comparable values derived for freshwater by the Ontario
Ministry of the Environment and the agreement was ex-
tremely good. Because of limitations of time and re-
sources, sampling stations in the NSI were not classified
by salinity regime, and further site-specific evaluations
are required to more definitively assess the toxicity at the
stations. However, the application of several different
screening values should provide a reasonable estimate
of probability of risk to aquatic life in freshwater, estua-
rine, and marine habitats.

Additional false positive and false negative classifi-
cations of risk to aquatic life from sediment contaminant
concentrations could occur when a default value for or-
ganic carbon content is applied. Draft SQCs and SQALs
are based on the partitioning of a chemical between or-
ganic carbon in the sediment and pore water at equilib-
rium. Because the organic carbon content of most
sediment samples in the NSI is unknown, these sediment
samples were assumed to contain 1 percent organic car-
bon. Total organic carbon (TOC) can range from 0.1 per-
cent in sandy sediments to 1 to 4 percent in silty harbor
sediments and 10 to 20 percent in navigation channel
sediments (Clarke and McFarland, 1991). Long et al. (1995)
reported an overall mean TOC concentration of 1.2 per-
cent from data compiled from 350 publications for their
biological effects database for sediments. Ingersoll et al.
(1996) reported a mean TOC concentration of 2.7 percent
with a 95 percent confidence interval of only 0.65 per-
cent. In contrast, the concentration ranges of contami-
nants normalized to dry weight typically varied by several
orders of magnitude. Therefore, normalizing dry-weight
concentrations to a relatively narrow range of TOC con-
centrations had little influence on relative concentrations
of contaminants among samples. Similar findings were
reported by Barrick et al., (1988) for AETs and Long et al.
(1995) for ERMs calculated using sediment concentra-
tions normalized to TOC concentrations.

Uncertainty associated with the equilibrium partition-
ing theory for developing draft SQCs and SQALs includes
the degree to which the equilibrium partitioning model

explains the available sediment toxicity data (USEPA,
1993d). An analysis of variance using freshwater and salt-
water organisms in water-only and sediment toxicity tests
(using different sediments) was conducted to support
development of the proposed sediment criteria. This
analysis indicated that varying the exposure medium (Le.,
water or sediment) resulted in an estimate of variability
that should be used for computing confidence limits for
the draft SQCs. The methodology used to derive the
octanol/water partitioning coefficient and the final chronic
value can also influence the degree of uncertainty asso-
ciated with the draft SQCs. Differences in the response
of water column and benthic organisms, and limitations
in understanding the relationship of individual and popu-
lation effects to community-level effects, have also been
noted (Mancini and Plummer, 1994). Site-specific modifi-
cations to screening values derived using the equilib-
rium partitioning model have been recommended to better
address chemical bioavailability and species sensitivi-
ties (USEPA, 1993b). Sediment chemistry screening val-
ues developed using the equilibrium partitioning
approach also do not address possible synergistic, an-
tagonistic, or additive effects of contaminants.

Based on the theoretical calculations used to com-
pute SQAL values, it is possible that SQALs might be
orders of magnitude larger or smaller than other screen-
ing values used for the analysis (ERLs/ERMs, PELs/TELs,
and AETs). This might be a result of the limited aquatic
toxicity data used to develop SQAL values for some of
the contaminants for which water quality criteria are un-
available. EPA did not develop SQALs for this analysis
in those cases where toxicity data were considered inad-
equate. The approach used to develop SQALs, and to
choose chemicals for which SQALs could not be devel-
oped, is presented in Appendix B.

Fish Tissue Screening Values

The approach used to assess sediment chemistry
data for the potential to accumulate in fish tissue also
represents a theoretical approach with field-measured
components. In addition to applying a site-specific or
default organic carbon content, the TBP calculation in-
cludes a field-measured biota sediment accumulation fac-
tor (BSAF) to account for the relative affinity of a chemical
for fish tissue lipids or sediment organic carbon. The
BSAF will account for the effects of metabolism and
biomagnification in the organism in which it is measured.
The primary limitation of this approach is the applicabil-
ity of a field-measured BSAF, or a percentile from a distri-
bution of values, at a variety of sites where the conditions
may vary.

5-13


-------
Conclusions and Discussion

TBPs were assumed to be equivalent to levels de-
tectable in fish tissue. However, this approach might not
completely account for biomagnification in the food chain,
especially when using a BSAF derived from a benthic
organism. In addition, it is assumed that sediment does
not move, that contaminant sources other than sediment
are negligible, that fish migration does not occur, and
that exposure is consistent. The TBP calculation assumes
that various lipids in different organisms and organic car-
bon in different sediments are similar and have distribu-
tional properties similar to the field-measured values used
to derive BSAFs. Other simplifying assumptions are that
chemicals are similarly exchanged between the sediments
and tissues and that compounds behave alike, indepen-
dent of site conditions other than organic carbon con-
tent, In reality, physical-chemical processes (e.g.,
diffusion through porous media and sediment mixing) can
vary and limit the rate at which chemicals can exchange
with bottom sediments. Uptake of contaminants by
aquatic organisms is also a kinetic (rate-controlled) pro-
cess that can vary and be slowed, for example, by awk-
ward passage of a bulky molecule across biological
membranes. Also, a BSAF of 1 (thermodynamic equilib-
rium) was used to estimate TBPs for many nonpolar or-
ganics. This BSAF might overestimate or underestimate
the bioaccumulative potential for certain nonpolar organic
chemicals because it is assumed that there is no meta-
bolic degradation or biotransformation of such chemi-
cals. Site-specific organic carbon content was often not
available, which leads to additional uncertainty concern-
ing the comparability of BSAFs among different loca-
tions. In addition, development of the BSAFs used in the
TBP evaluation relied on a large amount of data that have
not been published or peer-reviewed. Because of these
factors, actual residue levels in fish resulting from direct
and/or indirect exposure to contaminated sediment might
be higher or lower. There is therefore uncertainty regard-
ing sampling stations classifications based on compari-

son of estimated TBPs with FDA tolerance/action and
guideline levels and EPA risk levels.

TBPs could not be calculated for polar organic com-
pounds or heavy metals. Therefore, sampling stations
could not be classified using FDA levels or EPA risk lev-
els for those chemicals using a TBP approach (although
fish tissue monitoring data are often available for many
stations).

Uncertainties and numerous assumptions are asso-
ciated with exposure parameters and toxicity data used to
derive EPA risk levels and FDA tolerance/action and guide-
line levels. For example, the derivation of EPA risk levels
is based on the assumption that an individual consumes
on average 6.5 g/day of fish caught from the same site
over a 70-year period. Also, the TBP calculation for hu-
man health assessments assumes fish tissue contains 3
percent lipid. This value is intended to be indicative of
the fillet rather than the whole body. Generally, the expo-
sure assumptions and safety factors incorporated into
toxicity assessments might overestimate risks to the gen-
eral population associated with sediment contamination,
but might underestimate risks to populations of subsis-
tence or recreational fishers.

Other Limitations

Because a numerical score was not assigned to each
sampling station to indicate the level of contamination
associated with that station, it is not possible to deter-
mine which of the stations in Tier I should be considered
the "most" contaminated. Such a numerical ranking sys-
tem was intentionally not used for the NSI data evalua-
tion because EPA does not believe that such ranking is
appropriate for a screening-level analysis such as this,
given the level of uncertainty.

5-14


-------
National Sediment Quality Survey

Chapter 6

Recommendations

The following discussion presents EPA's recom-
mendations for addressing sediment con-
tamination throughout the country and for im-
proving the ability to conduct sediment quality assess-
ments. These recommendations relate to five activities
or information needs:

1.	Further investigate conditions in the 96 targeted
watersheds.

2.	Coordinate efforts to address sediment quality
through watershed management programs.

3.	Incorporate a weight-of-evidence approach and
measures of chemical bioavailability into sedi-
ment monitoring programs.

4.	Evaluate the National Sediment Inventory's
(NSI's) coverage and capabilities and provide
better access to information in the NSI.

5.	Develop better monitoring and assessment
tools.

Recommendation 1: Further
Investigate Conditions in the 96
Targeted Watersheds

To characterize the incidence and severity of sedi-
ment contamination in the United States, EPA has per-
formed a screening-level analysis of the information in
the NSI, the results of which are presented in Chapter 3.
As mentioned previously, the results of the NSI data
evaluation alone should not be used as justification for
taking corrective actions at potentially contaminated sites.
The initial evaluation of NSI data was performed as a
means of screening and targeting. Additional, site-spe-
cific data and information should be gathered to verify
the NSI evaluation results and to support a comprehen-
sive assessment of the incidence and severity of sedi-
ment contamination problems.

The primary recommendation resulting from the NSI
data analysis is to encourage further investigation and

assessment of contaminated sediment. States, in coop-
eration with EPA and other federal agencies, should pro-
ceed with further evaluations of the 96 watersheds
containing areas of probable concern for sediment con-
tamination (APCs). In many cases, it is likely that much
additional investigation and assessment has already oc-
curred, especially in well known areas at risk for contami-
nation, and some areas have been remediated. If active
watershed management programs are in place, these
evaluations should be coordinated within the context of
current or planned actions. Future monitoring and as-
sessment efforts should focus on areas such as the 57
water body segments (or river reaches) located within
the 96 watersheds containing APCs that had 10 or more
stations categorized as Tier 1. The purpose of these ef-
forts should be, as appropriate, to gather additional sedi-
ment chemistry data and related biological data and
conduct further assessments of the data to determine
human health and ecological risk, determine temporal and
spatial trends, identify potential sources of sediment con-
tamination and determine whether potential sources are
adequately controlled, and determine whether natural re-
covery is a feasible option for risk reduction. Additional
monitoring and analysis of data from the 96 watersheds
containing APCs will also be used to track and document
the effectiveness of management actions taken to ad-
dress sediment contamination problems over time. Trends
in sediment contamination in the 96 APCs over time will
be reported in future reports to Congress.

Available options for reducing health and environ-
mental risks from contaminated sediment include physi-
cal removal and land disposal; subaqueous capping; in
situ or ex situ biological, physical/chemical, or thermal
treatment to destroy or remove contaminants; and natu-
ral recovery through continuing deposition of clean sedi-
ment. Assuming further investigation reveals the need
for management attention to reduce risks, the preferred
means depends on factors such as the degree and extent
of contamination, the value of the resource, the cost of
available options, likely human and ecological exposure,
and the acceptable time period for recovery. If risk man-
agers anticipate a lengthy period of time prior to recovery
of the system, state and local authorities can consider

6-1


-------
Recommendations

consider options such as placing a fish consumption
advisory on water bodies or portions of water bodies
where a significant human health risk exists.

Many state and federal government monitoring pro-
grams already do a good job of gathering data at loca-
tions with known contamination problems (including

some of the 96 APCs), and additional monitoring at those
locations will probably not be necessary. However, for
other locations not previously targeted for focused moni-
toring, additional data might be required to adequately

assess potential sediment contamination problems, es-
pecially in areas where significant human health expo-
sures occur. In addition, in some cases it might be

necessary to conduct baseline studies to determine where
to focus monitoring activities.

Further investigation might reveal that risks are mini-
mal or that natural recovery has diminished risk or will
diminish risk in an acceptable time period, or it might
verify that current contamination is significant and un-
likely to sufficiently improve under existing conditions.
Following verification of sediment contamination prob-
lems based on these additional assessments, appropriate
actions (e.g., remediation, permit review, TMDL assess-
ment, best management practices for nonpoint sources,
or "no action") should be taken to address the problem.
In many cases, the mechanisms for corrective actions are
already in place (e.g., permit review, TMDL assessments)
and responsible parties have already been identified. In
other cases, the states should work with EPA to deter-
mine the best course of action.

Recommendation 2: Coordinate
Efforts to Address Sediment
Quality Through Watershed
Management Programs

The watershed approach is a community-based water
resource management framework that requires a high
level of interprogram coordination to consider all factors
contributing to water and sediment quality problems and
to develop integrated, science-based, cost-effective so-
lutions that involve all stakeholders. It is within the
watershed framework, therefore, that EPA recommends
that federal, state, and local government agencies pool
their resources and coordinate their efforts to address
their common sediment contamination issues. These
activities should support efforts such as selection of fu-
ture monitoring sites, setting of priorities for reissuance
of NPDES permits, permit synchronization, total maxi-
mum daily load (TMDL) development, and pollutant
trading between nonpoint and point sources.

The NSI provides an important tool for targeting
efforts to further investigate the 96 watersheds contain-
ing APCs. It is also useful for screening additional po-
tential areas of concern where there are known data gaps.
In addition, the targeting technique used for identifying
the APCs is directly applicable to local-level analysis
because it uses site-specific information. As the NSI is
expanded, it will provide further information to help
environmental managers better understand which of the
Nation's watersheds have sediment contamination prob-
lems that pose the greatest risk to aquatic life and human
health, and track progress in addressing those problems.

There are many active watershed management ef-
forts. EPA recommends strengthening and expanding
these efforts, as appropriate, to better address sediment
contamination issues. The majority of the NSI data were
obtained by local watershed managers from monitoring
programs targeted toward areas of known or suspected
contamination. NSI data and evaluation results can as-
sist local watershed managers by providing additional
data that they may not have, enabling them to compare
their sites to others throughout the region or country,
demonstrating the application of a weight-of-evidence
approach for identifying and screening contaminated
sediment locations, and allowing researchers to draw
upon a large data set of information to conduct new analy-
ses that ultimately will be relevant for local assessments
and responses.

An important component of watershed management
is to educate and engage all stakeholders in government,
industry, and the community. The NSI can help explain
the need to establish pollution prevention initiatives for
point sources and nonpoint sources that might go be-
yond current practices. For example, chemical use prac-
tices in industry and by landowners, homeowners, and
local governments might need to be changed to prevent,
reduce, or eliminate potential sources of sediment con-
taminants.

Recommendation 3: Incorporate a
Weight-of-Evidence Approach and
Measures of Chemical
Bioavailability into Sediment
Monitoring Programs

As stated in Chapter 2 of this volume, the ideal as-
sessment methodology would be based on matched data
sets of multiple types of sediment quality measures to
take advantage of the strengths of each measurement
type and to minimize their collective weaknesses. For
example, sediment chemistry can indicate the amount of

6-2


-------
National Sediment QuaiiH Survey

contaminant present, but cannot definitively indicate
an effect. On the other hand, toxicity tests or benthic
community surveys can indicate an effect, but cannot
definitively implicate a chemical cause. However,
matched sediment chemistry data and toxicity tests, es-
pecially linked through innovative toxicity identifica-
tion evaluation (TIE) approaches, can provide a
preponderance of evidence implicating a chemical cause
of a biological effect. This advocacy of a weight-of-
evidence approach is supported by the consensus of par-
ticipants in an expert workshop on sediment ecological
risk assessment sponsored by the Society of Environ-
mental Toxicology and Chemistry held in Pacific Grove,
California, in April 1995. These scientists concluded
that no single approach provides the best answer for risk
assessment, but each endpoint has strengths and weak-
nesses and the best approach is to use multiple endpoints
(Ingersoll et al., 1997). Toward this end, monitoring
programs should be planned and executed to support
weight-of-evidence assessments.

EPA recommends that future sediment monitoring
programs collect tissue residue, biological effects (i.e.,
toxicity, histopathology), and biological community
(e.g., benthic abundance and diversity) measurements.
These types of data are necessary to better assess actual
effects resulting from exposure to contaminated sedi-
ment. Matched sediment chemistry and tissue residue
data should be collected where human exposures are a
concern. In areas where aquatic life effects are a con-
cern, monitoring programs should collect matched sedi-
ment chemistry and biological effects data and biological
community measurements. There is a need to evaluate
matched sediment chemistry and toxicity data to deter-
mine the predictive ability of screening values to cor-
rectly classify toxicity and minimize both Type I (false
positive) and Type II (false negative) errors.

Collection of measures of chemical bioavailability
is critical to the success of weight-of-evidence assess-
ments. As noted in the previous chapter, a large number
of stations had elevated concentrations of metals. How-
ever, many of these stations could not be categorized as
Tier 1 because of a lack of acid volatile sulfiide (AVS)
and simultaneously extracted metals (SEM) data, which
were required to place stations in the Tier 1 category
based on sediment contamination from cadmium, cop-
per, nickel, lead, or zinc. AVS and SEM provide informa-
tion necessary to assess the bioavailability of metals in
sediment, and future sediment monitoring programs
should specify collection of AVS and SEM measurements
where metals are a concern.

Total organic carbon (TOC) data were also lacking
for many monitoring stations with data in the NSI. TOC,
like AVS and SEM, provides information related to the
bioavailability of contaminants—in this case, nonionic
organic chemicals. Because of the lack of site-specific
TOC data, a default TOC value was used in the NSI evalu-
ation in the comparison of measured sediment chemistry
values to screening values. This approach resulted in
the possible overestimation or underestimation of po-
tential impacts. Therefore, EPA recommends that future
monitoring programs also include TOC measurements
where organic chemicals are a concern.

Recommendation 4: Evaluate the
NSI's Coverage and Capabilities
and Provide Better Access to
Information in the NSI

The NSI is currently limited in terms of the number
of data sets it includes and the national coverage it pro-
vides. Over 50 percent of the monitoring stations evalu-
ated in the NSI are located in eight states (Washington,
Florida, Illinois, California, Virginia, Ohio, Massachu-
setts, and Wisconsin). In addition, only 11 percent of all
river reaches in the United States include one or more
sampling stations that were assessed as part of the NSI
data evaluation.

EPA should continue compiling sediment chemis-
try data and related biological data in the NSI to:

•	Obtain a greater breadth of coverage across the
United States.

•	Increase the number of water bodies evaluated.

•	Include additional data for more chemicals of
concern.

•	Provide more recent data for evaluation for fu-
ture reports to Congress.

During the course of developing and compiling the
NSI, commentators and reviewers identified several ad-
ditional databases that should be included in the NSI for
future evaluations. Those databases and others should
be evaluated and added to the NSI in the future as appro-
priate. EPA plans to obtain the most recent data from
databases currently in the NSI (e.g., STORET and ODES)
and add new data from recent monitoring efforts targeted
at specific water bodies, states, or other areas that are
currently underrepresented in the NSI.

6-3


-------
Recommendations

Although some historical trend information is avail-
able, a comprehensive assessment of temporal trends is
not presented in the current report to Congress. EPA
should consider whether to design future evaluations of
the NSI data to determine where and why sediment qual-
ity conditions are improving or worsening. EPA plans to
develop an approach for assessing temporal trends that
might include, for example, a statistical analysis of recent
and older data from national databases that are updated
on a regular basis, such as STORET, ODES, and the Na-
tional Oceanic and Atmospheric Administration's NS&T
database. In addition, in the search for additional data-
bases for use in future NSI data evaluations, EPA should
focus on obtaining sediment core data, which can pro-
vide valuable information concerning historical trends in
sediment contamination. An assessment of temporal
trends in sediment contamination will provide valuable
information concerning the effectiveness of measures
taken to control the release of sediment contaminants.

The NSI can be a powerful tool for water resource
managers at the national, regional, state, watershed, and
water body levels. It provides in a single place a wealth
of information that could be very useful, especially with
improved access and availability. Multiple agencies
should have access to the same data for decision makers
in regional management, state-level management, and
watershed-level management.

Plans are under development to make this happen.
By the summer of 1997 the NSI data, organized by water-
shed and including maps and summary tables, should be
available on EPA's mainframe computer for on-screen view-
ing and download. In addition, near future plans are to
make this information available on EPA's World Wide Web
site. EPA has also included the NSI data in its compre-
hensive GIS/modeling system, BASINS (Better Assess-
ment Science Integrating Point and Nonpoint Sources).
Future activities should include the addition of the NSI
evaluation tools to BASINS to allow users to query the
NSI evaluation results. For managers, this could be use-
ful for identifying watersheds, water bodies, or sampling
stations where various sediment chemistry and/or bio-
logical screening values have been exceeded. Identify-
ing potential point and nonpoint sources of sediment
contaminants is also critical.

Increased access to data and information in the NSI
has many implications. At the national level, the data
and information can:

* Demonstrate the need and provide impetus for
increased pollution prevention efforts.

•	Demonstrate the need for safer or biodegrad-
able chemicals.

•	Determine relative risk compared to other prob-
lems.

At the state and watershed level, better access to
NSI information can help in:

•	Educating and involving the public.

•	Setting goals and prioritizing activities and ex-
penditures.

•	Evaluating the adequacy and effectiveness of
control actions, clean-up activities, and other
management actions.

Related to source identification are plans under way
at the Agency for one-stop reporting of and access to
integrated information about the environmental perfor-
mance and emissions of major industrial facilities and other
pollution sources. States and EPA will give every major
industrial facility and other type of facility generating,
storing, and disposing of hazardous and toxic wastes a
unique identifying number. This number will be used by
states and EPA to link all environmental information re-
lated to the facility. NSI development will be linked to
these Agency-level efforts.

Interagency and intergovernmental cooperation is
essential for enhancing NSI information, coverage, and
comprehensiveness. Reporting of water quality informa-
tion and environmental indicator development at the Of-
fice of Water are important ongoing efforts related to the
collection of information from state agencies (through
305(b) reporting), other federal agencies, and the private
sector. Efforts for future data collection for the NSI should
be integrated into these related initiatives.

Recommendation 5: Develop Better
Monitoring and Assessment Tools

The National Sediment Quality Survey is the first
attempt to analyze sediment chemistry and biological data
from numerous databases from across the country in an
effort to identify the national incidence and severity of
sediment contamination. Because the data were not gen-
erated by a single monitoring program designed at the
outset to provide this national picture, numerous hurdles
had to be overcome to analyze the data with as little bias
and as much scientific validity as possible. This exercise
itself provided an opportunity to assess the needs to
develop better basic and applied science with respect to
sediment chemistry data and related biological data.

6-4


-------
National Sediment Quality Survey

To ensure effective quality control and quality as-
surance management, monitoring programs should adopt
standard sample collection, storage, analyses, and docu-
mentation procedures. Lack of available quality control
information and the recognized limitations of some past
sampling and analyses methods necessarily restricts the
interpretation of much of the historical data base. How-
ever, these limitations should be eliminated in the future
through current practices such as "clean" laboratory tech-
niques, lowered analytical detection limits, and better
record keeping. Modernization of federal and other data
repositories to accommodate the storage of much addi-
tional valuable and relevant information should help fa-
cilitate the process.

During the evaluation of information in the NSI, ana-
lysts continually came up against the limitations of avail-
able tools and techniques to assess the sediment
contaminant information. Although screening values were
adopted or developed for the NSI data evaluation wher-
ever feasible, many data for some potentially harmful con-
taminants were not evaluated. For example, many
contaminants included in the NSI, such as kcponc and
tributyl tin, could not be evaluated due to a lack of appro-
priate screening values for comparison with measured
values.

The sediment quality evaluation tools used for the
current NSI data evaluation should be used as the basis
for further methods development. As sediment quality
data become more available and the state of the science
for sediment assessment evolves, assessment methods
will also evolve. For example, new and better screening
values and laboratory tests for biological effects will be
developed. EPA should incorporate new sediment as-
sessment techniques into future NSI data evaluations as
they are developed, tested, and proven reliable. For ex-

ample, although biological community data were included
in the NSI, the data were not evaluated for this report to
Congress because there is little agreement among sedi-
ment assessment experts concerning biological commu-
nity conditions that can be directly related to sediment
quality problems, EPA should work to develop these and
other sediment assessment tools for future assessments,
EPA needs to evaluate the ecological relevance of the
assessment tools used to evaluate contaminated sedi-
ment.

Other relevant issues and science needs that should
be addressed to better characterize the sources, fate, and
effects of sediment contaminants include:

•	Methods to better predict the fate and transport
of sediment contaminants.

•	Methods to predict or track atmospheric sources
and cross-media transfers of sediment contami-
nants such as mercury, pesticides, PCBs, and
PAHs.

•	Bioavailability of compounds other than non-
ionic organics.

•	Estimates of land use impacts on sediment con-
ditions {predictive capabilities).

•	Methods for fingerprinting chemicals for source
identification.

In the context of the budget process, EPA and other
federal agencies should evaluate whether to request fund-
ing to support the development of tools to better charac-
terize the sources, fate, and effects of sediment contami-
nants.

6-5


-------
National Sviliment ()ui\lil\ Survey

Glossary

Acid-volatile sulfide (AVS): Reactive solid-phase
sulfide fraction that can be extracted by cold hydrochlo-
ric acid. Appears to control the bioavailability of most
divalent metal ions because of the sulfide ions' high af-
finity for divalent metals, resulting in the formation of
insoluble metal sulfides in anaerobic (anoxic) sediments.

Acute toxicity: Immediate or short-term response of
an organism to a chemical substance. Refers to general-
ized toxic response with lethality usually being the ob-
served endpoint.

Apparent Effects Thresholds (AElfc): Sediment
chemistry screening values based on a biological effects
correlation approach. The AET is the highest concen-
tration at which statistically significant differences in
oberseved adverse biological effects from reference con-
ditions do not occur, provided that the concentration also
is associated with observance of a statistically signifi-
cant difference in adverse biological effects. Based on
empirical data from Puget Sound, EPA defined the AET-
low as the lowest AET among applicable biological indi-
cators, and the AET-high as the highest AET among
applicable biological indicators.

Benthic abundance: The quantity or relative degree
of plentifulness of organisms living in or on the bottom
of streams, rivers, or oceans.

Benthic organisms: Species living in or on the bot-
tom of streams, rivers, or oceans.

Bioavailability: The fraction of chemical present that
is available for uptake by aquatic organisms.

Biological community: An assemblage of organ-
isms that are associated in a common environment and
interact with each other in a self-sustaining and self-regu-
lating relationship.

Biological effects correlation approach: A method
for relating the incidence of adverse biological effects to
the dry-weight sediment concentration of a specific chemi-
cal at a particular site based on the evaluation of paired
field and laboratory data. Exceedance of the identified
level of concern concentration is associated with a likeli-

hood of adverse organism response, but does not dem-
onstrate that a particular chemical is solely responsible.

Cataloging unit: Sometimes referred to as a hydro-
logic unit, corresponds to a watershed that was delin-
eated by the U.S. Geological Survey. A watershed is an
area that drains ultimately to a particular watercourse of
body of water. There are approximately 2,100 cataloging
units in the contiguous United States, which are, on av-
erage, somewhat larger than counties. Each cataloging
unit is uniquely identified with an 8-digit hydrologic unit
code (HUG).

Chronic toxicity: Response of an organism to re-
peated, long-term exposure to a chemical substance. Typi-
cal observed endpoints include growth and reproduction.

Combined sewer overflow: A discharge of a mixture
of storm water and untreated domestic wastewater that
occurs when the flow capacity of a sewer system is ex-
ceeded during a rainstorm.

Contaminated sediment: Sediment that contains
chemical substances at concentrations that pose a known
or suspected threat to aquatic life, wildlife, or human
health.

Demersal species: Swimming organisms that prefer
to spend the majority of their time on or near the bottom
of a water body.

Divalent metals: Metals that are available for reac-
tion in a valence state of two (i.e., carrying a positive
electric charge of two units).

Ecosystem: An ecological unit consisting of both
the biotic communities and the nonliving (abiotic) envi-
ronment, which interact to produce a system which can
be defined by its functionality and structure.

Effects range-median (ERM) and effects range-low
(ERL) values: Sediment chemistry screening values
based on a biological effects correlation approach. Rep-
resent chemical concentration ranges that are rarely (i.e.,
below the ERL), sometimes (i.e., between ERL and ERM),
and usually (i.e., above the ERM) associated with toxic-

Glossary-1


-------
Cilossarj

ity for marine and estuarine sediments. Ranges are de-
fined by the tenth percentile and fiftieth percentile of the
distribution of contaminant concentrations associated
with adverse biological effects.

Elutriate phase toxicity test: Toxicity test in which
sediments are mixed with test water for a fixed period of
time, the test water is then siphoned off, and test organ-
isms are introduced to the test water (the elutriate) in the
absence of sediments. Useful for representing the expo-
sure to chemicals that can occur after sediments have
been resuspended into the water column or after they
have passed through the water column as part of dredged
material disposal operations.

Equilibrium concentration: The concentration at
which a system is in balance due to equal action by op-
posing forces within the system. When the partitioning
of a nonionic organic chemical between organic carbon
and pore water and partitioning of a divalent metal be-
tween solid and solution phases are assumed to be at
equilibrium, an organism in the sediment is assumed to
receive an equivalent exposure to the contaminant from
water only or from any equilibrated phase. The pathway
of exposure might include pore water (respiration), sedi-
ment carbon (ingestion), sediment organism (ingestion),
or a combination of routes.

Equilibrium partitioning (EqP) approach: Approach
used to relate the dry-weight sediment concentration of
a particular chemical that causes an adverse biological
effect to the equivalent free chemical concentration in
pore water and to that concentration sorbed to sediment
organic carbon or bound to sulfide. Based on the theory
that the partitioning of a nonionic organic chemical be-
tween organic carbon and pore water and the partition-
ing of a divalent metal between the solid and solution
phases are at equilibrium.

Histopathology: The study of diseases associated
with tissue changes or effects.

Hydrology: A science dealing with the properties,
distribution, and circulation of water on the surface of
the land, in the soil, and in the atmosphere.

Interstitial water: Water in an opening or space, as
between rock, soil, or sediment (i.e., pore water).

Microbial toxicity test: Type of toxicity test in which
members of the microbial community (i.e., bacteria) are
used as the test organism. Microbial responses in toxic-
ity tests have been recommended as early warning indi-
cators of ecosystem stress. However, questions have

been raised concerning the sensitivity of sediment mi-
crobial toxicity testing.

Molar concentration: The ratio of the number of
moles (chemical unit referring to the amount of an ele-
ment having a mass in grams numerically equal to its
atomic weight) of solute (the substance being dissolved
or that present in the smaller proportion) in a solution
divided by the volume of the solution expressed in liters.

National Sediment Inventory (NSI): A national com-
pilation of sediment quality data and related biological
data. Results of the evaluation of data from the NSI serve
as the basis for the report to Congress on the incidence
and severity of sediment contamination across the coun-
try (i.e., the National Sediment Quality Survey). Eventu-
ally, all compiled NSI data will be incorporated into the
new, modernized STORET, where they will be permanently
stored.

Nonionic organic chemicals: Compounds that do
not form ionic bonds (bonds in which the electrical charge
between bonded atoms in the compound is unequally
shared). Nonionic compounds do not break into ions
when dissolved in water and therefore are more likely to
remain in contact with and interact with sediment com-
pounds or other compounds in water.

Nonpoint source pollution: Pollution from diffuse
sources without a single point of origin or pollution not
introduced into a receiving stream from a specific outlet
Such pollutants are generally carried off the land by storm
water runoff. Sources of nonpoint source pollution in-
clude atmospheric deposition, agriculture, silviculture,
urban runoff, mining, construction, dams and channels,
inappropriate land disposal of waste, and saltwater intru-
sion,

Nonpolar organic chemicals: Compounds that do
not exhibit a strong dipole moment (there is little differ-
ence between the electrostatic forces holding the chemi-
cal together). Nonpolar compounds tend to be less soluble
in water. In aquatic systems, nonpolar chemicals are more
likely to be associated with sediments or other nonpolar
compounds than with the surrounding water.

Point source pollution: Pollution contributed by any
discernible, confined, and discrete conveyance includ-
ing, but not limited to, any pipe, ditch, channel, tunnel,
conduit, well, discrete fissure, container, rolling stock,
concentrated animal feeding operation, or vessel or other
floating craft, from which pollutants are or may be dis-
charged.

GIossary-2


-------
Satioiml Svdinu nt Qutilily Snrvi-%

Pore water: See Interstitial water.

Probable effects levels (PELs) and threshold effects
levels (TELs): Biological effects correlation-based sedi-
ment chemistry screening values similar to ERMs/ERLs.
A generalized approach used to develop effects-based
guidelines for the state of Florida and others. The lower
of the two guidelines for each chemical (i.e., the TEL) is
assumed to represent the concentration below which toxic
effects rarely occur. In the range of concentrations be-
tween the two guidelines, effects occasionally occur.
Toxic effects usually or frequently occur at concentra-
tions above the upper guideline value (i.e., the PEL).
Ranges are defined by specific percentiles of both the
distribution of contaminant concentrations associated
with adverse biological efects and the "no effects" distri-
bution.

River Reach: A stream segment between the con-
secutive confluences of a stream. Most river reaches
represent simple streams and rivers, while some river
reaches represent the shoreline of wide rivers, lakes, and
coastlines. EPA's River Reach File 1 (RF1) was completed
for the contiguous United States in the mid-1980s and
includes approximately 68,000 river reaches. The average
length of a river reach is 10 miles. The more detailed
version of the Reach File (RF3) was not used for the Na-
tional Sediment Inventory.

Sampling Station: A specific location associated
with latitude/longitude coordinates where data have been
collected. Defined by the data source, sponsoring agency,
and station identification code. Multiple sampling sta-
tions can have the same latitude/longitude coordinates if
labeled with a different station identification code for sam-
pling performed on different dates or by different spon-
soring agencies.

Sediment quality advisory levels (SQALs): Equilib-
rium partitioning-based sediment chemistry screening val-
ues. Derived using the same approach used to develop
sediment quality criteria; however, SQALs may be based
on a limited set of aquatic toxicity data.

Sediment quality criteria (SQCs): Published draft
sediment quality criteria for the protection of aquatic life.
Based on the equilibrium partitioning-based approach
using the highest quality toxicity and octanol/water par-
titioning data, which have been reviewed extensively.
Draft SQCs have been developed by EPA for five noil-

ionic organic chemicals: acenaphthalene, dieldrin, en-
drin, fluoranthene, and phenanthrene.

Simultaneously extracted metals (SEM): Metal con-
centrations that are extracted during the same analysis in
which the acid-volatile sulfide (AVS) content of the sedi-
ment is determined.

Solid-phase toxicity test: A toxicity test in which
test organisms are exposed directly to sediments. Sedi-
ments are carefully placed in the exposure chamber and
the chamber is then filled with clean water. Resuspended
particles are allowed to settle before initiation of expo-
sure. Solid-phase toxicity tests integrate multiple expo-
sure routes, including chemical intake from dermal contact
with sediment particles as well as ingestion of sediment
particles, interstitial water, and food organisms.

Theoretical bioaccumulation potential (TBP): An
estimate of the equilibrium concentration of a contami-
nant in tissues if the sediment in question were the only
source of contamination to the organism. TBP is esti-
mated from the organic carbon content of the sediment,
the lipid content of the organism, and the relative affini-
ties of the chemical for sediment organic carbon and ani-
mal lipid content.

Total organic carbon (TOC): A measure of the or-
ganic carbon content of sediment expressed as a percent
Used to normalize the dry-weight sediment concentra-
tion of a chemical to the organic carbon content of the
sediment.

U.S. Environmental Protection Agency (EPA) risk
levels: Levels of contaminant concentrations in an expo-
sure medium that pose a potential carcinogenic risk (e.g.,
10"5, or a 1 in 100,000 extra chance of cancer over a life-
time) and/or noncancer hazard (i.e., exceeds a reference
dose). Used in this document to estimate human health
risk associated with the consumption of chemically con-
taminated fish tissue.

U.S. Food and Drag Administration (FDA) tolerance/
action or guideline levels: FDA has prescribed levels of
contaminants that will render a food "adulterated." The
establishment of action levels (the level of a food con-
taminant to which consumers can be safely exposed) or
tolerances (regulations having the force of law) is the
regulatory procedure employed by FDA to control envi-
ronmental contaminants in the commercial food supply.

Glossary-3


-------
Glossary-4


-------
Acronyms

AET: apparent effects threshold

APC: area of probable concern for sediment con-
tamination

AVS: acid volatile sulfide

BASINS: Better Assessment Science Integrating
Point and Nonpoint Sources (EPA model-
ing tool)

BSAF: biota-sediment accumulation factor

CAA: Clean Air Act

CAS: Chemical Abstract Service

COSED: Coastal Sediment Inventory

CWA: Clean Water Act

CZMA: Coastal Zone Management Act

DMATS: Dredged Material Tracking System

EMAP: Environmental Monitoring and Assessment
Program	!

EPA:	U. S. Environmental Protection Agency

ERL:	effects range-low value

ERM:	effects range-median value

FDA:	Food and Drug Administration

FIFRA: Federal Insecticide, Fungicide, and Roden-
ticide Act

MPRSA: Marine Protection, Research, and Sanctu-
aries Act

NEPA: National Environmental Policy Act

NOAA: National Oceanic and Atmospheric Admin-
istration

NPDES: National Pollutant Discharge Elimination
System

NSI:	National Sediment Inventory

NURP:	National Urban Runoff Program

ODES:	Ocean Data Evaluation System

OST:	Office of Science and Technology, U. S. En-
vironmental Protection Agency

PAH:	polynuclear aromatic hydrocarbon

PCB:	polychlorinated biphenyls

PCS:	Permit Compliance System

PEL:	probable effects level

QA/QC:	quality assurance/quality control

RCRA:	Resource Conservation and Recovery Act

RF1:	River Reach File 1

SEM:	simultaneously extracted metals

SQAL:	sediment quality advisory level

SQC:	sediment quality criteria

STORET: Storage and Retrieval System

TBP:	theoretical bioaccumulation potential

TEL:	threshold effects level

TIE:	toxicity identification evaluation

TMDL:	total maximum daily load

TOC:	total organic carbon

TRI:	Toxic Release Inventory

TSCA:	Toxic Substance Control Act

USACE:	U. S. Army Corps of Engineers

USGS:	U. S. Geological Survey

WRDA:	Water Resources Development Act of 1992

Acronyms-1


-------
Acn tin ins


-------
References

Adams, W.J., R.A. Kimerle, and R.G. Mosher.
1985. Aquatic safety assessment of chemicals
sorbed to sediments. In Aquatic Toxicology and
Hazard Assessment: Seventh Annual Symposium,
American Society for Testing and Materials,
Philadelphia, PA, ed. R.D. Cardwell, R. Purdy,
and R.C. Bahner, pp. 429-453.

Allen, H.E., G. Fu, and B. Deng. 1993. Analysis of
acid-volatile sulfide (AVS) and simultaneously
extracted metals (SEM) for the estimation of
potential toxicity in aquatic sediments. Environ.
Toxicol. Chem. 12:1441-1453,

API. 1994. User's guide and technical resource docu-
ment: Evaluation of sediment toxicity tests for
biomonitoring programs. API pub. no. 4607. Pre-
pared for American Petroleum Institute, Health and
Environmental Sciences Department, Washington,
DC, by PTI Environmental Services, Bellvue, WA.

Auerlich, R.J., R.K. Ringer, and S. Iwamoto. 1973.
Reproductive failure and mortality in mink fed on
Great Lakes fish. J. Rep rod. Fertil. Suppl. 19:365.

Baker, J.E., T.M. Church, S.J. Eisenreich, W.F.
Fitzgerald, and S.R. Scudlark. 1993. Relative,
atmospheric loadings of toxic contaminants and
nitrogen to the great waters. Prepared for U.S.
Environmental Protection Agency, Office of Air
Quality Planning and Standards, Pollution Assess-
ment Branch, Durham, NC. Cited in USEPA,

1994a.

Barrick, R., S. Becker, L. Brown, H. Beller, and R.
Pastorok. 1988. Sediment quality values refinement:
1988 update and evaluation of Puget Sound AET. Vol.
1. Prepared for the Puget Sound Estuary Program,
Office of Puget Sound.

Baudo, R., and H. Muntau. 1990. Lesser known in-place
pollutants and diffuse source problems. Chap. 1 in
Sediments: Chemistry and toxicity of in-place pollut-
ants, ed. R. Baudo, J. Giesy and H. Muntau. Lewis
Publishers, Chelsea, ML

Baumann, P.C., W.D. Smith, and W.K. Parland. 1987.
Tumor frequencies and contaminant concentrations in
brown bullheads from an industrialized river and a
recreational lake. Trans. Amer. Fish. Soc. 116:79-80.

Becker, D.S., T.C. Ginn, M.L. Landolt, and D.B. Powell.
1987. Hepatic lesions in English sole (Parophrys
vetulus) from Commencement Bay, Washington
(USA). Mar. Environ. Res. 23:153-173.

Berry, W., M. Cantwell, P. Edwards, J. Serbst, and D.
Hansen. 1996. Predicting the toxicity of silver-spiked
sediments using interstitial water metal and acid
volatile sulfide normalization. In Abstract Book,
SETAC 17th Annual Meeting, sponsored by the
Society of Environmental Toxicology and Chemistry,
Washington, DC, November 17-21,1996.

Bilyard, G.R. 1987. The value of benthic infaunain
marine pollution monitoring studies. Mar. Poll. Bull.
18:581-585.

Bondelid, T.R., and S.A. Hanson. 1990. Technical
description of the Reach File. Prepared for U.S.
Environmental Protection Agency, Assessment and

. Watershed Protection Division, Washington, DC.

Bopp, R.F., M.L. Gross, H. Tong, H.J, Simpson, SJ.
Monson, B.L. Deck, and F.C. Moser. 1991. A
major incident of dioxin contamination: Sediments
of New Jersey estuaries. Environ. Sci. Technol.
25(5): 951-956.

Bopp, R.F., H.J. Simpson, S.N. Chillrud, and D.W.
Robinson. 1993. Sediment derived chronologies of
persistent contaminants in Jamaica Bay, New York.
Estuaries 16(3B): 608-616.

Bricker, S.B. 1993. Historical trends in contamination of
estuarine and coastal sediments: The history of Cu, Pb,
and Zn inputs to Narragansett Bay, Rhode Island as
recorded by salt marsh sediments. Estuaries
16(3B):589-607.

Brown, R.C., R.H. Pierce, and S.A. Rice. 1985. Hydro-
carbon contamination in sediments from urban
storm water runoff. Mar. Poll. Bull. 16(6):236-240,

References-1


-------
Ri"IVri'»ccs

Burton, G.A., Jr., B.L. Stemmer, K.L. Winks, EE.
Ross, and L.C. Burnett. 1989. A multitrophic
level evaluation of sediment toxicity in
Waukegan and Indiana harbors. Environ.

Toxicol. Chem. 8:1057-1066.

Canfield, T.J., N.E. Kemble, W.G. Brumbaugh, F.J.
Dwyer, C.G. Ingersoll, and F.J. Fairchild. 1994.
Use of benthic invertebrate community structure
and the sediment quality triad to evaluate metal-
contaminated sediment in the Upper Clark Fork
River, Montana. Environ. Toxicol. Chem.
13:1999-2012.

Clarice, J.U., and V.A. McFarland. 1991. Assessing
bioaccumulation in aquatic organisms exposed to
contaminated sediments. Long-Term Effects of
Dredging Operations Program, misc. pap. D-91-2.
U.S. Army Corps of Engineers, Waterways Experi-
ment Station, Vicksburg, MS.

Cohn-Lee, R., and D. Cameron. 1991. Poison runoff in
the Atlanta Region. Natural Resources Defense
Council, Washington, DC.

Crawford, D.W., N.L. Bonnevie, and RJ. Weaning.
1995. Sources of pollution and sediment contamina-
tion in Newark Bay, New Jersey. Ecotoxicol. Environ.
Saf. 30(1): 85-100.

Dexter, R.N., DJE. Anderson, E.A. Quinlan, L.S.

Goldstein, R.M. Strickland, R.M. Kocan, M. Landolt,
J.P. Pavlou, and V.R. Clayton. 1981. A summary of
knowledge ofPuget Sound. NOAA tech. mem.
OMPA-13. National Oceanic and Atmospheric
Administration, Washington, DC.

Di Toro, D.M., J.D. Mahony, DJ. Hansen, K.J.
Scott, M.B. Hicks, S.M. Mays, and M.S.
Redmond. 1990. Toxicity of cadmium in
sediments: The role of acid volatile sulfide.
Environ. Toxicol. Chem. 9:1487-1502.

Di Toro, D.M., J.D. Mahony, D.J. Hansen, K.J.
Scott, A.R. Carlson, and GT. Ankley. 1992.

Acid volatile sulfide predicts the acute toxicity
of cadmium and nickel in sediments. Environ.
Sci. Technol. 26(1):96-101.

Eisler, R. 1995. Electroplating wastes in marine
environments. A case history at Quonset Point,

Rhode Island. In Handbook of ecotoxicology, ed. D.J.
Hoffman, B.A. Rattner, G.A. Burton, Jr., and J.

Cairns, Jr., pp. 609-630. Lewis Publishers, Boca
Raton, FL.

FDEP. 1994. Approach to the assessment of sediment
quality in Florida coastal water. Vol. 1. Development
and evaluation of sediment quality assessment
guidelines. Prepared for Florida Department of
Environmental Protection, Office of Water Policy,
Tallahassee, FL, by MacDonald Environmental
Sciences Ltd., Ladysmith, British Columbia.

Fox, R.G., and M. Tuchman. 1996. The assessment and
remediation of contaminated sediments (ARCS),
program. J. Great Lakes Res. 22:493-494.

Garton, L.S., J.S. Bonner, A.N, Ernest, and R.L.

Autenrieth. 1996. Fate and transport of PCBs at the
New Bedford Harbor Superfund site. Environ.
Toxicol. Chem. 15(5): 736-745.

Gilliom, R.J., and D.G. Clifton. 1990. Organochlorine
pesticide residues in bed sediments of the San Joaquin
River, California Water Res. Bull. 26(1): 11-24.

Ginn, T.C., and R.A. Pastorok. 1982. Assessment
and management of contaminated sediments in
Puget Sound. In Sediment toxicity assessment,
ed. G.A Burton, Jr., pp. 371-397. Lewis Publish-
ers, Chelsea, MI.

Hansen, D.J., 1995. Assessment tools that can be used
for the National Sediment Inventoiy. Memorandum
from D.J. Hansen, Environmental Research Labora-
tory, Narragansett, to C. Fox, USEPA Office of Water,
February 28,1995.

Hansen, D.J., W.J. Berry, J.D. Mahoney, W.S. Boothman,
D.M. DiToro, D.L. Robson, G.T. Ankley, D. Ma, Q.
Yan, and C.E. Pesch. 1996. Predicting the toxicity of
metal-contaminated field sediments using interstitial
concentrations of metals and acid-volatile sulfide
normalizations. Environ. Toxicol, Chem. 15(12):
2080-2094.

Helfrich, J., and D.E. Armstrong. 1986. Polycyclic
aromatic hydrocarbons in sediments of the southern
basin of Lake Michigan. J. Great Lakes Res.
12(3): 192-199.

Henny, C.J., L J. BIus, D.J. Hoffman, and R.A. Grove.
1994. Lead in hawks, falcons and owls downstream
from a mining site on the Coeur d'Alene River, Idaho.
Environ. Mo nit. Assess. 29(3):267-288.

Hoffman, E.J. 1985. Urban runoff pollutant inputs to
Narrangansett Bay: Comparison to point sources. In
Proc. Perspectives on Nonpoint Source Pollution,

References-2


-------
I

NuJiomil	Survey

sponsored by U.S. Environmental Protection Agency,
Kansas City, Missouri, May 19-22, pp. 159-164. EPA
440/5-85-001.

Horowitz, A J., K.A. Elrick, and R.B. Cook. 1993. Effect
of mining and related activities on the sediment trace
element geochemistry of Lake Coeur d'Alene, Idaho,
USA. Part 1: Surface sediments. Hydrol. Processes
7(4):403-423.

Huggett, R.J., and J.M. O'Connor. 1988. Aquatic j
" pollution problems, North Atlantic Coast, including
Chesapeake Bay. Aquat. 7o;u'co/.l 1:163-190.

IJC. 1987. Guidance on characterization of toxic
substances problems in areas of concern in the Great
Lakes Basin. International Joint Commission,
Surveillance Work Group, Windsor, Ontario.

Ingersoll, C.G., P.S. Haverland, E.L. Branson, T.J.
Canfield, F.J. Dwyer, C.E. Henke, and N.E. Kemble.

1996.	Calculation and evaluation of sediment effect
concentrations. J. Great Lakes Res. 22:602-623.

Ingersoll, C.G., G.R. Biddinger, and T. Dillon (editors).

1997.	Ecological risk assessments of contaminated
sediments. SETAC Press, Pensaeola, Ft: In press.

Johnson, N.D., M.T. Scholtz, V. Cassaday, K. Davidson,
and D. Ord. 1992. MOE toxic chemical emission
inventory for Ontario and eastern North America.
Report no. P92-T61-5429/OG. Final report to the Air
Resources Branch, Ontario Ministry of the Environ-
ment, prepared by Ortech International.

Keeler, G.J. 1994. Lake Michigan Urban Air Toxics
Study. U.S. Environmental Protection Agency, Office
of Research and Development, Atmospheric Research
and Exposure Assessment Laboratory, Research
Triangle Park, NC. Cited in USEPA, 1994a.

Keeler, G.J., J.M. Pacyna, T.F. Bidleman, and J.O.
Nriagu. 1993. Identification of sources contribut-
ing to the contamination of the great waters by
toxic compounds. Prepared for U.S. Environmen-
tal Protection Agency, Office of Air Quality
Planning and Standards, Pollution Assessment
Branch, Durham, NC.

Kennicutt, M.C., T.L. Wade, B,J. Presley, A.G. Requejo,
J.M. Brooks, and GJ. Denoux. 1994. Sediment
contaminants in Casco Bay, Maine: Inventories,
sources, and potential for biological impact. Environ.
Sci. Technol. 28(1):1-15.

Kleinman, R.L.P. 1989. Acid mine drainage, U.S. Bureau
of Mines researches and develops control methods for
both coal and metal mines. Energ. Min. J. 16I-16N.

Kubiak, T.J., H.J. Harris, L.M. Smith, T.R. Schwartz,
D.L. Stalling, J. A. Trick, L. Sileo, D.E. Docherty, and
T.C. Erdman.1989, Microcontaminants and reproduc-
tive impairment of the Forster's tern on Green Bay,
Lake Michigan. 1983. Arch. Environ. Contam.
Toxicol. 18:706-727.

Lake, J.L., R.J. Pruell, and F.A. Osterman. 1992. An
examination of dechlorination processes and pathways
in New Bedford Harbor sediments. Mar. Environ. Res.
33(1): 31-47.

Landis, W.G., and M-H. Yu. 1995. Introduction to
environmental toxicology: Impacts of chemicals
upon ecological systems. Lewis Publishers, Boca
Raton, FL,

Leigh, D.S. 1994. Mercury contamination and floodplain
sedimentation from former gold mines in north
Georgia. Water Res. Bull. 30(4):739-748.

Livingston, E.H., and J.H. Cox. 1985, Urban stormwater
quality management: The Florida experience. In Proc.
Perspectives on Nonpoint Source Pollution, sponsored
by U.S. Environmental Protection Agency, Kansas
City, Missouri, May 19-22, pp. 289-291. EPA 440/5-
85-001.

Long, E.R. 1982. An assessment of marine pollution in
Puget Sound. Mar. Pollut, Bull. 13:380-383.

Long, E.R., L.J. Field, and D.D. MacDonald.
Predicting toxicity in marine sediments with
numerical sediment guidelines. Submitted to
Environ. Toxicol. Chem.

Long, E.R., D.D. MacDonald, S.L. Smith, and F.D.
Calder. 1995. Incidence of adverse biological
effects within ranges of chemical concentrations
in marine and estuarine sediments. Environ.
Manage. 19(l):81-97.

Long, E.R., A. Robertson, D.A. Wolfe, I. Hameedi, and
G.M. Sloane. 1996. Estimates of the spatial extent of
sediment toxicity in major U. S. estuaries. ES&T.
30(12): 3585-3592

Long, E.R., and L.G. Morgan. 1990. The potential for
biological effects of sediment sorbed contaminants
tested in the National Status and Trends Program.

References-3


-------
NOAA tech. mem. NOA OMA 52. National Oceanic
and Atmospheric Administration, Seattle, WA.

Luoma, S.N. 1983. Bioavailability of trace metals
to aquatic organisms—A review. Sci. Tot.

Environ. 28:1-22.

Lyman, WJ., A.E. Glazer, J.H. Ong, and S.F. Coons.
1987. An overview of sediment quality in the United
States. Prepared for U.S. Environmental Protection
Agency, Office of Water Regulations and Standards,
Washington, DC.

Malins, D.C., B.B. McCain, D.W. Brown, A.K.
Sparks, and H.O. Hodgins. 1980. Chemical
contaminants and biological abnormalities in
central and southern Paget Sound. NOAA tech.
mem. OMPA-2. National Oceanic and Atmo-
spheric Administration, Washington, DC.

Mancini, J.L., and A.H. Plummer, Jr. 1994. A review of
EPA sediment criteria Proceedings of the Water
Environment Federation, 67th Annual Conference and
Exposition, October 15-19, Chicago, EL. In Surface
quality ecology, Vol. 4, pp. 681-694.

Mas (ran, T.A., A.M. Dietrich, DX. Gallagher, and T.J.
Grizzard. 1994. Distribution of polyaromatic hydro-
carbons in the water column and sediments of a
drinking water reservoir with respect to boating
activity. Water Res. 2S(ll):2353-2366.

Meyer, J.S., W. Davison, B. Sundby, J.T. Ores, D .J.
Lauren, U. Forstner, I. Hong, and D.G. Crosby. 1994.
Synopsis of discussion session; The effects of variable
redox potentials, pH, and light on bioavailability in
dynamic water-sediment environments. In
Bioavailability physical, chemical, and biological
interactions, proceedings of the Thirteenth Pellston
Workshop, ed. J.L. Hamelink, P.F. Landmen, H.L.
Bergman, and W.H. Benson, pp. 155-170. Lewis
Publishers, Boca Raton, FL.

Moore, J.N., S.N. Luoma, and D. Peters. 1991.
Downstream effects of mine effluent on an
intermontane riparian system. Can. J. Fish.

Aquat. Sci. 48(2):222-232.

Myers, M.S., CM. Stehr, O.P, Olseo, L.L. Johnson, B.B.
McCain, S-L Chan, and U. Varansi. 1994. Relation-
ships between toxicopathic hepatic lesions and
exposure to chemical contaminants in English sole
(Pleuronectes vetulus), starry flounder (Platichthys
stellatus), and white croaker (Genyonemus Uneatus)

from selected marine sites on the Pacific Coast, USA.
Envir. Health Perspect. 102(2):200-215.

Nichols, M.M. 1990. Sedimentologic rate and
cycling of kepone in an estuarine system: Ex-
ample from the James River Estuary. Sci. Tot.
Environ. 97-98:407-440.

NOAA. 1994. Inventory of chemical concentrations in
coastal and estuarine sediments. NOAA tech. mem.
NOS ORCA 76. National Oceanic and Atmospheric
Administration, National Ocean Service, Silver
Spring, MD.

NRC (National Research Council). 1989. Contaminated
marine sediments—Assessment and remediation.
National Academy Press, Washington, DC.

Pereira, W.E., F.D. Hostettler, and R.B. Rapp. 1994.
Bioaccumulation of hydrocarbons derived from
terrestrial and anthropogenic sources in the Asian
clam, Potamocorbula amurensis, in San Francisco Bay
Estuary. Mar. Poll. Bull. 24(2): 103-109.

Pitt, R.E. 1995. Effects of urban runoff on aquatic biota.
In Handbook of ecotoxicology, ed. D.J. Hoffman, B.A.
Rattner, G.A. Burton, Jr., and J. Cairns, Jr., pp. 609-
630. Lewis Publishers, Boca Raton, FL.

Power, E.A., and P.M. Chapmaim. 1992. Assessing
sediment quality. In Sediment toxicity assessment, ed.
G.A. Burton, Jr. Lewis Publishers, Ann Arbor, MI.

Prahl, F.G., E. Crecellus, and R. Carpenter. 1984.
Polycyclic aromatic hydrocarbons in Washington
coastal sediments: An evaluation of atmospheric and
riverine routes of introduction. Environ. Sci. Technol.
18:687-693.

Rand, G.M., and S.R. Petrocelll. 1985. Fundamentals of
aquatic toxicology. Hemisphere Publishing Corp.,
New York, NY.

Rice, D.W., S.P Seltenrich, R.B. Spies, and M.L. Keller.
1993. Seasonal and annual distribution of organic
contaminants in marine sediments from Elkhorn
Slough, Moss Landing Harbor and Nearshore
Monterey Bay, California. Environ. Pollut. 82:79-91.

Riley, R.G., E.A. Crecelius, M.L. O'Malley, K.H. Abel,
and D.C. Mann. 1981. Organic pollutants in water-
ways adjacent to Commencement Bay (Puget Sound).
NOAA tech. mem. OMPA-12. National Oceanic and
Atmospheric Administration, Rockville, MD.

References-4


-------
National Nctlijiitul y»sililv Survey

Ryan, I., and J.H. Cox. 1985. The influence of NPS
pollution in Florida estuaries: A case study. In Proc.
Perspectives on Nonpoint Source Pollution, sponsored
by U.S. Environmental Protection Agency, Kansas
City, Missouri, May 19-22, pp. 172-179. EPA 440/5-
85-001.

SAB. 1989. Review of the apparent effects threshold
approach to setting sediment criteria. Report of the
Science Advisory Board, Sediment Criteria Subcom-
mittee, U.S. Environmental Protection Agency,
Washington, DC.

Salomons, W„ N.M. de Rooji, H. Kerdijk, aid J. Bril.
1987. Sediment as a source for contaminants?
Hydmbiologia 149:13-30.

Schueler, T. Fall 1995. Urban pesticides: From the lawn
to the stream. Watersh. Prat. Tech. 2(1):247- 253.

Schueler, T.R. Summer 1994. Pollutant dynamics of pond
muck. Watersh. Prot. Tech. l(2):39-46.

Sorensen, J.A., G.E. Glass, K.W. Schmidt, l.K. Huber,
and G.R. Rapp, Jr. 1990. Airborne mercury deposition
and watershed characteristics in relation to mercury
concentrations in water, sediments, plankton, and fish
of eighty northern Minnesota lakes. Environ. Set
Technol. 24(11)1716-1727.

Swartz, R.C., W.A. Deben, K.A. Sereo, and J.O.

Lamberson. 1982. Sediment toxicity and distribution
of amphipods in Commencement Bay, Washington,
USA. Mar. Poll Bull. 13:359-364.

Swartz, R.C., D.W. Schults, l.O. Lamberson, R.J.
Ozretich, , and J.K. Stull. 1991. Vertical profiles
of toxicity, organic carbon, and chemical con-
taminants in sediment cores from the Palos
Verdes Shelf and Santa Monica Bay, California.
Mar. Environ. Res. 31:215-225.

Swartz, R.C., D.W. Schults, R.J. Ozretich, J.O.

Lamberson, EA. Cole, T.H. DeWitt, M J. Redmond,
and S.P. Fenaro. 1995. PAH: A model to predict the
toxicity of polynuclear aromatic hydrocarbon mixtures
in field-collected sediments. Environ. Toxicol. Chem.
14(11): 1977-1987.

USEPA. 1987. An overview of sediment quality in
the United States. EPA-905/9-88-002. U.S.
Environmental Protection Agency, Office of ,
Water, Washington, DC.

	. 1992a. Proceedings of EM's contaminated

sediment management strategy forums. Chicago, EL,
April 21-22; Washington, DC, May 27-28 and June
16,1992. EPA 823-R-92-007.

	. 1992b. Environmental impacts of stormwater

discharges: A national profile, U.S. Environmental
Protection Agency, Office of Water, Washington, DC.

	. 1993a. Framework for the development of the

National Sediment Inventory. U.S. Environmental
Protection Agency, Office of Science and Technology,
Washington, DC.

	. 1993b. Guidelines for deriving site-

specific sediment quality criteria for the protec-
tion of benthic organisms. EPA-822-R-93-017.
U.S. Environmental Protection Agency, Office of
Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1993c. Identification of sources contributing to

the contamination of the great waters by toxic
compounds. U.S. Environmental Protection Agency,
Office of Manning and Standards, Durham, NC.

	. 1993d. Technical basis for establishing

sediment quality criteria for nonionic organic
contaminants for the protection of benthic organisms
by using equilibrium partitioning. Draft. EPA 822-R-
93-011. U.S. Environmental Protection Agency, Office
of Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1994a. Deposition of air pollutants to the great

waters. U.S. Environmental Protection Agency,

Office of Air Quality, Research Triangle Park, NC

	. 1994b. Methods for assessing the toxicity of

sediment-associated contaminants with estuarine and
marine amphipods. EPA 600-R-94-025. US. Envi-
ronmental Protection Agency, Office of Research and
Development, Washington, DC.

	. 1994c. Methods for measuring the toxicity and

bioaccumulation of sediment-associated contaminants
with fresh water invertebrates. EPA 600/R-94/024,
Environmental Protection Agency, Duluth, MN.

——. 19944. Proceedings of the National Sediment
Inventory Workshop; April 26-27, 1994, Washing-
ton, DC. U.S. Environmental Protection Agency,
Office of Science and Technology, Washington, DC.

References-5


-------
I

Kcl'cmu't's

	. 1994e, National water quality inventory:

1992 report to Congress. EPA-841-R-94-001. U.S.
Environmental Protection Agency, Office of Water,
Washington, DC.

	, 1994f. Guidance for assessing chemical

contamination data for use in fish advisories, Vol.
II: Risk assessment and fish consumption limits.
EPA 823-B-94-004, U.S. Environmental Protection
Agency, Office of Water, Washington, DC.

	. 1994g. The Watershed Protection Approach:

1993/94 Activity Report. EPA 840-S-94-001. U.S.
Environmental Protection Agency, Office of Water,
Washington, DC.

	. 1995a. Great Lakes Water Quality Initiative

criteria documents for the protection of wildlife.
EPA-820-B-95-008. U.S. Environmental Protec-
tion Agency, Office of Science and Technology,
Washington, DC.

	. 1995b. A Phase I Inventory of Current EPA

Efforts to Protect Ecosystems. EPA 841-S-95-001.
U.S. Environmental Protection Agency, Office of
Water, Washington, DC.

	. 1995c. Guidance for Assessing Chemical

Contaminant Data for Use in Fish Advisories,
Volume 1: Fish Sampling and Analysis (Second
Edition). EPA 823-R-95007, U.S. Environmental
Protection Agency, Office of Water,. Washington,
DC.

	. 1996. Derivation of EPA's Sediment Quality

Advisory Levels (DRAFT). U.S. Environmental
Protection Agency, Office of Science and Technol-
ogy, Washington, DC.

USGS. 1993. Persistence of the DDT pesticide in the
Yakima River Basin Washington. USGS circular
1090. U.S. Geological Survey, Denver, CO.

Van Veld, P.A., D.J. Westbrook, B.R. Woodin, R.C.
Hale, C.L. Smith, R.J. Hugget, and J.J. Stegraan.
1990. Induced cytochrome P-450 in intestine and
liver of spot (Leiostomus xanthurus) from a
polycyclic aromatic contaminated environment.
Aquat. Toxicol. 17:119-132.

Voldner, E.C. and L. Smith. (1989). Production,
usage, and emissions of fourteen priority toxic
chemicals. Appendix 2 to the Proceedings of the
Workshop on the Estimation of Atmospheric
Chemicals to the Great Lakes Basin. Scarborough,
Ontario, Joint Water Quality Board/Science
Advisory Board/International Air Quality Advisory
Board of the International Joint Commission
Windsor, Ontario, ISBN 1-895085-20-9, 94 pp.

Wenning, R.J., N.L. Bonne vie, and S,L. Huntley.
1994. Accumulation of metals, polychlorinated
biphenyls, and polycyclic aromatic hydrocarbons in
sediments from the lower Passaic River, New
Jersey. Arch. Environ. Contam. Toxicol. 27(1):64-
81.

Willis, G.H., and L.L. McDowell. 1983. Environmen-
tal chemistry review: Pesticides in agricultural
runoff and their effects on downstream water
quality. Environ. Toxicol. Chem. 1:267-279.

Yousef, Y.A., H.H. Harper, L, Wiseman, and M.
Bateman. 1985. Consequential species of heavy
metals. FL-ER-29-85. University of Central
Florida, Department of Civil Engineering and
Environmental Sciences. Sponsored by the Florida
Department of Transportation.

References-6


-------
\sllt«>il:ll Si cliiiH iiI <_)u.ilil v Stiru-,\

Appendix A

Detailed Description of
NSI Data

Sources of the NSI Data

The scope of the data compilation component of the NSI was to collect, review, and compile readily available
data that could be used to evaluate the incidence of sediment contamination throughout the United States.
As a result, emphasis was placed on gathering data sets with sediment chemistry data since thosfe were the
most prevalent data available on a national basis. The minimum data elements for inclusion in the NSI were date of
sample collection, latitude/longitude, reliable units (e.g., mg/kg), and source of data. The electronic data sources
used for the NSI are listed below.	;

•	EPA's Storage and Retrieval System (STORET)

•	EPA's Ocean Data Evaluation System (ODES)

•	NOAA's Coastal Sediment Inventory (COSED)

•	EPA Region 4's Sediment Quality Inventory

•	EPA Gulf of Mexico Program's Contaminated Sediment Inventory

•	EPA Region 10/COE Seattle District Sediment Inventory

•	EPA's Great Lakes Data Base

•	EPA's Environmental Monitoring and Assessment Program (EMAP)

•	EPA Region 9 Dredged Material Tracking System (DMATS)

•	USGS Massachusetts Bay Data (metals only)

•	National Source Inventory (PCS and TRI)

In several cases, the readily available data sources for the NSI were compilations of existing data. For example, the

EPA Gulf of Mexico Program's Contaminated Sediment Inventory included data from ODES, STORET, and EMAP.
Since those data sources had been reviewed independently, they were deleted from the Gulf of Mexico Inventory
before that data set was added to the NSI. A similar screening of data was conducted for the other data sets included in
the NSI. Below is a summary of the remaining contributors to the individual data sets:

STORET	Numerous federal and state agencies

ODES	Boston Harbor	Tennessee

Masschusetts Bay

Cape Arundel
City of Gloucester

Mile 106
South Carolina
Alabama
Mississippi
Georgia
North Carolina
Encina 301(h)
Morro Bay 301(b)
Hyperion 301(h)

Kentucky
Florida

GLNPO/ARCS
Galveston Bay
San Diego Pre-301(h)

Orange County 301(h)
Oxnard 301(h)
Los Angeles 301(h)
Thums Ocean Dumping
Puget Sound
Anchorage
Endicott 403(c)

A-l


-------
Goleta 301(h)

San Francisco NEP
LA2 Ocean Dumping
LA5 Ocean Dumping

COSED	NOAA NS&T

Region 4	City of Tampa

Dept of Navy
EPA Region 4
Florida DER

South Florida Water Mgmt DisL
USACE

Gulf of Mexico ADEM (Mobile)

Army Corps Eng.

EPA-Houston
ERL-N

GCRL, Mississippi

Seattle COE Department of Social and Health Services
Department of Ecology
U.S. Fish and Wildlife Service
Puget Sound Water Quality Authority
Tetra Tech, Inc.

Department of Fisheries
Department of Natural Resources
Department of Wildlife
EPA Region 10

Batelle Northwest Sequim Laboratory

Environmental Systems Corporation

Department of Health

College of Ocean and Fisheries Science

PTI Environmental Services

National Oceanic and Atmospheric Admin. Fish

and Wildlife Health Consultants
City of Bellingham

U.S. Army Corps of Engineers, Seattle
Columbia Northwest, Inc.

Hulbert Mill
King County

Municipality of Metropolitan Seattle

Wildlife Health Consultants

U.S. Navy	'

City of Olympia, LOTT treatment plant

Port of Bellingham

Port of Everett

Port of Olympia

Port of Port Town send

Thurston County Dept of Public Health

U.S. Coast Guard

Great Lakes Heidelberg College, Tiffin, Ohio
Illinois EPA

Michigan Tech. Univ., Houghton, MI

Kuparuk STP 403(c)
Prudhoe Bay 403(c)
Port Valdez 403(c)

Ed Long

USACE, Jacksonville
USACE, Mobile
USACE, Savannah
USACE, Wilmington
USFWS

TVA

USACE (Mobile)

USEPA Region 6
USGS

Department of Parks and Recreation
Environmental Information Consultants
South. CA Coastal Water Research
Proj., Army Corps of Engineers, San
Francisco

Environmental Science Associates, Inc.
E.V.S. Consultants, Sausalito, CA
Marine Bioassay Labs, Watsonville,

CA

MEC Analytical Systems, Watsonville,
CA

San Francisco Port Commission
ToxScan, Inc., Watsonville, CA
Tetra Tech, Inc., Lafayette, CA
Port of Grays Harbor
Port of Tacoma

Tristar Marine	^

Morton Marine

Port of Seattle

South Park Marina

U.S. Oil and Refining Company

Weyerhauser

Day Island Yacht Club

Shell Oil

Capital Regional District, Victoria, BC
Environment Canada Greater

Vancouver Regional District
E.V.S. Consultants, Seattle, WA
E.V.S. Consultants, Vancouver, BC
British Petroleum Oil Company
American Petroleum Institute

US Army COE, Buffalo District

Beak Consultants, Inc

Ontario Ministry of the Environment

A-2


-------
\;ifion;iJ Sediment Quality Survey

EMAP

DMATS

USGS

Massachusetts
Bay

USGS

Massachusetts
Bay

Univ. of Wisconsin-Superior, WI
Michigan Dept. Natural Resources
Ohio EPA

Illinois Geological Survey

USEPA-GLNPO

USEPA-ERL-Duluth

Louisianian Province

USEPA Region 9

A.D. Little, 1990

ACE_NED permit file #29-91-00473E

ACE_NED permit file 199102068

ACE_NED permit file 09-89-2777

ACE_NED permit file 09-89-530

ACE.NED permit file 1989-2911

ACE_NED permit file 199101096

ACE_NED permit file 20-87-2002

ACE_NED permit file 20-89-2206

ACE_NED permit file 22-87-927

ACE_NED permit file 23-198902070

ACE_NED permit file 24-87-912

ACE_NED permit file 24-89-1180

ACE_NED permit file 25-81-374

ACE_NED permit file 25-86-1007

ACE_NED permit file 25-86-290E

ACE_NED permit file 25-86-641

ACE_NED permit file Boston Harbor

ACE_NED permit file Bridge marine- Salisbury, MA

ACE_NED permit file CENED-OR (1145-2-303b)

ACE_NED permit file HULL-72-CHA30

ACE_NED permit file Long Wharf Boston

ACE_NED permit file MA DPW Beverly-Salem Bridge

and By-Pass Project
ACE_NED permit file
MA-HULL-81 -180

ACE_NED permit file MA-HULL-84-210
ACE_NED permit file MWRA- Stoney Brook Conduit
ACE_NED permit file Massport Bird Island Flats -

Harborwalk phase III
ACE_NED permit file Navigation Improvement Study
Dredge Material Disposal Plan Supplement to Feasibility

Rep
USACOE.1981
Wong, 1983
USEPA MBDS, 1989
USACOE, 1990b (DAMOS)

Aqua Tech, Melmore, OHEG&G
Bionomics/Aqua Tech Environ. Cnstlt.
Applied Biology, Inc., Decatur, GA
Recra Research, Inc., Tonawanda, NY
USFWS, Columbia, MO - ARCS
Michigan State University

Virginian Province

ACE_NED permit file Navigation
Improvement Study Feasibility
Report and Environmental
Assessment; Mystic RI
ACE_NED permit file Navigation
Improvement Study Dredge
Material Disposal Plan Supplement
to Feasibility Rep
Boehm, 1983
Bajek, 1983

Battelle, 1984; 1987 a, b
Boehm & Farrington, 1984
Boehm etal., 1984
CDM, 1980
Cudmore, 1988
Enseco, 1987a
Enseco, 1987b
GCA Corp., 1982
Gardner et al., 1986
Gardner et al., 1988
Hubbard, 1987

Jason M. Cortell & Assoc., 1982

Jason Cortell, 1990

MA DEQE, 1985

MADEQE, 1986 MA DPW, 1991

MA DEQE, 1982

MacDonald, 1991

NET Atlantic, 1990

Nolan et al., 1981

Penney etal., 1981

Phillips, 1985

Pruell et al., 1989

Ryan et al., 1982

Robinson et al., 1990

Shea et al., 1991

Shiaris et al., 1986

Types of Data Included in the NSI

In addition to sediment chemistry data, tissue residue, benthic abundance, toxicity (solid-phase and elutriate),
histopathology, and fish abundance data have been gathered and included in the NSI, although only the sediment
chemistry, tissue residue, and toxicity data have been evaluated for this report to Congress. The NSI also includes

A-3


-------
\|>|H'ti
-------
Table A-l. Number of Sampling Stations at Which Various Types of Data Were Collected

Data Set

Number of Stations When Measured

Sediment
Chemistry

Tissue
Residue

Benthlc
Abundance

Toxicity

Histopath-
ology

Sediment
Chemistry
and Tissue
Residue

Sediment
Chemistry
and Benthlc
Abundance

Sediment
Chemistry
and Toxicity

Sediment
Chemistry

and
Hlstopath-
ology

Sediment
Chemistry,
Tissue
Residue,
and Toxicity

Sediment
Chemistry,

Beothk
Abundance,
and Toxicity

STORET

12,907

6,057







1,533











Region 4

1,024





















ODES

1317

1,722

2,592

296



37

664

70



2

49

COSED

1,104





















Gulf of
Mexico

210





82







6







Great Lakes

761

26

476

373



26

449

369



26

68

DMATS

213

202



245



169



188



163



Mass. Bay

979





















EMAP
LAProv.
VAProv.

260
200

199

259

212

259
212

259

198

259
202

259
202

259

198

259
202

Seattle
USCOE

2,116



365

876





365

707





270

Total

21,093

8,206

3,904

2,343

259

1,963

1,939

1,801

259

389

848


-------
Ibble A-2. Number of Sampling Stations With Data Included in the NSI

Measurement Parameters

Total Number of
Stations

Stations with Coordinates

Number

% of Total Number
of Stations
^Coordinates*

Sediment Chemistry

21,093

19,546

76

TOC

6,170

5,335

21

AVS

425

371

1

Tissue Residue

8,206

7,208

28

Toxicity

2,343

1,523

6

Elutriate Phase

630

—

, _

SoBd Phase

1,865

—

—

Benthic Abundance

3,904

1,844

7

Histopathotogy

259

259

1

Sediment Chemistry & Tissue

1,963

1,930

8

Sediment Chemistry & Toxicity

1,801

1,263

5

Sediment Chemistry & Abundance

1,939

1,340

5

Sediment Chemistry & Histopathotogy

259

259

1

Sediment Chemistry, Tissue, & Toxicity

389

359

1

Sediment Chemistry, Toxicity, & Abundance

848

733

3

Tola] number of stations with coordinates = 25,555.

evaluation of sediment chemistry data described in Chapter 2). Key changes to the data set from version 1.0
include the following:

•	Inclusion of Regional/state review codes. (See data element NSIREVCD in tables ALLSEDI and ALLTISS.)

•	Resolution of species codes for tissue residue data.

•	Inclusion of biotoxicity control data for EMAP programs.

•	Revised loadings data from Permit Compliance System (PCS) and Toxic Release Inventory (TRI). Facili-
ties with no loadings data are included as a separate table.

•	Inclusion of species information and toxicity phase for purposes of the NSI evaluation methodology.

The remainder of this section contains a listing of the field names and descriptions associated with each
table in the NSI.

A-6


-------
NalitinnI Sediment (,)u;ilitv Survey

Figure A-l. Organization of NSI Data.

A-7


-------
Thble A-3. Data TMbles Available In the NSI

Table Name

Table Description

ALLSTAT.DBF

Station

ALLSBDI.DBF

Sediment chemistry

ALLTISS.DBF

Tissue residue

ALLBIOT.DBF

Biotoxicity

ALLSEDM.DBF

Sediment grain size and miscellaneous sediment chemistry

ALLTISM.DBF

Miscellaneous tissue residue

ALLBLUT.DBF

Elutriate

LOADD.DBF

PCS/TRI loadings

LOADS.DBF

PCS/TRI facilities (have loadings data)

LOADO.DBF

Other PCS/TRI facilities (no associated loadings data)

BIOTCODE.DBF

Toxicity phase for biotoxicity table (ALLBIOT)

ELUTPARM.DBF

list of analytes for elutriate table (ALLELUT)

SEDJ?ARM.DBF

List of analytes for sediment tables (ALLSEDI, ALLSEDM)

TIS_CODE.DBF

List of species for tissue tables (ALLTISS, ALLTISM)

TIS.PARM.DBF

List of analytes for tissue tables (ALLTISS, ALLTISM)

SEACOB.DBF

EPA Region 10/COE Seattle District's Sediment Inventory Code file (important for
interpreting a large number of codes unique to this data source)

REMARK. WP

Text file on remark codes (important for remark codes other than "K" or "U")

ALLSUPR.DBF

Superfund facilities

ALLBENA.DBF

Benthic species abundance

ALLBENC.DBF

Benthic community

ALLHIST.DBF

Histopathology

ALLFISA.DBF

Fish abundance

SPEC-CD.DBF

Species codes for benthic data

FISH-CD.DBF

Species codes for fish abundance data

A-8


-------
Njiiiomil Nt ilinn iH Qu;ilil> Siirtcy

ALLSTAT.DBF

Station

SOURCE
AGENCY

STATION

COUNTY

DEPTH

DEPT.MAX

DEPT_MIN

DREDGESI

DRWATERB

GEOCODE

INSTIT

LAT

LAT_2

LNG

LNG_2

LOCATION

LOC_CODE

NSIREACH

ORIGIN

ORG.NAME

REFER

SR.SCI

STATE

WA1ERBOD

EPA_REG

FIPS

FIPS_DIS
HUCJDIS
RFl.DIS

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National
Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STNJCDII'' II STA-
TION II DATE. DMATS NOTE: STATION = ZD II " IISTATIONIII " II SERIES II' * II
SCAN.)

County

Water depth (m)

Maximum water depth (m)

Minimum water depth (m)

Dredged site
Dredged water body
Geologic code
Institution

Latitude (decimal degrees)

Latitude #2 forming a rectangle (decimal degrees)

Longitude (decimal degrees)

Longitude #2 forming a rectangle (decimal degrees)

Location

Location code

Reach File 1 reach

Origin

Organization name
Reference, literature citation
Senior scientist
State

Waterbody
EPA Region
FIPS code

Distance to nearest FIPS (mile)

Distance to nearest catologic unit (mile)

Distance to RF1 reach (mile)

ALLSEDI.DBF

SOURCE
AGENCY

STATION

DATE

SAMPLE

SUBSAMPL

REPLICAT

SEQ

CAS

CLEANUP

COMMENTS
DRY.WGT

Sediment chemistry

Identification of data origin (e.g,, REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National
Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STN_CD II " II STA-
TION II DATE. DMATS NOTE: STATION = ID II " II STATIONI li'' II SERIES II " II
SCAN.)

Date of sample collection
Unique sample identifier code
Unique subsample identifier code
Unique replicate identifier code

Computer-generated sequence number when multiple samples were taken; SOURCE,
AGENCY, STATION, and DATE were identical; and no SAMPLE, SUBSAMPL, or
REPLICAT codes were provided
CAS number for analyte

Sample cleanup code to indicate an additional step taken to further purify the sample

extracts or digestates

Comments

Percent of total sample remaining after drying

A-9


-------
\|>|K'ii
-------
Niitioiv.il Sediment Quality Survey

SAMPTYPE

Sample type

SEX

Sex code used to identify sex of sample

SMP_EQP

Sampling equipment code

SPECCODE

Species code

SPECIMEN

Unique identifier for the individual organism being analyzed

TOT_REP

Number of replicates

WEIGHT

Weight of organism

WET_WGT

Total weight of sample

LIPIDS

% Extractable lipids

SPEC_BIO

STORET taxonomic code

ALLBIOT.DBF

Biotoxicity

SOURCE
AGENCY

STATION

DATE
SAMPLE
REPLICAT
SEQ

AMMONIA

ABNORMAL

BIOASS_DA

BIOASSAY

BIOMASS

COMMENTS

COM_NAME

DIL_UNIT

DILUTION

DOX

ENDPOIN2

ENDPOINT

E_QUALIF

EMERGENC

EXT_MTHO

FEEDING

FLUSH

GENUS

HARDNESS

HOLD_TIM

LFSTG_EN

LFSTG_ST

MEASURED
NAME
NUM_ORGA
P

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STNjCD II ' ' II STA-
TION II DATE. DMATS NOTE: STATION = ID II " IISTATIONIII " II SERIES II " II
SCAN.)

Date of sample collection
Unique sample identifier code
Unique replicate identifier code

Computer-generated sequence number when multiple samples were taken; SOURCE,

AGENCY, STATION, and DATE were identical; and no SAMPLE, SUBSAMPL, or

REPLICAT codes were provided

Ammonia concentration (mg/L)

Abnormality

Bioassay date

Type of bioassay reported

Biomass

Comments

Common name

Concentration/Dilution units

Concentration/Dilution

Dissolved oxygen (mL/L)

Endpoint #2 of bioassay test

Endpoint of bioassay test

EMERGENC qualifier

Emergence after 10 days

Extraction method code to indicate the method used to extract or digest the sample matrix
and remove or isolate the chemical of concern
Feeding of species tested

Flushing rate in percent of chamber volume exchanged/24 hours

Organism genus

Hardness

Holding time of sample prior to analysis (weeks)

Life stage end—for bioassays that span more than one life stage, record predominant life
stage at the end of the bioassay

Life stage start—for bioassays that span more than one life stage, record predominant life
stage at the start of the bioassay
Measured (Y/N)

Genus and species name (linked to PHASE)

Number of organisms

Result associated with ENDPOINT

A-ll


-------
\|>|H>ii
-------
DRY.WGT

Percent of total sample remaining after drying

EXT_MTHO

Extraction method code to indicate the method used to extract or digest the sample matrix



and remove or isolate the chemical of concern

FINE_MTH

Method of analysis for analysis of fine particles. Left blank if sample was not split into



fractions.

INSTRUME

Instrument code to identify the final chemical analysis method(s) used for analyzing the



sample

meas_bas

Result is wet or dry weight basis (see also P)

P

Result associated with PARM

PARM

Analyte measured (see also P and R)

PHI_B

Phi boundaries in phi units, between the coarse and fine fractions

PHI_MAX

Phi boundary maximum at the fine end of the analyzed range

PHI_MIN

Phi boundary minimum at the coarse end of the analyzed range

R

Remark code associated with PARM and P

SAMP_DTL

Depth to bottom of sample interval (m)

SAMP_DTU

Depth to top of sample interval (m)

SMP.EQP

Sampling equipment code

SPHERE

Sphere (i.e., environment) code from which the sample came

TOT.WGT

Total weight of sample (g)

UNITS

Units associated with PARM, P, and R

WET_WGT

Total wet weight of sample (g)

P_ALP

Nonnumeric result associated with PARM

ALLTISM.DBF

Miscellaneous tissue residue

SOURCE
AGENCY

STATION

DATE

SAMPLE

SEQ

REPLICAT
ANAT_CD
CAS

CLEANUP

COMPOSIT

DRY_WGT
EXT_MTHO

INSTRUME

LENGTH

LIPIDS

LIFE_STA

MEAS_BAS

NUMB_IND

P

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STN_CD II " II STA-
TION II DATE. DMATS NOTE: STATION = ID II " IISTATIONIII " II SERIES II " II
SCAN.)

Date of sample collection
Unique sample identifier code

Computer-generated sequence number when multiple samples were taken; SOURCE,

AGENCY, STATION, and DATE were identical; and no SAMPLE, SUBSAMPL, or

REPLICAT codes were provided

Unique replicate identifier code

Organ/tissue sampled code

CAS number for analyte

Sample cleanup code to indicate an additional step taken to further purify the sample
extracts or digestates

A unique identifier to indicate a sample created by compositing tissues from several
individuals.

Percent of total sample remaining after drying

Extraction method code to indicate the method used to extract or digest the sample matrix
and remove or isolate the chemical of concern

Instrument code to identify the final chemical analysis method(s) used for analyzing the
sample

Length of specimen
Lipids (%)

Life stage code to identify the life stage of sample
Result is wet or dry weight basis (see also P)

Number of organisms in sample
Result associated with PARM

A-13


-------
Appendix A

PARM

Analyte measured (see also P and R)

R

Remark code associated with PARM and P

SEX

Sex code used to identify sex of sample

SMPJEQP

Sampling equipment code

SPECCODE

Species code

SPEC_SCI

Species scientific name

SPECIMEN

Unique identifier for the individual organism being analyzed

UNITS

Units associated with PARM, P, and R

WET_WGT

Total weight of sample

P_ALP

Nonnumeric result associated with PARM

ALLELUTJ3BF

SOURCE
AGENCY

STATION

DATE

SAMPLE

SEQ

SUBSAMPL

REPLICAT

CAS

EXT.MTHO
INSTRUME
P

PARM
R

SAMP_DTL
SAMP_DTU
SAMP_EQP

Elutriate

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STNjCD II ' ' II STA-
TION II DATE. DMATS NOTE: STATION = ID II " IISTATIONIII " II SERIES II " II
SCAN.)

Date of sample collection
Unique sample identifier code

Computer-generated sequence number when multiple samples were taken; SOURCE,

AGENCY, STATION, and DATE were identical; and no SAMPLE, SUBSAMPL, or

REPLICAT codes were provided.

Unique subsample identifier code

Unique replicate identifier code

CAS number for analyte

Extraction method code to indicate the method used to extract or digest the sample matrix
and remove or isolate the chemical of concern

Instrument code to identify the final chemical analysis method(s) used for analyzing the
sample

Result associated with PARM (|ig/L)

Analyte measured (see also P and R)

Remark code associated with PARM and P
Depth to bottom of sample interval (m)

Depth to top of sample interval (m)

Sampling equipment code

LOADD.DBF

PCS/TRI loadings

ID

Facility identification number

CAS

CAS number for analyte

CHEMICAL

Analyte name

SIC

SIC code for facility

E3KGY0

PCS loadings using below detection limit (dl) equal to 0.0 assumption

E3KGYE

PCS loadings using below detection limit equal to 0.5-dl assumption

E3KGY1

PCS loadings using below detection limit equal to dl assumption

E3FLOO

PCS flow using below detection limit equal to 0.0 assumption

E3FLOE

PCS flow using below detection limit equal to 0.5-dl assumption

E3FL01

PCS flow using below detection limit equal to dl assumption

E6KGYE

TRIPOTW transfers

E6KGY75

75 percent of TRI POTW transfers

A-14


-------
LOADS.DBF

PCS/TRI facilities (have loadings data)

ID	Facility identification number

CODE	"PCS" or "TRI"

SPC	State postal code

LAT	Latitude (decimal degrees)

LNG	Longitude (decimal degrees)

NSIREACH	Reach File 1 Reach

LOADO.DBF	Other PCS/TRI facilities (no associated loadings data)

ID	Facility identification number

SPC	State postal code

LAT	Latitude (decimal degrees)

LNG	Longitude (decimal degrees)

NSIREACH	Reach File 1 Reach

BIOTCODE.DBF	Toxicity phase for biotoxicity table (ALLBIOT)

NAME	Genus and species name

PHASE	Toxicity phase listed in source of data (when available)

SOURCE	Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

NSIPHASE	Toxicity phase used by NSI

ELUTPARM.DBF List of analytes for elutriate table (ALLELUT)

SOURCE
PARM
CAS
LNAME

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)
Analyte measured (see also P and R)

CAS number for analyte
Analyte long name

SED_PARM.DBF List of analytes for sediment tables (ALLSEDI, ALLSEDM)

SOURCE
PARM
CAS
LNAME

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)
Analyte measured (see also P and R)

CAS number for analyte
Analyte long name

TIS_CODE.DBF List of species for tissue tables (ALLTISS, ALLTISM)

SPECCODE

SPEC_SCI

SPEC_COM

RES_MIG

BOT.PEL

EDIBLE

Species code

Species scientific name

Species common name

Species resident, migratory, or either

Species benthic, pelagic, or either

Species considered edible by humans

TIS_PARM.DBF List of analytes for tissue tables (ALLTISS, ALLTISM)

SOURCE	Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

PARM	Analyte measured (see also P and R)

A-15


-------
CAS
LNAME

CAS number for analyte
Analyte long name

ALLSUPR.DBF

STATE
ID

NAME

COUNTY

CNTYJPIP

C0305

C0326

LAT

LNG

NSIREACH

Superfund facilities

State postal code

Superfund identification

Facility name

County name

3-digit county FIPS code

C0305

C0326

Latitude (decimal degrees)
Longitude (decimal degrees)
Reach FUe 1 Reach

ALLBENA.DBF Benthic species abundance

SOURCE
AGENCY

STATION

DATE

SAMPLE

REPLICAT

BOTTOM

AREAJ3AS

COMMJBAS

EXT_MTHO

GENUS

MESH_SZ

NJREP

NUMBJND

NUMB„SPE

ORDER

P

PARM
P_MEAN
P_STD
R

SAMP_DTL

SAMP_DTU

SPECIES

SPECCODE

UNITS

Identification of data origin (e.g.» REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STNjCD II' ' II STA-
TION II DATE. DMATS NOTE: STATION = ID II " IISTATIONIII»* II SERIES II " II
SCAN.)

Date of sample collection
Unique sample identifier code
Unique replicate identifier code
Bottom type

Area basis for reported data

Basis for community abundance measurements

Extraction method code to indicate the method used to extract or digest the sample matrix

and remove or isolate the chemical of concern

Organism genus

Seive mesh size

Number of replicate samples

Total number of individuals

Total number of unique species

Organism order

Result associated with PARM

Analyte measured (see also P and R)

MeanP

Standard deviation of P

Remark code associated with P and PARM

Depth to bottom of sample interval (m)

Depth to top of sample interval (m)

Organism species

Species code

Units associated with PARM, P, and R

ALLBENC.DBF Benthic community

SOURCE	Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

A-16


-------
AGENCY
STATION

DATE

SAMPLE

AMPHIPOD

AMPHMABN

AREA_BAS

ARTHROPO

BIOM_TOT

BIOMMEAN

BIV_MABN

BSPINDEX

BSP_GRAB

BSP_MABN

BSP_MDIV

BSP.MEAN

BSP_MEXP

BSP_TABN

BSP_TDIV

BSP.TOT

CAPIMABN

COMM.BAS

CRUSTACE

DECAMABN

DOMINANC

ECHINODE

EVENESS

ITI

MEDJDIAM

MISC.TAX

MOIST.M

MOLLUSCS

NEMATODE

OLIGOCHA

PABN_AMP

PABN_BIV

PABN_GAS

PABN.TUB

PLYC_MWT

PLYCMABN

P_SENSIT

P_TOLERA

POLYCHAE

QUARDVTM

Ql.PHI

Q3_PHI

RPDDEP.M

SICL_B_M

SKEWNESS

TUBIMABN

I

i	N:iXiofi;il ScrJiiueuf Qaitliir Siu voy

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National
Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STN_CD II " II STA-
TION II DATE. DMATS NOTE: STATION = JD II " IISTATIONIII " II SERIES II' * II
SCAN.)

Date of sample collection

Unique sample identifier code

Number of amphipod

Mean abundance of amphipods

Area basis for reported data

Number of arthropods in the sample

Total biomass (g)

Mean biomass per grab (g)

Mean abundance of bivalves (g)

Benthic species index

Number of pabs

Mean abundance per grab

Mean Shannon-Wiener diversity index

Mean number of species per grab

Expected mean number of species

Total abundance

Pooled Shannon-Wiener diversity index

Total number of species

Mean abundance of eapitellids

Basis for community abundance measurements

Number of crustaceans in the sample

Mean abundance of decapods

Numeric dominance in the sample

Number of echinoderms in the sample

Eveness

ITI

50% quartile diameter (phi)

Number of miscellaneous taxa in sample

Sediment moisture content (%)

Number of molluscs in the sample

Number of nematodes in the sample

Number of oligochaetes in the sample

Percent abundance amphipods

Percent abundance bivalves

Percent abundance gastropods

Percent abundance tubificids

Mean biomass per polychaete (g)

Mean abundance of polychaetes

Abundance of pollution sensitive organisms (%)

Abundance of pollution tolerant organisms (%)

Number of polychaetes in the sample

Phi quartile deviation

25% quartile diameter (phi)

75% quartile diameter (phi)

Mean RPD in mm

Mean silt/clay content (%)

Phi quartile skewness

Mean abundance of tubificids

A-17


-------
ALLfflSTJDBF

SOURCE
AGENCY

STATION

DATE

BODYPATH

BRNCPATH

BUCCPATH

FSP_ABN

FSP_TOT

MNMDTRSH

Histopathology

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STN_CD II " II STA-
TION II DATE. DMATS NOTE: STATION = ID II' * IISTATIONIII " II SERIES II4 ' II
SCAN.)

Date of sample collection

Number of fish with body pathologies

Number of fish with branchial pathologies

Number of fish with buccal pathologies

Abundance (number/trawl)

Number of species

Manmade trash (Y/N)

ALLFISA.DBF

SOURCE
AGENCY

STATION

DATE

LEN_MEAN

LEN_STD

P

PARM

SPECCODE

UNITS

Fish abundance

Identification of data origin (e.g., REG4 is the Region 4 Pilot Study)

Identification of group responsible for collecting data (e.g., NS&T is NOAA's National

Status and Trends Program)

Monitoring station identification code. (ODES NOTE: STATION = STNjCD II " II STA-
TION II DATE. DMATS NOTE: STATION = ID II " II STATIONI II " II SERIES II ' ' II
SCAN.)

Date of sample collection
Mean length (in)

Standard deviation length (in)

Result associated with PARM
Analyte measured (see also P)

Species code

Units associated with PARM and P

SPEC-CD JDBF

SPECCODE
SPEC_SCI
SPEC COM

Species codes for benthic data

Species code
Species scientific name
Species common name

FISH-CD.DBF	Species codes for fish abundance data

SPECCODE	Species code

SPECJSCI	Species scientific name

SPEC_.COM	Species common name

A-18


-------
National Sediment Quality Sur* cv

Appendix B

Description of Evaluation
Parameters Used in the NSI
Data Evaluation

Chapter 2 of this document presented the methodology used in the evaluation of the NSI
data. This appendix describes in greater detail the screening values and other parameters
used in the NSI data evaluation. The actual parameter values used are presented in Appendix D. For the
purpose of discussion, the sediment evaluation parameters have been placed into three groups: (1) those used to assess
potential impacts on aquatic life, (2) those used to assess potential impacts on human health, and (3) those used to assess
potential impacts on wildlife. The uncertainties associated with the use of these parameters in the NSI data evaluation arc
discussed in Chapter 5.

Aquatic Life Assessments

To evaluate the potential threat to aquatic life from chemical contaminants detected in sediments, measured concen-
trations of contaminants were compared to sediment chemistry screening levels. The results of toxicity tests to indicate
the actual toxicity of sediment samples to species of aquatic organisms, when available, were also evaluated for the NSI.

Sediment chemistry screening levels are reference values above which sediment contaminant concentrations could
pose a significant threat to aquatic life. Several different approaches, based on causal or empirical correlative method-
ologies, have been developed for deriving screening levels of sediment contaminants. Each of these approaches
attempts to predict contaminant concentration levels that could result in adverse effects to benthic species, which are
extrapolated to represent the entire aquatic community for this evaluation. For the purpose of this analysis, the
screening levels selected include the following:

•	EPA's draft sediment quality criteria (SQCs) for five nonionic organic chemicals, developed using an equilib-
rium partitioning approach (USEPA, 1992a, 1993a).

•	Sediment quality advisory levels (SQALs) for selected nonionic organic chemicals, developed using an
equilibrium partitioning approach (USEPA, 1992a, 1993a).

•	The sum of simultaneously extracted divalent transition metals concentrations minus the acid-volatile sulfide
concentration ([SEM] - [AVS]), also based on an equilibrium partitioning approach.

•	Effects range-median (ERM) and effects range-low (ERL) values for selected nonionic organics and metals
developed by Long et al. (1995).

•	Probable effects levels (PELs) and threshold effects levels (TELs) for selected nonionic organics and metals
developed for the Florida Department of Environmental Protection (FDEP, 1994).

•	Apparent effects thresholds (AETs) for selected organics and metals developed by Barrick et al. (1988).

The principles behind the development of each of these sediment chemistry screening values are discussed below.
The sediment toxicity tests are also briefly described in this section.

. B-l


-------
\|>|H'it
-------
anoxic sediments. However, models that can use these factors to predict the bioavailability of trace metals in sediments
are not fully developed (see below). Mechanisms that control the partitioning of nonionic and nonpolar organic
compounds with log Kows of less than 2.0 or greater than 5.5 and polar organic compounds in sediments, and affect their
toxicity to benthic organisms, are less well understood. Models for predicting biological effects from concentrations of
such compounds have not yet been developed; therefore, these chemicals have not been evaluated using equilibrium
partitioning approaches.

Draft Sediment Quality Criteria

The equilibrium partitioning model was selected for the development of sediment quality criteria because it can be
applied to predict sediment contaminant concentrations below which biological effects are not expected to occur based
on the toxicity of individual nonionic organic chemicals—and hence can protect benthic aquatic life in bedded, perma-
nently inundated, or intertidal sediments—while accounting for sediment characteristics that affect the bioavailability of
the chemical (Di Toro etal., 1991; USEPA, 1993a). The predominant phase for sorption of nonionic organic chemicals to
sediment particles appears to be organic carbon, for sediments in which the fraction of organic carbon (f^) is greater than
0.2 percent.

The partitioning of a chemical between the interstitial water and sediment organic carbon is explained by the
sediment/pore water partition coefficient for a chemical, Kp, which is equal to the organic carbon content of the sediment
(f^.) multiplied by the sediment particle organic carbon partition coefficient (K^). Kp is the ratio of the concentration of
the chemical in the sediment to the concentration of the chemical in the pore water. Normalizing the dry-weight
concentration of the chemical in sediment to organic carbon is as appropriate as using the interstitial water concentra-
tion of the chemical because organic carbon in the sediment can also bind the chemical and affect its bioavailability and
toxicity. The particle organic carbon partition coefficient (KJ is related to the chemical's octanol/water partition
coefficient (Kow) by the following equation (Di Toro et al., 1991):

logK^ = 0.00028 + 0.983(logKow)

Hie octanol/water partition coefficient for each chemical can thus predict the likelihood of the chemical to complex
or sorb to organic carbon, when measured with modern experimental techniques that provide the most accurate estimate
of this parameter. The concentration of the chemical on sediment particles (Cs) is then equal to the dissolved concentra-
tion of chemical (Cd) multiplied by the organic carbon content of the sediment (f^) and the particle organic carbon
partition coefficient (K^.), when f^ is greater than 0.2 percent (USEPA, 1993a), thus normalizing the dry-weight sediment
concentration of the chemical to the organic carbon content of the sediment.

The criterion for the dissolved concentration of chemical (Ct) is derived from the final chronic value (FCV) of EPA's
water quality criteria (USEPA, 1985). Freshwater and saltwater FCVs are based on the results of acceptable laboratory
tests conducted to determine the toxicity of a chemical in water to a variety of species of aquatic organisms, and they
represent the highest levels of a chemical to which organisms can be exposed without producing toxic effects. This level
is predicted to protect approximately 95 percent of aquatic life under certain conditions. An evaluation of data from the
water quality criteria documents and benthic colonization experiments demonstrated that benthic species have chemical
sensitivities similar to those of water column species (Di Toro et al., 1991). Thus, if the concentration of a chemical in
sediment, measured with respect to the sediment organic carbon content, does not exceed the sediment quality criterion,
then no adverse biological effects from that chemical would be expected (USEPA, 1992a, 1993a).

EPA has developed and published draft freshwater sediment quality criteria (SQCs) for the protection of aquatic life
for five contaminants: acenaphthene, dieldrin, endrin, fluoranthene, and phenanthrene. These draft SQCs are based on
the equilibrium partitioning approach (USEPA 1993b, c, d, e, f) using the aquatic life water quality criterion final chronic
value (FCV, in jig/L) and the partition coefficient between sediment and pore water (Kp, in L/g sediment) for the chemical

B-3


-------
Appendiv B

chemical of interest (Di Toro et al., 1991; USEPA, 1993a). Thus, SQC = Kp FCV. On a sediment organic carbon basis,
the sediment quality criterion, SQC^., is:

SQC,, (m I gx) = FCV(jig / L) X (L / kg) % dO"3 kgoc / goc)

where:

FCV	= EPA aquatic life water quality criterion final chronic value and

Koc	= organic carbon-water partitioning coefficient.

is presumed to be independent of sediment type for nonionic organic chemicals, so that the SQC^ is also
independent of sediment type. Using a site-specific organic carbon fraction, f^ (g^/g sediment), the SQC^, can be
expressed as a sediment-specific value, the SQC:

SQC = (SQCoc)(fK)

Sediment Quality Advisory Levels

EPA intends to develop sediment quality criteria for additional chemicals in the future. In the interim, EPA's
Office of Science and Technology developed equilibrium partitioning-based sediment quality advisory levels (SQALs)
using the following equation:

SQAL^g / g„) = [FCV, SCV^g/ L)1 % K„(L / kg) % (10"3kgoc /gx)

where:

SQAL^ = calculated sediment quality advisory level;

FCV, SCV = EPA aquatic life chronic criterion (final chronic value, FCV), or other chronic threshold
water concentration (secondary chronic value, SCV); and
= organic carbon-water partitioning coefficient.

As noted in Chapter 2, EPA has proposed sediment quality criteria (SQCs) for five chemicals based on the highest
quality toxicity and octanol/water partitioning (Kow) data, which have been reviewed extensively. This section de-
scribes the sources of data used to calculate the values used in the SQAL equations: log Kows (used to derive K^s) and
chronic threshold water concentrations. A detailed description of the methods and data used to develop SQALs for
specific chemicals using the equilibrium partitioning approach will be published by EPA as a separate document.

SQALs for use in the NSI data evaluation were developed in conjunction with other programs at EPA (established
under the Resource Conservation and Recovery Act, RCR A, and the Superfund Amendments and Authorization Act,
SARA) to provide the same values for conducting screening-level evaluations of sediment toxicity for these programs.
The SQALs (as well as the other sediment chemistry threshold levels) are meant to be used/or screening purposes only.
The screening values are not regulatory criteria, site-specific cleanup standards, or remediation goals. The screening
levels are set to be appropriately conservative, so samples that do not exceed the screen would not be expected to
exhibit adverse effects from the action of the specific chemical evaluated; exceeding the screening levels does not
indicate the level or type of risk at a particular site, but can be used to target additional investigations. EPA's Office of
Research and Development (ORD), including staff from Environmental Research Laboratory, Athens, Georgia; Envi-
ronmental Research Laboratory, Duluth, Minnesota; and Environmental Research Laboratory, Narragansett, Rhode
Island, provided guidance and assisted in the development of the necessary values. The SQALs used for the NSI data
evaluation are presented with other screening values in Table D-l of Appendix D.

Method for Determination of Log Kob$. Log Kow values were initially identified in summary texts on physical-
chemical properties, such as Howard (1990) and Mackay et al. (1992a, b) and accompanying volumes. Additional
compendia of log Kow values were also evaluated, including De Kock and Lord (1987), Doucette and Andren (1988),
Klein et al. (1988), De Bruijn etal. (1989), Isnard and Lambert (1989), Leo (1993), Noble (1993), and Stephan (1993).
To supplement these sources, on-line database searches were conducted in ChemFate, TOXLINE, and Hazardous
Substances Data Bank (HSDB) (National Library of Medicine); Internet databases such as CARL UNCOVER; and

B-4


-------
EPA databases such as ASTER, OLS, and the ORD BBS, Original references were identified for the values, and
additional values were identified. In cases where log Kow values varied over several orders of magnitude or measured
values could not be identified, detailed on-line searches were conducted using TOXLIT, Chemical Abstracts, and
DIALOG. Values identified from all of these sources and the method used to obtain each log Kow value were compiled
for each chemical. A few chemicals lacked experimentally measured log Kows, and no log Kow data were available from
any source for butachlor, DCPA/Dacthal, and Ethion/B laden.

The determination of Kow values was based on experimental measurements taken primarily by the slow-stir, gen-
erator-column, and shake-flask methodologies. The SPARC Properties Calculator model was also used to generate
Kow values, when appropriate, for comparison with the measured values. Values that appeared to be considerably
different from the rest were considered to be outliers and were not used in the calculation.

For each chemical, the available value based on one of these methods was given preference. If more than one such
value was available, the log Kow value was calculated as the arithmetic mean of those values (USEPA, 1994), Recom-
mended log Kows were finalized by ORD-Athens based on recommended criteria, and the justification for selection of
each value was included in the report (Karickhoff and Long, April 10,1995, report).

Selection of Chronic Toxicity Values. A hierarchy of sources for chronic toxicity values to develop the SQ ALs was
prepared. The following sources were identified and ranked from most to least confidence in the chronic values to be used:

1.	Sediment quality criteria (SQCs).

2.	Final chronic values from the Great Lakes Initiative (USEPA, 1995c),.

3.	Final chronic values from the National Ambient Water Quality Criteria documents.

4.	Final chronic values from freshwater criteria documents.

5.	Final chronic values developed from data in EPA's Aquatic Toxicity Information Retrieval database (AQUIRE)
and other sources.

6a. Secondary chronic values developed from data in AQUIRE and other sources.

6b. Secondary chronic values from Suter and Mabrey (1994)

EPA SQCs were available for five chemicals: acenapthene, dieldrin, endrin, fluoranthene, and phenanthrene. There
were no final chronic values (FCVs) obtained by the aquatic life criteria methodology (referred to as "Her I") de-
scribed in USEPA (1995c) available for the remaining chemicals in the NSI. Two SQALs were based on the FCVs
from National Ambient Water Quality Criteria documents, for gamma-BHC/Lindane and toxaphene. No FCVs were
available from criteria documents.

Thirteen SQALs were based on work conducted by Oak Ridge National Laboratories (Suter and Mabrey, 1994)
using the USEPA (1995c) methodology for obtaining secondary chronic values ("Tier II"). This methodology was
developed to obtain whole-effluent toxicity screening values based on all available data, but the SCVs could also be
calculated with fewer toxicity data than are required for the criteria methodology. The SCVs are generally more
conservative than those which can be produced by the FCV methodology, reflecting greater uncertainty in the absence
of additional toxicity data. The minimum requirement for deriving an SCV is toxicity data from a single taxonomic
family (Daphnidae), provided the data are acceptable. Only those values from Suter and Mabrey (1994) that included
at least one daphnid test result in the calculation of the SCV were included for the NSI. SCVs from Suter and Mabrey
(1994) were used to develop SQALs for the following chemicals:

benzene	napthalene

chlorobenzene	1,1,2,2-tetrachloroethane

delta-BHC	tetrachloroethene

dibenzofuran	, toluene

diethyl phthalate	1,1,1-trichloroethane

di-n-butyl phthalate
ethylbenzene

trichloroethene

B-S.


-------
A preliminary search of data records in EPA's AQUIRE database indicated that the following chemicals might
have sufficient toxicity data for the development of SCVs:

Insufficient toxicity test data were found in AQUIRE for acenapthylene, endosulfan sulfate, heptachlor epoxide,
and trichlorofluoromethane. In addition, review of AQUIRE data records indicated that no daphnid acute toxicity
tests had been conducted for hexachlorobutadiene. These chemicals were dropped from further development of
SQALs.

Acid-Volatile Sulfide Concentration

The use of the total concentration of a trace metal in sediment as a measure of its toxicity and its ability to
bioaccumulate is not supported by field and laboratory studies because different sediments exhibit different degrees
of bioavailability for the same total quantity of metal (Di Toro et al., 1990; Luoma, 1983). These differences have
been reconciled by relating organism toxic response (mortality) to the metal concentration in the sediment pore water
(Adams et al., 1985; Di Toro et al., 1990). Metals form insoluble complexes with the reactive pool of solid-phase
sulfides in sediments (iron and manganese sulfides), restricting their bioavailability. The metals that can bind to these
sulfides have sulfide solubility parameters smaller than those of iron sulfide and include nickel, zinc, cadmium, lead,
copper, and mercury. Acid-volatile sulfide (AVS) is one of the major chemical components that control the activities
and availability of metals in the pore waters of anoxic sediments (Meyer et al., 1994).

AVS is operationally defined as the sulfide liberated from a sediment sample to which hydrochloric acid has been
added at room temperature under anoxic conditions (Meyer et al., 1994). The metals concentrations that are extracted
during the same analysis are termed the simultaneously extracted metals (SEM). SEM is operationally defined as
those metals which form less soluble sulfides than do iron or manganese (i.e., the solubility products of these sulfides
are lower than that of iron or manganese sulfide) and that are at least partially soluble under the same test conditions
in which the AVS content of the sediment is determined (Allen et al., 1993; Di Toro et al., 1992; Meyer et al., 1994).

Laboratory studies using spiked sediments and field-collected metal-contaminated sediments demonstrated that
when the molar ratio of SEM to AVS [SEM]/[AVS] was less than 1 (excess AVS remained), no acute toxici ty (mortal-
ity greater than 50 percent) was observed in any sediment for any benthic test organism. When [SEM]/[AVS] was
greater than 1 (excess metal remained), the mortality of sensitive species (e.g., amphipods) increased in the range of
1.5 to 2.5 |imol of SEM per |imol AVS (Casas and Crecelius, 1994; Di Toro et al., 1992).

Experimental studies indicate that the lower limit of applicability for AVS is approximately 1 pmol AVS/g sedi-
ment and possibly lower; other sorption phases, such as organic carbon, probably become important for sediments
with smaller AVS concentrations and for metals with large partition coefficients and large chronic water quality
criteria (Di Toro et al., 1990), In addition, studies indicate that copper, as well as mercury, might be associated with
another phase in sediments, such as organic carbon, and AVS alone might not be the appropriate partitioning phase
for predicting its toxicity. Pore-water concentrations of metals should also be evaluated (Allen et al., 1993; Ankley et
al., 1993; Casas and Crecelius, 1994), However, the AVS approach can be used to predict when a sediment contami-
nated with metals is not acutely toxic (Ankley et al., 1993; Di Toro et al., 1992).

There are several important factors to consider in interpreting the [SEMHAVS] difference. First, all toxic SEMs
present in amounts that contribute significantly to the [SEM] sum should be measured. However, because mercury
presents special problems, it is not included in the current SEM analysis. Second, if the AVS content of sediment is

biphenyl

4-bromophenyl phenyl ether
butyl benzyl phthalate
diazinon

1.2-dichlorobenzene

1.3-dichlorobenzene

1.4-dichlorobenzene
endosulfan mixed isomers
alpha-endosulfan
beta-endosulfan

fluorene

hexachlorethane

malathion

methoxychlor

pentachlorobenzene

tetrachl oro methane

tribromomethane

1,2,4-trichlorobenzene

trichloromethane

m-xylene

B-6


-------
(Adams et al., 1992; Zhuang et al., 1994). Most benthic macroorganisms, including those used in toxicity tests, survive
in sediments that have a thin oxidized surface layer and then an anoxic layer. The anoxic layer can have significant AVS
concentrations that would reduce the metal activity to which these organisms are exposed (Di Toro et al., 1992). Third,
AVS varies spatially in sediment—vertically with depth and horizontally where patches of an appropriate carbon source
occur under low oxygen conditions for the sulfate-reducing bacteria. Lastly, AVS can vary when sediments are oxgenated
during physical disturbance and seasonally as changes in the productivity of the aquatic ecosystem alter the oxidation
state of sediment and oxidize metal sulfides; therefore, the toxicity of the metals present in the sediment also changes
over time (Howard and Evans, 1993).

Selection of an [SEM] - [AVS] difference sufficiently high to place a sediment in the Tier 1 classification requires
careful consideration because the relationship between organism response and the [SEM] - [AVS] difference of sediment
depends on the amount and kinds of other binding phases present. Using freshwater and saltwater sediment amphipod
toxicity data, researchers at EPA's Environmental Research Laboratory in Narragansett, Rhode Island, plotted [SEM] -
[AVS] versus the percentage of sediments with a higher [SEM] - [AVS] value that were toxic. For this analysis, the
researchers defined toxicity as greater than 24 percent mortality. Analysis of these data reveals that between 80 percent
and 90 percent of the sediments were toxic at [SEM] - [AVS]=5. The running average mortality at this level was between
44 percent and 62 percent (Hansen, 1995), EPA's Office of Science and Technology selected [SEM] - [AVS] = 5 as the
demarcation line between the higher (Tier 1) and intermediate (Tier 2) probability categories.

Biological Effects Correlation Approaches

Biological effects correlation approaches are based on the evaluation of paired field and laboratory data to relate
incidence of adverse biological effects to the dry-weight sediment concentration of a specific chemical at a particular
site. Researchers use these data sets to identify level-of-concern chemical concentrations based on the probability of
observing adverse effects. Exceedance of the identified level-of-concern concentrations is associated with a likelihood
of adverse organism response, but it does not demonstrate that a particular chemical is solely responsible. Conse-
quently, correlative approaches do not indicate direct cause-and-effect relationships. In fact, a given site typically
contains a mixture of chemicals that contribute to observed adverse effects to some degree. These and other potentially
mitigating factors tend to make screening values based on correlative approaches lower than screening values based on
effects caused by a single chemical. However, correlative procedures differ from one another by design and, subse-
quently, in how they relate to sediment toxicity. For example, ERMs are levels usually associated with adverse effecs,
whereas AETs are levels intended to always be associated with adverse effects. Thus, when in error, ERMs minimize
false negatives relative to AETs and AETs minimize false positives relative to ERMs (Ingersoll et al., 1996).

Effects Range-Medians and Effects Range-Lows

The effects range approach for deriving sediment quality guidelines involves matching dry-weight sediment con-
taminant concentrations with associated biological effects data. Long and Morgan (1990) originally developed informal
guidelines using this approach for evaluation of NOAA's National Status and Trends (NS&T) data. Data from equilib-
rium partitioning modeling, laboratory, and field studies conducted throughout North America were used to determine
the concentration ranges that are rarely, sometimes, and usually associated with toxicity for marine and estuarine
sediments (Long et al., 1995). Effects range-low (ERL) and effects range-median (ERM) values were derived by Long et
al. (1995) for 28 chemicals or classes of chemicals: .9 trace metals, total PCBs, 13 individual polynuclear aromatic
hydrocarbons (PAHs), 3 classes of PAHs (total low molecular weight, total high molecular weight, and total PAH), and
2 pesticides (p,p'-DDE and total DDT), For each chemical, sediment concentration data with incidence of observed
adverse biological effects were identified and ordered. The authors identified the lower lOth-percentile concentration as
the ERL and the 50th-percentile concentration as the ERM. In terms of potential biological effects, sediment contami-
nant concentrations below the ERL are defined as in the "minimal-effects range," values between the ERL and ERM are
in the "possible-effects range," and values above the ERM are in the "probable-effects range." Data entered into this
biological effects database for sediments (BEDS) were expressed on a dry-weight basis.

The accuracy of these guidelines was evaluated using the data in the database not associated with adverse effects
and noting whether the incidence of effects was less than 25 percent in the minimal-effects range, increased consistently
with increasing chemical concentrations, and was greater than 75 percent in the probable-effects range. Long et al.

B-7


-------
Appendix 15

tently with increasing chemical concentrations, and was greater than 75 percent in the probable-effects range. Long et
al. (1995) reported that these sediment quality guidelines were most accurate for copper, lead, silver, and all classes of
PAHs and most of the individual FAHs; however, accuracy was low for nickel, chromium, mercury, total PCBs, and
DDE and DDT. The guidelines generally agreed within factors of 2 to 3 with other guidelines, including the freshwater
effects-based criteria from Ontario. The authors attributed variability in the concentrations associated with effects to
differences in sensitivities of different taxa and physical factors that affect bioavailability, but they argued that because
of the synergistic effects of multiple toxicants, the inclusion of data from many field studies in which mixtures of
chemicals were present in sediments could make the guidelines more protective than guidelines based on a single
chemical. The authors also emphasized that ERLs and ERMs were intended to be used as informal screening tools
only.

Although the ERL and ERM guidelines were not based upon deterministic or cause-effects studies, their accuracy
in correctly predicting nontoxicity and toxicity has been determined empirically among field-collected samples (Long
et al., in press). Analyses were performed with matching laboratory bioassay data and chemical data from 989 samples
collected in regions of the Atlantic, Pacific, and Gulf coasts. Data were gathered from results of amphipod survival
tests (Ampelisca abdita and Rhepoxynius abronius) for all 989 samples. Data from a battery of sensitive bioassays
(fertilization success of urchin gametes, embryological development of mollusc embryos, and microbial biolumines-
cence) were gathered for 358 of these samples. The percentages of samples indicating non-toxicity (not significantly
different from controls, p > 0.05), significant toxicity (p < 0.05), and high toxicity (p < 0.05 and mean response >20
percent difference from controls) were determined for the results of the amphipod tests alone and for the results of any
one of the tests performed.

Results of the analyses (summarized in Table B-l) suggest that highly toxic responses occurred in 12 percent of
the samples in the amphipod tests and 28 percent of the samples in any one of the tests performed when all chemical
concentrations were less than their respective ERL values. These samples were analogous to those classified as Her 3
in this report (i.e., all chemical concentrations less than the screening values). When one or more chemicals exceeded
ERL concentrations, but all concentrations were lower than the ERM concentrations (analogous to Tier 2), the percent-
ages of samples indicating high toxicity were 19 percent in the amphipod tests and 64 percent in any one of the tests
performed. The incidence of high toxicity in the amphipod tests increased from 10 percent when only one ERL value
was exceeded to 58 percent when 20-24 ERLs were exceeded. The incidence of toxicity in any one of the tests
increased from 29 percent when only one ERL was exceeded to 91 percent when 20-24 ERLs were exceeded. In
samples analogous to those classified as Tier 1 (one or more ERMs exceeded), the incidence of high toxicity was 42
percent in amphipod tests and 80 percent in any one of the battery of tests performed. If both the significant and highly
toxic results were combined in the Tier 1 samples, the percentage of samples indicating toxicity increases to 55 percent
in amphipod tests and 87 percent in any one of the tests. As with the ERLs, the incidence of toxicity increased with
increasing number of chemicals that exceeded the ERMs.

Probable Effects Levels and Threshold Effects Levels

A method slightly different from that used by Long et al. (1995) to develop ERMs and ERLs was used by the
Florida Department of Environmental Protection (FDEF, 1994) to develop similar correlative, effects-based guidelines

Table B-l. Incidence of Toxicity in Amphipod Survival Tests Alone and Any One of 2-4 Tests Performed in
Samples Analogous to Those Classified as Tier 1,2, or 3 (from Long et al., in press)





Amphipod Tests Alone

Any Test Performed

Chemical
Concentrations

Analogous
Tier

% Not
Toxic

% Signif.
Toxic

% Highly
Toxic

% Not
Toxic

% Signif.
Toxic

% Highly
Toxic

ail < ERLs

Tier 3

64

23

12

67

5

28

> I or more ERLs

Tier 2

59

22

19

20

15

64

> 1 or more ERMs

Tierl

45

13

42

13

7

80

B-8


-------
Nsiliomil Sediment Qiisililv Survey

for Florida's coastal waters. Modifications to the Long et al. (1995) approach increased the relevance of the resultant
guidelines to Florida's coastal sediments by making information in the database more consistent and by expanding the
information used to derive sediment quality assessment guidelines with additional data from other locations in the
United States and Canada, particularly Florida and the southeastern and Gulf of Mexico regions (FDEP, 1994). Three
effects ranges were developed with a method that used both the chemical concentrations associated with biological
effects (the "effects" data) and those associated with no observed effects (the "no-effects" data). In this method, the
threshold effects level (TEL) is the geometric mean of the lower 15th-percentile concentration of the effects data (the
ERL) and the 50th-pereentile concentration of the no-effects data. The probable-effects level (PEL) is the geometric
mean of the 50th-percentile concentration of the effects data (the ERM) and the 85th-percentile concentration of the
no-effects data. Essentially, the PEL and TEL reflect the ERM and ERL values adjusted upward or downward depend-
ing on the degree of overlap between the distributions of "effects" and "no effects" data. TELs and PELs have been
developed for 33 chemicals: 9 trace metals, total PCBs, 13 individual polynuclear aromatic hydrocarbons (PAHs), 3
classes of PAHs (total low molecular weight, total high molecular weight, and total PAH), 6 pesticides (chlordane,
dieldrin, p.p' -DDD, p.p* -DDE, p,p' -DDT), and total DDT (FDEP, 1994).

As was the case with the Long et al. (1995) approach, in the FDEP (1994) approach the lower of the two guidelines
for each chemical (i.e., the TEL) was assumed to represent the concentration below which toxic effects rarely occurred.
In the range of concentrations between the TEL and PEL, effects occasionally occurred. Toxic effects usually or
frequently occurred at concentrations above the upper guideline value (i.e., the PEL). TEL and PEL values were
developed on a sediment dry-weight basis.

Although the extensive database and evaluation of effects data make this approach applicable to many areas of the
country, the available data still have limitations. For example, FDEP (1994) noted that there is a potential for
underprotection or overprotection of aquatic resources if the bioavailability of sediment-associated contaminants and
other factors affecting toxicity are not included. Most of the TELs and PELs were within a factor of 2 to 3 of other
sediment quality guideline values. Most were deemed reliable for evaluating sediment quality in Florida's coastal
waters, with less confidence in the values for mercury, nickel, total PCBs, chlordane, lindane, and total DDT. An
evaluation of independent sets of field data from Florida, the Gulf of Mexico, California, and New York showed that
TELs and PELs correctly predict the toxicity of sediment in 86 percent and 85 percent of the samples, respectively.

As with ERLs and ERMs, the accuracy of fEL and PEL guidelines to correctly predict nontoxicity and toxicity
has been determined empirically among field-collected samples (Long et al., in press). Analyses were performed with
matching laboratory bioassay data and chemical data from 989 samples collected in regions of the Atlantic, Pacific,
and Gulf coasts. Data were gathered from results of amphipod survival tests (Ampelisca abdita and Rhepoxynius
abronius) for all 989 samples. Data from a battery of sensitive bioassays (fertilization success of urchin gametes,
embryological development of mollusc embryos, and microbial bioluminescence) were gathered for 358 of these
samples. The percentages of samples indicating nontoxicity (not significantly different from controls, p > 0.05),
significant toxicity (p < 0.05), and high toxicity (p < 0.05 and mean response >20 percent difference from controls)
were determined for the results of the amphipod tests alone and for the results of any one of the tests performed.

Results of the analyses (summarized in Table B-2) suggest that highly toxic responses occurred in 10 percent of
the samples in the amphipod tests and 5 percent of the samples in any one of the tests performed when all chemical
concentrations were less than their respective TELvalues, These samples were analogous to those classified as Tier 3
in this report (i.e., all chemical concentrations less than the screening values). When one or more chemicals exceeded
TEL concentrations, but all concentrations were lower than the PEL concentrations (analogous to Tier 2), the percent-
ages of samples indicating high toxicity were 17 percent in the ampipod tests alone and 59 percent in any one of the
tests performed. The incidence of high toxicity in the amphipod tests increased from 13 percent when only one TEL
value was exceeded to 52 percent when 20-27 TELs were exceeded. The incidence of toxicity in any one of the tests
increased from 31 percent when 1-5 TELs were exceeded to 63 percent when 20-27 TELs were exceeded. In samples
analogous to those classified as Tier 1 (one or more PELs exceeded), the incidence of high toxicity was 38 percent in
amphipod tests and 78 percent in any one of the battery of tests performed. If both the significant and highly toxic
results were combined in the Her 1 samples, the percentage of samples indicating toxicity increases to 51 percent in
amphipod tests and 86 percent in any one of the tests. As with the TELs, the incidence of toxicity increased with
increasing number of chemicals that exceeded the PELs.

B-9


-------
Thble B-2. Incidence of Toxicity in Amphipod Survival Tests Alone and Any One of 2-4 Tests Performed in
Samples Analogous to Those Classified as Tier 1,2, or 3 (from Long et al., in press)

Chemical
Concentrations

Analogous
Her

Amphipod Tests Alone

Any Test Performed

% Not
Toxic

% Signif.
Toxic

% Highly
Toxic

% Not
Toxic

% Signlf.
Toxic

% Highly
Toadc

all 1 or more TELs

Tier 2

62

21

17

22

19

59

> 1 or more PELs

Tierl

49

13

38

14

8

73

Apparent Effects Thresholds

The AET approach is another empirical data evaluation approach to defining concentrations in sediment associ-
ated with adverse effects, Barrick et al. (1988) reported that AETs can be developed for any measured chemical
(organic or inorganic) with a wide concentration range in the field. The AET concept applies to matched field data for
sediment chemistry and any observable biological effects (e.g., bioassay responses, infaunal abundances at various
taxonomic levels, bioaccumulation). By using these different biological indicators, application of the resulting sedi-
ment quality values enables a wide range of biological effects to be addressed in the management of contaminated
sediments. Using sediment samples from Puget Sound in Washington State, AET values have been developed for 52
chemicals: 10 trace metals, 15 individual polynuelear aromatic hydrocarbons (PAHs), 3 pesticides (p,p'-DDD, p,p'-
DDE, p,p'-DDT), 6 halogenated organics, and 18 other compounds.

The focus of the AET approach is to identify concentrations of contaminants that are associated exclusively with
sediments exhibiting statistically significant biological effects relative to reference sediments. AET values were based
on measured chemical concentrations per dry weight of sediment, AETs for each chemical and biological indicator
were developed using the following steps (Barrick et al., 1988).

1.	Collected "matched" chemical and biological effects data—Conducted chemical and biological effects test-
ing on subsamples of the same field sample.

2.	Identified "impacted" and "nonimpacted" stations—Statistically tested the significance of adverse biologi-
cal effects relative to suitable reference conditions for each sediment sample and biological indicator.

3.	Identified the AET using only "nonimpacted" stations—For each chemical, the AET was identified for a
given biological indicator as the highest detected concentration among sediment samples that did not exhibit
statistically significant effects.

4.	Verified that statistically significant biological effects were observed at a chemical concentration higher than
the AET; otherwise, the AET was only a preliminary minimum estimate.

5.	Repeated steps 1-4 for each biological indicator.

For a given data set, the AET value for a chemical is the sediment concentration above which a particular adverse
biological effect for individual biological indicators (amphipod bioassay, oyster larvae bioassay, Microtox bioassay,
and benthic infaunal abundance) is always significantly different statistically relative to appropriate reference condi-
tions. Two thresholds were recognized in the evaluations conducted in this report, when possible, based on the differ-
ent indicators. EPA defined the AET-low as the lowest AET among applicable biological indicators, and the AET-high
as the highest AET among applicable biological indicators. The use of the high/low AET values is not a recommenda-
tion of the authors of the approach; rather it was developed for the NSI evaluation. The two thresholds were used in
this evaluation to give a range of effects values (as with the ERL/ERMs and TEL/PELS). AET values based on
Microtox bioassays were not used for the NSI evaluation.

B-10


-------
Sediment toxicity tests provide Important information on the effects of multiple chemical exposures to assist in the
evaluation of sediment quality. Methods for testing the acute and chronic toxicity of sediment samples to benthic
freshwater and marine organisms have been developed (see reviews in API, 1994; Burton et al„ 1992; Lamberson et al.,
1992; USEPA, 1994b, c) and used primarily for dredged material evaluation (USEPA and USACOE, 1994), The NSI data
contain acute sediment toxicity results from tests in which organisms were exposed to field-collected sediments and
mortality was recorded. Results of whole sediment and elutriate toxicity tests were used in the evaluation of the NSI.

Variations in observed toxicity from tests of the same sediment sample may be attributed to the relative sensitivities
of the species used in the tests; disruption of geochemistry and kinetic activity of bedded sediment contaminants during
sampling, handling, and bioturbation; and laboratory-related confounding factors (Lamberson et al„ 1992). Recent
studies indicate that aqueous representations of whole sediment (e.g., elutriate) do not accurately predict the bioavail-
ability of some contaminants compared to whole-sediment exposures (Harkey et al., 1994). Acute sediment toxicity tests
have been widely accepted by the scientific and regulatory communities and the results can be readily interpreted,
although more work is needed on chronic testing (Thomas et al,, 1992), Appendix G presents the methodology for
evaluating sediment toxicity tests as applied in the NSI data evaluation.

Human Health Assessments

In the evaluation of NSI data, two primary evaluation parameters were used to assess potential human health
impacts from sediment contamination: (1) sediment chemistry theoretical bioaccumulation potential and (2) tissue levels
of contaminants in demersal, nonmigratory species.

Theoretical Bioaccumulation Potential

The theoretical bioaccumulation potential (TBP) is an estimate of the equilibrium concentration of a contaminant in
tissues if the sediment in question were the only source of contamination to the organism (USEPA and USACOE, 1994),
The TBP calculation is used as a screening mechanism to represent the magnitude of bioaccumulation likely to be
associated with nonpoiar organic contaminants in the sediment. At present, the TBP calculation can be performed only
for nonpoiar organic chemicals; however, methods for TBP calculations for metals and polar organic chemicals are under
development (USEPA and USACOE, 1994).

The environmental distribution of nonpoiar organic chemicals is controlled largely by their solubility in various
media. Therefore, in sediments they tend to occur primarily in association with organic matter (Karickhofif, 1981) and in
organisms they are found primarily in the body fate or lipids (Bierman, 1990; Geyer et al., 1982; Konemann and van
Leeuwen, 1980; Mackay, 1982), Bioaccumulation of nonpoiar organic compounds from sediment can be estimated from
the organic carbon content of the sediment, the lipid content of the organism, and the relative affinities of the chemical
for sediment organic carbon and animal lipid content (USEPA and USACOE, 1994). It is possible to relate the concentra-
tion of a chemical in one phase of a two-phase system to the concentration in the second phase when the system is in
equilibrium. The TBP calculation focuses on the equilibrium distribution of a chemical between the sediment and the
organism. By normalizing nonpoiar organic chemical concentration data for lipid in organisms, and for organic carbon
in sediment, it is possible to estimate the preference of a chemical for one phase or the other (USEPA and USACOE,
1994).

The TBP can be calculated relative to the biota-sediment accumulation factor (BSAF), as in the following equation
(USEPA and USACOE, 1994):

TBP = BSAF(C,/f,Jf,
where TBP is expressed on a whole-body basis in the same units of concentration as C1 and

TBP	= theoretical bioaccumulation potential (ppm);

C	= concentration of nonpoiar organic chemical In sediment (ppm);

B-ll


-------
Appendix B

C,	= concentration of nonpolar organic chemical in sediment (ppm);

BSAF = biota-sediment accumulation factor (ratio of the concentration of a chemical in tissue,

normalized to lipid, to the concentration of the chemical in surface sediment, normalized to
organic carbon (in kg sediment organic carbon/kg lipid));

= total organic carbon (TOC) content of sediment expressed as a decimal fraction (i.e., 1
percent = 0.01); and

fj	= organism lipid content expressed as a decimal fraction (e. g., 3 percent = 0.03) of fillet or

whole-body dry weight.

BSAF values used in the TBP evaluation are discussed in Appendix C. If TOC measurements were not available
at a site, f^ was assumed to be 0.01 (1 percent).

For the evaluation of NSI data, EPA selected a 3 percent lipid content in fish fillets for the TBP calculation for
assessing human health effects from the consumption of contaminated fish. Lipid normalization is now part of the EPA
guidance on bioaccumulation, and the current national methodology uses a 3 percent value for human health assess-
ments. The Great Lakes Water Quality Initiative Technical Support Document for the Procedure to Determine
Bioaccumulation Factors (USEPA, 1995b) uses a 3.10 percent lipid value for trophic level 4 fish and 1.82 percent for
trophic level 3 fish in its human health assessments.

As part of the NSI TBP evaluation, EPA also evaluated percent lipid measurements included in the STORET

database, the, National Study of Chemical Residues in Fish (NSCRF; USEPA, 1992b), and other published sources, and
compared those values to the value selected for the NSI evaluation (Appendix C). The mean fillet percent lipid content
for various groups of fish species in the STORET database ranged from 0.753 to 4.49 percent; in the NSCRF, mean
fillet values ranged from 1.6 to 4.9 percent. The mean whole-body percent lipid content for various groups of fish
species in the STORET database ranged from 3.757 to 6.33 percent; in the NSCRF, mean whole-body values ranged
from 4.6 to 8.8 percent

In the NSI data evaluation approach, TBP values were compared to U.S. Food and Drug Administration tolerance/
action/guidance levels and EPA risk levels. These parameters an® discussed below.

FDA Tolerance/Action/Guidance Levels

Hie U.S. Food and Drug Administration (FDA) is responsible for the safety of the Nation's commercial food
supply, including fish and shellfish, for human consumption. Under the authority of the Federal Food, Drug and
Cosmetic Act (FFDCA), FDA ensures that regulated products are safe for use by consumers. The FFDCA authorizes
FDA to conduct assessments of the safety of ingredients in foods. The key element of the FFDCA, and the source of
FDA's main tools for enforcement, is the prohibition of the "adulteration" of foods. FDA can prescribe the level of
contaminant that will render a food adulterated and, therefore, can initiate enforcement action based on scientific data.
The establishment of guidance and action levels (informal judgments about the level of a food contaminant to which
consumers can be safely exposed) or tolerances (regulations having the force of law) is the regulatory procedure
employed by FDA to control environmental contaminants in the commercial food supply.

During the 1970s, the available detection limits were considered to demonstrate elevated contamination and were
used as action levels. Since that time, FDA has focused on using risk-based standards. These standards have been
derived by individually considering each chemical and the species of fish it is likely to contaminate. FDA also
considered (1) the amount of potentially contaminated fish eaten and (2) the average concentrations of contaminants
consumed. FDA has established action levels in fish for 10 pesticides and methylmercury, tolerance levels for poly-
chlorinated biphenyls (PCBs), and guidance for 5 metals.

EPA Risk Levels

Potential impacts on humans are evaluated by estimating potential carcinogenic risks and noncarcinogenic haz-
ards associated with the consumption of chemically contaminated fish tissue. In this assessment it was assumed that
the only source of contamination to fish is contaminated sediment. The procedures for estimating human health risks
due to the consumption of chemically contaminated fish tissue are based on Risk Assessment Guidance for Superfund

B-12


-------
XiiliajJiiJ Si'fJiim-nt Quiilit t .S>/ri< »

(USEPA, 1989) and Guidance for Assessing Chemical Contamination Data for Use in Fish Advisories, Volume II;
Development of Risk-Based Intake Limits (USEPA, 1994a).

EPA human health risk assessment methods were used in this assessment to determine the levels of contamination in
fish that might result in a 10 5 cancer risk (1 in 100,000 extra chance of cancer over a lifetime) or a noncancer hazard in
humans. A 10 s risk level exceeds the lower bound (i,e„ 10'6) but is lower than the upper bound (i.e., 10'4) of the risk range
accepted by EPA (USEPA, 1990).

Human health cancer risks and noncancer hazards are based on the calculation of the chronic daily intake (CD!) of
contaminants of concern:

CDI= (EPC)(IR)(EF)(ED)

(BW)(AT)

where:

CDI	=	chronic daily intake (mg/kg/day);

EPC	=	exposure point concentration (contaminant concentration in fish);

IR	=	ingestion rate (6.5 g/day);

EF	=	exposure frequency (365 days/year);

ED	=	exposure duration (70 years);

BW	=	body weight (70 kg); and

AT	=	averaging time (70 years x 365 days/year).

These are the same parameter values used by EPA to develop human health water quality criteria. Carcinogenic
risks are then quantified using the equation below:

Cancer ris^ =001,2 SF,

where:

Cancer risk, = the potential carcinogenic risk associated with exposure to chemical i (unitless);
CDI.	= chronic daily intake for chemical i (mg/kg/day); and

SF.	= slope factor for chemical i (mg/kg/day)"1.

The hazard quotient, which is used to quantify the potential for an adverse noncarcinogenic effect to occur, is
calculated using the following equation:

CDI

110 _

' RfD,

where:

HQ.	= hazard quotient for chemical i (unitless);

CDI.	= chronic daily intake for chemical i (mg/kg/day); and

RfD,	= reference dose for chemical i (mg/kg/day).

If the hazard quotient exceeds unity (i.e., 1), an adverse health effect might occur. The higher the hazard quotient,
the more likely that an adverse noncarcinogenic effect will occur as a result of exposure to the chemical. If the
estimated hazard quotient is less than unity, noncarcinogenic effects are unlikely to occur.

Using these formulas, the fish tissue concentration (EPC) of a contaminant that equates to a cancer risk of 10 s or
a hazard quotient that exceeds unity can be back-calculated.

Cancer risk:

B-13


-------
V|»|H'ii(ii\ 1?

EPC =

(1Q-5)(BW)(AT)(C,)

(rR)(EF)(ED)(SFi)
Noncancer hazard:

EPC =

(BW)(AT)(RfDj )(C,)

(IR)(EF)(ED)
whore:

C,	= conversion factor (10' g/kg).

Tissue Levels of Contaminants

In addition to sediment chemistry TBP values, measured levels of contaminants in the tissues of resident aquatic
species were used to assess potential human health risk. As was the case with the evaluation of TBP values, the NSI
evaluation approach compared contaminant tissue levels to FDA tolerance/action/guidance levels and EPA risk levels.
Each of these parameters was discussed in the previous section. In such a comparison it is assumed that contaminant
concentrations in tissue result from bioaccumulation of contaminants in the sediment

Wildlife Assessments

In addition to the evaluation parameters described above for the assessment of potential aquatic life and human
health impacts, EPA also conducted a separate analysis of potential wildlife impacts resulting from exposure to sediment
contaminants.

Wildlife criteria based on fish tissue concentrations were derived using methods similar to those employed for
deriving EPA wildlife criteria, as presented in the Great Lakes Water Quality Initiative Criteria Documents for the
Protection of Wildlife (USEPA, 1995a). EPA has developed Great Lakes Water Quality Wildlife Criteria for four chemi-
cals: DDT, mercury, 2,3,7,8-TCDD, and PCBs. A Great Lakes Water Quality Wildlife Criterion (GLWC) is the concentra-
tion in the water of a substance that, if not exceeded, protects avian and mammalian wildlife populations from adverse
effects resulting from the ingestion of surface waters and aquatic prey (USEPA, 1995a). Wildlife values are calculated
using the equation:

(NOAEL ,gSSF) jWtA

WA + (Fa ^BAF)

where:

WV	= wildlife value (mg/L);

NOAEL = no-observed-adverse-effect level, as derived from mammalian or avian studies (mg/kg-d);

WtA	= average weight for the representative species identified for protection (kg);

WA	= average daily volume of water consumed by the representative species identified for protec-

tion (L/d);

SSF	= species sensitivity factor, an extrapolation factor to account for the difference in toxicity

between species;

Fa	= average daily amount of food consumed by the representative species identified for protec-

tion (kg/d); and

BAF	ss bioaccumulation factor (L/kg), the ratio of the concentration of a chemical in tissue, normal-

ized to lipid, to the concentration in ambient water. Chosen using guidelines for wildlife
presented in appendix B to part 132, Methodology for Development of Bioaccumulation
Factors {Federal Register, Vol. 58, No. 72, April 16,1993).

In the development of the four GLWCs, wildlife values for five representative Great Lakes basin wildlife species
(bald eagle, herring gull, belted kingfisher, mink, and river otter) were calculated, and the geometric mean of these values
within each taxonomic class (birds and mammals) was determined. The GLWC is the lower of two class-species means
(USEPA, 1995a).

B-14


-------
National Sediment Quality Suim'v

The wildlife values are considered to be generally protective of wildlife species. However, it should be noted that
the approach is not based on the most sensitive wildlife species, but rather a typical class of either avian or mammalian
piscivores. Despite this limitation, this approach is still considered appropriate and conservative because of the many
conservative assumptions used to derive these wildlife values (e.g., species sensitivity factors, assumption that animals
consume only contaminated fish).

Proposed EPA wildlife criteria are based on surface water contaminant levels protective of potential wildlife
exposure. Thus, the proposed EPA wildlife criteria cannot be compared directly to the NSI fish tissue concentrations
(either the calculated TBPs or fish tissue monitoring data). Therefore, it was necessary to develop an approach for
estimating wildlife criteria for fish tissue based on the same toxicity and exposure parameter assumptions that were
used to derive the surface water wildlife criteria; First, wildlife values (i.e., fish tissue concentrations protective of
wildlife) were derived for the most sensitive mammalian species (ie., otter and mink) and avian species (i.e., king-
fisher, herring gull, and eagle)—the same species used to derive the proposed EPA wildlife criteria. The equation used
to estimate wildlife values for fish tissue is presented below. (Exposure assumptions used for each species are pre-
sented in USEPA, 1995a.)

where:

WV rNOAEI^SSFj^Wt,

Fa

WV

vnih

NOAEL

SSF

Wt.

wildlife value for fish tissue (mg/kg);
no-observed-adverse-effect level (mg/kg-day);
species sensitivity factor
average weight of animal in kilograms (kg); and
average daily amount of food consumed (kg/day).

Secondly, the geometric mean of the wldlife values was calculated for the mammal group, as well as for the avian
group. Finally, the lower of the two geometric mean values was considered the wildlife criterion for fish tissue for a
given chemical.

It should be noted that direct ingestion of surface water was included when developing proposed EPA wildlife
criteria for surface water. This exposure route, however, was not considered when evaluating NSI data, even though
sediment contamination might result in contamination of surface water available for wildlife consumption. A sensitiv-
ity analysis was conducted to evaluate the impact of excluding the surface water ingestion exposure route. Based on
this analysis, ingestion of surface water contributes less than 0.0001 percent of the total exposure (i.e., ingestion of fish
and water). Therefore, excluding the water ingestion exposure route had no significant impact on the evaluation of
NSI data with regard to potential wildlife impacts.

Wldlife criteria derived for DDT, mercury, 2,3,7,8-TCDD, and PCBs based on fish tissue concentration are presented below.

Fish Tissue

Chemical	Criterion fmp/kg)

DDT	3.93E-2

Mercury	5.73E-2

2,3,7,8-TCDD	5.20E-7

PCBs	1.60E-1

The wildlife criteria were compared to measured fish tissue residue data contained in the NSI and to TBPs calcu-
lated for DDT, 2,3,7,-TCDD, and PCBs. Mercury is not a nonpolar organic chemical, and thus a TBP for mercury was
not calculated. A whole-body lipid value of 10.31 was assumed for the TBP evaluation of potential wildlife impacts,
based on the Great Lakes Water Quality Technical Support Document for the Procedure to Determine Bioaccumulation
Factors (USEPA, 1995b).

B-15


-------
A|)|R'i)(li\ B

References

Adams, W.J., R.A. Kimerle, and R.G. Mosher. 1985. Aquatic safety assessment of chemicals sorbed to sediments.
In Aquatic Toxicology and Hazard Assessment: Seventh Annual Symposium, American Society for Testing and
Materials, Philadelphia, PA, ed. R.D. Cardwell, R. Purdy, and R.C. Bahner, pp. 429-453.

Adams, WJ., R.A. Kimerle, and J.W. Barnett, Jr. 1992. Sediment quality and aquatic life. Environ. Sci. Technol.,
26:1865-1875.

Allen, H.E., G. Fu, and B. Deng. 1993. Analysis of acid-volatile sulfide (AVS) and simultaneously extracted metals
(SEM) for the estimation of potential toxicity in aquatic sediments. Environ. Toxicol. Chem. 12:1441-1453.

Ankley, G.T., V.R. Mattson, E.N. Leonard, C.W. West, and J.L. Bennett 1993. Predicting the toxicity of copper in
freshwater sediments: Evaluation of the role of acid-volatile sulfide. Environ. Toxicol. Chem. 12:315-320.

API. 1994. User's guide and technical resource document: Evaluation of sediment toxicity tests for biomonitoring
programs. API pub. no. 4607. Prepared for American Petroleum Institute, Health and Environmental Sciences
Department, Washington, D.C., by.PTI Environmental Services, Bellvue, WA.

Barrick, R., S. Becker, L. Brown, H. Beller, and R. Pastorok. 1988. Sediment quality values refinement: 1988
update and evaluation ofPuget Sound AET. Vol. 1. Prepared for the Puget Sound Estuary Program, Office of
Puget Sound.

Bierman, V.J. 1990. Equilibrium partitioning and magnification of organic chemicals in benthic animals. Environ.
Sci. Technol. 24:1407-1412.

Burton, G.A., Jr., J.K. Nelson, and C.G. Ingersoll. 1992. Freshwater benthic toxicity tests. In Sediment toxicity
assessment, ed. G.A. Burton, Jr., pp. 213-240. Lewis Publishers, Chelsea, MI.

Casas, A.M., and E.A. Crecelius. 1994. Relationship between acid-volatile sulfide and the	toxicity of zinc,

lead and copper in marine sediments. Environ. Toxicol. Chem. 13(3):529-536.

de Bruijn, J., F. Busser, W. Seinen, and J. Hermens. 1989. Determination of octanol/water partition coefficients for
hydrophobic organic chemicals with the "slow-stirring" method. Environ. Toxicol. Chem. 8:499-512.

De Kock, A.C., and D.A. Lord. 1987. A simple procedure for determining octanol-water partition coefficients using
reverse phase high performance liquid chromatography (RPHPLC). Chemosphere 16(1):133-142.

Di Toro, D.M., J.D. Mahony, D.J. Hansen, K.J. Scott, M.B. Hicks, S.M. Mays, and M.S. Redmond. 1990. Toxicity
of cadmium in sediments: The role of acid-volatile sulfide. Environ. Toxicol. Chem. 9:1487-1502.

Di Toro, D.M., J.D. Mahony, D.J. Hansen, K.J. Scott, A.R. Carlson, and G.T. Ankley. 1992. Acid-volatile
sulfide predicts the acute toxicity of cadmium and nickel in sediments.	Environ. Sci. Technol.

26(1):96-101.

Di Toro, D.M., C.S. Zarba, D.J. Hansen, W.J. Berry, R.C. Swartz, C.E. Cowan, S.P. Pavlou, H.E. Allen, N.A.
Thomas, and P.R. Paquin. 1991. Technical basis for establishing sediment quality criteria for nonionic organic
chemicals using equilibrium partitioning. Environ. Toxicol. Chem. 10:1541-1583.

Doucette, W.J., and A.W. Andren. 1988. Estimation of octanol/water partition coefficients: evaluation of six
methods for highly hydrophobic aromatic hydrocarbons. Chemosphere 17(2):345-359.

FDEP. 1994. Approach to the assessment of sediment quality in Florida coastal waters, Vol. 1. Development and

B-16


-------
.National hVcliiilonI Quality .Niiim'V

evaluation of sediment quality assessment guidelines. Prepared for Florida Department of Environmental
Protection, Office of Water Policy, Tallahassee, FL, by MacDonald Environmental Sciences Ltd., Ladysmith,
British Columbia.

Geyer, H., P, Sheehan, D, Kotzias, and F. Korte. 1982. Prediction of ecological behavior of chemicals: Relation-
ship between physico-chemical properties and bioaccumulation of organic chemicals in the mussel Mytilus
edulis. Chemosphere 11:1121-1134.

Hansen, DJ. 1995. Assessment tools that can be used for the National Sediment Inventory, Memorandum from
DJ. Hansen, Environmental Research Laboratory, Narragansett, to C. Fox, USEPA Office of Water, February 28,
1995.

Harkey, G.A., P.F. Landrum, and SJ. Klaine. 1994. Comparison of whole-sediment, elutriate and pore-water
exposures for use in assessing sediment-associated organic contaminants in bioassays. Environ. Contam. Toxicol.
13(8): 1315-1329.

Howard, P.H. 1990. Handbook of environmental fate and exposure data for organic chemicals. Vol. II. Solvents.
Lewis Publishers, Chelsea, MI.

Howard, D.E., and R.D. Evans. 1993. Acid-volatile sulfide (AVS) in a seasonally anoxic mesotrophic lake:
Seasonal and spatial changes in sediment AVS. Environ. Toxicol. Chem. 12:1051-1057.

Ingersoll, C.G., P.S. Haverland, E.L. Branson, TJ. Canfield, F.J. Dwyer, C.E. Henke, and N.E. Kemble. 1996.
Calculation and evaluation of sediment effect concentrations. J. Great Lakes Res. 22:602-623.

Isnard, P., and S. Lambert. 1989. Aqueous solubility and n-octanol/water partition coefficient correlations. Chemo-
sphere 18:1837-1853.

KarickhofF, S. 1981. Semi-empirical estimation of sorption of hydrophobic pollutants on natural sediments and
soils. Chemosphere 9:3-10.

KarickhofF, S. W., and J.M. Long. 1995. Internal report on summary of measured, calculated, and recommended log
Kow values. Prepared for U.S. Environmental Protection Agency, Office of Water, Washington, DC.

Klein, W., W. Kordel, M. Weis, and HJ. Poremski. 1988. Updating of the OECD test guideline 107 "partition
coefficient n-octanol/water": OECD laboratory intercomparison test on the HPLC method. Chemosphere
17(2):361-386.

Konemann, H„ and K. van Leeuwen. 1980. Toxicokinetics in fish: Accumulation and elimination of six chlo-
robenzenes by guppies. Chemosphere 9:3-19.

Lamberson, J.O., T.H. DeWitt, and R.C. Swartz. 1992. Assessment of sediment toxicity to marine benthos.
In Sediment toxicity assessment, ed. G.A. Burton, Jr., pp. 183-211. Lewis Publishers, Chelsea, MI.

Leo, A.J. 1993. Calculating log PM from structures. Chem. Rev. 93:1281-1310.

Long, E.R., D.D. MacDonald, S.L. Smith, and F.D. Calder. 1995. Incidence of adverse biological effects within
ranges of chemical concentrations in marine and estuarine sediments. Environ. Manage. 19(l):81-97.

Long, E.R., and L.G. Morgan. 1990. The potential for biological effects of sediment-sorbed contaminants tested in
the National Status and Trends Program. NOAA tech. mem. NOS OMA 52. National Oceanic and Atmospheric
Administration, Seattle, WA.

Long, E.R., L J. Field, and D.D. MacDonald. In press. Predicting toxicity in marine sediments with numerical

B-17


-------
sediment quality guidelines. Submitted to Environ, Toxicol, Chem.

Luoma, S.N. 1983. Bioavailability of trace metals to aquatic organisms—A review. Sci, Tot, Environ. 28:1-22.

Mackay, D. 1982. Correlation of bioconcentration factors. Environ. Sci. Technol. 5:274-278.

	. 1991. Multimedia environmental models: The fugacity approach. Lewis Publishers, Boca Raton, FL.

Mackay, D„ W.Y. Shiu, and K.C. Ma. 1992a. Illustrated handbook of physical-chemical properties and environmen-
tal fate for organic chemicals. Volume II, Palynuclear aromatic hydrocarbons, poly chlorinated dioxins and-
dibenzofurans. Lewis Publishers, Boca Raton, FL.

Mackay, D., W.Y. Shiu, and K.C. Ma. 1992b. Illustrated handbook of physical-chemical properties and environmen-
tal fate for organic chemicals. Volume I, Monoaromatic hydrocarbons, chlorobenzenes, and PCBs. Lewis
Publishers, Boca Raton, FL

Meyer, J.S., W. Davison, B. Sundby, J.T. Ores, D.J. Lauren, U. Forstner, J. Hong, and D.G. Crosby. 1994. Synopsis
of discussion session: The effects of variable redox potentials, pH, and light on bioavailability in dynamic water-
sediment environments. In Bioavailability physical, chemical, and biological interactions, proceedings of the
Thirteenth Pellston Workshop, ed. J.L. Hamelink, P.F. Landrum, H.L. Bergman, and W.H. Benson, pp.155-170.
Lewis Publishers, Boca Raton, FL.

Noble, A. 1993. Partition coefficients (w-octanol-water) for pesticides. J. Chromatography 642:314.

Stephan, C.E. 1993. Derivation of proposed human health and wildlife biaccumulation factors for the Great Lakes
initiative. U.S. Environmental Protection Agency, Office of Research and Development, Duluth, MN.

Suter, G.W.H, and J.B. Mabrcy. 1994. Toxicological benchmarks for screening potenital contaminants of concern
for effects on aquatic biota: 1994 revision. ES/ER/TM-96/R1. Oak Ridge National Laboratory, Environmental
Sciences Division, Oak Ridge, TN.

Thomas, N., J.O. Lamberson, and R.C. Swartz. 1992. Bulk sediment toxicity test approach. In Sediment classifica-
tion methods compendium, pp. 3-1-3-10. EPA 823-R-92-006. U.S. Environmental Protection Agency, Office of
Water, Washington, DC.

USEPA. 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.

	. 1989. Risk assessment guidance for Superfund. Volume I: Human health evaluation manual (Part A).

Interim final. OSWER Directive 9285.7-01a. U.S. Environmental Protection Agency, Office of Solid Waste and
Emergency Response, Washington, DC. December 1989.

	, 1990, National contingency plan. Federal Register, March 8,1990,55:8666.

	. 1992a. Sediment classification methods compendium. EPA 823-R-92-006. U.S. Environmental Protec-
tion Agency, Office of Water, Washington, DC.

	. 1992b. National study of chemical residues in fish. 2 vols. EPA 823-R-92-008a,b. U.S. Environmental

Protection Agency, Office of Science and Technology, Washington, DC.

	. 1993a. Technical basis for establishing sediment quality criteria for nonionic organic contaminants for

the protection of bent hie organisms by using equilibrium partitioning. Draft. EPA 822-R-93-011. U.S. Environ-
mental Protection Agency, Office of Science and Technology, Health and Ecological Criteria Division, Washing-
ton, DC.

B-18


-------
National Sediment	Survey

	. 1993b. Proposed sediment quality criteria for the protection ofbenthic organisms: Acenapthene. EPA

822/R93-013. U.S. Environmental Protection Agency, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1993c. Proposed sediment quality criteria for the protection ofbenthic organisms: Dieldrin. EPA 822/

R93-015. U.S. Environmental Protection Agency, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1993d. Proposed sediment quality criteria for the protection of bent hie organisms: Endrin. EPA 822/R93-

016. U.S. Environmental Protection Agency, Office of Science and Technology, Health and Ecological Criteria
Division, Washington, DC.

	. 1993e. Proposed sediment quality criteria for the protection of benthic organisms: Fluoranthene. EPA

822/R93-012. U.S. Environmental Protection Agency, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1993f. Proposed sediment quality criteria for the protection ofbenthic organisms: Phenanthrene. EPA

822/R93-015. U.S. Environmental Protection Agency, Office of Science and Technology, Health and Ecological
Criteria Division, Washington, DC.

	. 1994a. Guidance for assessing chemical contamination data for use in fish advisories, Volume II: Develop-
ment of Risk - Based Intake Limits. U.S. Environmental Protection Agency, Office of Science and Technology,
Washington, DC.

	. 1994b. Methods for measuring the toxicity and bioaccumulation of sediment-associated contaminants with

estuarine and marine amphipods. EPA 600/R-94/025. U.S. Environmental Protection Agency, Office of
Research and Development, Washington, DC.

	. 1994c. Methods for measuring the toxicity and bioaccumulation of sediment-associated contaminants with

freshwater invertebrates. EPA 600/R-94/024. U.S. Environmental Protection Agency, Office of Research and
Development, Duluth, MN.

	. 1995a. Great Lakes Water Quality Initiative criteria documents for the protection of wildlife, EPA-820-

. B-95-008. U.S. Environmental Protection Agency, Office of Science and Technology, Washington, DC.

	. 1995b. Great Lakes Water Quality Initiative technical support document for the procedure to determine

bioaccumulation factors. EPA-820-B-95-005. U.S. Environmental Protection Agency, Office of Water, Wash-
ington, DC.

	. 1995c. Water quality guidance for the Great Lakes System: Supplementary information document (SID).

EPA-820-B-95-001. U.S. Environmental Protection Agency, Office of Water, Washington, DC.

USEPA and USACOE. 1994. Evaluation of dredged material proposed for discharge in waters of the U.S.—

Testing manual (draft). EPA-823-B-94-002. U.S. Environmental Protection Agency, Office of Water, and U.S.
Army Corps of Engineers, Washington, DC.

Zhuang, Y., H.E. Allen, and G. Fu. 1994. Effect of aeration of sediment on cadmium binding. Environ. Toxicol.
Chem. 13(5):717-724.

B-19


-------
\p|H*ncli\ H


-------
National .^i-dSmciii Qiralil.v Survey

Appendix C

Method for Selecting Biota-
Sediment Accumulation
Factors and Percent Lipids in
Fish Tissue Used for Deriving
Theoretical Bioaccumulation
Potentials

Theoretical bioaccumulation potentials (TBPs) are empirically derived potential concentrations that might
occur in the tissues of fish exposed to contaminated sediments. TBPs are computed for nonpolar organic
chemicals as a function of sediment concentrations, fish tissue lipid contents, and sediment organic carbon
contents. Four separate pieces of information are required to compute the TBP for nonpolar organic chemicals:

1.	Concentration of nonpolar organic compound in sediment.

2.	Organic carbon content of the sediment.

3.	Biota-sediment accumulation factor (BSAF).

4.	Lipid content in fish tissue.

The details of the TBP calculations and related assumptions are found in Appendix B of this report to Congress.
This appendix describes the approach used to develop the BSAFs used in the NSI TBP evaluation and to evaluate fish
tissue lipid content data from selected information sources for comparison to the values used in the NSI TBP evalu-
ation. The BSAF values used for each chemical evaluated are presented in Appendix D.

Chemicals considered for fish tissue residue evaluation as part of the NSI data evaluation have at least one
screening value available, and the sum of positive sediment results and positive tissue results is greater than 20
observations. BSAF values were assigned to all nonpolar organic chemicals in the NSI having available screening
values. These screening values are risk-based concentrations (RBCs) developed either from carcinogenic potency
slopes or from oral reference doses. Carcinogenic potency slopes and reference doses were obtained from IRIS
(USEPA, 1995) and HEAST (USEPA, 1994b). Other screening values used for comparison to TBP values and tissue
data are U.S. Food and Drug Administration (FDA) tolernnce/action/guidance levels and EPA wildlife criteria. The
BSAF values used in the analysis are presented in Appendix D along with the screening values discussed above.

Method for Selecting BSAFs

Biota-sediment accumulation factors (BSAFs) are transfer coefficients that relate concentrations in biota to con-
centrations in sediment. They are calculated as the ratio of the concentration of nonpolar organic chemical in fish
tissue (normalized by lipid content) to the concentration of nonpolar organic chemical in sediment (normalized by
organic carbon content). At equilibrium, BSAFs are in theory approximately 1.0. In practice, BSAFs can be greater
than or less than 1.0 depending on the disequilibrium between fish and water, and that between water and sediment
Although based on partitioning theory, field measured BSAFs empirically account for factors such as metabolism and

C-l


-------
food chain biomagnificaiton. BSAFs can vary depending on the biota, dynamics of chemical loadings to the water
body, food chain effects, and rate of sediment-water exchange. Thus, measured BSAF values will depend on many
site-specific variables including hydraulic, biological, chemical, and ecological factors that affect bioavailability.
The accuracy of a BASF, measured at one location at a point in time, when applied to another location at another point
in time depends on two factors: (1) the degree to which variation from a theoretical BSAF of 1.0 is controlled by
inherent properties of the chemical as opposed to environmental conditions of the locale, and (2) the degree of
similarity between environmental conditions at the place of measurement and place of application.

BSAF values were assigned only to nonpolar chemicals in the NSI. This section describes how the BSAF values
used for the TBP assessment were selected from recommended values for specific chemicals.

Sources of Recommended BSAFs

BSAFs used in the NSI TBP evaluation were obtained from the EPA Office of Research and Development (EPA/
ORD) Environmental Research Laboratories at Duluth, Minnesota (Cook, 1995) and Narragansett, Rhode Island
(Hansen, 1995). In some cases (i.e., EPA/ORD-Duluth), BSAFs were provided for specific chemicals; in other cases
(i.e., EPA/ORD-Narragansett), BSAFs were provided by chemical class. Recommended BSAFs from each laboratory
are described below.

EPA Environmental Research Laboratory, Duluth

BSAF recommendations obtained from EPA/ORD-Duluth included mainly chemical-specific values for:

•	PCB congeners

•	Pesticides

•	Dioxins/Furans

•	Chlorinated benzenes

The recommended values from EPA/ORD-Duluth were based on BSAF data compiled from various sites and studies.
Data were selected based on the following criteria (Cook, 1995):

•	The primary source of chemical exposure to food webs was through release of chemicals in sediments.

•	The BSAF was derived for pelagic organisms (i.e., fish).

•	Chemicals in sediments and biota were at roughly steady state with respect to environmental loadings of the
chemical.

Pelagic BSAF data which predict relative bioaccumulation potentials of different chemicals are available for
ecosystems in which sediments are a primary source of the chemicals to pelagic food webs through release of chemi-
cals to the water. Little or no BSAF data exist for sites in which water and sediments are at steady-state with respect
to external chemical loadings. The best BSAF data for fish are those measured for Lake Ontario and used to estimate
BAFs in the Technical Support Document (TSD) for the Great Lakes Water Quality Initiative (GLWQI) (Cook, 1995;
Cook et al., 1994; USEPA, 1994a). The lake Ontario BSAFs are based on a large set of sediment and fish samples
collected in 1987 (USEPA, 1990). The BSAFs for PCDDs, PCDFs and co-planar PCB congeners are available from
ORD-Duluth data. Additional BSAFs for PCBs and pesticides are available from the data of Oliver and Niimi
(1988). These contemporary BSAFs are estimated to be approximately 20 to 25 percent of BSAFs when Lake
Ontario surface sediments and water are at steady-state with chemical loading to the ecosystem; a condition which
probably existed in the 1960s. EPA has measured BSAFs in the Fox River and Green Bay in Wisconsin and find
similar values despite much different species and exposure conditions (Cook, 1995).

EPA Environmental Research Laboratory, Narragansett

EPA/ORD-Narragansett provided a second source of information for selecting BSAF values. Probability distri-
bution curves for selecting BSAFs were presented by EPA/ORD-Narragansett for three chemical classes:

C-2


-------
N;t(ion;il Scdiiiiciif (_>«';iIil_\ S*11 \cv

•	PAHs

•	PCBs

•	Pesticides

EPA/ORD-Narragansett researchers developed cumulative probability curves for each chemical class from their da-
tabase of BSAFs (Hansen, 1995). The database from which general BSAF recommendations were summarized in-
cluded data from laboratory and Held studies conducted with both freshwater and marine sediments. Data must be
from species that directly contact sediments or feed on organisms that live in sediments, i.e., benthic invertebrates and
benthically coupled fishes.

Overall the database contained more than 4,000 BSAF observations. Cumulative probability curves summariz-
ing the BSAF data in the database were provided by Hansen (1995) for PAHs, PCBs, and pesticides. BSAF values
were tabulated for several probability percentiles. These findings have been published in Tracey and Hansen, 1996.

Approach for Selecting BSAFs from Recommended Values

The general approach for selecting a BSAF for a chemical follows:

•	Use a chemical-specific value for the BSAF, if available.

•	If no chemical-specific value is available, use a BSAF derived for a chemical category.

•	For chemicals having no specific information on the BSAF, use a default value of 1.

The EPA/ORD-Narragansett values for the BSAF were selected as the 50th percentile of the distribution of
BSAFs by chemical class (Table C-l). The BSAF values from EPA/ORD-Duluth were averages of individual data
points for specific chemicals. The preference for central tendency measures reflects risk management that imples an
approximate 50 percent chance of bioaccumulation to a predicted level. Other components of the EPA risk levels for
fish tissue chemical residues and FDA action/tolerance/guidance, such as toxic potency (cancer potency factor and
oral reference doses) and exposure frequency, reflect more precautionary and protective risk management.

Because there was some overlap between the categories of chemicals for which BSAF values were recommended,
the following approach was used to assign BSAFs to specific chemicals in the NSI (Table C-2). For dioxins and
furans, chemical-specific values recommended by EPA/ORD-Duluth were applied; for PCBs, the value for total
PCBs recommended by EPA/ORD-Duluth was used. When using BSAFs from USEPA (1994a), values from the
study by Cook et al. (1994) were preferred over values reported by Oliver and Niimi (1988).

Pesticides received recommendations from both laboratories. The BSAFs developed by EPA/ORD-Narragansett
were for benthic organisms and demersal (bottom-dwelling) fishes. The BSAFs developed by EPA/ORD-Duluth, on

Table C-l. EPA/ORD-Narragansett Data BSAF Distributions (kg sediment organic carbon/kg lipid)

Probability Percentile

Chemical Class

PAHs

PCBs

Pesticides

50

0.29

1.11

1.80

70

0.55

2.26

3.34

80

0.94

3.66

4.61

90

1.71

5.83

7.31

95

2.84

9.15

10.61

100

4.19 ,

16.46

22.63

C-3


-------
\|>|H-ii(li\ C

Table C-2. Conventions for Assigning BSAFs to Nonpolar Organic Compounds in NSI

Category of Chemical

Source of BSAF

BSAF Value Used in
Evaluation

Dioxins

EPA/ORD-Duhith' "pelage" chemical-specific BSAF

0.059

PCBs

EPA/ORD-Duluth' "pelagfc" BSAF for total PCBs

1.85

Pesticides

bgKw<5.5

EPA/ORD-Narragansett6 "benthic" class-specific BSAF for
50th percentile protection level

1.80

logKow^5.5

EPA/ORD-Duluth' "pelagfc" chemical-specific BSAF
if available; otherwise, use EPA/ORD-Narragansettb value

See chemical-specifc BSAF
given in Appendix D

PAHs

EPA/ORD-Narragansettb "benthic" class-specific BSAF for 50th
percentile protection level

0.29

Halogenated and other
compounds

Default value of 1 unless chemical-specific value available from
EPA/ORD-Duluth1

1.0

•Cook, 1995.
'Hansen, 1995.

the other hand, were for benthically coupled pelagic (open-water) fishes. BSAFs from EPA/ORD-Narragansett were
used for pesticides having log Kow values less than 5.5. For pesticides having log Kow values greater than or equal to
5.5, the BS AF values from EPA/ORD-Duluth were used. BSAF values selected by this approach are more appropri-
ate because food web transfer to pelagic fishes is considered to be a more important process for chemicals having
high log Kw values. Exposure through environmental media, as in direct contact with sediments by benthic organ-
isms, is a more important process for chemicals having low log Kow values. Chemicals having no recommended
BSAF values available were assigned a default BSAF of 1.

Evaluation of Tissue Lipid Content

Fish tissue lipid content enters the risk screening assessment as the normalizing factor in the numerator of the
TBP equation. Normalizing by organic carbon content removes much of the site-to-site variation in the sorption of
nonpolar organic chemicals by sediments (Karickhoff et al., 1979). In a similar manner, normalizing by lipid content
can eliminate much site and species variation in the tendency of organisms to bioaccumulate nonpolar orgjinic com-
pounds (Esser, 1986). Lipid contents can vary naturally with species, site, season, age and size of fish, and trophic
level. In addition, reported lipid contents can vary significantly depending on the analytical method (Randall et al.,
1991).

The purpose of this section is to evaluate the percent fish lipid content data from various sources and compare
these values to those selected for use in the NSI evaluation (i.e., 3.0 percent for fillets for human health TBP evalua-
tions and 10.31 for whole body wildlife TBP evaluations).

The remainder of this section describes the lipid data sources evaluated and analysis of the lipid content data.
Sources of Lipid Data

Lipid data used for comparison with the percent lipid values selected for the NSI evaluation were obtained from
three major sources:

•	EPA's water monitoring database, STORET.

•	National Study of Chemical Residues in Fish, or NSCRF (USEPA, 1992).

•	U.S. Department of Agriculture's (USDA's) Composition of Foods (Dickey, 1990).

C-4


-------
Additional sources included examples of whole fish and fillet lipid contents taken from the recent literature.

Each of the three major sources is described in the following paragraphs.

STORET

The STORET database was the single largest source of reported data on fish tissue lipid contents. Data stored
under various parameter codes for lipid content in STORET were converted into units of percentage. Some screening
of the data was performed as follows:

•	Records were retrieved from January 1990 to March 1995.

•	Reported lipid contents greater than 35 percent were eliminated because they were significantly greater than
the 90th percentile.

•	Only records having an anatomy code of "whole organism" or "fillet" were included. Records with a code of
"fillet/skin" or "edible portion" were excluded.

•	Data that appeared to be reversed (i.e., fillet percent lipid was greater than whole organism lipid) were also
not considered.

•	Also not considered were records in which the minimum and maximum were equal, or very nearly equal,
when the number of observations was large.

There is less consistency in the data obtained from STORET relative to the NSCRF data because the analyses in
STORET were conducted by numerous laboratories around the Nation. Data reported under different parameter
codes (i.e., different methods for lipids) were grouped for the analysis. Moreover, the quality of the data in STORET
is unknown. STORET data are compiled by species in Table C-3. The fishes are divided by trophic level and habitat
into four subtables (Tables C-3a through C-3d) for the combinations of trophic levels 3 and 4 and epibenthic (bottom-
dwelling) and pelagic (water column-dwelling) habitat.

National Study of Chemical Residues in Fish

The second largest database on fish tissue lipid content was available from the NSCRF (USEPA, 1992) (Table C-3).
This set of lipid analysis data was taken in conjunction with analyses for dioxins/furans. An advantage of this data-
base is that all of the lipid measurements were performed by the same laboratory using the same method. The data
were screened to exclude data for fish species for which two or fewer observations were made.

USDA Report on Composition of Foods

A summary of a relatively small database on the composition of fish and shellfish foods and food products was
available from USDA (Dickey, 1990). The section on fish and shellfish in the report coordinated by Dickey (1990)
came from an earlier USDA report by Exlcr (1987). Data presented by Exler (1987) for various fish species were
summarized from the USDA's Nutrient Data Bank (NDB). Records in the NDB are based primarily on published
scientific reports and technical journal articles. To a lesser extent, the NDB contains unpublished data from indus-
trial, government, and academic institutions under contract with the Human Nutrition Information Service. Lipids
data are given in percentage of edible portion, where "edible portion" is the part of food customarily considered
edible in the United States. Records were available for 32 fishes.

C-5


-------
Vppomlk ("

Table C-3a. Lipid Contents of Trophic Level 3, Epibenthic Fishes

Spec&a Name

Common Name

mole Fish Lipid
Content,
Percent (size)

Fillet Lipid Content,
Percent (stee)

Refereocf,
Comments

Aplodtnotus
grunntcns

freshwater drum



mean as 1.9
(1,3 to 2.5, 3 obs)

EPA (1992)

Apfodinotus
grunrtiens

freshwater drum



mean m 4.93, standard
(error = 0.103, 905
obs)

Exler (1987)

Carpoides carpio

river carpsucker

mean * 5.8
(0.5 to 15.0, 3865

obs)

mean * 4.4

(1.8 to 9.2, 184 obs)

STORET

Carpoides cyprmus

quflfcack

mean as 5.1

(0,3 to L3,0, 780
obs)

mean = 3.2

(0.4 to 4,89, 78 obs)

STORET

Catostomus aniens

Utah sucker

mean = 3.5
(1.1 to 8.2, 356
obs)

mean » 1.6

(0.1 to 6.7, 695 obs)

STORET

Colostomas
catostomus

kmgnose sucker
(FW)



0.8 to 3.8 (not given)

Owens et aL (1994)

Catostomus
catostomus

Iongnose sucker

mean = 3.9

(2.5 to 7.2, 298
obs)

mean = 7.05
(6.4 to 7.7, 32 obs)

STORET

Catostomus
columbkmus

bridgclip sucker

mean = 4.6
(0.7 to 10.4, 309
obs)



STORET

Catostomus
commersoni

white sucker



5.41 ± i.18
1.07 ± 0,23
1.36 ±0.17
0.99 ± 0.22
2.25 ± 0.65
(not given)

Servos et aL (1994)

Catostomus
commersoni

white sucker

mean = 6.1
(1.4 to 21.8, 39
obs)



USEPA (1992)

Catostomus
commersoni

white sucker

mean = 4.3

(0.2 to 12.0, 4102

obs)

mean = 1.7

(0.2 to 9.1, 586 obs)

STORET

Catostomus
commersoni

white sucker



mean = 2.32
(standard error =
0.069, 157 obs)

ExJcr (1987)

Catostomus
macrochetius

largescafc sucker

mean = 6.7

(0.3 to 13.0, 752
obs)

mean = 1.6

(0.1 to 5.26, 482
obs)

STORET

Catostomus
occidentals

Sacramento sucker

mean = 9.8

(1.7 to 18.5, 3 obs)



USEPA (1992)

Cottus cognatus

scutpin (FW)

8 (5.4 g)



USEPA (1994a)

Cyprinus carpio

carp

9 (15 g)



Cook et aL (1991)

Cyprtnus carpio

carp

18.7 (69.5 g)
15.7 (56.0 g)
13.0 (37.5 g)
16.6 (36.5 g)

17.5 (29.0 g)



Kuehlet aL (1987)

€-6


-------
Nutioiuil .Sediment Quality JS«r\cy

Table C-3a. (Continued)

Species Name

Common Name

Whole Fish Lipid
Contenfl,
Percent (siae)

Fillet Lipid Content,

Percent (size)

Reference,
Comments

Cyprinus carpio

carp

18.7 (69.5 g)
15.7 (56.0 g)
: 13.0 (37.5 g)
16.6 (36.5 g)
17.5 (29.0 g)



Kuehl et aL (1987)

Cyprinus carpio

carp

mean = 9.3
(0.5 to 25.1, 145
obs)

mean = 9.0
(2.0 to 19.6, 6 obs)

USEPA (1992)

Cyprinus carpio

carp

¦ mean = 6.5
(0.3 to 17.0,70002
obs)

msan = 4.3

(0.02 to 21.6, 16139

obs)

STORET

Cyprinus carpio

carp



mean = 5.60
(standard error =
0.207, 163 obs)

Exler (1987)

Ctenophyaryngodo-
n idclla

grass carp



msan = 5.2
(3 obs)

USEPA (1992)

Erimyzon oblongus

creek chubsucker

mean = 3.9
(3.9 to 4.0, 3 obs) .



USEPA (1992)

Hypentelium
nigricans

northern hogsucker

mean = 4.4
(0.8 to 8.98, 637
obs)

mean = 0.7

(0.S to 0.99, 70 obs)

STORET

Ictalurus furcatus

blue catfish

mean = 7.3
(5.3 to 10.4, 5 obs)

mean = 2.7
(2.0 to 3.0, 4 obs)

USEPA (1992)

Ictalurus furcatus

blue catfish



mean = 6.0

(1.5 to 12.0, 56 obs)

CTAOCT
O lUKbi

fctalurus melus
(Ameiurus melus)

black bullhead

mean = 2.9

(0.9 to 6.2, 911 obs)

mean = 1.4

(0.15 to 5.1, 573 obs)

STORET

Ictalurus natalis
(Ameiurus natalis)

yefow bullhead

mean = 2.8

(0.5 to 7.5, 235 obs)

mean = 0.96
(0.1 to 3.2, 294 obs)

STORET

Ictalurus nebulosus
(Ameiurus
nebulosus)

brown bullhead

mean = 2.2

(1.3 to 4.1, 133 obs)

mean= 1.5

(0.4 to 3.3, 107 obs)

STORET

Ictalurus punctatus .

channel catfish

mean = 9.8

(3.4 to 23.0, 22 obs)

mean = 5.1

(1.1 to 11.5, 17 obs)

USEPA (1992)

Ictalurus punctatus

channel catfish

mean = 7.1

(0.3 to 15.0,7512

obs)

msan = 5.1

(0.2 to 17.3, 20655

obs)

STORET

Ictalurus punctatus

channel catfish



mean = 4.26
(standard error =
0.417, 59 obs)

Exler (1987)

Ictiobus bubalus

smaltouth bufiafo

mean = 5.7
(2.2 to 11.9, 6 obs)



USEPA (1992)

C-7


-------


Table C-3a. (Continued)

Spedes Name

Common Name

Whole Fish Lipid
Content,
Percent (she)

Fillet lipid Content,
Percent (size)

Reference;
Comments

fcliobus bubalus

smalhmoush buffaJo

mean = 9.7
(2-8 to 17.3, 886
obs)

mean = 4.8

(0.2 to 14.5, 595 obs)

STORET

Ictiobus cyprinettus

bigmouth buffalo

mean = 15.1
(5.7 to 22.6, 3 obs)



USEPA (1992)

la lob us cyprincllus

bigmouh buffalo

mean = 5.8
(0,4 to 16.2, 675
obs)

mean = 4.1

(0.3 to 15, 1678 obs)

STORET

Ictiobus niger

black buflaJo



trean = 3.5
(1.2 to 7.1, 42 obs)

STORET

Minytmma
melmops

spotted sucker

mean = 4.5
(0.9 to 7.4, 9 obs)



USEPA (1992)

Mwytrcma
mclanops

spotted sucker

mean = 3.7
(0.7 to 5.9, 188

obs)

mean = 1.5

(0.9 to 3.2, 197 obs)

STORET

Moxostoma
artisurum

silver redhorse

mean = 8.2
(6.2 to 8.5, 180
obs)

mean = 2.1
(1.3 to 2.7, 7 obs)

STORET

Moxostoma
carina! um

river redhorse

mean = 5.1
(1.9 to 5.9, 193

obs)

mean = 1.3

(0.5 to 2.4, 170 obs)

STORET

Moxostoma
duquesnei

black redhorse

mean = 5.0
(0.3 to 9.7, 1774
obs)

mean = 0.97
(0.7 to 1.8, 58 obs)

STORET

Moxostoma
erythrurum

golden redhorse

mean = 6.0
(0.8 to 16.1, 2018

obs)

mean =1.8

(0.6 to 2.8, 154 obs)

STORET

Moxostoma
macrolepidotum

shotthead redhorse

mean = 19.8
(10.8 to 31.9, 4
obs)



USEPA (1992)

Moxostoma
macrolepidotum

shorthead redhorse

mean = 6.5
(0.4 to 10.9, 683
obs)

mcan = 3.0

(1.4 to 13.5, 342 obs)

STORET

Mugil cephalus

striped muBet



mean = 3.79
(standard error =
0.357, 43 obs)

ExJer (1987)

Mylochcilus

caurmus

peanciah

mean = 11.0 (9.36
to 12.91, 162 obs)



STORET

Ptychochetlus
ottgoni

northern squawfish

mean = 5.6 (0.8
to 12.0, 812 obs)

mean = 1.3

(0.7 to 3.0, 117 obs)

STORET

Ptychochetlus

squawfeh



mean = 2.2
(0.5 to 3.0, 7 obs)

USEPA (1992)

Scaphirhynchus
platorhynchus

shovelnose sturgeon



mean = 7.4

(1.1 to 20.3, 392 obs)

STORET

C-8


-------
National Sejtliim>nt Quality Suvma

Table C-3b. Lipid Contents of Ttophie Level 3, Pelagic Fishes

Species Name

Common Name

Whole Fish lipid
Content,
Percent (siae)

Fillet Lipid Content,
Percent (sia*>)

Reference,
Comments

Acipenser sp.

sturgeon (unknown)



mean = 4.04
(7 obs)

Exler (1987)

Acmcheilus
atutaceus

chiselmoulh

mean = 5,0
(3.2 to 6.8, 47 obs)

mean = 0,55
(0,19 to 1.00, 91 obs)

STORET

Atosa

pseudoharengus

ale wife

7 (32 g)



USEPA (1994a)

Alosa

pseudoharengus

ale wife

mean = 8.9
(3.7 to 15.2, 128
obs)



STORET

Alosa sapidissima

Amerfcan shad

mean = 6.55
(5.9 to 7.6, 270
obs)



STORET

Alosa sapidissima

American shad



mean = 13.77
(standard error = 1.00,
11 obs)

Exler (1987)

Anguilla mstrata

American eel



mean = 11.66
(standard error =
0.885, 14 obs)

Exler (1987)

Aphdinotus
grunniens

freshwater dnim

mean = 5,5
(1.0 to 19.7, 574
obs)

mean = 4.8

(0.3 to 21.2, 459 obs)

STORET

Archosargus
pmbalocephatus

sheepshcad



mean = 2.41
(standard error =
0.040, 5 obs)

Exler (1987)

Coregonus artedii

cisoo (lake herring)



mean = 1.91
(standard error =
0.149, 69 obs)

Exler (1987)

Coregonus
clupeaform

lake whitefish



mean = 5,86
(standard error =
0.451, 68 obs)

Exler (1987)

Coregonus hoyi

bloater

mean = 21.1
(16 to 25.5, 52 obs)

msan = 8.3

(3.2 to 17.0, 98 obs)

STORET

Domsoma
cepedianum

gizzard shad

mean = 7.4
(1.3 to 18.0, 189
obs)



STORET

Domsoma
petenense

threadfin shad

mean = 3.0
(0.5 to 18.0, 9 obs)



STORET

Gadus

macmcephalus

true or Pacific cod



mean = 0.63
(standard error =
0.031, 18 obs)

Exler (1987)

Hiodon absoides

goldcye

msan = 3.2
(3.5 to 2.8, 74 obs)



STORET

C-9


-------
VppcndK ('

Table C-3b. (Continued)

Spedes Name

Common Name

Whole Hsh Lipid
Content,
Percent (size)

Fillet Lipid Content,
Percent (she)

Reference,
Comments

Plalygobia

(Hybopsis h
database) gracilis

fbthead chub



mean = 3,3

(0.68 to 8.14, 75 obs)

STORET

Lepomis auritis

redbreast sunfish

mean = 3.6
<1.3 to 8.1, 550
obs)



STORET

Lepomis cyanettus

green stnfish

mean = 3.2
(2.2 to 7.8, 376
obs)



STORET

Lepomis gibbosus

punpkinsccd

mean = 3.9
(2.2 to 7.7, 126
obs)



STORET

Lepomis gibbosus

purrpkircced



mean = 0.70
(standard error =
0.071, 8 obs)

Exler (1987)

Lepomis megalotis

longear sunfish

mean = 2.8
(1.0 to 7.2, 536
obs)



STORET

Osmerus mordax

rainbow smelt

4 (16 g)



USEPA (1994)

Osmerus mordax

rainbow smelt



mean = 2.42
(standard error =
0.107, 52 obs)

Exler (1987)

Pimephales
promelas

fathead minnow

19 (1 8)



Cook et al (1991)

Lepomis
macmchirus

bliegffl sunfish

mean = 3.5
(2.4 to 4.6, 4 obs)



USEPA (1992)

Lepomis
macmchirus

bliegill smfish

mean = 4.4
(0.1 to 8.7, 1034
obs)



STORET

Lota lota

bubot



0.35 to 0.7

Owens et al. (1994)

Lota lota

burbot



mean = 0.2
(0.1 to 0.3, 18 obs)

STORET

Lota lota

burbot



mean = 0.81
(standard error =
0.059, 13 obs)

Exler (1987)

Oryzias latlpes

medaka

8 (0.175 g)



Schirieder et al
(1992)

Phoxinus
etythrogaster

southern redbeDy
dace

mean = 5.6
(2.2 to 10.0, 762
obs)



STORET

C-IO


-------
N:ili»ii:il Svriimonl Qu:iJit\ Surrey

Table C-3b. (Continued)

Species Name

Common Name

WlMjIe Fish Lipid
Content,
Percent (sire)

Fillet Lipid Content,
Percent (size)

Reference,
Comments

Pomoxis annularis

white crappie



mean = 1.0

(0.5 to 2.0, 7 obs)

USEPA (1992)

Pomoxis annularis

white crappb

mean = 2.1
(0.4 to 5.8, 622
obs)

mean = 0,4

(0.08 to 2.6, 936 obs)

STORET

Pomoxis
nigmmacuhuus

black crappie



mean = 1.1
(0.5 to 1.5, 3 obs)

USEPA (1992)

Pomoxis
nigmmaculatus

black crappie

mean = 2.7
(0.7 to 8.4, 457
obs)

mean = 1.4

(0.13 to 5.3, 118 obs)

STORET

Pmsopium
williamsoni

mountain whitefch

mean = 8.5
(0.5 to 13.8, 327
obs)

mean = 1.6,

(0.2 to 4.1, 532 obs)

STORET

Pmsopium
williamsoni

mountain whitefish



3.4 to 11.8
(not given)

Owens et al (1994)

Hichanhonius
baUeatus

rcdsidc shiner



mean = 0.9

(0.85 to 0.96, 50 obs)

STORET

Sebastes
auriculatus

brown rockfish



mean = 1.57
(81 obs)

Exler (1987)

Sebastes marinus

redfish



mean = 1.63
(standard error =
0.092, 208 obs)

Exler (1987)

Semotiius
atmmacula

creek chub

mean = 3.9
(1.0 to 5.0, 815
obs)



STORET

Semotiius
corporalis

feBfish

mean = 1.9
(0.25 to 3.9, 100
; obs)



STORET

Table C-3c. Lipid Contents of Trophic Level 4, Epibenthic Fishes

Species Name

Common Name

Whole Fish Lipid
Content,
Percent (size)

Fillet Lipid
Content, Percent

Reference,
Continents

PyloMctis olivaris

flathead catfish

mean = 3.1 (0.5 to
8.1,829 obs)

mean = 3.0 (0.2 to
21,1,1315 obs)

STORET

Pylodictis olivaris

flathead catfish

mean = 6.0
(1.6 to 8.7,3 obs)

mean= 1.9
(0.6 to 3.1,4 obs)

USEPA (1992)

C-ll


-------


Table C-3d. Lipid Contents of Trophic Level 4, Pelagic Fishes

Species Name

Common Name

Whole Fish Lipid
Content,
Percent (size)

Fillet Lipid Content,
Percent (size)

Reference,
Comments

Ambtoplites
rupestris

rock bass



mean = 1.0
(0.8 to 1.2, 3 obs)

USEPA (1992)

Ambtoplites
rupestris

rock bass

mean = 2.3
(0.6 to 4.4, 759
obs)

mean = 0.7

(0.4 to 0.98, 129 obs)

STORET

Amia caiva

bowfin



mean = 0.5

(0.04 to 1.4, 230 obs)

STORET

Ceniropristis
striata

black sea bass



mean — 2.00
(standard error =
0.221, 40 obs)

Exler (1987)

Esox lucius

northern pike



mean = 1.4
(0.6 to 2.6, 5 obs)

USEPA (1992)

Esox lucius

northern pike

mean = 1.9
(0.1 to 9.8, 810
obs)



STORET

Esox lucius

northern pike



mean = 0.69
(standard error =
0.005, 224 obs)

Exler (1987)

Esox niger

chain pickerel



mean= 1.3
(0.6 to 2.0, 5 obs)

USEPA (1992)

Leiostamus
xanthurus

spot

mean = 5.2
(3.3 to 7.9, 300
obs)



STORET

Lciastomus
xanthurus

spot



mean = 4.90
(standard error = 2.93,
10 obs)

Exler (1987)

Littjanus
campechanus

red snapper



1.34(55 obs)

Exler (1987)

Micropogonias
undulatus

Altaic croaker



3.17

(standard error =
0.529, 8 obs)

Exler (1987)

Micropterus
dolomteu

smaflmouth bass



mean = 1.6
(0.8 to 4.4, 19 obs)

USEPA (1992)

Micropterus
dolomicu

smaUmouth bass

mean = 3.4
(0.3 to 8.8, 1166
obs)

mean = 0.6

(0.01 to 2.3, 848 obs)

STORET

Micropterus
punctulatus

spotted bass



msan = 2.8
(0.9 to 4.5, 4 obs)

USEPA (1992)

Micropterus
punctualtus

spotted bass

mean = 2.4
(0.6 to 4.9, 322
obs)

mean = 0.7

(0.1 to 1.8, 353 obs)

STORET

C-12


-------
National Sediment Quality Svmev

Table C-3d. (Continued)

Species Name

Common Name

Whole Fish lipid
Content,
Percent (size)

Fillet lipid Content,
Percent (size)

Reference,
Comments

Micmpterus
sal mo ides

largemouih bass



mean = 1.6
(0.4 to 7.6, 54 obs)

USEPA (1992)

Micmpterus
salmoides

large mouth bass

mean = 4.1

(0.3 to 10.6, 2924

obs)

msan = 0,7
(0.04 to 9.2, 4548
obs)

STORET

Momne amcricana

white perch

mean = 4.5
,(2.6 to 7.1, 249
obs)



STORHT

Momne chrysops

white bass



mean = 2.7
(0.7 to 4.8, 11 obs)

USEPA (1992)

Momne chrysops

white bass

mean = 4.6
(0.3 to 15.4, 615
obs)

mean — 3.9

(0.01 to 8.1, 847 obs)

STORET

Momne saxatitis

striped bass



mean = 2.33
(standard error =
0.381, 14 obs)

Exler (1987)

Oncorhynchus
gorbuscha

pink salmon



mean = 3.45
(standard error =
0.141, 144 obs)

Eifer (1987)

Oncorhynchus
kisutch

coho salmon



mean = 2.7

(0.4 to 10.7, 383 obs)

STORET

Oncorhynchus
kisutch

coho salmon



rrean = 5.92
(standard error =
0.162, 217 obs)

Exler (1987)

Oncorhynchus
mykiis

rainbow trout

11 (35 g)



Branson et aL
(1985)

Oncorhynchus
my kiss

rakibow trout



mean = 5.0
(4.1 to 5.6, 3 obs)

USEPA (1992)

Oncorhynchus
nsrka

sockeye salmon



mean = 8.56
(standard error =
0.392, 48 obs)

Exler (1987)

Oncorhynchus
tshawytscha

Chinook salmon

mean = 3.7
(2.4 to 5.1, 52 obs)

mean = 2.2

(0.04 to 17.7, 1957

obs)

STORET

Oncorhynchus
tshawytscha

Chinook salmon



mean = 10.44
(standard error =
1.692, 10 obs)

Exler (1987)

Petva ftavescens

yelfow perch

irean = 3.6
(1.2 to 9.1, 112
obs)

mean = 0.5

(0.1 to 4.6, 280 obs)

STORET

Pomatomus
saltatrix

blue fish



mean = 4.27
(3 obs)

Exler (1987)

C-13


-------


Table C-3d. (Continued)

Specks Name

Common Name

Whole Fish Lipid
Content,
Percent (size)

Fillet lipid Content,
Percent (size)

Reference,
Comments

Salmo clarki
(Onchorhynchus

clarki)

cutthroat trout



mean = 1.0

(0.2 to 1.7, 378 obs)

STORET

Salmo gairdneri
(Onchorhynchtu
mykiss)

rainbow trout



mean = 3.36
(standard error =
0.256, 24 obs)

Exfcr (1987)

Salmo salar

Atlantic salmon



mean = 6.34
(standard eiror = 1.72,
7 obs)

Exler (1987)

Salmo trutta

brown trout



mean = 4.0
(1.6 to 8.1, 6 obs)

USEPA (1992)

Salmo trutta

brown trout

mean = 6.0
(1.5 to 8.9, 112
obs)

mean = 5.0
(0.14 to 14.8, 741
obs)

STORET

Salvelinus
namaycush,
Oncorhynchus mykiss,
Oncorhynchus spp.

salmo nkls

11 (2410 g)



USEPA (1994a)

Satvelinus matma

DoBy Varden

mean = 7.1
(2.1 to 9.9, 3 obs)



USEPA (1992)

Salvelinus namaycush

lake trout

mean = 15.9
(12.6 to 18.3, 42
obs)

mean = 7.8

(2.5 to 20.0, 1883

obs)

STORET

Scom.bcromorus
cavall

king mackerel



mean = 2.00
(standard error =
0.188, 6 obs)

Exler (1987)

Scombemmorus
macula

Spanish mackerel



mean = 6.30
(standard error=3.810,

3 obs)

Exfcr (1987)

Stizostedion
canadense

sauger

mean ss 6.0
(0.8 to 16.3, 139
obs)

mean s 1.7

(0.3 to 10.0, 195 obs)

STORET

Stizostedion vitreum

walleye



0.6 to 0.7

Owe® et al (1994)

Stizostedion vismum

walleye

mean = 6.2
(0.3 to 15, 1089
obs)

mean =1.3

(0.3 to 6.0, 440 obs)

STORET

Stizostedion vitreum

walleye



mean = 1.22
(standard error =
0.162, 14 obs)

Exler (1987)

Stizostedion vixreum

walleye



msan = 1.6
(0.7 to 2.6, 13 obs)

USEPA (1992)

C-14


-------
"S;i(i
-------
Table C-4a. Lipid Analysis - STORET

Analysis

Matrix of Fishes Included in Avenge

Tissue/
Organ

Lipid Content, %

"Ifcophlc
Level

Position in Water
Column

Mobility

Habitat

Mca-
n

Standard
Error

Number of
Observatio-
ns

Range

3

4

Demersal

Pelagic

Resident

Migratory

Freshwat-
er

Saltwater

A

•

•

•

•

•

•

•

•

whole

5.97



113,978

0.1-26.7

B

•

•

•

•

•



•



whole

5.97

0.010

110,998

0.1-26.7

C



•



•

•

•

•

•

fillet

2.5



13,293

0.01-20

D



•



•

•



•



fillet

0.753

0.010

6793

0.01-10

E

•

•

•



•



•



whole

6.33

0.011

91867

0.22-26.7

F

•

•



•

•



•



whole

3.757

0.020

13025

0.10-16.3

G

•

•

•



•



m



Sllet

4.49

0.018

42687

0.02-24

H

•

•



•

•



•



fillet

1.06

0.021

9378

0.01-21.-
07


-------
Table C-4b. Lipid Analysis - NSCRF

Analysis

Matrix of Fishes Included in Average

Tissue/
Organ

lipid Content, %

TVophic
Level

Position in Water
Coliam

Mobility

Habitat

Mean

Standard
Em»r

Number of
Observations

Range

3

4

Demeisal

Pelagic

Resident

Migratory

fteshmter

Saltwater

A

•

•

•

•

•

•

•

•

whole

8,5



249

0.5-31.9

B

•

m

•

•

•



•



whole

8.6

0.328

246

0.5-31.9

C



m



•

•

•

•

•

fillet

1.9



122

0,4-8.1

D



•



•

•



•



fillet

1.6

0.116

103

0.4-7.6

E

•

•

•



•



•



whole

8.8

0.338

233

0,5-31.9

F

•

•



•

•



m



whole

4.6

1.02

7

1.6-8.7

0

•



•



•



•



fflet

4.9

0,697

34

0.5-19.6

H

•

•



•

•



•



ffllet

1.6

0,106

117

0.4-7.6

Data for fillets and whole fish were evaluated separately. All analyses except MA" were of fishes in the NSI exclusively. Summary statistics reported include the mean, stao<£ard error, range, and number of
observations. The matrices in Tables C-4a and C-4b indicate the categories of fisbes averaged. The average of edible portions from USDA data was 4.1 percent lipid.

The mean fillet percent lipid content for various groups of fish species is the S TO RET database ranged from 0.753 to 4.49 percent; in the NSCRF, mean fillet values ranged from 1.6 to 4.9 percent. The mean
whole-body percent lipid content for various groups of fish species in the S TO RET database ranged from 3.757 to 6.33 percent; in the NSCRF, mean whole-body values ranged from 4.6 to 8.8 percent.


-------
References

Branson, D.R., I.T. Takahashi, W.M. Parker, and G.E. Blau. 1985. Bioeoncentration of 2, 3,7,8-
tetrachlorodibenzo-p-dioxin in rainbow trout. Environ. Toxicol. Chem. 4:799-788.

Cook, P.M. 1995. Pelagic BSAFs for NSI methodology. Memorandum from P.M. Cook, ERL, Duluth, to C. Fox,
USEPA Office of Water, March 29,1995.

Cook, P.M., G.T. Ankley, R.J. Erickson, B.C. Butterworth, S.W. Kohlbry, P. Marquis, and H. Corcoran. 1994. The
biota-sediment accumulation factor (BSAF): Evaluation and application to assessment of organic chemical
bioaccumulation in the Great Lakes. In preparation, (cited in: USEPA, 1994a)

Dickey, L.E. 1990. Composition of foods, raw, processed, prepared—1990 supplement. Agriculture Handbook 8,
1990 Supplement U.S. Department of Agriculture, Human Nutrition Information Service, Washington, DC.

Esser, H.O. 1986. A review of the correlation between physicochemical properties and bioaccumulation. Pestic. Sci.
17:265-276. (as cited by Randall et al., 1991).

Exler, 1.1987. Composition of foods: Finfish and shellfish products. Agriculture Handbook No. 8-15. U.S. Depart-
ment of Agriculture, Human Nutrition and Information Service, Washington, DC.

Flint, R.W. 1986. Hypothesized carbon flow through the deep water Lake Ontario food web. J. Great Lakes Res.
12:344-354.

Hansen. 1995. Assessment tools that can be used for the National Sediment Inventory. Memorandum from D.J.
Hansen, ERL, Narragansett, to C. Fox, USEPA Office of Water, February 28,1995.

KarickhofF, S.W., D.S. Brown, and T.A. Scott. 1979. Sorption of hydrophobic pollutants on natural sediments.
Water Res. 13:241-248.

Kuehl, D.W., P.M. Cook, A.R. Batterman, D.B. Lothenbach, and B.C. Butterworth. 1987. Bioavailability of
polychlorinated dibenzo-p-dioxins and dibenzofurans from contaminated Wisconsin River sediment to carp.
Chemosphere 18:1997-2014.

Oliver, B.G., and A.J. Niimi. 1988. Trophodynamic analysis of polychlorinated biphenyl congeners and other
chlorinated hydrocarbons in the Lake Ontario ecosystem. Environ. Sci. Technol. 22:388-397.

Owens, J.W., S.M. Swanson, and D.A. Birkholz. 1994. Bioaccumulation of 2, 3,7, 8-tetrachlorodibenzo-p-dioxin,
2,3,7,8-tetrachlorodibenzofuran and extraetable organic chlorine at a bleached kraft mill site in a northern
Canadian river system. Environ. Toxicol. Chem. 13:343-354

Randall, R.C., H. Lee II, R J. Ozretich, J.L. Lake, and RJ. Praell. 1991. Evaluation of selected lipid methods for
normalizing pollutant bioaccumulation. Environ. Toxicol. Chem. 10:1431-1436.

Tracey, G.A. and D.J. Hansen. 1996. Use of biota-sediment accumulation factors to assess similarity of nooionic
organic chemical exposure to benthically-coupled organisms of differing trophic mode. Arch, Environ. Contam.
Toxical. 30:467-475.

Schmeider, P., D. Lothenbach, R. Johnson, R. Erickson, and J. Tietge. 1992. Uptake and elimination kinetics of
'H-TCDD in medaka. Toxicologist 12:138.

Servos, M.R., S.Y. Huestis, D.M. Whittle, G. J. Van Der Kraak, and K.R. Munkittrick. 1994. Survey of receiving-
water environmental impacts associated with discharges from pulp mills. 3. Polychorinated dioxins and furans
in muscle and liver of white sucker (Catostomus commersoni). Environ. Toxicol. Chem. 13:1103-1115.

C-18


-------
Niitinir.il St'tliinciH Quality Survey

USEPA. 1990. Lake Ontario TCDD bioaccumulation study—Final report. U.S. Environmental Protection Agency,
Region 2, New York, NY.

USEPA. 1992. National study of chemical residues infish. 2 vols. EPA 823-R-92-008a,b. U.S. Environmental
Protection Agency, Office of Science and Technology, Washington, DC.

USEPA. 1994a. Great Lakes Water Quality Initiative technical support document for the procedure to determine
bioaccumulation factors—July 1994. EPA-822-R-94-002. U.S. Environmental Protection Agency, Office of
Water, Office of Science and Technology, Washington, DC.

USEPA, 1994b. Health effects assessment summary tables FY 1994. Supplement number 2. EPA/540/R-94/114,
NTIS PB94-921102. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response,
Washington, DC. November.

USEPA, 1995. Integrated Risk Information System (IRIS). Online. U.S. Environmental Protection Agency, Office
of Health and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH.

C-19


-------
Appendix ('


-------
Appendix D

Screening Values for
Chemicals Evaluated

Sediment Concentrations

Table D-l presents the screening values used in the evaluation of NSI sediment chemistry data. Values listed
in this table are in parts per million (ppm) except for the values for EPA draft sediment quality criteria
(SQC^) and sediment quality advisory levels (SQALoc), which are in micrograms per gram (fig/g) organic
carbon. These values were multiplied by the organic carbon content (f^) of the sediment sample, when known, or
the default value if unknown (f^. = 0.01). SQALs used in this analysis were calculated specifically for use in the
screening analysis of NSI data. Effects range-low (ERL) and effects range-median (ERM) values were taken from
Long et al. (1995). Apparent effects threshold-low (AET-L) and apparent effects threshold-high (AET-H) values
listed are values that have been normalized to dry weight. AET-Ls and AET-Hs were taken from Barrick et al.
(1988). Threshold effects levels (TELs) and probable effects levels (PELs) were taken from FDEP (1994).

Fish Tissue Concentrations

Fish tissue concentrations are presented in the right columns of Table D-l. EPA risk levels were calculated for
both a human health cancer risk of 10 s and a noncancer hazard quotient of 1 (USEPA, 1995a, b). Other available
EPA sources were consulted as necessary for risk-based concentrations to be used in a screening analysis, including
the Environmental Criteria and Assessment Office (as cited in USEPA, 1995c). FDA guidance/action/tolerance
levels were obtained from the FDA Office of Seafood (DHHS, 1994; 40 CFR 180.213a and 180.142; USFDA, 1993a,
b, c, d, e).

Biota-Sediment Accumulation Factors

The final column in Table D-l presents the biota-sediment accumulation factors (BSAFs) used in the analysis.
The BSAFs were adopted for use in the theoretical bioaccumulation potential (TBP) calculations that represent
potential concentrations that might occur in tissues of fish exposed to contaminated sediments. The methodology
used in deriving BSAFs and other parameters used in the TBP calculations are described in Appendix C of this
document.

Methodology for Combining Chemical Data Using a Risk-Based Approach

Several screening values, as provided in the original source documents, refer to groups of chemicals. The
majority of the data included in the NSI exist as specific chemicals. To perform a screening analysis that accommo-
dates the way the data exist in the NSI and provides a reasonably conservative risk-based approach, chemical data
were combined in particular cases.

Two of the chemical groups affected by this approach are polychlorinated biphenyls (PCBs) and dioxin com-
pounds. The data for PCBs in the NSI occur in three ways: (1) total PCBs, (2) PCB congeners, and (3) PCB aroclors.
The data for the PCB congeners were summarized (excluding as appropriate the lower chlorinated homologs that
may be present as laboratory artifacts) to provide a total PCB value where one was not provided by the original
database. This summarization enabled comparisons to the screening values available for total PCBs. Aroclor-spe-

D-l


-------
Table D-l. Screening Values for Chemicals Evaluated

GUTDEUNE VAUJESINTENDED ONLY FOR SCREENC-LEVEL HAZARD COMPARISON AMONG CHEMICALS
Ma^BcQretverUadeffcvtectire efS*daxBt*taGh*aLo«ftJfe«Ofcratrik

1

















0.2

a



1.0

1597260$

AkchbtfLawo

1

















1.3

110





116063

AHkarts/Tcnisk





















11





309002

Aidrin

1,3

















0.0063

0,32

0.3

1.80*

62533

AnSnc



















19







120127

Anthracene

!



.0853

1.1

.96°

13*



0.0469

0.245



3200



0.29s

999999933

Anthracene & Phemtnlhttae

1

180

.0853

1.1

.96*

6.9*

180

0.0469

0.245



3200



0.29*

7440360

Antta»ny









150*

200*









4.3





7440382

Ancrae

2



8.2

70

57b

700°



7.24

41.6

0.062

3.2

68



1912249

Atraane



















0.49

380





7440393

Dirkori





















750





92875

Bcraadjoc



















0,00047

32





71432

Benzene

1,6











5.7





3.7





1.0

56553

Bcrau{B)anthraa:nc

1



.261

1,6

1,6s

5,1**



0.0748

0.693

0.15





0.29*

999999955

Bcmo(a)anthniccne.'Cbfyse-
ne

1



.261

1.6

l£*

5.1**



0,0748

0.693

0.15





0.29"*

50328

Benaj(i»)pyreae

1



.43

1.6

1.6°

3.6"



0.0888

0.763

0.015





0.29s

205992

Berne (b)&ioRinthf dc

1







3.6®

9.9*







0.15





0.29*

191242

Beamfj^Tperyfct*

1







.72*

2.6*















207089

QetQofk)lbofiusbrnB

1







3.6°

9.9*







!.5





0.291"

65850

Qcszoicacid









.65°*

.76*









43000





98077

IkmotricWoride

1

















0.0083








-------
Table D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR SCREENING-LEVEL HAZARD COMPARISON AMONG CHEMICALS
Ma; Be Orer» or Uiiieiprfcdhe of Seincii at * Given Location Depodkie on Site-Specific CtaStlon

CAS Number

Name

Code

Sediment Cotxcttratfon

Fbh Tissue CooceoCnitk-o (fpo)

BSAF
(unities*)

SQC-

ER-L

ER-M

{ppuj

AET-L

(P»)

AET-n
(f!*^

SQM^
WzJ

TEL
)

PEL
<&**$

Cbd»&
a EPA
Risk I4H

EPA
Neacaac*-
r

Haxsitl
Qnotteet

= 1

FDA
Guidance/

Action/
Tokmocc
Level

100516

Benzyl alcohol









.073"

Ml*









3200





100447

Benzyl cfck>ridt

1

















0.63







744041?

BeryEaim



















.025

54





319846

BHC, *3pha-

u













0,00032

0.00099

0.017



0.3

1.80"

319857

BHC, bcta-

u













0.00032

0.00099

0.060



03

1,80s

319868

BHC, dtKa»

U,6











13

0.00032

0.00099

0,060



03

IJQ*

58899

BHC, gamra- (Lindane)

13. 6











0.37

0,00032

0.00099

0.083

3.2

03

1.80*

60873I

BHC, technical grade

1,3











0.37

0.00032

0.00099

0,060

3.2

03

1.81?

92524

Blphenyl

1,6











110







540



0.29*

111444

Bs(2-chiorocthyl)elhcr

1

















0.098







108601

Bis(2-c^orois»prDpy3)cther

1

















1.5

430





11781?

Bts(2-ethylh5xyf)phthala'i

1.6







1.3*

1.9°



0.182

2.65

7.7

220



1,0

542881

Bb(chioromcthy!}ether



















0.00049







7440428

Bvrun





















970





75274

Bro mcxlkhloro methane

I

















1.7

220





74839

Broisoiaethaoe

1



















15





101553

Broraophenvi phenyl ethen 4-

1,6











130







620



1,0

1689845

Brortojcynil





















220





85687

Biayl berasl ph?hatate

1.6







.9*

.9°

1100







2200



1.0

7440439

Cadmium

2



1.2

9,6

5.1*

9.6°



0.676

4,21



5.4

3



63252

CaxttaryVSevin





















1100





1563662

Ca^ftrarvlanuUn





















54





75150

Caibon disulfide





















1100





133904

CMorarcben





















160





57749

Chkutiane

13













0.00226

0.00479

0.083

0.65

03

4.77*

5103719

C Mordant, a^pha(ds>-

13













0.00226

0,00479

0.083

0.65

03

4.77*


-------
Table D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR 6CRTO%TNC-LEVEL HAZARD COMPARISON AMONG CHEMICALS
May Be Onr» «eVtArftvUtAre 01SttiatdL •! • Given Lae»lfea Dcpeedac mSUt-Sf**16c GMdtfens

CAS Number

Chtnbl Nmbb

Oxfc

Sedkseot C»se«aCmtioa

flibTluue CoBccatiafibB (ffcci

BSAF
(uafOtfts)

5QC„
(WfeJ

ER-L

ER-M
(pp™)

AET-L

AET-H

(Jr*»)

SQAL„

TEL
(PP»»

PEL
(n*!0

Cboetft
» EPA
Risk 104

EPA
Neaesaec*
r

QuBtkui
-1

FDA
CtfdtBee/
Artfarf
TOeiwec*
Level

5103742

ChJortUns, bet*(tma>-

U













0.00226

0.00479

0.083

0.65

03

2*

5566347

Chterdww, gsjrmi(6ra«it)-

w













0.00226

0.00479

0.083

0.65

03

2.22*

999999247

CHonJ«j»-NonBchiof(crij>-

13













0.00226

0,00479

0.083

0.65

03

4.77*

999999248

ChfcrdJU*-Nor«cHor•

U













0.00226

0.00479

0.083

0.65

0.3

4.77*

108907

Chbrobcrecac

1,6











82







220



1.0

510156

CKbfobersrskte



















0.40

220





750Q3

Chksme thane

!



















4300





75014

CHbmeihefle

1

















0.057







110758

Chbrocthyfviriyi ctfwr, 2-

1



















270





74873

Chfarorottbane

I

















8.3







91587

ChkHoruphthtknc, 2-

I



















860





93578

Chkjrophenol, 2-





















54





2921882

Chbjpyrifos'DwsbKi

1



















32



1.80*

7440473

Chromium

2



81

370

260?

270-



52.3

160



54

n



218019

Chryseae

1



-384

2.8

Z&

9,2**



0-108

0.846

15





0.29s

7440508

Copper





34

270

390"

1300*



18.7

108



400





IGS394

Cresol n>-









.63*®

,72*









540





95487

Creaol 0-









,63^

.72"









540





106445

Crcsot, p-









.67
-------
Table D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR SCREENING-LEVEL HAZARD COMPARISON AMONG CHEMICALS
May Be Orer* or Underptotective of Sediment at a Gtren Location Depending on Site-Specific Conditions

CAS Nunter

QwttJcal Name

Code

Sedbnent Concentration

Fish tissue Concentration (ppn)

BSAF
(uaftless)

SQC,
(MSfeJ

ER-L
(Pf*n)

ER-M

(ppra)

AET-L
(Pl*n)

AET-n
(ppm)

SQAL^

TEL
(ppra)

PEL
(Piro)

Concea
= EPA
Rbk It*

EPA
None* nee-
r

Hazard
Quotient
= 1

FDA
Guidance/

Action'
Ibfenacc
Level

3424826

DDE, o,p'-

13



.0022

.027

.009s

.015'



0.00207

0.374

0.32



5

7.7*

72559

DDE, p, p'-

13



,0022

.027

.009s

.015*



0.00207

0.374

0.32



5

7.7*

789026

DDT, o,p'-

1,3



.00158

.027

.034b

.034*



0.00119

0.00477

0.32

5.4

5

1.67*

50293

DDT, p, p'-

13



.00158

.027

.034*

.034*



0.00119

0.00477

0.32

5.4

5

1.67*

999999300

DDT (Total)

13



.00158

.0461

.009*

.015*



0.00389

0.0517

0.32

5.4

5

7.7*

1163195

Deeabroraodipbcnyi oxide

i



















110





. . $4742

Di-rv butyl phttabte

1.6







1.4M

1.4"

..1100





... .

1100



1.0

117840

Di-n-octyl phihalaie

1







6.2b

6.2*









220



1.0

333415

DiazinorVSpectracide

1,6











.019







9.7



1.80*

53703

Dibenzo(aJi)anthracene

1



.0634

.26

,23«

.97*



0.00622

0.135

0.015





0.29*

132649

Dibcnzofuran

1.6







.54®

1.7'

200







43



1.0

96128

Dtbromo-3-chbropropane,
1,2-

I

















0.077







124481

DiroroocWoromcihax*

1

















1.3

220



1.0

1918009

Dicarrfca





















320





95501

Dichfcrobcnzcne, 1,2-

1.6







0.058*

0.05*6

34







970



1.0

541731

Dchbrobenzene, 13-

1.6











170







960



1.0

106467

Dk^lorobenzrne, 1,4-

1.6







.11*

.12**

35





4.5





1.0

25321226

Dichlorobenzenes

1







0.05**

0.05**

34





4.5

960



1.0

91941

Dichbrobenadine, 33'-



















0.24







75718

Dchbrodifluoronethane

1



















2200





75343

Dichbroethane 1,1-

1



















1100



1.0

107062

Dfchb roe thane 1,2-

1

















1.2





1.0

75354

Dfchloroetheoe, 1,1-

1

















0,18

97





156605

Dichbroethene, trans-1,2-

1



















220



1.0.

156592

DicWoroethyfene, cs-1,2-

1



















110





75092

Dichbro methane

1

















14

650



1.0


-------
Ifcble D-l. (Continued)

CUH>EUNE VALUES EXTENDED ONLYFO* SCREENIN'G-LEVTL UAZ^D COMPARISON AMONG CHEMICALS
May Ik Orcr- «r UWetfreteefrr* ef SeAerel af a Chts UoiiM Dtfiirfbi mi SHc*S?edBc CMMnt

CASNwHber

Chrakal Naax>

OxSo

SHhneat Gmxacm&m

FWiUmbd CwMifUwi

B5AF
)

SQCm

W

ER-C.

m-M
(Km*

AET-L '
(ftwd

AET-Q
(ppw)

SQAL„

o>»y

TEL

(fftnj

PEL
1

FDA
GiMikc/

Actioa/
Ifeicruec

Lerci

120832

DfeMorof»I)«iQi 2,4-





















32





94737

DichbropiEaDX>*cetk: add, 2,4-

5



















110

1



94826

DidNompheooxybocanob *dd,
2,4-





















86





78875

Dk^hropfcpane, 1,2-

1

















1.6





1,0

542756

DjchfaropropeEe, 1,3-

1

















0,62

3.2





62737

EScUorvoj

1

















0.37

5.4





115322

DicofbVKckbane



















0.24







60571

Diddm

I J,6

U









u

7.15E-4

0.0043

.0067

.54

.3

1.80*

84662

Dkohyi pbthaiair

1,6







0.2"

0.2s

63







8600



1.0

I19904

Dmthcxybenadine.3,3'-



















7.7







131113

Dimethyl phtahie

1







0.16*

0.16°









110000



1.0

105679

Dimtihyiphcnol, 2,4-









.029*

.21"









220





528290

DiirrobeiHtne, 1,2-





















4.3





99650

Dntrobeszene, 13-





















1.1





100254

Dinfero benzene, 1,4





















4.3





51285

Dhfcrophenol, 2,4-





















22





121142

DHtictolaene, 2,4-





















22





606202

DiraEroEol*nc, 2,6-





















11





88857

DinoscWDNBP





















1!





12266?

D^henyih)dJrt«iDe. 1,2-

















.

0.13







298044

Dsu&hob

I



















0.43





959988

Endosul&n, *lphi-

1.6











.29







65



1 M8>

33213659

Endosylfea bets-

1,6











1.4







65



1.80s

115297

Ecdosul&a maed isomer*

1.6











34







65



1.80*

72208

Endria

1,6

4.2









4J







3.2



1.80*

563122

Btaen/BSadeB

I



















5.4



1.80*


-------
Table D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR SCREENING-LEVEL HAZARD COMPARISON AMONG CHEMICALS
May Be Over- or Uncle (protective of Sediment at • Given Location DepemSng on S He-Specific Condfttons

CAS Nunier

Qxidcal Name

Code

Sedhmtf Concentration

FbhTbsue Concentration (ppra)

BSAF
(uohlets)

SQC.

(M8/8,e>

ER-L
(ppa)

ER-M
(ppa)

AET-L
(pf*n)

AET-H
(Pl*n)

SQAL^

TEL

(ppn)

PEL
(PI*")

ConcciL
= EPA
Risk 1«H

EPA
Nomine-
r

Hianl
Quotient

= 1

EDA
Cttfdaoee/

Action/
Tolerance
Level

141786

Ethyl acetate

1



















9700





100414

Ethyfcenzene

1.6







.0Ib

.037°

480







1100



1.0

106934

Flhytene dihromide

I

















.0013







206440

Fiioraniheoe

I

620

.6

5.1

2.5®

30*

620

0.113

1.494



430



0.29*

86737

Fluorene

1.6



.019

.54

.54®

3.6*

54

0.0212

0.144



430



0.29*

944229

Fonofos

1



















22





76448

Heptachbr

1.3

















0.024

5.4

.3

1.80*

1024573

Heptachbr epoxide

U

















0.012

0.14

.3

1.80b

118741

Hexachloro benzene

1







.022*

.23°







0.067

8.6



0.09*

87683

Hexachbrobutadicne

1







.011*

.27®







1.4

2.2



1.0

77474

HexachbrocycbpenUdiene

1



















75





67721

Hexachbroethane

1.6











100





7.7

11



1.0

51235042

Hexazinone

1



















360





123319

Hydroqienone





















430





193395

lndeno(l,2P3-ed)pyrere

1







.69°

X6*







0.15





0.29"

78591

Isophorone

1

















no

2200



1.0

33820530

Isopropain





















160





7439921

Lead

2



46.7

218

4506

660-



30.2

112





1.3



121755

Malathton

1.6











.067







220



1.80*

108316

Mafeic anhydride





















1100





7439965

Manganese





















54





7439976

Mercwy





.15

.71

.59*

2.1**



0.13

0.696



1.1

1



72435

Methoxychbr

1.6











1.9







54



1.80*

78933

Methyl ethyl ketone

1



















6500



1.0

108101

Methyl isobutyl ketone

1



















860





22967926

Methyl nrrcwy

3



















1.1

1




-------
Table D-l. (Continued)

GUIDELINE VALDE3 ^tended only for screening-level hazard comparison among chemicals
M»| B« Or*** or IMttfntcdn ef Stdfanrnt at ¦ Ctrci Lw»M&9m

CASNisrbcr

OxmiadNum

Code

ScdkscotCoactfltnUMt

lUtTiutx GMntsbMiMQfiid

BSAF
(unities*)

SQC„

er.l
{«¦**

ER-M

AEF»L
(PP*J

AET-II
(WW#

SQAL„

W&J

TEL
(R**)

PEL

Ceoccsv
m EPA
Rbklfr*

EPA

Nwapincc»
r

ELuMfd
Quotient
-1

FDA
Grfduee/

Tbfc rente
Level

91576

Mc&^phthoboe, 2-

1



.07

.67

.67*

1,9*



0.0202

0501









21087649

Metribum





















270





2385855

Mircx/Dechlomne

U

















0.060

2.2

0.1

UI*

7439987

Molybdenum





















54





mm

NtpkJalcne

1,6



,16

2,1

XI-

%T>

4?

0.0346

0391



430



0,29*

91598

NspbtbyfcmA*, 2-



















0.00083







7440020

Nickc!

2



20.9

51.6







15.9

42.8



220

70



9S953

Nitrobenzene





















5.4





100027

Njerophesoi, 4





















670





924163

Nnosod^r^bix^Kisirc, N-



















0.020







621647

Nirosodi-ii-piopyluDine, N-



















aois







55185

NbtHOdinsthyiHnins, N-



















0.0021







86306

Nirosodipbtnytaniie, N-









.028*

,13°







22







999999484

PAK$ (faS$) mofecuh* weight)





1.7

9.6

n*»

69*Aa



0.655

6676









999999502

PAHs (faw a»fecul« weight)





,552

3,16



24^



0312

L442









56382

P*r&£h*3n ethyl





















65





12674111

PCB (Arocbr-1016)

1.4



.0227

.180

t&

3.1*



0,0216

aiss

aoi4

0.75

2

L85*

11104282

PCB (Arocbr-1221)

1.4



.0227

.180

1.0°

3.1*



0,0216

0.189

0.014

0.22

2

1.85*

11141165

PCB (Aroctor-1232)

M



.0227

.180

U?

3.1*



0.0216

0.189

aoi4

0.22

2

1.85*

53469219

PCB (Arocbr-1242)

1.4



.0227

.180

1.04

3.t*



0.0216

0.189

0,014

0.22

2

1.85*

12672296

PCB (Aroclor-1248)

1,4



.0227

.180

I.06

3.1*



0.0216

0.189

0.014

0.22

2

1.85*

11097691

PCB (Aroclor-1254)

1.4



.022?

.180

L0>

3.1*



0.0216

0.189

0.014

0.22

2

1.85*

11096825

PCB (Arocbr-1260)

1.4



.0227

.180

1.0*

3.1*



0.0216

0.189

0.014

0.22

2

1.85*

608935

PenticWorobenzetrc

1.6











69







8.6



0.04*

82688

PentacWoroniTOb
-------
Table D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR SCREENING-LEVEL HAZARD COMPARISON AMONG CHEMICALS
May Be Over- or Uode (protective of Sc dental al • Ghtn Location Dependng en SL3«SpedSe CondMooi

CASNimter

Chemical Nan

Code

Scdbaent Cooccotxvtfcm

Rib JHtat Concentration (ppm)

BSAF
(tablets)

SQC,
(vstej

ER-L

(ffOl)

Bt-M
(ppm)

AET-L

(ffOl)

AUT-H
(ppm)

sqalk

TEL

(pH

PEL

(FPU)

Coacea
¦ EVA
Risk 104

EPA
r

Hazard
Quotient
« 1

FDA
Guidance/

Actlonf
Inference
Level

85018

Pheoutbrene

1

180

0.240

1.5

1.5"

6.9*

180

0.0867

0.544









108952

Phenol









.42b

1.2**









6500





298022

Phonte/Fannphoa^nimet

1



















2.2





8S449

PMtaic anhydride





















22000





1336363

Polychbrsaied bpbesyfe

1,4



0.0227

0.180

1.0»

3.1'



0.0216

0.189

0.014

0.22

2

1.85*

1610180

Proraeton'Pnunitol





















160





7287196

Prometyn/Capvol





















43





23950585

Proaanade





















810





1918167

Propachlor





















140





129000

Pyrene

1



.665

2.6

3J®

16**



0.153

1.398



320



0.29*

91225

Quinofine

1

















0.009







7782492

Sekimim





















54





7440224

Stiver





1

3.7

6.1*

6.1*



0.733

1.77



54





122349

Stnnzne

5

















0.90

54

12



7440246

Strontium





















6500





100425

Styrene

1



















2200





13071799

Terbufbs/Cooter

1



















0.27





886500

Terbutryn





















11





95943

Tetrnchlorobeincne, 1,2,4,5-

1



















3.2



1.0

1746016

Tetrechloro<12>en20-p-dbxvi,23>7-
.8-

1

















6.9E-7





0.059*

79345

Tctrachloroethane, 1,1,2,2-

1.6











160





0.54





1.0

127184

TctracUoroetheoo

1.6







.057*

.14®

53





2.1

110



1.0

56235

TcCrachloromcthano

1.6











120





0.83

7.5



1.0

58902

TctrachlorophfcooL 2,3,4,6-





















320





961115

Tctrnchbrvinphos/Gardoni/Stiof

1

















4.5

320





7440315

Tm





















6500






-------
Tabic D-l. (Continued)

GUIDELINE VALUES INTENDED ONLY FOR SCREENING-LEVEI, HAZARD COMPAX1SON AMONG CHEMICMS
May B« Ore* «r IMrtpetfetln at fetfmrnf at i Gtvca Leettfe* Dcptadtag m Ske4Ffe46e OmUom

CAS Ni*riber

Genticst! N*me

Code

Scdbxret Cwmft'itlwi

fkbTluK

BSAF
(atfes*)

SQC„

ER-L

(?|Nt$

ER-M

Afrr-L
(pproj

AET-n
(w»>

SQAI*

TEL
(&**$

PEL
Cp?e$

Co see B.
-EPA
Rkk It*

ETA
Neoatxsce-

r

BtBHtt

Qnolfcnt
-1

SUA
CMdftnec/
Actio ts?
Ifeteranee
Level

108883

Toluene

1,6











89







2200



1.0

8001352

ToxAphcno

1.6











10





.098





I JO

75252

TribromottrthAne (Brcrrofenn)

1,6











65





14

220



1.0

120821

Tfiiiofobancne, 1,2,4-

Ifi







.051*

.064*

920







no



1.0

71556

TrichbrDcthine, 1,1.1-

1,6











17







970



1.0

79905

Trichbrocthaoe, 1,1,2-

1

















1.9

43



1.0

79016

Trfchbrocthmc

1.6











210





9.8

65



L0

75694

TnchbroSjoroar thane

1



















3200



1.0

67663

Triddoronjsiiuuac (Chiorolbmj

1

















18

no



1.0

95954

Trichlorophenal, 2,4,5-





















1100





88062

Itichlorophenol, 2,4,6*



















9.8







93765

TricMorophenoxyacedc t&d, 2,4,5-





















no





93721

TrkH»roplc4K>x>pivpbnb acid.
2,4,5-





















86





1582098

Tnftaafic^Tccfiaa



















14

31





95636

Trinxthyifoeizeae, 1,2,4-

1



















5.4





118967

TnatrotDiHeHe



















3.6

5.4





7440622

V^mdaim





















75





108054

Vinyl tcctaic

i



















11000





108383

Xylene, to

1,6







.04*

.12*

2.5







22000



1.0

95476

Xyfeoe, o-

1







.04*

.12"

2.5







2200©



1.0

106423

Xyfenc, p-

1







.04*

.12*

2.5











1.0

1330207

Xyfcacs

1







.04*

.12*

Z5







22000



1.0

7440666

Zbc





150

410

410^

1600s



124

271



3200





S88888881

Dbsa-taxie «|iav*kn3

1

















«.9E-7





0025*


-------
Table D-l. (Continued)

Codes:

1.	Chemical is a nonpolar organic.

2.	FDA criterion is a guideline,

3.	FDA criterion is an action level.

4.	FDA criterion is a tolerance level, with the force of law,

5.	Fish tissue action level set by USEPA, 40 CFR Part 180,

6.	Preliminaiy SQALM developed for this chemical is under technical review.

AET Criteria:

' Sediment concentration based on am phi pods.

6 Sediment concentration based on benthic organisms.

•Sediment concentration based on oysters.

BSAF Sources:

'Cook, 1995.

'Hansen, 1995,


-------
Appendix I)

cific data were analyzed separately. In addition, the dioxin congeners were evaluated using the toxicity equivalence
factor (TEF) approach (USEPA, 1989). This approach involves summarizing specific dioxin congeners based on
their toxicity as compared to 2,3,7,8-tetrachlorodibenzo-p-dioxin, for which screening values are available. PCBs
and dioxin represent the only cases where chemical data were actually combined for the NSI evaluation.

Because EPA typically performs risk-based screening by analyzing closely related chemicals with the same risk-
based concentrations, this methodology was applied to the NSI evaluation. If no screening values were available for
a certain chemical, but were available for a closely related chemical or group of chemicals, the lower or more
conservative screening values of the closely related chemicals were used in analyzing the chemicals without screen-
ing values. This methodology was applied only for chemicals or chemical groups with more than 20 positive results.
The following chemicals and chemical groups were affected by this methodology; BHCs, chlordanes, cresols, DDT
and metabolites, dichlorobenzenes, endosulfans, methylmercury, anthracene and phenanthrene, benzo(a)anthracene/
chrysene, xylenes, and PCBs (in applying screening values to aroclors with no available screening values).

Frequency of Detection

The frequency at which a given chemical or chemical group is responsible for sites in the NSI being categorized
as Tier 1 or Her 2 is often a reflection of the number of times that chemical is measured and detected in sediment
samples. Thus, chemicals that are measured and detected less frequently might not often be identified as posing a
potential risk to aquatic life or human health, even though the chemical is highly toxic. Table D-2 lists the number
of times each chemical included in the NSI evaluation was measured and detected (i.e., a positive result) in sediment
and fish tissue and the number of times each chemical was responsible for Tier 1 or Her 2 sampling stations being
classified.

D-12


-------
Table D-2, Frequency of Detection of Chemicals in Sediment and Fish Tissue and Number of Detections
Resulting in Risk (Tier 1 or Tier 2)*-k

CAS Number

Chemical Name

N umber of

Times
Measured
ill SedjmeM

Number or
FosKfte
Sediment
Results

Ntmber
of Times
Measur-
ed

In
Tissue'

Number

of
Positive
Tissue
Results'

Her I
Level
Results

Tier 2
Level
Results

83329

Acenapbshene

6126

1567

777

41

144

359

208968

Aceraphthylene

5774

1286

-

-

74

958

67641

Acetone

547

48

22

16

-



107028

Acrofesi

-

-

464

-

-

-

107131

ActybnMc

1034

9

464

-

-

7

15972608

Ahchbr/Lasso

-

-

976

1

-

-

309002

AMrio

14311

658

8029

612

2

712

62533

Anilte

-

-

10

-

-

-

120127

Anthracene

5211

1798

748

63

168

728

999999933

Anthracene & Phenantbrene

260

199

4

-

82

95

7440360

Antimony

S923

2980

1275

99

-

56

7440382

Arsenic

22281.

18791

5528

2113

189

8613

1912249

Atrazme

-

-

880

-

-

-

7440393

Bariim 1

-

-

986

837

-

-

71432

Benzene

2248

136

976

90

-

16

92875

BenzHi*

-

-

537

-

-

-

56553

BeniD(a)H)tt>raeene

6718

3236

820

153

241

1540

999999955

Benzo(a)antaaeene/Ctarysetie

272

243

-

-

146

76

50328

Beim>{a)pyrcne

7011

3263

831

58

317

2292

205992

Benjo(b)fliKjranihene

4179

1249

717

26

-

441

191242

BenzaCgbOpeiyfene

6034

2016

-

-

-

259

207089

fknroOOfluoraritbeiie.

4192

1093

651

21

-

113

65850

Beraoic ackl

1724

247

121

5

-

41

100516

Benzyl alcohol

1910

90

120

-

-

13

7440417

Berlin

-

-

1301

81

-

39

92524

B phenyl

1215

873

564

138

-

2

542881

B B(ehtareaeth5tl)elher

-

-

76

-

-

-

111444

Bis(2-cbbroethyl)etber

-

-

636

3

-

3

108601

BS(2-chbroEopropyI)ether

-

-

34

1

-

-

117817

Bis(2-ethyliexyl)phthakte

4606

1998

647

91

401

1109

7440428

Boron

-

-

44

21

-

-

75274

Bromodichbromethare

-

-

560

4

-

-

74839

Bro mo methane

-

-

491

3

-

-

101553

Bron»phenyl phenyl ether, 4-

2698

20

656

1

-

7

85687

Butyl benzyl phthabte

4069

333

_ 634

4

1

51

319846

BHC, alpha-

9109

219

8148

1670

11

461

319857

BHC, bcta-

6761

241

3060

209

-

257

319868

BHC, deta-

4891

99

2156

65

1

94

D-13


-------
\p|H*ndi\ I)

Table D-2. (Continued)

CAS Number

Chemical Name

Number of

Times
Measured
in Sediment

Number of

Positive
Sediment
Results

Number
of Times
Measured

io Tissue'

Number

of
Positive
Tissue
Results'

Tier 1
Level
Results

Tier 2
Level
Results

58899

BHC, pimm-/L«xla(ie

14442

999

8750

1391

101

527

608731

BHC, teetotal grade

169

166

115

31

3

66

7440439

Cadmfam

27919

15176

6743

3321

-

7206

75150

Carbon dMUe

-

-

24

21

-

-

57749

Chbrdane

12432

2170

7316

4568

116

4228

999999247

Chbniair-Nonachfer(cE>

1476

9

4468

2101

-

268

999999248

Chtaiane-Nonachbr(!ians)-

1992

31

4569

2764

-

556

5103719

CUoidare, a(pha(cis)-

44-16

1516

6092

3659

3

1157

5103742

Chtordanc, bcta(trans).

2833

443

5841

3045

3

847

5566347

Chlordane, gmtm(trans)-

967

334

85

19

-

207

108907

Chbroberaens

2111

58

819

18

•

4

510156

Chbrobenabte

-

-

22

-

-

-

75003

Cbfaroctkirc

-

-

557

1

-

-

75014

Chbroelbene

-

-

706

2

-

2

110758

ChtonjclbyMnyl ether, 2-

-

-

534

-

-

-

74873

Cbbromettiane

-

-

744

12

-

-

91587

Chbronapbttakne, 2-

-

-

655

I

-

-

95578

Chbmphciiol 2-

-

-

629

1

-

-

2921882

Chlorpynfcsfflursban

305

5

793

143

-

-

7440173

Chcomlm

27504

25216

5508

3283

426

4126

218019

Ckysene

6975

3580

893

149

185

1618

7440508

Copper

27956

25452

6284

5533

-

11213

108394

Ctesol, m-

988

780

-

-

-

41

95487

Crcsol o

1993

745

51

-

-

22

106445

Cresol p-

985

84

49

3

-

31

1319773

Cresois

18

1

-

-

-

1

21725462

Cyamate

-

-

326

-

-

-

57125

Cyanic

-

-

14

3

-

-

84742

Di-n-butyi phtbaMe

4651

986

637

55

9

112

117840

Di-ri-octyl phlMhle

4179

435

650

6

-

23

333415

DkzinoiVSpcctracHe

3712

249

172

-

-

188

53703

Dibemo(aii)aiatacene

7564

2431

824

16

419

1732

132649

Dibcnzofuran

2564

416

126

-

25

51

124481

Dirornochlorometharc

2033

18

562

I

-

-

95501

Dchbrobccnene, 1,2-

4402

107

892

2

38

23

541731

Dfchbrobenasns, 13-

4315

132

797

2

-

22

106467

Dchbrobcraiix, 1,4-

4352

268

887

3

53

41

25321226

Dichferabenzenes

27

12

-

-

6

3

91941

Dfcllbroberradire, 3,3'-

-

-

639

1

-

-

D-14


-------
Nnlioiuil Scd'um-nt OustVit.v Sur\i>.\

liable D-2. (Continued)

CAS Number

Chemical Name

Number of

Urns
Mewived
in Sediment

Number of
Positive
Sediment
Results

Number
of Times
Measured
In Tissue"

Number

of
Positive
Tissue
Results'

Tier 1
Level
Results

Tier 2
Level
Remits

75718

Dichbrodifboroirelhar*

-

-

174

-

-

-

75343

Dlchloroethane 1,1.

1918

19

561

-

-

-

107062

Dchtoroelhane 1,2-

1981

20

972

8

-

-

156605

Dchbroethene, trans-1,2-

1393

33

793

2

-

-

75354

Dfchbroelhene, 1,1-

-

-

973

2

-

-

75092

DiehlDromethaifc

2177

576

532

112

-

11

120832

Dishbrophenol 2,4-

-

-

642

1

-

-

94757

DichloropheiKHtyacetk: acid, 2,4-

-

-

39

-

-

-

78875

Dfchbropropaic, 1,2-

2015

15

563

2

-

-

542756

Dfchbtopropene, 1,3-

-

-

107

-

-

-

115322

DicofoVKellhane

-

-

400

26

-

-

60571

DicUrin

14702

3113

10243

5583

89

6709

84662

Diethyl phlhalate

4188

367

654

2

34

48

131113

Dmiliyl phihalatc

4118

135

653

-

-

38

105679

Dime fbyip tic rwl 2,4-

4541

80

640

1

-

54

51285

Binkrophenol, 2,4-

-

-

631

-

-

-

121142

Dinitrotoluene, 2,4-

-

-

636

1

-

-

606202

DUtrolotieiK, 2,6-

-

-

636

1

-

-

122667

Dip he ny[hydrazine, 1,2-

-

-

509

-

-

-

298044

Disulfbton

-

-

23

-

-

-

1861321

DCPA/Daclhal

129

76

827

586

-

3

53190

DDD, o,p'-

6349

977

3397

428

73

502

72548

DDD, p, p'-

15311

4411

6252

2481

572

2574

3424826

DDE, o,p'-

5434

632

3427

401

118

222

72559

DDE, p, p'-

15961

5980

7656

5715

823

3501

999999300

DDTflbla!)

3710

736

5750

4183

122

860

789026

DDT, o,p'-

6056

567

3479

368

25

268

50293

DDT, p, p'-

16028

3288

5843

1677

371

1839

115297

EndosuKm ttixed booths

2606

80

49

12

-

20

959988

Endosuifan, afcha-

5581'

84

2832

53

-

45

33213659

Endosufin, beta-

5886

260

2157

10

-

42

72208

Endrm

12694

289 '

8192

893

-

8

563122

Ethion/Bladen

2953

38

170

-

-

-

100414

Ettiytoenzenc

2543

118

807

50 .

1

42

206440

Fkxtramhene

7562

4563

953

216

234

1074

86737

Fluorene

6652

2280

797

14

231

1141

944229

Forofos

-

-

288

-

-

-

76448

Hcptachlor .

11952

673

7369

1006

-

210

1024573

Hcptachbr epoxide

12829

986

7480

2896

-

1431

D-15


-------
Table D-2. (Continued)

CAS Number

CbemlcalName

Ntmberof

Times
Measured
In Sediment

Number of
Positive
Sediment
Results

Number
or Times
Measured
in Tissue'

Number

of
Positive
Tissue
Results

Tier 1
Level
Results

Her 2
Level
Results

118741

Henachbrobenzsne

10044

1445

6970

1519

-

224

87683

Hexachbrobutadtoie

4198

128

1161

14

-

81

67721

Heme hb methane

3801

4

636

-

-

1

193395

IndenotlAS-cdJpyrene

5874

1913

7S6

20

-

559

78591

feopbarone

3400

40

635

4

-

8

33820530

ItopiopaBn

-

-

392

15

-

-

7439921

Lead

29979

24971

6654

3008

-

8883

121755

Mahlhton

4041

38

500

1

-

26

108316

Mack: anhydride

-

-

2

-

-

-

7439965

Mangansse

-

-

1000

971

-

5

7439976

Mcrcuiy

26142

16632

9752

8424

1951

5049

72435

MethojQchfor

9183

154

5912

63

-

33

78933

Methyl ethyl ketone

519

7

20

11

-

-

108101

Methyl isGbutyi ketone

-

-

26

-

¦

-

22967926

Methyl meicury

-

-

9

8

-

-

91576

Metl^taphthakne, 2-

2629

973 .

-

¦

71

522

21087649

Metribuzh

-

-

289

-

-

-

2385855

Mrex/Dechbrane

5794

544

4800

915

-

40

7439987

Molybdenum

-

-

707

169

-

-

91203

Naphthalene

6823

2820

803

22

291

1247

7440020

Nickel

21519

18550

3120

974

-

9260

98953

Nitoberasne

-

-

635

-

-

-

100027

Niropbenol, 4

-

-

606

1

-

-

621647

NtosodJ-n-piopytaifce, N-

-

-

645

I

-

1

86306

NkrosodipbenyiiiTwie, N-

3730

66

661

3

-

45

999999484

PAHs (high niobctihr weight)

1566

885

-

-

93

383

999999502

PAHs (bw molecular weight)

1604

895

-

-

112

382

56382

Parathbn ethyl

-

-

499

4

-

-

608935

PentachbrobetEBne '

114

54

404

30

-

4

82688

PenQchbrantobenane/Quiruozene

-

-

390

2

-

-

87865

Pentachbropbenol

5622

195

1756

149

-

26

85018

PiwBMhitne

7067

4078

.

-

335

694

108952

Ftercl

4595

864

647

12

-

155

1336363

Polychbriravtx) b|>henyls

11296

4183

10642

7379

8151

2620

1610180

Promton/Paniol

-

-

289

-

-

-

1918167

Propachtor

-

-

1

-

-

-

129000

Pjitne

7558

4555

952

187

482

1896

12674112

PCB (Aiocbr-1016)

5098

46

3161

12

19

39

11104282

PCB (Arocbr-1221)

5627

7

3568

2

4

5

D-I6


-------
National Sediment Quality Survey

Table D-2. (Continued)

CAS Numfwr

Ok mica I Name

Number of

limes
Measured
to Sediment

Number or
HjilUve
Sediment
Result?

Number of

Times
Measured
fa Tissue1

Number

sf
Positive
Tissue
Results'

Tterl
Lewi
Results

Tier 2
level
Results

11141165

PCB (Arocbr-1232)

5417

13

3195

1

4

10

53469219

PCB (Arocbr-1242)

6375

435

4446

220

355

270

12672296

PCB (Arocbr-1248)

6314

559

4464

688

916

230

1109769!

PCB (Arocbr-l2J4)

7178

1305

5871

3343

3664

765

11096825

PCB (Atocbr-1260)

6885

890

6035

3611

3866

531

7782492

Selenium

-

-

2559

2079

-

4

7440224

SDwf

11082

6256

1139

515

350

1083

122349

Sirazsne

-

-

289

-

-

-

7440246

Sttonttan

-

•

45

45

-

-

100425

Stjirene

-

-

191

-

-

-

8S8888882

SFA1 est flSEMMAVS])

335

335

-

-

8

161

95943

TcttachbnDbenzenc, 1,2,4,5-

97

1

398

12

.

-

1746016

TeErachbrod iber*zn- p- d b xii. 2,3,7,8-

631

38

908

391

353

23

79345

litrachbroettaMs, 1,1,2,2-

1683

49

978

33

-

2

127184

•fttrachbrociheiK

2429

109

973

49

2

17

56235

Tettachbromeitans

2010

15

979

4

-

-

58902

TfctraeMjtophenol, 2,3,4,6-

-

-

71

-

-

-

7440315

Tta

-

-

382

264

-

-

108883

Tohicrc

2338

325

814

116

-

28

8001352

Hsxaphene ;

10912

75

6566

643

-

684

75252

TltwromoixietteJW/Broniofcrai

2078

44

818

7

-

-

120821

HiMotobiaaene, 1,2,4-

4256

87

10S2

46

6

49

71556

Hfchbroetfai*:, 1,1,1-

2083

63

815

23

-

10

79005

TWchtoroeUane, 1,1,2-

2035

14

879

7

-

-

79016

TftHbroefene

2494

75

975

19

,

1

75694

Uxhbroftoromelhane

1096

9

288

15

-

-

67663

HfcUoromeltaneClkMofcim

2277

76

972

37

-

-

95954

Ubhbropheno!. 2,4,5-

-

-

73

-

-

-

88052

H-fchbropbenoI, 2,4,6-

-

-

658

-

-

-

93765

"Dxhbfophenoxyacetfc acid, 2.4,5-

-

-

3

-

-

-

93721

IttUDrophtnoxjpropijrt: acid, 2,4,5

-

-

36

-

.

-

1582098

THtaaiimelhti

-

-

925

193

-

-

7440622

\Sinadium

-

-

768

465

-

-

108054

Vinyl acetate

-

-

21

-

-

-

108383

Xylene, m-

55

31

-

-

4

6

95476

Xyfcre, o-

61

1

-



.

1

106423

Xylene. p-

14

2

-

-

-

2

1330207

Xylercs

922

48

22

13

5

11

7440666

Zinc

27065

26473

4580

4553

-

5176

D-17


-------
\|>|H'IUli\ I)

¦Bible D-2. (Continued)











Number









Number of

Number of

Number

or









Times

Positive

or lines

Positive

Tier 1

Tier 2





Measured

Sediment

Measured

Tissue

Level

Level

CAS Number

Chemkal Name

In Sediment

Results

in Tissue*

Results'

Results

Results

833888881

Dbxxi toxic equhakns

56

56

590

590

459

45

¦Results presented at observation level. Multiple observations may have occurred at a given station.
^Observations recorded here correspond only to stations with available latitude/longitude coordinates.
'Fish tissue results are presented for demersal, resident, and edible species only.

D-18


-------
National Sediment Quality Survey

References

Barrick, R., S. Becker, L. Brown, H, Beller, and R. Pastorok. 1988. Sediment quality values refinement: 1988
update and evaluation of Puget Sound AET. Vol. 1. Prepared for the Puget Sound Estuary Program, Office of
Puget Sound.	•

Cook, P.M. 1995. Pelagic BSAFs for NSI methodology. Memorandum from P.M. Cook, ERL Duluth, to C. Fox,
EPA, Office of Water, March 29, 1995.

DHHS. 1994. Action levels for poisonous or deleterious substances in human food and animal feed. U.S. Food
and Drug Administration, Department of Health and Human Services, Public Health Service, Washington, DC.

FDEP. 1994. Approach to the assessment of sediment quality in Florida coastal waters, Vol. 1. Development and
evaluation of sediment quality assessment guidelines. Prepared for Florida Department of Environmental
Protection, Office of Water Policy, Tallahasee, FL, by MacDonald Environmental Sciences Ltd., Ladysmith,
British Columbia.

Hansen. 1995. Assessment tools that can be used for the National Sediment Inventory. Memorandum from D, J.
Hansen, ERL, Narragansett, to C. Fox, Office of Water, February 28, 1995.

Long, E.R., D.D. MacDonald, S.L. Smith, and F.D. Calder, 1995. Incidence of adverse biological effects within
ranges of chemical concentrations in marine and estuarine sediments. Environ. Manage. 19(l);81-97.

USEPA. 1989. Interim procedures for estimating risks associated with exposures to mixtures of chlorinated
dibenzo-p-dioxins and dibemofurans (CDDs and CDFs) and 1989 update. EPA/625/3-89/016 U.S. Environ-
mental Protection Agency, Risk Assessment Forum, Washington, DC.

	. 1995a. Integrated Risk Information System (IRIS). Online. U.S. Environmental Protection Agency,

Office of Health and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati,
OH.

	. 1995b. Health effects assessment summary tables FY 1995. EPA/540/R-95/036. U.S. Environmental

Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC.

	. 1995c. Risk-based concentration table, January-June 1995. USEPA Region 3.

USFDA. 1993a. Guidance document for arsenic in shellfish. U.S. Food and Drug Administration, Center for
Food Safety and Applied Nutrition, Washington, DC.

	. 1993b. Guidance document for cadmium in shellfish. U.S. Food and Drug Administration, Center for

Food Safety and Applied Nutrition, Washington, DC.

—	. 1993c. Guidance document for chromium in shellfish. U.S. Food and Drug Administration, Center for

Food Safety and Applied Nutrition, Washington, DC.

-——. 1993d. Guidance document for lead in shellfish. U. S. Food and Drug Administration, Center for Food
Safety and Applied Nutrition, Washington, DC.

	. 1993e. Guidance document for mercury in shellfish. U.S. Food and Drug Administration, Center for

Food Safety and Applied Nutrition, Washington, DC.

D-19


-------
Vppi'iulK I)


-------
Appendix E

Cancer Slope Factors and
Noncancer Reference Doses Used
to Develop EPA Risk Levels

T able E-l presents the cancer slope factors and noncancer reference doses that were used to calculate the EPA
risk levels and hazard quotients used in the analysis. The calculations for the EPA risk levels and hazard
quotients used in the analysis appear in Appendix B. The slope factors and reference doses were obtained
from the following sources:

•	Health Effects Assessment Summary Tables FY 1995. EPA/540/R-95/036. U.S. Environmental Protection
Agency, Office of Solid Waste and Emergency Response, Washington, DC.

•	Integrated Risk Information System (IRIS). Online. U.S. Environmental Protection Agency, Office of Health
and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH.

•	Risk-Based Concentration Table, January-June 1995. U.S. Environmental Protection Agency, Region 3,
Philadelphia, PA.

E-l


-------
Apprntlis K

Tfrble E-l. Cancer Slope Factors and Noncancer Reference Doses Used to Develop EPA Risk Levels

CAS Number

Chemical Names

Cancer Slope factor

((nig/kg/d)*1)
(Followed by source;
see footnotes)

Noncancer Reference Dose
(mg/kg/d)
(Followed by sources
see footnotes)

Surrogate 'Chemical
Used (if neccessary)

83329

Acen&phthene



6.00E-21



67641

Acetone



l.OOE-l'



98362

Acetophenone



1.00E-11



107028

Acrolein



2,00E-2S



107131

Acjylonitrile

5.40E-11

1.00E-3"



15972608

Alachlor/Lasso

8.00E-2"

1.00E-21



116063

Aldlcarb/Temik



1.00E-3'



309002

Aldrin

1.70E+11

3.00E-5'



62533

Aniline

5.70E-31





120127

Antthacene



3.00E-11



999999933

Anthracene & Phenanihrene



3.00E-1

anthracene

7440360

Antimony



4.00E-4'



7440382

Arsenic

1.75E+01

3.00E-41



1912249

Airazine

2.22E-1*

3.50E-2'



7440393

Barium



7.00E-2'



92875

Benzidine

2.30E+21

3.00E-3'



71432

Benzene

2.90E-21





56553

Benzo(a)anthracene

7.30E-1®





999999955

Benzo(a)anthracene/Chiysene

7.30E-1



benzo(a)artthracene

50328

Benzo{a)p3irene

7.30E+01





205992

Benzo(b}fluoranthene

7.30E-1'





2070S9

Benzo(k)fluoranthene

7.30E-25





65850

Benzoic acid



4.00E+01



98077

Benzatrichloride

1.30E+11





100516

Benzyl alcohol



3.00E-lb



100447

Benzyl chloride

1.70E-11





7440417

Beryllium

4.30E+tf

5.00E-3'



319846

BHC, alpha-

6.30E+01





319857

BHC, bctt-

1.80E+Q1





319868

BHC, delta-

1.80E+0



beta-BHC

58899

BHC, gamma- (Lindane)

1.30E+0®

3.00E-4'



608731

BHC, technical grade

1.80E+01





E-2


-------
Niilioiiii) Sediimiil	Kurvcv

j		*	*

Table E-l. (Continued)

CAS Number

Chemical Name

Cancer Slope Factor

((mg/kg/d)'1)
(Followed by source;
see footnotes)

Noncancer Reference

Dose (mg/kg/d)
(Followed by source;
see footnotes)

Surrogate Chemical
Used (If neeessaiy)

608731

BHC, technfcal gptde

1.80E+01





92524

Biphenyl '



5.00E-2'



111444

Bis(2-chforoethyl)ether

1.1OE+0





108601

Bis(2-chbroisopropyl)ether

7.00E-2"

4.00E-2'



117817

Bis(2-ethylhex5%>hthalate

1.40E-2'

2.00E-21



542881

Bis(chtaromethyI)ether

2.20E+2'





7440428

Boron



9.00E-21



75274

BrornodicMoromethaiiB

6.20E-21

2.00E-2'



74839

Bro it» methane !



1.40E-3'



101553

Bromuphsnyl phenyl ether, 4-



5.80E-2'



1689845

Bromoxyml



2.00E-21



8S687

Butyl benzyl phthalate



2.00E-11



7440439

Cadmium



5.00E-41



63252

CarbaiyVSevin



1.00E-1'



1563662

Carbofaan/furadan



5.00E-3'



75150

Carbon disulfide



1.00E-1'



133904

Chloramben



1.50E-2'



57749

Chlordane

1.30E+O

6.00E-51



5103719

CMordane, alpha(cis)-

1.30E+0

6.00E-5

chlordane

5103742

CHordane, beta(trans)-

1.30E+0

6.00E-5

chbrdane

5566347

Chlordane, gamma(trara>

1.30E+0

6.00E-5

chtordane

999999247

Chlofda«-r»nachlor(cis)-

1.30E+0

6.00E-5

chbrdane

999999248

Chtordane-nonachbr(trans)-

1.30E+0

6.00E-5

chlordane

108907

Chlorobenzene



2.00E-2'



510156

Chtorobenrilate

2.70E-1'

2.00E-2'



75003

Chtoroethane



4.00E-1®



75014

CWoroethene

1.90E+0"





110758

Chloroethytvinyl ether, 2-



2.50E-2'



74873

Chloromethane

1,30E-2h





91587

Chloconaphthalcre, 2-



8.00E-21



95578

Chlorophenol, 2-



5.00E-31



2921882

Chkapyrifbs/Dursban



3.00E-3'



E-3


-------
Vj)|K'luli\ 1%

liable E-l. (Continued)

CAS Number

Chemical Name

Cancer Slope Factor

((mg/kg/d)-')
(Followed by source;
see footnotes)

Noncancer Reference

Dose (mg/kg/d)
(Followed by source;
see footnotes)

Surrogate Chemical
Used (if nectssaty)

7440473

Chrotrium



5.00E-3'



218019

Chrysene

7.30E-3*





7440508

Copper



3.71E-2"



108334

C re sol n>



5.00E-21



95487

Cresol, o-



5.00E-21



106445

Cresot p-



5.00E-3"



1319773

Cresob



5.00E-3

p-Cresol

98828

Current



4.00E-2'



21725462

Cyanazix

8.40E-1"

2.00E-03"



57125

Cyanide



2.0OE-21



1861321

DCPA/Daclhal



1.00E-2'



53190

DDD, o,p'-

2.40E-1



p,p'-DDD

72548

DDD, p,p -

2.40E-1'





3424826

DDE, o,p-

3.40E-1



p,p'-DDE

72559

DDE, p,p'-

3.40E-11





789026

DDT, o,p'-

3.40B-1

5.00E-4

p,p'-DDT

50293

DDT, p,p-

3.40E-1'

5.00E-4i



999999300

DDT (Total)

3.40E-1

5.00E-4

p.p'-DDT

1163195

DecabroiTodpheny] oxide



1.00E-21



84742

Di-ivbutyl phthalate



1.00E-11



117840

Di-i>-oct>1 pllMalc



2.00E-2"



3334515

Diazinon/S pectiacide



9.00E-4"



53703

DjbeiHD(a4i)arthracene

7.30E+0*





132649

Dibemofuran



4.00E-3®



96128

Dibromo-3-chloropropane, 1,2-

I.40E+0"





124481

D&raniocHloronE thane

8.40E-2

2.00E-21



1918009

Dfcairf»a



3.00E-2'



95501

Dichtorobenzfine, 1,2-



9.00E-2'



541731

DichtorobereEne, 1,3-



8.90E-2'



106467

Dichlorobenzene, 1,4-

2.40E-2»





25321226

DkWorobenzEncs

2.40E-2

8.90E-2

1,3-and 1,4-
dicUorobeiizene

91941

Dichbrobenzidu;, 3,3-

4.50E-1'





E-4


-------
National Sdlimcut Quality Survey

Table E-l, (Continued)

CAS Number

Chemical Name

Cancer Slope Factor

(Cme/kg/d)'}
(Followed by source;
see footnotes)

Noncancer Reference

Dose (mg/kgtf)
(Followd by source;
see footnotes)

Surrogate Chemical
Used (if necessary)

75718

DchlorodltoroiTeihaiie



2.006-1'



75343

Dicbkxoethime 1,1-



1.00E-1'



S07062

Dfchkwoethane 1,2-

9.10E-2'





75354

Dfchkxoethene, 1,1-

6.00E-1'

9.00E-3'



156605

DichtoroetheiB, trans-1,2-



2.00 E-2'



156592

Dfchkjiocthylene, cis-1,2-



1.0GE-2*



75092

rHchlwomethane

7.50E-3'

6.00 E-2'



120832

Dchloropihenol, 2,4-



3.00E-3(



94757

McltaophetioxyaceUc acid, 2,4-



1.00E-2'



94826

DfcWorophenoxybutaiuic acid, 2,4-



8.00E-31



78875

Dfchloropropanc, 1,2-

6.80E-2"





542756

Dichkwopropene, 1,3-

1.75E-1'

3.00E-4'



62737

Dichtorvos

2.90E-1'

5.00E-*



115322

DicofeWCeMiaiie

4.40E-1*





60571

Dfeldrm

1.60E+1"

5.00E-5"



84662

Dclhyl phihalate



8.00E-1'



119904

Djmethoxyberajdinc,3,3'-

1.4GE-21"





131113

Dkrethyl phihalate



1.00E+1"



105679

Dimelhylphenol, 2,4-



2.00B-2*



528290

DiiiUobcnzB», 1,2-



4.a>E-4"



99650

Dinteobcnzcne, 1,3-



t.OOE-4'



100254

DWttobetHene, 1,4-



4.00E-4'



51285

Dintaophenol, 2,4-



2.00E-3'



121142

DMtrotoIuene, 2,4-



2.00E-3'



606202

DWtrotoluenc, 2,6-



1.00E-3"



88857

Dkioseb/DNBP



1.00E-3'



122667

Dqphenylhydruiie, 1,2-

8.0CE-1'





298044

Disufibton



4.00E-5'



959988

EndosuKan, alpha-



6.00E-3

erxiosulfan

33213659

EndosuKan, beta-



6.00E-3

endosufan

115297

Endosulfim mixed isomers



6.00E-3'



72208

Endrin



3.00E-41



E-5


-------
\|>|)on
-------
NitfiorinJ SV
-------
Appendix K

Table E-l. (Continued)

CAS Number

Chemical Name

Cancer Slope Factor

((mg/kg/d)-1)
(Followed by source;
see footnotes)

Noncancer Reference

Dose (mg/kg/d)
(Followed by sounce;
see footnotes)

Surrogate Chemical
Used (if necessaiy)

13071799

Terbufos/Counter



2.50E-5"



886500

Terbutryn



1.00B-31



95943

Tetrachlorobenzene, 1,2,4,5-



3.00E-4'



1746016

Tetrachtorodfoenzo-p-dhxin, 2,3,7,8-

1.56E+5h





79345

Tetrachloroe thane, 1,1,2,2-

2.00E-11





127184

Tetrachloroethene

5.20E-2®

1.00B-21



56235

Tetrachlororre thane

1.30E-11

7.00E-4'



58902

Tetrachlorophenol, 2,3,4,6-



3.00E-2'



961115

Tetrachtorvinphos/Gardona/S tirof

2.40E-2"

3.00B-21



7440315

Hn



6.00B-1"



108883

Toluene



2.00E-11



8001352

Toxaphene

1.10E+01





75252

TVfcromorre thane (Bromofbrm)

7.90E-"

2.00E-21



120821

IVichlorobenzeiK, 1,2,4-



1.00E-21



71556

THchkjroethane, 1,1,1-



9.00E-2"



79005

Ttichloroethane, 1,1,2-

5.70E-2'

4.00E-31



79016

Urichloroe there:

1.10E-2"

6.00E-3'



75694

TticHorofluoroirethane



3.00E-11



67663

TYichloro methane (Chloroform)

6.10E-3'

1.00E-2i



95954

HichJorophenol, 2,4,5-



1.00E-1'



88062

Ttichlorophenol, 2,4,6-

1.10E-21





93765

Ttichlorophenoxyacetic ackl, 2,4,5-



1.00E-21



93721

TYichtoropteroxypropionic acid,
2,4,5-



8.00E-31



1582098

HifluraSn/Iteflan

7.70E-3'

7.50E-3'



95636

THnrthylbenzEne, 1,2,4-



5.00E-4®



118967

THnkrotokjene

3.00E-21

5.00E-4'



7440622

Vanadium



7.00E-3"



108054

Vinyl acetate



1.00E+0"



108383

Xylene, m-



2.00E+0h



95476

Xylene, o-



2.00E+0"



1330207

Xylenes



2.00E+01



7440666

Zinc



3.00E-11



E-8


-------
Nillioiliil Si'fliiiiriif	Shims

Codes:

'Integrated Risk Information System (IRIS).

'Health Effects Assessment Summary Tables (HEAST).

'Environmental Criteria and Assessment Office (ECAO, as cited in Risk-Based Concentration Table).

"Other EPA documents, as cited in Risk-Based Concentration Table.

"Withdrawn from HEAST, but use continued for screening assessments (USEPA, Risk-Based Concentration Table).

E-9


-------
Appendix

E-10


-------
Appendix F

Species Characteristics
Related to NSI
Bioaccumulation Data

TableF-l presents the species used in tissue residue analyses whose results are included in the NSI, For each
species listed, Table F-l identifies the species as resident or migratory (or either) and demersal or pelagic (or
either) and specifies whether the species might be consumed by humans (i.e., recreational or subsistence
anglers). A species is considered either resident or migratory if it stays predominately in one location as long as food
and habitat are available but is capable of traveling long distances to find food and suitable habitat. A species is
considered either demersal or pelagic if it spends much of its time in the water column but is likely to feed off the
bottom. If a species is identified as either resident or migratory, it is considered resident for the purpose of this
analysis. If a species is identified as either demersal or pelagic, it is considered demersal.

F-l


-------
Appendix I'

Thble F-l. Species Characteristics Related to Tissue Residue Data

Specks Ode

Scientific Name

Common Name

Resident/Migratory"

Demersal/Pelagic'

Potenialiy Eatable

615301010-100

Acmlfiomysis mscmpsb

Mysid shrimp

E

E



611829010000

Acartia spp.

Copepod (unknown species)

M

P



872901010000

Acipenserspp.

Sturg»n (unknown Specfes)

M

D

Y

872901010600

Acipeaser fulvescens

Lake sturgeon

R

D

Y

872901010500

Aclpenser oxyrhynchus

Atlantic sturgeon

M

D

Y

872901010300

Acipenser tranmonlanus

White sturgeon

M

D

Y

877601200100

Acrocheihts alutaceus

Chisetaotlh

R

P



87S503060100

Albsmerus ?Lor,go!us

Whitebait smelt

M

P

Y

874701010200

Absa aestivalis

Bijeback herring

M

P

Y

874701010600

Absa chrysochhris

Skipjack herring

M

P

Y

874701010300

Absa mediocris

Hickosy shad

M

P

Y

874701010500

Absa pseudohairngus

Afcwift

M

P

Y

874701010100

Absa sapidissima

American shad

M

P

Y

383516020200

Ambbplites cavifmtts

Roanoke bass

R

P

Y

883516020100

Ambbplites rupestris

Rock bass

R

P

Y

877702060100

Ameiurus brimneus

Snal bullhead

R

D

Y

877702060200

Ameiurus coins

White cattish

R

D

Y

877702060300

Amriunis mehs

Black bullhead

R

D

Y

877702060400

Ameiurus nataUs

YeBow bullhead

R

D

Y

877702060500

Ameiurus nebubsus

Brawi billhead

R

D

Y

877702060600

Ameiurus platycephaius

Flat bullhead

R

D

Y

877702060700

Ameiurus serracanthus

Spo&d bullhead

R

D

Y

873401010100

Amia caiva

Bow&t

R

E

Y

884202010200

Anarhichas denticulatus

Northern woKfth

R

D

Y

874101010100

Anguilb rostmia

Ametfcaneel

M

P

Y

883544260100

Apbdinotus gmmtkns

Freshwater dram

M

E

Y

883516090100

ArchopUles imermptta

Sacramento perch

R

P

Y

883543030100

Archosargus pmbatocephaha

Sheepshead

M

P

Y

551539010100

Arctica isbndica

Ocean quabog

R

D

Y

877718020200

Aritisfelis

Hardhead catfish

M

D

Y

883102040500

Artedius notospilotus

Bonehead sculpm

R

D



618102000000

Astacidae

Crayfish (family)

R

D

Y

F-2


-------
Table F-l. (Continued)

Species Code

Scientific Name

Common Name

Resident/Migratory"

Demersal/Pelagic*

Potentially Eatable

551519010000

Astarte spp.

Astarte clam (Unknown species)

R

D



551519011300

Astarte undata

Waved astarte

R

D



883561010100

Astro not us oceUatus

Oscar

R

P

Y

810601051100

Aslropecten verrilli

Margined seas tar

R

D



877718010100

Bagre marinus

GafitopsaQ catfish

M

E

Y

883544030100

BairdieUa chrysoura

Silver perch

M

P

Y

550000000000

Bivatvia

Class of molluscs

R

D

Y

550701160100

Brachiodontes recurvus

Hooked mussel

R

D

Y

874701040000

Brevoortia spp.

Menhaden (unknown species)

M

P

Y

874701040100

Brevoortia tyrannus

Atlantic menhaden

M

P

Y

618901030100

Callinectes sapid us

Btoe crab

M

D

Y

618105010600

Cambarus bartoni

Crayfish

R

D

Y

877601140100

Campostoma anomalum

Central stone roller

R

E



618803010400

Cancer magister

Dungcness crab

M

D

Y

883528030300

Caranx hippos

CrevaQc jack

M

P

Y

877601030100

Carassius aural us

Goldfish

R

E



870802050100

Carcharhinus obscurus

Dusky shark

M

E

Y

870802050300

Carcharhinus plumbeus

Brown shark (sandbar)

M

E

Y

877604020000

Carpiodes spp.

Carpsucker (unknown species)

R

D

Y

877604020200

Carpiodes carpio

River carpsucker

R

D

Y

877604020100

Carpiodes cyprinus

QuiQback

R

D

Y

877604020300

Carpiodes velifer

High fin carpsucker

R

D

Y

877604010000

Catostomus spp.

Sucker (unknown sp)

R

D

Y

877604010500

Colostomas aniens

Utah sucker

R

D

Y

877604010100

Catostomus catostomus

Longnose sucker

R

D

Y

877604010400

Catostomus columbianus

BridgeHp sucker

R

D

Y

877604010200

Catostomus commersoni

Whfie sucker

R

D

Y

877604011200

Catostomus latipinnis

Flannebrouth sucker

R

D

Y

877604010300

Catostomus macrocheilus

Largescale sucker

R

D

Y

877604011500

Catostomus occidentalis

Sacramento sucker

R

D

Y

877604011600

Catostomus platyrhynchus

Mountain sucker

R

D

Y

877604012000

Catostomus snyderi

Klamath largescale sucker

R

D

Y

F-3


-------
Appendix. F

Ibble F-l. (Continued)

Speclej Code

Scientific Name

Common Name

Resident/Migratory*

Demcrsal/ft lagle1"

Potentially Eatable

8776W012100

Colostomas tahoensis

"fthoesiKker

R

D

Y

883516000000

Cenlrarchidae

Sut&h

R

P

Y

883516030100

Centramhus macmplerus

FHer

R

P

Y

883501010500

Ctnneptsmus undecimalis

Common snook

M

P

Y

883502030100

CeMmpristis striata

Black sea bass

M

P

Y

900201010100

Chefydra serpentina

Snapping turtle

R

E

Y

648933000000

Chinnomidae

MHgs 6m%

R

D



648960063300

Chimnomus riparius

Midge

R

D



883561090100

Cichla oceitaris

Peacock cicMd

R

P

Y

885703010100

Cilkarichshys sordidus

Pacific sanddab

E

D



88570301U00

Citharichtkp xamhostigma

Longlki sanddab

E

D



877712010200

Cichla Clarias fuscus

Whiespotted clarias

M

D

Y

877601070100

CBmstomus funduloides

Rosyside dace

R

P



551545020100

Corbkuta manilensis

Asiatic clam

R

D

Y

875501010800

Corrganus artedii

Cisco (lake herring)

M

P

Y

875501010600

Cottgonus ciupeafcrmis

Lake wijiefish

M

P

Y

875501010900

Cottgonus hoy!

IS baler

M

P

Y

883102000000

Cottidae

Scu^in family

R

D

Y

883102080000

Coitus spp.

Sculp in (unknown species)

R

D



883102080100

Callus aleulicus

Coastraitge scii^r

R

D



883102080700

Cottus bairdi

Moiled sculpm

R

D



883102080900

Cottus camiinae

Banded scuipii

R

D



883102080200

Coitus cognatus

Slimy sculpii

R

D



551002010000

Crassastrta spp.

Oysters (unknown species)

R

D

Y

551002010100

Crassostira gigas

Pacific oyster

R

D

Y

551002010200

Crmsostrea virginica

Eastern oyster

R

D

Y

877601230100

Ctenopharyngodon idtUa

Glass carp

R

E

Y

877604060100

Cycleptm ebngatus

Bbe sucker

M

D

Y

883544010200

Cynoscion nebubsus

Spotted sea trout

R

P

Y

883544010300

Cynoscion nothus

Silver sea trout

M

P

Y

883544010400

Cyrtoscion tvgalis

Weakfish

M

P

Y

877601761400

Cyprinetla lutrensis

Red shiner

R

P



F-4


-------
Table F-l. (Continued)

Species Code

Scientific Name

' Connnn Name

Resident/Migratory"

Demersal/Pelagic"

Potentially Eatable

877601761900

CyprineUa spiloptera

Spotfin shiner

R

P



877601000000

Cyprbshlae

Carp/goldfish (hybrid)

R

E

Y

877601010100

Cyprinus carpb

Common carp

R

D

Y

871305010500

Dasyatis sabina

Atlantic stingray

M

D

Y

874701050100

Domsoma cepedianum

Gizzard shad

M

P



874701050200

Dorosoma pesemnse

Throadfin shad

M

P



551202030100

Etiiplio compkmata

Freshwater clam

j

D

Y

885704040300

Eopsetta ex ills

Slender sole

E

D

Y

883544120500

Equetus punctata:

Spotted dram

R

D



877604030000

Erimyzon spp.

Chubsucker (unknown speces)

R

E



877604030200

Erimywn obbngus

Creek chubsucker

R

E



877604030100

Erimyzon sucetla

Lake chubsucker

R

E



875801000000

Esocidae

Pike

R

P

Y

875801010201

Esox americanus americanus

Red fin pickerel

R

P

Y

875801010202

Esox americanus vermiculalus

Crass pickerel

R

P

Y

875801010100

Esox lucius

Northern pike

R

P

Y

875801010400

Esox masquinongy

Musketage

R

P

Y

875801010300

Esox niger

Cham pickerel

R

P

Y

883520016700

Elheosioma radiosum

Orangebely darter

R

D



883520010900

Etheostoma spectabile

Orangethroat darter

R

D



883520017600

Etheostoma stigmaeum

Speckled darter

R

D



883520018700

Etheostoma whipplei

Redfil darter

R

D



883520018800

Etheostoma zpnale

Banded darter

R

D



880404021000

Fundulus zebrinus

Plate kSBfish

R

P



880404021100

Fundulus olivaceus

Blackspotted topminnow

R

P



879103040100

Gad us macmcephalus

Paci& cod

M

E

Y

870802020100

Gainocenio cuyier

Tiger shark

M

E

Y

880408010100

Gambusia qffinis

Western mosqutofish

R

P



883544020100

Genyonemus tineas us

While croaker

M

E

Y

877601260000

Gila spp.

Chub (unknown species)

R

E



877601261500

Gila mbusta

Round tail chub

R

E



883551020100

Girelfa nigricans

Opaleye

M

P



F-5


-------
Append!* !•"

Table F-l. (Continued)

Species Code

Scientific Name

Common Name

Resident/Migratory"

Demersal/Pelagic"

Potentially Eatable

8SS704350100

Gfyplacephalus zachi

Rex sole

E

D



S51202060100

Gonidea angulata

Freshwater mussel

R

D

Y

874701(KXHX)0

Glupetdae

Herring family

M

P

Y

622003030000

Htxagenm spp.

Bunowiig mayfty (unknown species)

R

D



622003030700

Hexagtnia limbata

Mayfly

R

D



875101010100

Hiodon absoides

GoHcyc

M

P

Y

t7S1010102X)

Hiodon lergiius

Mooreye

M

P

Y

885703110200

Hippoghssina stcmata

Bigmouth sole

M

D

Y

885704060100

Hippogbssoides das

Flathead sole

M

D

Y

885704060300

Hippogbssoides platessoides

American plaice

M

D

Y

6169230-10100

Hyalella aztcca

Freshwater amphipod

R

E



877601050300

Hybognaihus placUus

Plains minnow

R

P



871602010100

Hydmlagus colliei

Spotted rat fish

M

D



877604050100

Hyptnteiium nigricans

Northern hog sucker

R

D

Y

875503010100

Hypomesus pitliosus

Surf smelt

M

P

Y

885704220100

Hypsopseua guUubia

Diamond turbot

7

D

Y

877702000000

Iclalaridae

Bullhead catfish family

R

D

Y

877702010000

Iclalurus spp.

Catfish (unknown specks)

R

D

Y

877702010200

Ictahtrus furcmus

Bkie catfish

R

D

Y

877702010500

let alums ptwclalus

Channel catfish

R

D

Y

877604070100

IcSbbus buhalus

Stralkiouth buflato

R

E

Y

877604070200

lesiobus cyprinelius

Bgrouth buffab

R

E

Y

877604070300

Ictbbus niger

Black buflab

R

E

Y

8S3543020100

Lagodon rkomboides

Pinfish

E

P



870600000000

Lamntfermes

Shark

M

P

Y

877601300100

Lavinia exiiimuda

Hitch

R

P



8S3S44040100

Lefoslomus xanthums

Spot

M

P

Y

884701030100

Lepidogobius lepidus

Bay gaby

R

P



873201010000

Lepisosteus spp.

Gar (unknown species)

E

P

Y

873201010200

Lepisosteus oculatus

Spotted gar

E

P

Y

873201010100

Lepisosteus osseas

Longnosc gar

E

P

Y

873201010300

Lepisosteus platostomus

Shortnose gar

i

P

Y

F-6


-------
Table F-l. (Continued)

Species Code

Scientific Name

, Cotrwncm Name

Residert/Migratoiy*

Demersal/Pelagic1

Potentially Eatable

873201010400

Lepisosteus spatula

Alligator gar

E

P

Y

883516050000

Lepomis spp.

Sunfeh (unknown species)

R

P

Y

883516050100

Lepomis auritus

Redbreast sunfish

R

P

Y

883516050200

Lepomis cyanettm

Green sunfish

R

P

Y

883516050500

Lepomis gibbosus

Pumpkins eed

R

P

Y

883516050300

Lepomis gulosus

Warmouth

R

P

Y

883516050600

Lepomis hum:Us

Orangespotted sunfish

R

P

Y

8S3516050400

Lepomis macmchirus

Blue gill

R

P

Y

883516050700

Lepomis marginalus

Dollar sunfeh

R

P

Y

883516050800

Lepomis megabits

Longear sunfeh

R

P

Y

883516050900

Lepomis micmlophus

Redearsunfish

R

P

Y

883516051000

Lepomis punctaius

Spotted sunfish

R

P

Y

879103080100

Lota lota

Burbot

M

E

Y

618701150200

Loxorhynchus gnmdis

Decorator crab

R

D



500501010300

Lumbriculus variegatus

Aqauatic worm

R

D



88353601Q700

Lutjanus campechanus

Red snapper

M

D

Y

877601780400

Lwcilus chrysocephalus

Striped shiner

R

P



877601780600

hixitus comutus

Common shiner

R

P



885704110100

Lyopsella exitis

Slender sob

M

D

Y

814802010600

Lytechinus anamesus

Litle gray sea urehii

R

D



551531013600

Macoma ims

Clam (macoma)

R

D

Y

551531011400

Macoma nasuta

Bent-nosed macoma

R

D



877601800200

Macrhybopsis gelida

Surgeon chub

R

E



551202430300

Megatonaias gigantea

Washboard mussel

R

D

Y

551547110100

Mercenaria meivenaria

Quahog

R

D

Y

883544070100

Micmpogonias mduktts

Atlantic croaker

M

P

Y

883516060000

Micmpterus spp.

Bass (unknown species)

R

P

Y

883516060500

Micmpterus coosae

Redeye bass

R

P

Y

883516060100

Micmpterus dolomieu

SmaCmoiith bass

R

P

Y

883516060600

Micmpterus notius

Swameebass

R

P

Y

883516060300

Micmpterus punctulaus

Spotted bass

R

P

Y

883516060200

Micmpterus salmoides

Laigemoulh Bass

R

P

Y

F-7


-------
\|)|RMl(li\ I''

TkbleF-1. (Continued)

Speclei Code

Scientific Name

Conrnun Name

Resident/Migratory*

Dc me rsaVPc Ingle'

Potentially Eatable

8776O408O100

Minytnma mekmops

Spotted sucker

E

D

Y

883502010000

Momne spp.

Tfemperate bass (unknown species)

E

P

Y

883502010100

Momm amcricana

While perch

M

P

Y

883502010400

Momne chrysops

White bass

M

P

Y

883502010300

Morom chrysops x saxatilis

Hybrid striped bass (whie/strfied)

E

P

Y

8S3502010500

Momne mlssissippiensis

YeSow bass

M

P

Y

S83502010200

Monme saxatilis

Striped bass

M

P

Y

877604040000

Moxostoma spp.

Redborse (unknown species)

R

D

Y

877604040400

Moxostoma anisumm

Stiver red horse

R

D

Y

877604040700

Moxostoma carinaium

River redborse

R

D

Y

877604040200

Moxostoma conges!um

Gtayredhorse

R

D

Y

877604040900

Moxostoma duquesnei

Black redborse

R

D

Y

877604041000

Maxostoma etythrumm

Golden redborse

R

D

Y

877604040100

Moxostoma macmlepidotum

Shorthead rcdboree

R

D

Y

877604041400

Moxostoma papptUosum

V-^j redhoree

R

D

Y

877604040300

Moxostoma poectturum

Blacktail redborse "

R

D

Y

877604041700

Moxostoma rupiscarles

Striped jumprock

R

D

Y

883601010100

Mugil cephahts

Striped millet

M

E

Y

883601010200

Mugil currma

White mulfct

M

E

Y -

870802040100

Musteius amis

Smooth dogfish

M

E

Y

551701020100

Mya armaria

Soft clam

R

D

Y

87760U70100

Mylocheilus caurinus

Peanwuih

R

E



877601350100

Mybpharodort conacephabts

Hard lead

R

E



350701010000

Mytilus spp.

Mussel (unknown species)

R

D

Y

550701010200

Mytilus califamimus

CaBbmia mussel

R

D

Y

550701010100

Mytilus edutis

Bkie mussel

R

D

Y

500124030500

Necmlhes arenaceodensala

Sand worm

R

D



500168040100

Neoampkllrile robusta

TfecrebeBd worm

R

D



500125011900

Nepktys caecoides

Sand worm

R

D



500168040100

Neoamphitrite robusta

TferrebeJki worm

R

D



500125011900

Nephtys caecoides

Sand worm

R

D



500125011500

Nepktys mcisa

Red-lined worm

R

D



F-8


-------
I

Table F-l. (Continued)

Species Code

Scientific Name

Common Name

Reside nl/Migratoiy'

Demersal/Pelagic1

Potentially Eatable

877601100300

Nocomis asper

Redspot chub

R

E



877601100200

Nocomis leptocephalus

Bbebead chub

R

E



877601100100

Nocomis micropogon

River chub

R

E



877601060100

Notemigonus crysoleucas

Golden shiner

M

P



877601501000

Notrvpis amblops

Bigcyc chub

R

E



877601114100

Notrvpis boops

Bigeye shher

R

P



877601111400

Notrvpis buchanani

Ghost shiner

R

P



877601110600

Notrvpis hudsonius

Spottail shiner

R

P



877601118100

Notrvpis nubilus

Ozark minnow

R

E



877601112300

Notrvpis stramineus

Sand shiner

R

P



877702020200

Noturus insignis

Margined madtrom

R

D



877702021800

Noturus miurus

Brindled madtom

R

D



877702022000

Noturus phaeus

Brown madtom

R

D



870703010100

Odoruaspis taurus

Sand tiger

M

E

Y

500300000000

OUgochaeles

Aquatk worms

R

D



875501020800

Oncorhynchus clarki

Cutthroat trout

E

P

Y

875501020100

Oncorhynchus gorbuscha

Pink salmon

M

P

Y

875501021100

Oncorhynchus mykiss

Rainbow trout

E

P

Y

875501020300

Oncorhynchus kisutch

Coho salmon

M

P

Y

875501020500

Oncorhynchus nerka

Sockeye salmon

M

E

Y

875501020600

Oncorhynchus tshawytscha

Chinook salmon

M

E

Y

878301020000

Opsanus spp.

Toadfish (unknown species)

R

. D



618105030000

Orconectes spp.

Crayfish

R

D

Y

877601360100

Orthodon micrvlepidotus

Sacramento blackfish

R

P



883540020100

Orthopristis chrysoptera

Pigfish ;

R

P

Y

875503000000

Osmeridae

Smelt (species unknown)

M

P

Y

875503030200

Osmerus mordax

Rainbow smelt

M

P

Y

618102020100

Pacifastacus leniusculus

Crayfish

R

D

Y

617918010100

Pandalus borealis

Maine shrimp

R

D



883502160400

Paralabrax nebulifer

Barred sand bass

E

D

Y

885703030900

ParaOchthys californicus

California halibut

M

D

Y

885703030100

Paraiichthys dentatus

Summer flounder (fluke)

M

D

Y

F-9


-------
A|)|)t'll(li\ I'

¦

"IbbIeF-1. (Continued)

Specks Code

Scientific Name

Convnon Name

Resident/Migratory"

Demersal/Pelagic*

Potentially Eatable

885703030400

Paralichihys lethostigma

Sotfhem flounder

M

D

Y

817502010100

Parasiichopus califomicus

California sea cucumber

R

D



500166030400

Pectinaria califomiensis

Sand worm

R

D



617701010000

Penaeus spp.

Shrimp

R

D

Y

617701010100

Pcnaeus aztecus

Brown shrimp

R

E

Y

617701010300

Penaeus setiferus

Whie shrimp

R

E

Y

883520020100

Perca flavescens

Yellow perch

R

P

Y

883520030900

Percina copelandi

Channel darter

R

D



883560050100

Phanerodon furcaius

White seaperch

R

P

Y

877601370300

Phoxlnus erythrogaster

Soiahern redbeDy dace

R

P



877601160200

Pimephales promelas

Fathead minnow

R

P



811703050100

Pisaster brevispirtus

Starfish

R

D



550905090100

Placopectert mageltanlcus

Atlantic deep-sea scaHup

R

D



885704140100

Ptailchthys steUaius

Starry flounder

M

D *

Y

877601840100

Platygobio gracilis

Flathead chub

R

E



885704151000

Pleuronectes bilineaius

Rock sole

E

D

Y

885704130100

Pleuronectes vetulus

English sob

M

D

Y

885704000000

Pleuronectidae

Rigjiteye flounder family

M

D

Y

885704160200

Pleuronichihys decurrens

CurBn sole

M

D

V

885704160400

Pleuronichihys verticalis

Hornybead turbot

M

D

Y

880408110200

Poecilia viltata

Cuban limia

E

P



883544080100

Pogonlas cromis

Black dmm

M

P

Y

872902010100

Pofyodon spaihula

Paddb&b

M

P

Y

883525010100

Pomatomus sallairix

B be fish

M

P

Y

883516070000

Pomoxis spp.

Crappie (unknown species)

R

P

Y

883516070100

Pomoxis annularis

White crappie

R

P

Y

883516070200

Pomoxis nigromaculatus

Black crappie

R

P

Y

882602010100

Pricnotus carvlinus

Northern searobin

R

D

Y

875501060100

Prosoplum cylindraceum

Round whie fish

M

P

Y

875501060200

Prosopium williamsoni

Mountain whiefish

M

P

Y

551547070100

P/otothaca staminea

Clam (Pacific littleneck)

R

D

Y

885704150400

Pleuronectes americanus

Winter flounder

M

D

Y

F-10


-------
Table F-l. (Continued)

Specks Code

Scientific Name

Common Name

Reside nt/Mijra tor)"

Demersal/Pelagic'

Potentially Eatable

885704150400

Pkumuectes americanus

Winter flounder

M

D

Y

877601180000

Ptychocheiius spp.

Sqwwfisb

. R

E

Y

877601180100

Ptychocheiius omgommis

Northern squawfish

R

E

Y

877702030100

Pylodictis olivaris

Flathead catfish

R

E

Y

871304010300

Raja binocuhta

Winter skate

I

M

D

Y

890302010600

Ram catesbeiana

Bullfrog

?

P

Y

551525040100

Rangia cuneota

Brackish water clam

R

D

Y

877601090000

Rhmkhl'nys spp.

Dace (unknown species)

R

D



877601190100

Richatdsomm ball eat us

Redside shiner

R

P



875501030000

Salmo spp.

Th>ut (unknown species)

E

P

Y

875501030500

Salmo salar

Atlantic sahen

M

P

Y

875501030600

Salrrw trutta

Brown trout

E

P

Y

875501000000

Salmantdae

Hmr (family)

E

P

Y

875501040000

SaJvefcus hybrid

Splake (hybrid)

E

P

Y

875501040400

SalvcBnus fontmalis

Brook trout

E

P

Y

875501040100

Satvelinus malma

Dolly vaidea

E

P

Y

S75501040300

Salvelinus namaycush

Lake trout

E

P

Y

551547020100

Saxidomus gigtmlem

Clam (smooth Washington)

R

D

Y

872901020200

Scaphirhynchus plmorynchus

Staovehose sturgeon

M

D

Y

883544000000

Scimnidoe

Drum 6n%

M

E

Y

883544090100

Sciaenaps oceBatus

Red drum

M

E

Y

885003030100

Scomber japonlcus

Chub mackerel

M

P

Y

885003050100

Scomberomoms cavalla

Khg mackeral

M

P

Y

885003050200

Scombcmmoms maculatus

Spanish mackerel

M

P

Y

885703040100

Scophlhalmus aquosus

Windowpane

M

D

Y

882601061600

Scorpaena guttata

CaKbmia scorpionfish

R

D

Y

883102310100

Scorpaenichtkys marmomtus

Cabezon

R

D



882601010300

Sebastes auricukttm

Brown rock&h

M

P

Y

882601012000

Sebastes maliger

Qulback rockfish

M

P

Y

882601012100

Sebastes melanops

Black rockfish

M

P

Y

882601013900

Sebastes nonegicus

Goto redfisb

M

P

Y

882601012700

Sebastes paacispinis

Bocaccio

M

P

Y

F-ll


-------
Appendix F

Tbble F-l. (Continued)

Specks Code

ScientilSc Name

Common Name

Resident/Migratory'

Demersal/Pelagic'

Potentially Eatable

882501012700

Scbastcs paucispinis

Bocaccb

M

P

y

882601013000

Sebastes pmrigcr

Redstripe rockfish

M

P

Y

877601080200

Seawlilus almmaculatus

Creek chub

R

E



877601080100

SemotUus corporate

Falffish

R

E



877601080300

Semelltus Umbee

Sandhills crab

R

E



617704010900

Sicycnta mgentis

Rock shitap

R

D



5S1529020100

Solrn sicarius

Razor clam

R

D



871001020100

Squalus aeanthias

Spiny dogfish

M

E

Y

883520040200

Stizpstedion canadcnse

Sauger

R

P

Y

883520040100

Slisostedion vitmum

Walleye

R

P

Y

880302020100

Strongylura marina

Atlantic needlefish

M

P



88S703130300

Syaciam papillosum

Dusky fbunder

M

D

Y

885802011600

Symphurus atricauda

CaEfomia tonpiefish

M

D



876202010100

Sywdus fattens

Inshore izartfisb

R

D



885003040400

Thumus atlanticus

Blackfti tuna

M

P

Y

87S50I070100

Thymallus aKticus

Arctic gray&g

E

P

Y

883561400100

Tilapia mossambica

Mozambique tilapia

R

E

Y

883561040500

1llapia zilUi

RedbeDy tilapia

R

E

Y

551525020100

Tmsus capax

Honse clam

R

D

Y

870802090200

Trbikis semfasciala

Leopard shark

M

E

Y

884701300100

Tridtnligcr trigonocephaly

Chamefcon goby

R

D



88SS01010100

Trinectes maculalus

Hagehoker

M

D



880302030200

fylosurus crocodibts

HoureJfisb

M

E

Y

875802010200

Umbra limi

Central roodmhnow

R

E



050601010000

Vaucheria

Mactoal^e

?

E



*Filh spedei is cooiidcrtd: R =» red dent, M » migratory, E = either resident or migratory, ? = unknown.
^FWi ipcdets is coQitderedi D =* demersal, P =- pelagic, E = either, ? = unknown.

F-12


-------
!
I

NiilioiuiJ	Qti:iJify Survey

Appendix G

Notes on the Methodology for
Evaluating Sediment Toxicity
Tests

Results of sediment toxicity tests conducted around the United States were submitted with several databases
for evaluation in the NSI. Additional processing of records was required for most of the data. Because
test results were reported differently in each database, appropriate interpretation of the test results was
sometimes confusing. This section explains how the toxicity test data were handled for the NSI evaluation with
respect to issues related to sampling date, type of test, sample location identification, and results of control or refer-
ence tests conducted during the toxicity tests.

Sampling Date

Only those tests in the databases for which the sediment samples were obtained between January 1, 1980, and
December 31, 1993, were evaluated. Tests before and after that period were eliminated.

Sample Location

Records were examined to determine whether the sampling station from which the sediment sample was col-
lected had been identified by latitude and longitude coordinates. Samples that were not referenced to a specific
location were not considered in this study. Tests from the Great Lakes Sediment Inventory (GLSI) database were not
considered because sample locations were not appropriately identified. Sediment samples in the EPA Region 10/lJ.S.
Army Corps of Engineers Seattle District's Sediment Inventory (SEACOE) from sampling stations located in British
Columbia were also not considered in the analysis.

Type of Test

Data from seven databases (Table G-l) were reviewed to determine whether they had reported the results of
sediment (solid-phase) and elutriate nonmicrobial toxicity tests in which the endpoint was mortality. Records per-
taining to chronic toxicity tests, microbial toxicity tests, tests that were not conducted with sediment or elutriate, and
tests in which the endpoint was not percent mortality (or percent survival, which could be converted to percent
mortality) were excluded from further consideration.

Only the DMATS and GOM databases clearly reported the phase (solid, elutriate, particulate) of sediment sample
used in the bioassays conducted; ODES provided this information for some of the tests. If the phase was not indi-
cated, this information was obtained or best professional judgment was used to identify the phase used in the tests.
For some tests, comparison of species with those used in standard EPA test protocols or with species used in other
sediment toxicity tests in the databases permitted assignment of phase with certainty. Other species might be used in
sediment-, elutriate-, and particulate-phase tests, and the phase was assigned with uncertainty. Table G-2 presents a
list of species used in toxicity tests whose results are included in the NSI. Table G-2 also presents the type of toxicity
test for which each species is generally used (i.e., liquid-phase, elutriate-phase, suspended particulate-phase, sedi-
ment/solid-phase). The data presented in Table G-2 are the basis for determining whether the toxicity test of concern
was conducted using the solid or elutriate phase. A "Y" entered in Table G-2 indicates that the phase was given with
the test results; an "E" indicates that the phase was estimated using best professional judgment based on the species
used in the toxicity test.

. .	' '	G-l


-------
Table G-l. Toxicity Test Database Characteristics

Database

Sample
Locations
Identified

by
Lat/Long

¦type
of Ttest

Laboratory
Control Tests

Reference Sediment
Tests

Comments

U.S. Army Corps of
Engineers, Dredged
Material Tracking
System (DMATS)

Yes, all 74

Solid and Elutriate
(identified in database)

Replicate control
test results provided

Replicate reference
sediments tested with
each batch of
sediment samples

Used means of reference
sediment replicates in the
evaluation (contact: Alan Ota, EPA
Region 9)

HPA's Environmental
Monitoring and
Assessment Program,
Louisianian Province
(EMAP-LA)

Yes, all 259

Solid Phase
(not identified in
database, provided)

Not provided In D3
database, provided on
request

No

Sediment sample test results were
calculated from the additional data
provided (contact: Kevin Summers,
EPA/ERLGB)

EPA's Environmental
Monitoring and
Assessment Program,
Virginian Province
(EMAP-VA)

Yes, all 179

Solid and Elutriate
(not identified in
database, provided)

Not provided in D3
database, provided on
request

No

Sediment sample test results were
calculated from the additional data
provided (contact: Daryl Keith, EPA/
ERLN)

Guif of Mexico
Program's

Contaminated Sediment
Inventory (GOM)

Yes, all 42

Solid Phase
(identified in database)

ERL-N: Yes
USACE: No
GCRL: No,
provided on request

ERL-N: Yes
USACE: Yes
GCRL: No

Long Island Sound reference sediment
was used to generate control data for tests
done by ERL-N (contacts: Phil Crocker,
EPA; John Scott, SAIC) and control data
obtained for GCRL (contact: Julia Lyle,
GCRL); for USACE tests used mean of
the reference test results as control

EPA's Great Lakes
Sediment Inventory
(GLSI)

No

Not identified in database

Not provided in database

No?

Sample location IDs and control test reults
were not provided; therefore, these data
were not evaluated for the NSI (contact:
Bob Hoke, SAIC)

EPA's Ocean Data
Evaluation System
(ODES)

Only 18 out
of 68

Solid Phase
(not identified in
database)

Yes

No

Used controls (contact:
Tad Deschler, Tetra Tech)

EPA's Region 10/U.S.
Army Corps of Engineers
Scsttlc District's
Sediment

Inventory (SEACOE)

Only 18 out
of 68

Solid Phase
(not identified in
database)

Yes, some had to be
provided on request

Yes

Used controls (contact: Roberts Feins,
Environmental Information Consultants;
John Armstrong, EPA Region 10; and
Gary Braun, Tetra Tech, for Puget Sound
Estuary Program Reports, 1988)


-------
Table G-2. Test Species Used in Sediment Bioassay Test Results Included in the NSI

0

1

W





Type of Toxicity Test



Species

Liquid

Elutriate

Particulate

Solid

C

(L most Common)

D

A



Species Code

Name

(L)

(E)

(P)

(S)

(L or E)

(L,E, or P)

(L,EJP,or S)

Unknown

80509070600













•



E

615301010900

Acanthomysis costata



Y



Y









615301010400

Acanthomysis macropsis

Y

Y



Y









615301010700

Acanthomysis sculpta

Y







E







611829010000

Acartia spp. spp.

Y

Y













616902010800

Ampelisca abdita







Y,E









616800000000

Amphipods







Y









610401010100

Anemia salina

Y

Y













616302070900

Asellus intermedius







E









650508331700

Chironomus riparius







E









650508330100

Chironomus tentans







E









885703010200

Citharichthys stigmaeus

Y





Y









616915021500

Corophium spinicome







Y,E









617922010000

Crangon spp. spp.

Y

Y



Y









551002010100

Crassostrea gigas



Y



Y



E





551002010200

Crassostrea virginica







Y









880404010100

Cyprinodon variegatus

Y



Y











610902010900

Daphnia magna









E







610902010100

Daphnia pulex









E







815501010100

Dendraster excentricus









E








-------
"Ikble G-2. (Continued)





Type of Toxicity Test

Species

Species

Liquid

Elutriate

Particulate

Solid

C

(L most Common)

D

A



Name

Name

(L)

m

(P)

(S)

(LorE)

(L,E, or P)

(L,E,P,®rS)

Unknown

880404020700

Fundulus grandis

¥



Y











881801010100

Gasterosteus aculeatus









E







616915090200

Grandidierella japonica







Y









622003030700

Hexagenia limbata







E









615301010700

Holmesimysis sculpta

Y

Y



Y

E







616923040100

Hyallella azteca







E









500501010300

Lumbriculus variegatus







B









814802010200

Lytechitms pktus

Y

Y













551531011600

Macoma balthica







E









551531011400

Macoma nasuta



Y



YJS









551531010000

Macoma spp.







E









615303140600

Metamysidopsis elongata

Y

Y



Y









651530100000

Mysid shrimp

Y



Y

Y









615301210200

Mysidopsis bahia





Y

Y









550701010100

Mytilus edulis

Y

Y









E



500124030500

Neanthes arenaceodentata







Y,E









500124030000

Neanthes spp.







E









500125011900

Nephtys caecoides







Y»E









500124030200

Nereis virens







Y









551706040100

Panopea generosa







E










-------
Table G-2. (Continued)





Type of Toxicity Test

Species .

Species

Liquid

Elutriate

Particulate

Solid

C

(L most Common)

D

A



Code

Name

(L)


-------
Vl»JK«ll(li\ (J

Only DMATS contained elutriate test results in addition to sediment test results; all other tests evaluated were
sediment (solid- phase) test results.

Test Controls

Toxicity data were screened to determine whether control data were reported. Sediment toxicity test laboratory
or performance controls are usually clean sand or sediment run under the same conditions in which the same test
organisms are exposed at the same time as those exposed to the sediment samples tested. Controls are used to
determine whether observed mortality might be the result of the quality of test organisms used or other factors, and
not the result of exposure to possible toxics in the sediment samples.

The databases were screened to locate control test data for each sediment sample tested. The GLSI database did
not contain any control test data; because of this, as well as the lack of station-identifying coordinates, the GLSI
database was eliminated from evaluation for the NSI. For the other databases, control test results were matched to the
sediment test results and were treated as follows:

•	Multiple control and reference sample test results were reported for each sediment tested in the DMATS
database. These were determined to be replicate test results. Because the sediment samples tested in DMATS
were usually fine-grained and the laboratory performance controls were sand, the reference sediment samples
were used as "controls" to evaluate toxicity of sediment samples. The percent mortality for the reference
replicates were averaged for each reference site to obtain the mean percent mortality for the reference sedi-
ment for comparison with the sediment sample test result.

•	The D3 version of both the EMAP-LA and EMAP-VA databases contained control-corrected results for the
sediment samples tested. The control-corrected results were obtained using the following equation:

percent survival of organisms in sediment sample test = control-corrected percent survival

percent survival of organisms in control test percent survival

•	EMAP-LA provided a revised database on request that contained the percent survival of the controls. The
sediment sample test results were calculated according to this equation:

percent survival of organisms in sediment sample test =

control-corrected percent survival X percent survival of organisms in control test

100

•	EMAP-VA provided a revised database on request that contained the mean percent mortality of controls and
the mean percent mortality of the sediment sample tests for each station, as well as the control-corrected
percent survival.

•	The GOM database reported control test results for tests conducted by EPA's Environmental Research Labo-
ratory in Narragansett. A low-salinity control test performed at the same time was not used in the evaluation.
The single reference sediment sample was treated as a sediment toxicity test result. No control tests were
available from the USACE data set within this database; the mean of reference sediment toxicity test results
was used as the "control" for these test data. No control test results were found in the GOM database for the
GCRL data set. Total percent mortality of pooled control test replicates were provided by Julia Lytle of
GCRL and entered into the database for the NSI analysis.

•	The ODES database reported single-value control results for the ARSR and OSE data sets. (Whether these
were means of replicate tests is unknown.) One sediment test result in ARSR was matched to two different
control test results; however, the one control test result that was not matched elsewhere in the data set was
eliminated for the analysis.

G-6


-------
National Sediment Quality Survey

*	The SEACOE database contained single-value control test results for the ALCTRAZ data set and several
series of control test results for other data sets (e.g., EVCHEM and EBCHEM). Information on the correct
control series was obtained, and the proper control test results were evaluated in the computer program.
Means were calculated for replicates in the series Mid used to evaluate the sediment sample test results.

Results of control tests reported as "percent survival" were converted to "percent mortality" by the following
calculations:

percent mortality = 100 - percent survival
percent mortality = number of surviving organisms/total number of organisms in test

Sometimes entries in databases reversed "mortality" and "survival" (e.g., PSE data set in the ODES database).
Any questions concerning the designation were checked and corrected if necessary. If replicate sediment toxicity test
results were provided for a sampling site in the database, a mean was calculated and compared to the mean control
mortality. (Some databases provided only the means, e.g., EMAP-LA, EMAP-VA.) For the purpose of the NSI
evaluation, if the control had greater than 20 percent mortality (less than 80 percent survival), that test was excluded
from further consideration.

Reference Sediment Stations

Some data sets included data for reference sediments that were run simultaneously with the control and sediment
samples. Reference sediment is sediment collected from a field site that is appreciably free of toxic chemical con-
taminants and has grain size, total organic carbon, sulfide and ammonia levels, and other characteristics similar to the
sediment samples to be tested for toxicity. Because reference sediments should match the characteristics of the
sediment samples more closely than the sand or sediment used for the laboratory (performance) control, they should
provide information on the appropriateness of using a particular test organism since the suitability and survival of
different species can be affected by these other physical and chemical characteristics of the sediment.

*	As noted previously, DMATS provided several reference sediment samples for each toxicity test, along with
control test results. The number of such reference sediment samples varied for different test dates, and these
sediment samples were determined to represent replicates. The average percent mortality was determined
from each set of replicates and this was used as a "control" to evaluate the toxicity of sediment samples in
this database. If percent mortality of the mean reference test result exceeded 20 percent, the sediment
toxicity tests that were run with that reference sediment were not used in the evaluation.

*	Reference sediment test results were not identified in the EMAP-LA, EMAP-VA, or ODES databases.

*	In the GOM database, a reference sediment test was run in tests conducted by EPA's Environmental Re-
search Laboratory in Narragansett. This single reference sediment sample was treated as a sediment toxicity
test result. Reference sediment tests in the USACE data set were averaged and used as the control for
analysis since other control test data were not provided in the data set.

*	Reference sediment toxicity test results in the SEACOE database were treated as a sample site.

Because reference toxicity test results were not available for all of the sediment toxicity tests, reference sediment
sample test results were not used as "controls" in the evaluation of sediment toxicity test data in the NSI, with the
exception of the DMATS data and the USACE data in the GOM database. The remaining reference sediment test
results were compared with the control results to determine whether significant toxicity was indicated at that field
site; i.e., they were treated like a sediment toxicity test result (see below).

It should be noted, however, that careful examination of such reference test results could improve the interpreta-
tion of sediment toxicity tests; i.e., they might indicate that test organisms were adversely affected by sediment
characteristics, not by toxic chemicals. Thus, the classification of some sites using the sediment toxicity tests might

G-7


-------
be inappropriate because the control test result did not adequately explain the result, based on the test organism's
health or sensitivity to test conditions.

Test Results

For the NSI evaluation protocol for sediment toxicity test data, significant toxicity was indicated if there was a
difference of 20 percent survival from control survival (e.g., if control survival was 100 percent and 80 percent or less
of the test organisms survived, or if control survival was 80 percent and 60 percent or less of the test organisms
survived, significant toxicity was indicated). Although a number of different test species and protocols were used in
the tests evaluated, this threshold provides a preliminary indication of sediment toxicity for classifying sampling
stations for the NSI.

G-8


-------
Appendix H

Additional Analyses for PCBs
and Mercury

To perform the screening analysis for the National Sediment Quality Survey using NSI data, EPA selected
reasonably conservative screening values, including theoretically and empirically derived risk-based screen-
ing levels. The limited number of sediment criteria available for use in this type of evaluation, however, contribut-
ed to the possibility of over- and underestimation of potential adverse effects associated with sediment contaminated for
some chemicals. Two chemicals where this issue is particularly relevant are PCBs and mercury. EPA conducted further
analyses on PCBs and mercury to determine the effect of using different assessment parameters on the number of sampling
stations where these chemicals were identified as associated with a probability of adverse effects.

Because of the tendency for PCBs to bind to sediment and because of the relative toxicity of these chemicals to

humans, EPA selected a precautionary approach for the analysis of PCBs in the NSI evaluation. The approach was
precautionary because (1) it did not require matching sediment chemistry data and tissue residue data for Tier 1
classification and (2) it used the cancer risk level of 10 s for all congener, aroclor, or total PCB measurements to
evaluate human health effects related to PCB contamination. EPA applied the cancer slope factor for aroclor 1260,
the most potent commercial mixture, to all measures. It should be noted that there were only 542 sampling stations
where matching sediment chemistry data and tissue residue data were available for analysis. In the foEowing evalu-
ation, the amount of PCB sediment and fish tissue data exceeding screening values other than those used in the NSI
analysis is compared to the number of sampling stations classified as Tier 1 or Her 2.

Figure H-l is a cumulative density function graph depicting the maximum PCB concentration at each sediment sam-
pling station where PCBs were detected. The various screening values that could be used to indicate adverse effects levels

1

*!

•s
*

1.00
0.90
0.80
0.70
0.80
F(x> 0.50
0.40





0.30
0.20
0.10
0.00

1,006-04 1.0QE-Q3 1.00E-O2 1.00E-01 1.00E-HJO 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06

PCB In Sedlmont Concontrallon (ppb)

I:

Note: Lofiara kknllfiod on saury# eeftetpcnd to PCB effect* icrwMitfjg yalwe* hi Table H-1,

Figure H-l. Cumulative Frequency Distribution of PCB Sediment Concentration Data (All Aroclors
and Total PCB).

H-l


-------
ApjH'ndh II

ofPCBs in sediment are plotted as A through S in the figure and described in Table H-l. The top two sections of Table H-
1 present the screening values ofPCBs in sediment that are protective of human or wildlife consumers. The levels shown
were derived using the theoretical bioaccumulative potential (TBP) analysis with the default lipid content (3 percent),
default organic carbon content (1 percent), and BSAFs with and without the safety factor of 4. (See Appendices B and C for
further explanation.) Depending on the screening value, the number of sediment chemistry sampling stations with detect-
able PCBs exhibiting potential human health or aquatic life effects varies from under 1 percent to over 99 percent The
screening values selected for the NSI evaluation classify approximately 85 percent of sediment chemistry sampling stations
in Her 2 for human health effects (Point D). For aquatic life effects, the selected screening values classify 25 percent of
sampling stations as Tier 1 (Point O) and 57 percent of sampling stations as Tier 2 (Point H).

Ibble H-l, Sediment Sampling Stations with Detectable Levels of PCBs That Exceed Various Screening
Values*,b

TVpc of Screening Value

Associated Level

(PPb)

Level Plotted in
Figure H-l
Corresponds to Letter

Number of Stations
with Detected PCBs
Exceeding Level

Percentage of Stations
with Detected PCBs
Exceeding Level

Protection of Consumers

Cancer Risk Level









lO6

0125

B

3,772

98.2

I0M

2.5

D

3,290

85.6

1»4

25

J

2,076

54.0

Nonoancer Hazard Quotient of 1

4°

L

1,761

45.8

FDA "Iblcraiicc Level

360

P

652

17.0

Wliifc Criteria

29

K

1,977

51.5

Protection of Consumers Using BSAF with Safety Factor'

Cancer Risk Level









10s

0,063

A

3,828

99.6

10'

0.63

C

3,648

95.0

10*

6.3

E

2,921

76.0

Noncancer Hazard Quotient of 1

9.9

G

2,699

70.2

FDA Tblcrance Level

90

M

1,330

34.6

Widife Criteria

7.2

F

2,849

74.2

Protection of Aquatic Life

ER-L

22.7

I

2,150

56.0

ER-M

180

N

976

25.4

AET-L

1,000

Q

353

9.2

AET-H

3,100

R

165

4.3

TEL-

21.6

H

2,182

56.8

PEL'

189

O

962

25.0

Other Protection Levels

TSCA« Level

50,000

s

21

0.55

tots) or uoclor-jpecific value si a gives station was used.

IPCB» wc*e detected at 3,842 (41%) of the 9,401 stations where collected samples were analyzed foe them.

•For LhU presentation, me Mured level* were compared to risk level* using a default organic carbon content (! %) and default organism lipid content (3%), Use of site-specific organic carbon
would yield slighUy different results.

'Levels used ia the current National Sediment Quality Survey evaluation for human health.

•Levels ujed ia the current National Sediment Quality Survey evaluation for aquatic life (Tier 2).
tovtU uied ia the current National Sediment Quality Survey evaluation for aquatic life (Tier I).

•Tbxte Subaance* Cootrol Act 40 CFR Part 761, Subpart B, f 761.20.

H-2


-------
Figure H-2 and Table H-2 present the comparison of different screening values and the corresponding number of
fish tissue sampling stations with detected levels of PCBs exceeding the screening values. The 10"5 cancer risk level
(Point B) was one of the most conservative thresholds: concentrations exceeded this level at approximately 95
percent of tissue residue sampling stations where PCBs were detected. These sampling stations were clssified as Tier
1 for potential human health risk.

m
o

CL .

Ii

¦B

ss

1.00
0.90
0.80
0.70

0.60

F{x) 0.50
0.40
0.30
0.20
0.10
0.00

I

1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.0QE+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E-K36

PCB In Fish Tlmua Concentration (ppb)

Net«: Lagan Idsrrtfted on Mw 
-------
Appendix II

111 contrast to the PCB evaluation, the evaluation of mercury detected in fish tissue residue in the NSI analysis was
substantially less conservative than that which would result from use of different screening values. To determine the
possible outcomes of different data evaluations, EPA performed additional analyses of mercury fish tissue data included in
the NSI. Figure H-3 aid Table H-3 present six screening values that could be applied for the protection of consumers
ingesting mercuiy-contaminated fish. As shown in these displays, both EPA's current noncancer reference dose recom-
mended for general use (Point E) and the FDA action level (Point D), the screening value used in the current NSI analysis,
result in only about 4 percent of sampling stations with detectable levels classified as posing potential risk to human health.

1.00E-Q2 1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.0OE+O3 1.00E+04 1.00E+05

Concentration (ppb)

NetK Uttart WarSffcd on tfw eurvs eorrMpond to affects ecrwitofl vafcrn in Ta«* H-3.

Figure H-3. Cumulative Frequency Distribution of Mercury Fish Tissue Data for Demersal, Resident, and
Edible Species.

Table H-3. Fish Tissue Sampling Stations with Detectable Levels of Mercury in Demersal, Resident,
Edible Fish Species That Exceed Various Screening Values*-"

Type of Screening Value

Associated Level
(ppb)

Level Plotted in
Figure H-3
Corresponds to Letter

Number of Stations

with Defected
Mercury Exceeding
Level

Percentage of Stations
with Detected Mercury
Exceeding Level

Protection of Consumers

Canadian Guideline*

200

B

908

35.1

Noncanxr Hazard Quotient of 1 (1995)°

1,100

E

91

3.5

Noncancer Hazard Quo tent of 1 (pre-
199 sy

3,231

F

15

0.6

Noncancer Hazard Quotient of 1 (pre-
1995 for Wants)"

646

C

204

7.9

FDA Action Level'

1,000

D

103

4.0

Wadlifc Criteria*

57.3

A

2,150

83.0

•Mercury wat detected u 2489 (90%) of the 2,861 stations where collected samples were analyzed for mercury.

^Canadian guideline limit for mercury in fish that are part of a subsistence diet (Health and Welfare Canada, 1979),

~Methyl mercury reference date ihit was available in IRIS in 1995 (1x10** mg/kg-day).

^Comjpoflds to mercury reference dose available in IRIS prior to 1995 (3x!0"* mg/kg-day).

•Corrtspoods ta mercury reference date available In IRIS prior to 1995 divided by a factor of 5 to protect against developmental effects among infants (6x10-® mg/kg-day). This value was
formerly u*d by the EPA Office of Water.

level uied in Ihe current National Sediment Quality Survey evaluation for human health.

The result* of the wildlife analysis shown In Table 3-5 are slightly different because the data set used for that analysis included demersal, resident species (could be considered edible or not).

H-4


-------
The NSI evaluation restricted the data analyzed to demersal, resident, and edible species. Figure H-4 and
Table H-4 present the same six mercury screening values with the data for all fish species considered edible by
humans with detectable levels of mercury in the NSI. If all edible fish species were analyzed using selected
screening values, 9 percent of sampling stations would be classified as Tier 2 because of mercury contamination
(Point D). However, the proportion of sampling stations with detectable levels of mercury that exceed some
other human health levels ranges from 20 percent to over 55 percent of sampling stations.

I

II

61
*

I

I I

1
*

1.00
OJO
0.80
0.79
0.60
F(x) 0.S0
0.40
0.30
0.20
0.10

0.00
1.0QE-02

t



1.00E-01 1.00E+00 1.00E+01 1.00H+02
Concentration (nib)

1.00E-HS3

1.00E+04 1.00E+OS

No»: Mrtm* ktentPMl oft ifei cemnpend 6a mercury •cfwjrtfta vrtJt* In Tafeia W

Figure H-4. Cumulative Frequency Distribution of Mercury Fish Tissue Data for All Edible Species.

Table H-4. Fish Tissue Sampling Stations with Detectable Levels of Mercury in Edible Fish Species That
	Exceed Various Screening Values**1'	

Type of Screening Value

Associated Level •

(ppb)

Level Hotted in

Figure H-4
Corresponds to Letter

Number of Stations

with Detected
Mercury Exceeding
Level

Percentage of Stations
•with Detected Mercury
Exceeding Level

Protection of Consumers

Canadian Guidefine1'

• 200

B

2,308

55.8

Noncancer Hazard Quotient of 1 (1995)®

1,100

E

353

7.8

Noncancer Hazard Quotient of I (pre-
1995)'

• 3,231

F

3?

0.9

Noncancer Hazard Quotient of I (jpre-
1995 for infants)"

646

C

821

19.9

FDA Action Level'

1,000

D

374

9.0

Wildlife Criteria'

: 57,3

A

3.623

87.6

~Mercury was delected at 4,135 (93%) of the 4,426 stations when? collected samples were analyzed for mercury.

"Canadian guideline limit for mercury in fish that are part of a subsistence diet (Health end Welfare Canada, 1979).

•Methyl mercury reference dose that was available ia IRIS in 1995 (lxl0J mg/kg-day).

Corresponds to mercury reference dose available in IRIS prior to 1995 (SxlO4 mgfltg-day).

¦Corresponds to mercury reference dose available in IRIS prior to 1995 divided by a factor ef 5 to protect against developmental effects among infants (SxlO-1 mg/Vg-doy). This value was
formerly used by the EPA Office of Water.

'Level used in the current National Sediment Quality Survey evaluation for human health.

•The results of ihe wildlife analysis shown in Table 3-5 are slightly different because the data set used for that analysis included demersal, resident species (could be considered edible or not).

H-5


-------
\ |>|)tluli\

H-6


-------
Appendix I

NSI Data Evaluation
Approach Recommended at
the National Sediment
Inventory Workshop,

April 26-27, 1994

T[he original proposed approach for the integration and evaluation of NSI sediment chemistry and biological
data was developed at the Second National Sediment Inventory Workshop held on April 26 and 27, 1994, in
Washington, D.C. The proposed workshop approach was modified, however, to address inconsistencies
found in trying to implement the approach and to address the concerns of the many experts in the field of sediment
quality assessment who commented on the workshop approach. This appendix presents the NSI data evaluation
approach developed by the April 1994 workshop participants. The actual approach that EPA used in the NSI data
evaluation is presented in Chapter 2. A list of workshop participants is provided at the end of this appendix.

Using the approach recommended by workshop participants, sediment sampling stations could be placet! into one
of the following five categories based on an evaluation of data compiled for the NSI:

*	High probability of adverse effects to aquatic life or human health

*	Medium-high probability of adverse effects to aquatic life or human health

*	Medium-low probability of adverse effects to aquatic life

*	Low probability of adverse effects to aquatic life or human health

*	Unknown probability of adverse effects to aquatic life or human health.

Using the workshop approach, contaminated sediment sampling stations could be placed into one of the five
categories based on an evaluation of the following types and combinations of data:

*	Sediment chemistry data alone

*	Toxicity data alone

*	Tissue residue data alone

*	Sediment chemistry and tissue residue data

*	Sediment chemistry and histopath-ological data

*	Sediment chemistry, sediment toxicity, and tissue residue data.

The overall approach developed by workshop participants is summarized in Table 1-1 and is described below.

High Probability of Adverse Effects to Aquatic Life or Human Health

Based on the evaluation approach proposed by the April 1994 workshop participants, a sampling station could be
classified as having a high probability of adverse effects to aquatic organisms or human health based on sediment
chemistry data alone, toxicity data alone, tissue residue data alone, or a combination of sediment chemistry and tissue
residue or histopathological data.

1-1


-------
Appviwlix

"Bible 1-1. Original Approach Recommended by NSI Workshop (April 1994)

Category of
Sanding
Station
Classifications

Data Used to Determine Classifications

Sediment Chemlstiy
(santing station is identified
by

any one of
the following characteristics)



Tissue Residue/
His to pathology



Tbxicity

High Probabffity
of Adverse

Effects lo
Aquatic Life or
Hunan Health

Sediment chemistry values exceed
sediment draft quaky criteria for
any one of the five chemicals for
which criteria have been
dewbped by EPA (based on
measured TOC)

OR

Human health thresholds
for dtoxin or PCBs are
exceeded in resident
species (not a consensus
agreement—participants
evenly divided on this
issue)

OR

Toxicity den»nsirated by
t wo or ma re acute
toxicity tests (one of
which must be a solid-
phase nonnicrobial test)

OR

Sediment chenistiy values exceed
a! relevant AEft (high), ERMs,
PELs, and SQALs for any ore
chemical (can use default TOG)

OR

Sediment chemistry values >50
ppmfbr PCBs

OR

Sediment chemistry TOP exceeds
FDA action levels, EPA risk
levels, or wfldtfe criteria

AND

Tissue bvels in resident
species exceed FDA
action levels or EPA risk
levels, or wiMlfe criteria

	



OR

Elevated sediment chenistiy
concentrations of PAHs

AND

Presence offish tumors

—



Medium-High

Pwbablftyof
Adverse Effects
Jo Aquatic Life
or Hunan
Healh

Sediment chemistry values exceed
at least two of the sediment tipper
screening vahss (Le., ERM,
SQAL, PEL, high AED (can use
default TOG)

OR

Tissue levels in resident
species exceed FDA
action bvels or wildlife
criteria

OR

Toxicity detnansirated by
a single-species toxicity
test (sold-phase,
nonnicrabiaf)

OR

Sediment chemistry IBP exceeds
FDA action levels or wfldfife
criteria

Medium-Low
Probability of

AdwtseEfifects
to Aquatic Life

Sediment chemistry values exceed
one of the lower scteerfng valies
(ERL, SQAL, TEL, tower AET)
(can use defeul TOC and AVS)

OR



	

Toxicity demonstrated by
a singb species toxicity
test (elutriate-phase,
nonnicrobial)

Low Probability
of Adverse

EfTccts to
Aquatic Life or
Human Health

No exceedance of tower
screening values

AND

No sediment chemistry IBP
exceedances of FDA action bvels
or wiMlfe criteria

AND

Tissue bvels in resident
species are tower than
FDA action bvels or
wili life criteria

AND

No toxicity demonstrated
in tests using at bast two
species and at least one
soSd-phase test using

amphipods

Unknown

Not enough data to place a site in any of the other categories.

1-2

I


-------
i\si(i«mil Scdinioiil Qiuilily Survey

For a sampling station to be classified as one with a high probability of adverse effects based on sediment chem-
istry data alone, at least one of three criteria must be met: (1) sediment chemistry values exceed the sediment quality
criteria (SQCs) developed by EPA for acenaphthene, dieldrin, endrin, fluoranthene, or phenanthrene; (2) sediment
chemistry values exceed all appropriate screening values for a given chemical (i.e., high apparent effects thresholds
(AETs), effects range-medians (ERMs), probable effects levels (PELs), and sediment quality advisory levels (SQALs));
and/or (3) sediment chemistry values exceed 50 ppm for polychlorinated biphenols (PCBs). When comparing sedi-
ment chemistry values to the SQCs, measured total organic carbon (TOC) must be used. Workshop participants sug-
gested using default TOC values in the comparison of sediment chemistry values to SQALs if actual measured TOC
values are not available. However, if default TOC values are used in a comparison of sediment chemistry measure-
ments to.SQCs, the highest that a sampling station could be classified would be medium-high potential for adverse
effects.

For a sampling station to be classified as having a high probability of adverse effects based on a combination of
sediment chemistry and tissue residue data, sediment chemistry theoretical bioaccumulation potential (TBP) and tissue
levels in resident, nonmigratory species must exceed FDA tolerance/action/guidance levels, EPA risk levels, or EPA
wildlife criteria. Workshop participants also recommended that a sampling station be classified as having a high
probability of adverse effects if fish tumors are present in resident species and elevated sediment chemistry concentra-
tions for polynuclear aromatic hydrocarbons (PAHs) are present.

The workshop participants were evenly divided on whether a sampling station could be classified as having a high
probability of adverse effects based solely on the exceedance of human health screening values for dioxins or PCBs in
resident fish species. Participants did agree that benthic community data in combination with sediment chemistry data
could be used in the future, but not for the current evaluation, to classify sediment sampling station. Methods are
currently not adequate to establish a direct causal relationship between benthic community changes and sediment
contamination at specific sampling stations without additional data.

For a sampling station to be classified as having a high probability of adverse effects based on toxicity data alone,
toxicity must be demonstrated by two or more acute toxicity tests, at least one of which must be a solid-phase, nonmi-
crobial test.	:

Medium-High Probability of Adverse Effects to Aquatic Life or Human Health

Workshop participants suggested that a sampling station could be classified as having a medium-high probability
of adverse effects on aquatic life or human health based on sediment chemistry data alone, toxicity data alone, or tissue
residue data alone.

For a sampling station to be classified as having a medium-high probability of adverse effects based on sediment
chemistry data alone, the station must meet at least one of two criteria: (1) sediment chemistry values exceed at least
two of the sediment chemistry upper screening values (i.e., appropriate ERMs, SQALs, PELs, or AET-highs) or (2)
sediment chemistry TBP values exceed FDA tolerance/action/guidance levels or EPA wildlife criteria. In the compari-
son of sediment chemistry values to SQALs, default TOC values can be used.

A sampling station could also be classified as having a medium-high probability of adverse effects if toxicity is
demonstrated by a single-species, nonmicrobial toxicity test using the solid phase as the testing medium or if actual fish
tissue residue levels exceed FDA tolerance/action/guidance levels or EPA wildlife criteria.

Medium-Low Probability of Adverse Effects to Aquatic Life

Workshop participants suggested that a sampling station could be classified as having a medium-low probability
of adverse effects to aquatic life based on either sediment chemistry data alone or toxicity data alone. A sampling
station could be classified as having a medium-low probability of adverse effects if sediment chemistry values exceed
at least one of the lower sediment chemistry screening values (i.e., ERL, TEL, SQAL, or AET-low). Workshop
participants suggested that default TOC and AVS values could be used. To classify a sampling station as having a
medium-low probability of adverse effects, toxicity would be demonstrated by a single-species, nonmicrobial toxicity

1-3


-------
test using the elutriate phase as the test medium. Workshop participants did not propose any human-health-related
criteria for placing a sampling station in the medium-low probability of adverse effects category.

Low Probability of Adverse Effects to Aquatic Life and Human Health

Using the workshop approach, for a sampling station to be classified as having a low probability of adverse effects
on aquatic life and human health, all of the following criteria must be met: (1) there are no exceedances of the lower
sediment chemistry screening values (i.e., ERL, TEL, SQAL, or AET-low); (2) there is no toxicity demonstrated in
tests using at least two species and at least one solid-phase test using amphipods; (3) there are no TBP exceedances of
FDA tolerance/action/guidance levels and EPA wildlife criteria; and (4) tissue levels of resident species are below
FDA levels and EPA wildlife criteria.

Unlmown Probability of Adverse Effects

Sampling station of unknown probability for causing adverse effects are those stations for which there are not
enough data to place them in any of the other categories. Sediments at the sampling stations might or might not cause
adverse impacts to aquatic life or human health.

Modifications to Workshop Approach

The approach for evaluating NSI data recommended by the April 1994 workshop participants provides the frame-
work for the final evaluation approach actually used to evaluate the NSI data. Workshop participants had less than 4
hours to reach consensus on their recommendations for the approach following a day and a half of debate covering
many challenging issues. As a result, some of the specific issues concerning how data were to be evaluated to place
sampling stations into the five categories remained unresolved. For example, "elevated sediment chemistry concentra-
tions of PAHs" together with the presence of fish tumors is one criterion for placing a sampling station in the high
probability of adverse effects category. However, how "elevated" do sediment chemistry concentrations of PAHs have
to be to meet this criterion? As another example, sediment chemistry values that exceed all relevant AETs, ERMs,
PELs, and SQAL values for any one chemical are sufficient to place a sampling station in the high probability category,
and exceedance of any two of these values is sufficient to place a sampling station in the medium-high probability
category. But what if there are only two relevant screening values for comparison for a given contaminant? Does a
sampling station at which both values are exceeded for a given chemical belong in the high or medium-high probability
category?

A significant modification in the final approach used to evaluate the NSI data was the reduction in the number of
categories from five to three, eventually combining the medium-high and medium-low categories and the low and
unknown categories proposed in the workshop approach. In addition, the following evaluation parameters were dropped
from the final approach:

•	Sediment chemistry values > 50 ppm for PCBs

-	Expert reviewers of the methodology believed that this parameter was not necessary; i.e., a sampling
station that was targeted as a higher probability for adverse effects by this parameter would already have
been targeted at a much lower concentration using other parameters.

•	Elevated sediment chemistry concentrations of PAHs and presence of fish tumors

-	Available fish liver histopathology data in the NSI are very limited; therefore, this evaluation parameter
was not considered further.

In the final approach adopted for the evaluation of the NSI data, the EPA wildlife criteria were not included in the
TBP and fish tissue residue parameters. Reviewers of the methodology felt that the wildlife criteria values were overly
conservative for this screening assessment and thus could not be used to distinguish potentially highly contaminated
sampling stations from only slightly contaminated station. A separate analysis of wildlife criteria was, however,
conducted.

1-4


-------
National Ni 
-------
Appendix I

i

i

Dom DiToro
Manhattan College
Environmental Engineering
Bronx, NY 10471

(718) 920-0276; Fax (718) 543-7914

Bob Engler
COE-WES

3909 Halls Ferry Road
Vicksburg, MS 39180-6199
(601) 634-3624

Jay Fields
NOAA/HAZMAT
7600 Sand Point Way, NE
Seattle, WA 98115
(206) 526-6404

Catherine Fox

EPA/OST (4305)

401 M Street, SW

Washington, DC 20460

(202) 260-1327; Fax (202) 260-9830

Tom Fredette

COE New England District

424 Trapels Rd.

Waltham, MA 02254

(617) 647-8291; Fax (617) 647-8303

Marilyn Gower
EPA Region 3
2530 Riva Rd., Suite 300
Annapolis, MD 21401
(410) 224-0942

Dave Hansen

EPA ERL-Narragansctt

27 Tarzwell Dr.

Narragansett, RI 02882

(401) 782-3027; Fax (401) 782-3030

Jon Harcum

Tetra Tech, Inc.

10306 Eaton PL, Ste.340

Fairfax, VA 22030

(703) 385-6000; Fax (703) 385-6007

Rick Hoffmann

EPA/OST (4305)

401 M Street, SW

Washington, DC 20460

(202) 260-0642; Fax (202) 260-9830

Bob Hoke
SAIC

411 Hackensack Ave.

Hackensack, NJ 07601

(201)	489-5200; Fax (201) 489-1592

Chris Ingersoll
NBS

Midwest Science Center
4200 New Haven Rd.

Columbia, MO 65201
(314) 875-5399

Doug Johnson

EPA Region 4

345 Courtland Street, NE

Atlanta, GA 30365

(404) 347-1740; Fax (404) 347-1797

Ken Klewio

EPA Region 5 (WS-16J)
77 W. Jackson Blvd.

Chicago, 1L 60604

(312)	886-4679; Fax (312) 886-7804

Fred Kopfler

Gulf of Mexico Program, Bldg. 1103
S tennis Space Center, MS 39529
(601) 688-3726; Fax (601) 688-2709

Paul Koska
EPA Region 6
1445 Ross Ave.

Dallas, TX 75115
(214) 655-8357

Mike Kravitz
EPA/OST
401 M Street, SW
Washington, DC 20460

(202)	260-8085

Peter Landrum
Great Lakes ERL
2205 Commonwealth Blvd.
Ann Arbor, MI 48105

(313)	741-2276

Matthew Liebman

EPA Region 1

JFK Federal Bldg., WQE

Boston, MA 02203

(617) 565-4866; Fax (617) 565-4940

email: bays@epamail.epa.gov

1-6


-------
Niilioiiiil Sc
-------
Jerry Stober

ESD-Athens

College Station Rd.

Athens, GA 30113

(706) 546-2207; Fax (706) 546-2459

Rick Swartz
EPA ERL-Newport
Hatfield Marine Science Center
Marine Science Drive
Newport, OR 97365
(503) 867-4031

Nelson Thomas
EPA ERL-Duluth
6201 Congdon Blvd.

Duluth, MN 55804
(218) 720-5702

Rachel Friedman-Thomas
Washington Dept. of Ecology
Mail Slot 47703
Olympia, WA 98504-7703
(206) 407-6909; Fax (206) 407-6904

Bumell Vincent

EPA/ORD

401 M Street, SW

Washington, DC 20460

(202) 260-7891; Fax (202) 260-6932

Mark Wildhaber
NBS

Midwest Science Center
4200 New Haven Rd.
__ Columbia, MO 65201
(314) 876-1847

Craig Wilson
California SWRCB
901 P Street
Sacramento, CA 95814
(916)657-1108

Drew Zacherle

Tetra Tech, Inc.

10306 Eaton PL, Ste. 340

Fairfax, VA 22030

(703) 385-6000; Fax (703) 385-6007

Chris Zarba
EPA/OST (4304)

401 M Street, SW
Washington, DC 20460
(202) 260-1326

Xiaochuo Zhang, WR/2

Wisconsin DNR

P.O. Box 7921

Madison, WI 53707

(608) 264-8888; Fax (608) 267-2800

1-8

rtu.s. OOVERHffiOT PRINTING OFFICE: 1996-619-098/9(1685


-------